This electronic theses or dissertation has been

downloaded from the King’s Research Portal at

https://kclpure.kcl.ac.uk/portal/

The musical language of Elliott Carter : analysis of selected works from the transitional
period (1945-55).

King, Samantha Jayne

The copyright of this thesis rests with the author and no quotation from it or information derived from
it may be published without proper acknowledgement.

END USER LICENSE AGREEMENT

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported

License. http://creativecommons.org/licenses/by-nc-nd/3.0/

You are free to:

 Share: to copy, distribute and transmit the work

Under the following conditions:

 Attribution: You must attribute the work in the manner specified by the author (but not in any
way that suggests that they endorse you or your use of the work).
 Non Commercial: You may not use this work for commercial purposes.
 No Derivative Works - You may not alter, transform, or build upon this work.

Any of these conditions can be waived if you receive permission from the author. Your fair dealings and
other rights are in no way affected by the above.

Take down policy

If you believe that this document breaches copyright please contact librarypure@kcl.ac.uk providing
details, and we will remove access to the work immediately and investigate your claim.

Download date: 08. Nov. 2015

 THE MUSICAL LANGUAGE OF ELLIOTT CARTER:
ANALYSIS OF SELECTED WORKS FROM THE
TRANSITIONAL PERIOD (1945-55)

Thesis submitted in partial fulfilment of the examination requirements for the degree
of DOCTOR OF PHILOSOPHY
1998

S.LT. King
(King's College London)
S.J.T.King

King's College, London

Description of PhD thesis:

THE MUSICAL LANGUAGE OF ELLIOTT CARTER: ANALYSIS OF

SELECTED WORKS FROM THE TRANSITIONAL PERIOD (1945-55)

The music of Elliott Carter is held in high regard, but is noted for its complexity.
Recent analytical studies have attempted to explain Carter's compositional method,
often taking the composer's own writings as a starting point. These studies have
concentrated on works from the later part of his career, which appear to be the
products of consistently applied ideas and techniques. This "mature style" may be
regarded as beginning with the composition of the Second String Quartet (1959). The
purpose of the present study is to complement work on the later music by examining
the works of the ten-year period from the Piano Sonata (1945) to the Variations for
Orchestra (1955), during which the composer's style underwent radical change from
"American neoclassical" to "avant-garde". The study begins with a discussion of the
historical and aesthetic context impinging upon Carter's work in the earlier part of his
career, including the concepts of neoclassicism, modernism and populism. This
discussion will explore the reasons for the composer's dissatisfaction with the musical
idiom of his pre-1945 works. This is followed by a chapter placing the study of
Carter's work in the context of analytical studies of twentieth-centuiy music,
particularly those which examine the notion of transitional music. The purpose of this
is to identify those analytical methods and techniques appropriate to the repertoire
under consideration. The methods discussed here will play a part in the analyses
which follow. These begin with an investigation of various aspects of Carter's
technique, using relatively brief examples, and culminate in studies of complete
movements, namely the first movement of the Piano Sonata, the second movement of
the Cello Sonata (1948) and the first movement of the Sonata for Flute, Oboe, Cello
and Harpsichord (1952).

2
ACKNOWLEDGEMENTS

During my initial period of research, I received the following financial support, for
which I am extremely grateful: a grant from the British Academy, a Studentship from
the Music Department at King's College London and grants from the Worshipful
Company of Musicians and the Sir Richard Stapley Memorial Trust.

The interest and support of colleagues, contemporaries and members of staff, first at
King's College London and later at Chetham's School of Music, have been invaluable,
but my special thanks are due to two individuals. Dr.Philip Grange first introduced
me to the music of Elliott Carter as a compositional model and a subject for analysis.
Professor Arnold Whittall supervised my work, broadened my analytical outlook and
made innumerable suggestions for the improvement of this thesis.

Finally I must thank my wife Nicola and other members of my family for their
assistance, perseverance and tolerance during the long gestation of this work.

3
CONTENTS

Abstract 2

Acknowledgements 3

List of Tables 8

List of Musical Examples 9

Part One: Background Issues 13

Chapter 1: Introduction 14

Chapter 2: Historical and Aesthetic Background 21

1. Expressionism and Neoclassicism 22

2. Modernism and Conservatism 28

3. Elitism and Populism 36

4. Cosmopolitanism and Americanism 43

Chapter 3: Analytical Context 47

1. Analysis of Transitional Music 47

2. Analysis of Neoclassical Music 56

3. Analysis of Carter's Music 64

Part Two: Elements of Carter's Musical Language in the 70
Transitional Period

Chapter 4: Keys, Modes, Fields 71

1. Diatonic fields 71

(a) The Rose Family 77
(b) Piano Sonata, second movement, bb. 1-26 81
(c) Piano Sonata, second movement, bb.27-52 84

2. Chromatic complementation 87

Chapter 5: Chords and/or Sets 89

1. A note on pitch-class
set genera 93
2. "Key-chords" 96

3. Diatonic sets 100

(a)The Rose Family 103

(b) Dust of Snow 106
(c) Pastoral, bb. 1-45 109

4. Non-diatonic harmonic principles 113

(a) Triadic alteration and superimposition 114

(b) Combination of intervals 118

4
 (i) Piano Sonata, second movement,

bb. 1-26 121
(ii) Cello Sonata,
second movement,
bb. 91-1 12 123
(iii) Eight
Etudes and a Fantasy,
Etude 1 123

(c) The Kh(6-20) subcomplex 124

(i) Cello Sonata, first movement 126

(ii) Quartet Sonata, first movement 129

(d) The all-interval tetrachords 130

Chapter 6: Voice-leading
and Linear Motion 134
1. Foreground 136

(a) The tonic-dominant relationship 137

(b) The Neighbour Note 141

(c) Linear Progression 145

(d) Arpeggiation 149

(e) Octaves 153
2. Middleground and Background 156

Chapter 7: Themes, Motives, Rows 160

1. Introduction 160

2. Exposition 168
(a) The Rose Family, bb.3-15 169
(b) Piano Sonata, second movement, bb. 1-26 170
(c) Piano Sonata, first movement, bb. 1-32 172
(d) Quartet
Sonata, second movement, bb.69-99 174
3. Development 178

(a) Interval expansion and contraction 178
(i) Piano Sonata, first movement 178

(ii) Piano Sonata, second movement 179
(iii) Cello Sonata, second and fourth movements 180
(iv) First String Quartet, Part ifi 181
(v) Quartet Sonata, first movement 181

(b) Polyrhythmic combination 182
(i) First String Quartet, Part ifi 182

(ii) Quartet Sonata, third movement 183

(c) Acceleration and Deceleration 183
(i) First String Quartet, Part ifi 184

(ii) Variations for Orchestra 185

(d) Quasi-serial procedures 186
(i) Piano Sonata, second movement, bb. 104-329 187

(ii) Cello Sonata, first movement 188

(iii) Quartet Sonata, third movement 188

4. Recapitulation 189

5

Part Three: Analytical Studies 193

Chapter 8: Piano Sonata, first movement 194

1. Populism and other
influences 195
2. "Sonata Form" 201

3. Harmonic materials: structural consequences 207
4. Exposition, bb. 1-82 212
(a) Bars 1-32 213
(i) Maestoso bb. 1-14 214
(ii) Legato correvoler.bb.14-23 217

(iii) A tempo, maestoso
bb.24-32 219
(b) Scorrevo1e bb.32-82 221

(i) Register 221

(ii) Linear motion 223
(iii) Tonality and diatonic field 227

(iv) Thematic process 230

(c) Summary 233

Chapter 9: Sonata for Cello and Piano, second movement 235

1. Genre and Character 235

2. Form and Thematic Process 239
A (bb.1-57) 239
B (bb.57-129) 242
A' (bb.130-213) 243

3. Harmony, Tonality and Sets 247

4. Conclusion 261

Chapter 10: Sonata for Flute, Oboe, Cello and Harpsichord,

first movement 262

1. Introduction 262

2. Form: processes 263

(a) Dynamics 265

(b)Pulse 267

(c) Textural continuity, density and register 270

(d) Summary 274
3. Motivic and harmonic material 276

(a) Rhythmic and pitch-contour cells 277

(b) The "key-chord" and its relatives 279

4. Large-scale pitch-organization
285
(a) Thematic recall 285

(b) Pitch fields 286

(c) Linear and registral connections 287

5. Conclusion 288

Chapter 11: Conclusion 290

6

Appendices 293

Appendix 1: Cross-Reference of Carter's Listing of Three- to

Six-Note chords with Forte's List of Prime Forms of Pitch-Class Sets 293

Appendix 2: Classification of Diatonic Sets 297

Appendix 3: Formation of Tetrachords from pairs of non-overlapping

dyads 298

Bibliography 304

7
LIST OF TABLES

Table

2.1 Aesthetic axes 21

3. 1 Whole-tone invariance/dominant prolongation in Scriabin 51

5.1 Similarity relations between pentachords appearing in The Rose Family 105
5.2 Similarity relations between "quasi-seventh" chords 116
5.3 Comparison of interval vectors for tetrachords formed from id +ic3 and
ic3+ic5 120

6.1 Intervallic structure of Cello Sonata, III, bb. 1-4 152

7.1 Acceleration of "Passacaglia" idea in String Quartet no.1, Part HI 185

7.2 Acceleration of scalic "row" idea in String Quartet no.1, Part III 186

9.1 Relationship between material in A and A' sections of Cello Sonata, II 244
9.2 Tetrachordal vocabulary of Cello Sonata, II 250
9.3 Trichordal vocabulary of Cello Sonata, II 251
9.4 Similarity relations between tetrachords in Table 9.2 253
9.5 Similarity relations between trichords in Table 9.3 254
9.6 Succession of Diatonic Fields in A section of Cello Sonata, H 257

10.1 Set-class vocabulary of Quartet Sonata, I 281

10.2 Statistical summary of trichords and tetrachords shown in Ex.10.11 284
10.3 Pitches occurring in Ex. 10.11 286

8
LIST OF MUSICAL EXAMPLES

(CS=Cello Sonata, PS=Piano Sonata, QS=Quartet Sonata)

4.1 The Rose Family diatonic fields 310
4.2 PSIIII1-26 diatonic fields 313
4.3 P5/1111-2, 25-6 melodic contour 314
4.4 (a) PSIIIJ1-15 interaction of modes on d and g 314
(b)PSIIII1-15 pitch relationships in upper voice melody 314
4.5 PS111127-52 diatonic fields 315

4.6 PSII/83-5 chromatic complementation 317
4.7 CS/I/l -6 chromatic complementation 317

5.1 The Rose Family pc set vocabulary 318

5.2 Dust of Snow pc set vocabulary 319

5.3 Pastoralil-45 pc set vocabulary 320
5.4 (a) PSIII/412-4 appearances of set 5-27 and subsets 321
(b) PSII/20-32 321
5.5 (a) PS11J245-51 use of set 5-21 322
(b) PSII/262-3 322
5.6 Voyage final chord 322

5.7 Warble for Iilac-time/90-127 quasi-triadic sets 323

5.8 PS/I/i 09-116 quasi-triadic sets 324
5.9 QSII/9-16 pc set vocabulary 324
5.10 PSIIII1-26 set vocabulary/harmonic motives 325
5.11 CS/11191-1 12 set vocabulary/harmonic motives 327
5.12 Eight Etudes/I set vocabulary/harmonic motives 328

5.13 "Chart 9" from Schiff The Music of Elliott Carter
330
5.14 CSII/118-129 use of Kh(6-20) 330
5.15 QS/I/1 use of Kh(6-20) 331
5.16 QS/I/23-4 use of Kh(6-20) 331
5.17 QS/I/30-68 use of Kh(6-20) 332
5.18 (a) First Quartet/ffl1430-43 use of 4-Zi5 332
(b) First Quartet/ffl1449-55 332
(c) First QuartetJllhJ4l6-7 332
5.19 (a) First Quartetllll 1-12 use
of 4-18 and 4-Z29 333
(b) First QuartetJIJ22-30(V1n2) 333
5.20 (a) First Quartetlffl/86-8 use of 4-18 334
(b) First Quartet/ffl/122-4 334
(c) First Quartet/1111146-8 334
(d) First Quartet/ffl1228-30 334
(e) First Quartetfffli276-9 334

5.21 First Quartetfffl/392-402 pc set vocabulary 335
6.1 (a) The Rose Family7'4-7 alternation of I and V 336
(b) Pastoral/5-8 336
(c) Voyage/1-3 336

(d) Dust of Sno123-4 336
6.2 (a) Pastoral/275-9
V-I cadence 337

6.2 (b) PSII/78-83, 101-2 v-I cadence 337

(c) CS/Ill, 67, 74-5, 97-8 338
6.3 (a)PSIII/23-6, 1-3 divider V 338
(b)CS/IT/35-44 339
6.4 (a) Dust of Snowf28-32 neighbour notes 340
(b) PS/IJ1-7 340
(c) CSIIHI12-21, 29-39 340
6.5 (a) PS/11297-303 fifth-relationships/neighbour notes 342
(b) CS11168-70 342
(c) CS/1J33-6 342
(d) CS/I/118-129 343
6.6 (a) The Rose Family/22-25 linear motion in parallel 344
(b) Warble for Lila c-time/90-1 06 344
(c) Voyage 345
(d) PS/11i15-26 346
(e) PS/V 163-6 347
(f) PSII/232-243 347
(g) CSII/19-21, 27-9 348
(h) First QuartetJllhl397-400 348
(i) CS/I/55-7 348
(k) QSIII/176-7 348
6.7 (a) Warble for Iilac-time/1-4 quasi-scalic patterns 348
(b) PS/L'44-8 348
(c) CS/ITI/4-5 349
(d) CS/IV/26-9 349
(e) First Quarteilffl/443-445 349
(f) QSuffl/432-5 349
6.8 (a) Warble for Lila c-time/53-6 arpeggiation 349
(b) PS/L1279-282
349
6.9 (a) PS/1183-5 intervallic process 350
(b) CS/ll/68-72 350
(c) CS/fflul-4 351
6.10 (a) CS/111136-140 quasi-arpeggio patterns 352
(b) First Quartetlfflhl95-8 352
(c) QS/I116-17 352
(d) QS/1111329-3 1 352
6.11 (a) PS/I/32-4 foreground octaves 352
(b) CS/1l83-6 353
(c) CS/111197-9 353
(d) Eight Etudes/Ill 6-18 353
(e) QSIHI143-50 353
(f) First Quartet/ffl189-90 354
(g) QSIL/1 8-20 354
6.12 PS/111103-112 octave progression 354
6.13
PS/Ill 1-52 middleground 355
7.1 The Rose Familyf3-15 phrase structure 356
7.2 PS/Hhl-26 phrase structure 357

10


7.3 PSIIJ1-32 phrase structure 358
7.4 QSIII/69-99 phrase structure 360

7.5 (a) PSII/83-8 intervallic alteration 361
(b) P5/1/156-60 361
(c) PS/L1205-13 361

7.6 (a) PSIIIJ68-73, 76-105 intervallic alteration 362
(b) PSIIII1-2, 362-3, 381-2 362
(c) P5/1113-5, 21-3, 364-96 363

7.7 CSIH/61-3, 67-9,
74-6 intervallic alteration 363
7.8 CSIIVI1 14-7 363

7.9 First QuartetJilhJ28l-9, 463-8 intervallic alteration 364
7.10 First Quartetfffl/469-73, 477-84 intervallic alteration 364

7.11 QS/J/1-7 (hpschd.) intervallic alteration 365
7.12 (a) First Quartet/111213, 111/77-80, 82-4 polyrhythmic combination 365
(b) First Quartet/ffl1181-4, 209-11 366
(c) First Quartetlffll263-9, 294-5 366

7.13 QSIffl/209-1 1, 347-50 polyrhythmic combination 367
7.14 First Quartetlffll2-12 vc., etc. acceleration of "Passacaglia" 367
7.15 First Quartetlfflhl99-218 ye., etc. acceleration of scalic "row" 368
7.16 Variations for Orchestra "Ritomello B" and "Ritomello A" 369
7.17 (a) PSIIIJ1O3-12 fugue subject as row 371
(b) 169-74 371
(c) 290-4 371
7.18 P5/11/169-89 stretto/serial manipulation 372
7.19 PS/il/209-26 white-note episode 373
7.20 (a)CS/L16-21 thematic "row" 374
(b) 22-32 374
(c) 78-82 374
(d) 105-115 374
7.21 QSIffl/205-213, 349-358 thematic "row" 375
8.1 (a) Copland PS/IT/1-15 376
(b) Carter PS/11176-103 376
8.2 (a) Copland PS/1/206-14 377
(b) Carter P5/11243-55 377
8.3 (a) Copland PSII/123-130 378
(b) Carter PSII/102-4 378
8.4 (a) PS/1/123-8 harmonics 379
(b) PS/11/388-92 379
8.5 Symmetrical partitioning of set 6-32 380
8.6 PS/I thematic appearances of 4-23 380
8.7 (a) PS11183-5 contrasting uses of 4-23 381
(b) PS/1/102-4 381

8.8 PS/1J1-123 middleground manifestation of 3-3 381
9.1 CS/il/A main material 382
(i) 1(vc.) 382
(ii) 10-11, 35-6 (piano) 382
(iii)12 (piano, RH) 382

11
9.2 CS/11h15-35 phrase structure 383
9.3 CS111135-50 phrase structure 385
9.4 CS/1115 1-7 transition to B section 384
9.5 CS/IT/B main material 386
(iv) 57-8 (vc.) 386
(v) 61-3 (piano, RH) 386
(vi) 60 (piano, RH) 386
9.6 CS/il/i 13-122 pc set structure 386
9.7 CS/W1204-7 pc set structure 387
9.8 (a) CS/1117 (vc.) use of ic3+ic3 tetrachords 389
(b) 10-11 (piano) 389
(c) 11-14 (piano, RH) 389
(d) 208-13 (vc.) 389
9.9 CS/1114-10 use of ic5+ic5 tetrachords 389
9.10 (a) CS/11178-81(vc.) use of 8-28 390
(b) 83-5 (vc.) 390
9.11 CS/il190-1 12 focal use of 4-28 392
10.1 QS/1/1-8 (hpschd.) main rhythmic/pitch-contour motive 393
10.2 QS/1Ji-8 (fit., ob., ye.) manipulation of harpsichord motive 393
10.3 QSIIJ1-8 thematic use of melodic semitone 394
10.4 QS/1I1-8 " " " perfect 5th 395
10.5 QSII/1-8 " " major 10th 395
10.6 QSIIII-8 pc set 3-2 395
10.7 Schiff 1983: "Chart 12" (p.165) 395
10.8 QS/IIl-8 (hpschd.) pc set structure 396
10.9 QS/1/1-8 (fit., ob., ye.) pc set structure 397
10.10 (a) QS/112-3 (ob.) development of 3-2 and 6-Z10 as 399
(b) 5-6 (ob.) thematic feature 399
(c) 16 (hpschd.) 399
(d) 21-2 (oh., fit.) 399
(e) 30-2 (hpschd.) 399
10.11 QSIII9-15 (hpschd.) trichordal and tetrachordal vocabulary 400
10.12 QSIII9-15 (fit., oh., ye.) " 401
10.13 (a) QSIII1-2 (b) 16-17 thematic recall 401
10.14 (a) QSIIJ19-20 (b) 25-6 402
10.15 (a) QS/1142-3 (oh.), 1-2 (fit., ob.) 402
(b) 53-7 (fIt.), 4-6 (fit.) 402
(c) 59 (ye.), 2-3 (ob.) 402
10.16 QSIIJ18-20 (fit., ob.) octave transfer 402
10.17 QSIII18-20 (fit., ob., ye.) octave transfer 403
10.18 (a) QSIII 1-7 (fit., ob.) middleground use of 3-2 and 4-1 403
(b) 16 (fit., ob.) 403

12
PART ONE: BACKGROUN) ISSUES

13
CHAPTER 1: INTRODUCTION

The purpose of this study is the examination of a crucial period in the

development of one of the major composers of the twentieth century. The sheer

range and quality of Carter's achievement is a recommendation for attentive study. In

a compositional career of approximately seventy years, Carter has written works in all

the established genres of western music, has tackled (and synthesized) many of the

central ideas of modernism and has written extensively on a broad range of musical

and artistic topics.

Since the 1950s and '60s, Carter has held a unique position in contemporary

musical life, playing a dual role of forward looking radical and traditionally rooted

elder. He represents a line of continuity with the mainstream of musical modernism

(Schoenberg, Stravinsky, Debussy), but he is by no means the bunkered conservative

that this characterization might imply since he has engaged with some of the principal

issues of radical modernism in the arts, such as the plural nature of experience and the

importance of discontinuity and elliptical logic. He has attempted to address these in

his music while maintaining a fundamentally traditional concept of the integrity of a

work of art and the role of the author. That Carter has managed to produce works

displaying a comprehensive command of large-scale form, without recourse to what he

would see as the regressive easy options of minimalism or neo-romanticism (that is to

say, without sacrificing his modernism), is possibly his most significant achievement.

It is also significant that he has achieved this on his own. Characterized as an

"individualist" by Arnold Whittall [Whittall 1977: 2121, Carter is neither the product

14
of, nor the founder of a compositional "school". Although often bracketed with

Olivier Messiaen and Michael Tippett because of the similarity in age and range of

achievement between him and his two great contemporaries, Carter's aesthetic

background is quite different. Unlike Messiaen's, Carter's achievements have not

been rooted in the expression of fundamental religious belief, but in a consistent

humanism. In this respect, Carter is more typical of later twentieth-century artists

than the Frenchman. However, unlike Tippett, Carter has not sought to develop

philosophical or psychological ideas explicitly in his works, but has expressed himself

in purely musical terms. This has perhaps made him the "composer's composer" of

the late twentieth century; a figure whose professional consistency and ever-fertile

creative imagination provide an inspirational model for his contemporaries and for the

younger generation.

It is, as Whittall states, "the extent and direction of Carter's stylistic evolution"

[Whittall 1977: 2121 which is particularly remarkable, prompting Pierre Boulez to

describe Carter as "an astonishing example of an individual musician, a personality

who found his own way, quite unexpectedly - as it always should be" [Boulez 1978:

8 1. This unexpectedness results from the extreme dissimilarity between Carter's

stylistic origins, which owe much to Copland and the neoclassical Stravinsky, and his

subsequent development, which placed him at the forefront of the post-war avant-

garde, alongside such figures as Boulez himself. Since the 1960s, Carter has

developed a personal repertoire of compositional techniques and resources with

increasing consistency. These include large-scale form-defining polyrhythms, twelve-

note all-interval "key-chords", stratification of harmonic and rhythmic vocabularies,

cross-cutting of formal sections and "metric modulation". However, during the 1940s

15
and '50s, Carter's ideas and methods were in a state of flux; principles of composition

deriving from various sources were brought into conjunction, resulting in a fascinating

interplay between the traditional and the innovative. This is not to say that the works

of 1945-55 are significant only because they exhibit "transitional" features. Works

such as the Piano Sonata (1945), the First String Quartet (1951) and the Variations for

Orchestra (1955) are triumphant solutions of the compositional problems Carter set

himself at the time and are masterpieces in their own right, not merely harbingers of

the later Carter.

The period under consideration is dominated by chamber works. Like

Schoenberg and some other composers during periods of radical stylistic

transformation, Carter seems to have felt the intimate world of the sonata and string

quartet to be a more appropriate medium for intense self-examination and

experimentation than were the public domains of orchestral or stage music. The

ballet The Minotaur (1947), may be regarded as a relatively conservative throwback

to, or summation of, Carter's earlier style, while the Variations for Orchestra come

towards the end of the period and may be seen as standing somewhat outside the main

line of Carter's development, again because of its use of a relatively conventional form

and also because of its reference to the styles of other composers. Other works

include "etude" types [Harvey: 201 in which Carter concentrated on the smallest

individual elements of musical language. The Eight Etudes and a Fantasy (1949) for

Woodwind Quartet and the Six Pieces for Timpani (1950) exemplifr this category.

The massive First String Quartet (1951) is the linchpin of the period; a remarkable

synthesis of the composer's ideas to that date and his first mature masterpiece.

However, this too is a unique work in Carter's output by virtue of its 45 minute length

16
(most of Carter's subsequent major works occupy approximately 20-25 minutes.)

This study will examine a range of chamber works from the period, including

the Pastoral for viola and piano (1940), the songs "Dust of Snow" and "The Rose

Family" from Three Poems by Robert Frost (1942), Warble for Lilac-time (1943),

Voyage (1943), the Eight Etudes and a Fantasy and the First String Quartet, but will

focus principally on the three Sonatas (for Piano (1945), Cello (1948) and Flute, Oboe,

Cello and Harpsichord (1952), hereafter referred to as the Quartet Sonata.) The

Sonatas present a convenient unit for study as continuity of genre and scale is

combined with an even chronological distribution and a clear stylistic sequence. The

Piano Sonata has been described by Schiff as a work which "looks in two directions"

[Schiff 1983: 123], being both a summary of elements of Carter's earlier style, and a

transitional work in that it prefigures elements of his mature style. The Cello Sonata

is more obviously transitional; it is the last of Carter's works to employ a key

signature, the first to employ the device of metric modulation and the first to exploit

the contrast between instrumental "characters". The Quartet Sonata was written in

the immediate wake of the First Quartet and embodies some of Carter's newly-forged

techniques, but in a more relaxed manner.

Carter's use of the title "Sonata" would seem to denote a multi-movement

instrumental work, free of any specific literary or programmatic elements - at least,

these are the features that Carter's three Sonatas have in common. However, their

individual relationships with the body of tradition that the term "sonata" invokes are

quite different. The Piano Sonata maintains links with a Beethovenian ideal - the two-

movement form, with a turbulent sonata structure balanced by a serene slow-

movement, perhaps suggests a parallel with Op. 111, while the use of fugue and the

17
 opposition of tonal poles of b flat and b natural evoke the "Hammerkiavier" Sonata.

The Cello Sonata, like its predecessor, employs cyclic themes, but aims at a

neoclassical balance of movements (perhaps recalling Hindemith) - fast (ternary),

scherzo, slow, fast (rondo). The Quartet Sonata, however, leaves classical models

further behind. In the relationship of its three movements, the emphasis is on

evolution and growth rather than balance and symmetry. Each movement is longer

and formally, rhythmically and texturally more complex than its predecessor.

Furthermore, there are no cyclic recurrences of thematic material between movements.

Indeed, the Quartet Sonata seems freer in form than almost any of Carter's other

works from the transitional period. An acknowledged influence was that of Debussy,

whose Sonata for Flute, Viola and Harp may be the work's closest "ancestor" [CEL':

229J.

An essential part of the study will be the evolution of an analytical method

flexible enough to enable the demonstration of both contrasts and continuities between

the three works; in other words, a method suitable to the study of "transitional" music.

This will necessarily involve a discussion of the nature of transitional music and of

current analytical approaches to it. The work of other analysts of Carter's music will

be considered in this context, as will Carter's relationship to other composers of

transitional music.

1
CEL denotes a reference to Elliott Carter: Collected Fssays and Lectures, 1937-1995
edited by Jonathan W.Bernard (Rochester, N.Y.: University of Rochester Press, 1997).
This volume updates the previous collection of Carter's writings, The Writings of Elliott
Carter: An American Composer Looks at Modern Music, compiled, edited and annotated
Else and Kurt Stone (Bloomington and London: Indiana University Press, 1977),
represented by the abbreviation WEC in the current text. References to Carter's writings
will identify their appearance in the later collection except in the case of some shorter
articles and programme notes which Bernard excludes.

18
 An important preliminary to the theoretical and analytical part of the study will

be an examination of the historical and aesthetic background to Carter's stylistic

transition. An investigation of formative influences and the contemporary context will

enable the analyses to be carried out with due sensitivity to the composer's unique

situation and his response to it. The structure of this thesis will, therefore, be as

follows, tracing a line from the general to the particular:

Part One: a discussion of the historical and aesthetic background to Carter's

transitional period and of analytical and theoretical issues germane to it; this will

include an investigation of the nature of transitional music and of appropriate analytical

approaches to it, a comparison of Carter's transitional period with those of other

twentieth-century composers and an evaluation of the work of other analysts of

Carter's music.

Part Two: an overview of aspects of musical vocabulary and syntax in works of

the period, focusing primarily on the role of harmonic and thematic techniques and

processes in the creation of large and small scale structures. This section will deal

with individual aspects of Carter's musical language (pitch fields, voice-leading and

thematic structure), using brief examples from a wide range of works from the period.

The areas of rhythm and form will not receive separate attention here for the following

reasons. Carter's rhythmic innovations, probably the most obviously new aspect of

his musical language, have already been discussed more than adequately elsewhere

(see Schiff 1983 and Bernard 1988). It is difficult, perhaps meaningless, to discuss

"form" in isolation from the elements of musical "content", and it will become obvious

in investigating the various techniques Carter employs that they all have formal

implications.

19
 Part Three: analytical studies of movements from the three Sonatas, namely the

first movement of the Piano Sonata, the second of the Cello Sonata and the first of the

Quartet Sonata. This section will draw on the concepts outlined in Part Two, and

show how particular techniques function in the context of specific works.

20
 CHAPTER 2: HISTORICAL AND AESTHETIC BACKGROUND

An investigation of Carter's stylistic transition should begin with an overview

of the foundations of his musical language in the mid-1940s, which, in turn, requires

an understanding of the historical and aesthetic background. This is a large and

complex area. The remarkable course of Carter's development was the result of the

interaction of a wide variety of forces, none of which was experienced by Carter

alone, but whose combination produced a unique reaction in his creative

consciousness, causing him to metamorphose from being merely a child of his time to

become one of its prime movers. The main sources of information regarding Carter's

reactions to his milieu are his collected writings (especially the earlier WEC, see

Chap. 1, footnote 1), which include many of the reviews he wrote for the periodical

M,dern Music in 1937-1944, and the series of extened interviews with Allen

Edwards1 , compiled during 1968-70 and published in 1971.

The aesthetic issues affecting Carter's development may be summarized by

representing them as opposing poles on axes of artistic principles (see Table 2.1).

Table 2.1 Aesthetic axes

1. expressionism neoclassicism
2. modernism conservatism
3. elitism populism
4. cosmopolitanism Americanism

'Edwards, Allen Fla ued Words and Stubborn Sounds: A Conrsation with Elliott
Carter (New York: W.W.Norton, 1971).

21
Such a representation necessarily demands explanation and qualification. For

example, it is by no means justifiable to equate neoclassicism exclusively with

populism; Stravinsky hardly saw himself as an educator of the masses. However, this

schematic representation provides a useftil framework for considering Carter in

relation to the major artistic issues of the inter-war years.

1. Expressionism and Neoclassicism

Carter was born in 1908, just as the expressionist movement in the arts was

beginning to reach its height. He therefore came to maturity during a period when

various forms of reaction against expressionism, whether artistically or politically

motivated, were prevalent. His own writings and reminiscences reveal an equivocal

relationship with expressionism which was not resolved until the late 1950s.

The polemical confrontation which developed between senalism and

neoclassicism, with Schoenberg and Stravinsky as the central protagonists, may, by the

1950s and 'óOs, have come to seem the result of opposing solutions to common

compositional problems (especially after Stravinsky's late "conversion" to serialism),

but for composers of Carter's generation, that confrontation was the central issue of

the 1920s and '30s. During Carter's youth, artistic life in New York was

characterized by an atmosphere of high-minded devotion to art for its own sake. The

cause of modern music was well served by the adventurous programming of Stokowski

and the activities of the International Composers' Guild and the League of Composers.

As a young man, it was music of the "ultra-modern" school which first fired his

imagination and prompted his decision to become a composer. Most of this music

22
may be regarded as "expressionist" or "experimental":

During this whole period.. .1 was much more interested in the very
advanced music of the time than I was in anything else and thus came to
know the music of Ruggles and Ives, particularly, and Bartok and
Stravinsky and the three Viennese - all the new music that was being
done in New York in this, a very active period. [Edwards: 411

I can't give a date, [for the decision to become a composer] but

certainly The Rite of Spring was a very important and meaningful work,
as were several of the works of Varèse like Intégrales and Octandre,
and certainly the later works of Scriabin, particularly Le Poeme de
I'extase, Prométhde, and the last preludes and sonatas, as well as Ives's
Concord Sonata and some of his songs. They were all very exciting
and beautiful to me, and it was as a result of hearing and thinking about
them that I decided to try composing. [Edwards: 45]

Carter's interest in the music of Scriabin is symptomatic of the popularity of

mystical and theosophical ideas among the artistic community during this period, and

was the result of his acquaintance with the pianist and mystic Katherine Ruth Heyman,

whose devotion to the Russian composer produced something of a vogue for Scriabin

among American musicians of her generation (including the composers Charles Griffes

and Dane Rudhyar). [Schiff 1983: 14]

In 1924, Carter was introduced to another influential figure, with whom he

developed a lasting, if somewhat equivocal, relationship. Initially Charles Ives acted

as a mentor to Carter, introducing him to much new music, encouraging him to

compose in a complex avant-garde idiom and effecting an introduction to the Dean of

Harvard for him. However, the balance of the relationship was later challenged by

the younger man's increasing realization that the visionary radicalism of Ives could not

sustain meaningful musical communication unless disciplined by a more practical

approach to performance. Ives's influence on Carter consisted less in his

transcendental mysticism, than in his critical approach to music which appeared too

23
simple, an approach which, ironically, led Carter to regard Ives's own attitude as

naïve.

The music of the composers of the Second Viennese School did not become

familiar to Carter until a business trip with his father to Vienna in 1925 gave him the

opportunity to purchase as many of the scores of Schoenberg, Berg and Webern as

were available [Edwards: 43]. From this point, however, he developed a lasting

interest which survived his later anti-expressionist period:

I had my own sort of very early "Expressionist" or avant-garde period,

against which I reacted at the time of the depression, and to which I
have since returned in a certain sense... ever since the beginning I've
always liked the music of Schoenberg, Berg and Webern - as much of it
as I knew. I was the head of the American ISCM when we gave the
first all-Webern concert in New York - largely at my behest, because I
was especially interested in Webern at that time (about 1952). i've
always been interested in these three composers and have never had any
"reservations" about them. My "anti-Expressionist" attitude had to do
only with what I myself wished, for a certain period, to do as a
composer. Later I began to see the Viennese music as more relevant
and suggestive in this respect. [Edwards: 61-621

This continuing interest is borne out by reference to some of Carter's reviews

of this music in the 1930s and '40s, for example:

Hearing again von Webern's Five Pieces for String Quartet with their
rarefied, delicate atmosphere, and Berg's Four Pieces for Clarinet and
Piano, or Bartók's Second ñolin Sonata, we experience not the sterility
which is the easy and complacent brand-term now so frequently applied
to music of that period, but the great beauty of imagination and very
special feeling. [WEC 77-78 ("Stravinsky and Other Moderns in
1940", Modern Music, 17, 3 (Mar.-Apr. 1940))]

In an article on "Expressionism and American Music", published in

Perspectives of I*w Music, 4,1 (Fall-Winter 1965) [CEL 72-83], Carter engages in

his most thorough discussion of the nature of expressionism and draws parallels

between the music and ideas of the Second Viennese School and those of their

24
American contemporaries such as Ives, Cowell, Varèse and Ruggles. One of the

main functions of the essay is the rehabilitation of the reputations of these composers,

a task to which Carter, whose aesthetic standpoint traces a curve, at first veering away

from theirs and later returning, may have felt himself particularly suited. In this

essay, Carter focuses on two characteristics of expressionism, drawing on the

following quotation from a paper by L.Mittner:

The two main artistic procedures of expressionism are the primordial utterance
(Urschrei [..])and the imposition of an abstract structure, often specifically
geometric, on reality. [CEL 78]

The desire to evoke the "primordial utterance" is a symptom of the "urge for [. .1

intensification of expression", which Carter finds strongly represented in the music of

Ruggles and Rudhyar. It is also reflected in Ives's .Ecsayc Before a Sonata, which has

several points of contact with Schoenberg's writings, particularly in Ives's desire to

exalt the inner world and his emphasis on "substance" or content, rather than on form

or superficial beauty. The second of these characteristics, "the imposition of an

abstract structure", finds a close parallel in Berg's complex numerological plans and in

the serial method in general. However, the emphasis among the Americans on

experimental techniques and their rejection of conventional means of expression relates

their aesthetic more strongly to the radical modernism of the Futurists and other

mavericks rather than the extended traditionalism of the Second Viennese School.

The anti-expressionist period to which Carter refers was as much politically as

artistically motivated. He admits that he found himself in sympathy with those who

regarded expressionism as a symptom of the supposed hysteria of German culture:

The whole Expressionist point of view had come, at a certain point, to

seem as if it were part of the madness that led to Hitler.. .Many people
felt - and I certainly was one of them (perhaps not rightly) - that the

25
 whole German cult of hyper-trophic emotion could have been held
responsible for the kind of disaster we were witnessing then in front of
our noses (certainly Brecht came to hold this view). This is why, in
my opinion, many of us became interested for a time in neoclassicism as
a way of "returning to reason" and to a more moderate point of view
about expression, as well as to a more accessible vocabulary.
[Edwards: 60-611

The rise of fascism in Germany and Italy, combined with a long-standing

francophilia, led Carter to feel that Paris was the natural place for him to continue his

musical studies after completing his Master's degree at Harvard in 1932. In

particular, Carter was drawn by the reputation of Nadia Boulanger who taught at the

Ecole Normale de Musique [see CEL. 281-2921. Boulanger gained this reputation

through the success of her American pupils, especially the first of these, Aaron

Copland, who studied with her between 1921 and 1924. During the 1920s, Boulanger

also taught Virgil Thomson, Walter Piston and Roy Harris, composers of widely

varying styles and techniques who nevertheless shared a broadly neoclassical

orientation. The outstanding gifts Boulanger had to offer as a teacher were a fervent

belief in the role of the artist as craftsman and a remarkable understanding of the

techniques of modern music. The former, which clearly placed her in the

neoclassical camp, eventually led her to reject music which did not subscribe to this

aesthetic. Thus while Copland was studying with her in the early 1920s, Berg's

Wozzeck was considered worthy of serious interest, but by the time Carter came to

study with her a decade later, such music was looked upon with distaste [Edwards:

511. Boulanger's favoured model for young composers was Stravinsky, whose latest

works were discussed exhaustively in her classes. Thus Boulanger became, in

Charles Rosen's words, Stravinsky's pedagogical surrogate IRosen: 72}. Carter's

reminiscences regarding her continually emphasize her high-principled devotion to the

26
craft of composition and, in particular, the importance which she accorded to a

thorough study of counterpoint [Edwards: 52-551.

The extent to which Carter absorbed the values of the Boulanger circle is seen

in a 1946 article on the American composer who perhaps embodied her ideals to the

highest degree, Walter Piston ["Walter Piston", The Musical Quarterly, 32, 3 (July

1946), also in CEL 158-1751:

To have helped to establish a deep understanding of the value of

craftsmanship and taste here and to have given such persuasive
exemplifications of these in his works is highly important for our future.
For, not having as ingrained a respect and love for high artistic ideals as
Europeans have had, we have often slipped into the trivial, chaotic, and
transitory. Piston's work helps us to keep our mind on the durable and
the most satisfying aspects of the art of music and by making them live
gives us hope that the qualities of integrity and reason are still with us.
[CEL 174]

As is evident from some of the passages already quoted, Carter's anti-

expressionism did not last. In the context of the Second World War, the disciplined

restraint of emotion which was such an important principle of the neoclassical

aesthetic, came to seem more like repression and denial, and its reliance on borrowed

material and attitudes began to appear shallow and inadequate. A rereading of Freud

during the mid-1940s reinforced the notion that the civilizing and rational aspirations

of neoclassicism were merely avoiding the consequences of the grisly self-knowledge

which mankind had begun to achieve at the turn of the century:

After a while [...] it became clear to me [...] that we were living in a

world where this physical and intellectual violence would always be a
problem and that the whole conception of human nature underlying the
neoclassic esthetic amounted to a sweeping under the rug of things that,
it seemed to me, we bad to deal with in a less oblique and resigned
way. [Edwards: 611

Of course, this passage represents the composer's later rationalization and

27
justification of his viewpoint and actions and it should not, therefore, be taken as a

comprehensive explanation of the matter. Other, personal and pragmatic reasons for

Carter's stylistic evolution may be found of equal importance to psychological and

political ones.

2. Modernism and Conservatism

The axis of modernism and conservatism, although running parallel to a certain

extent with that of expressionism and neoclassicism, is by no means identical to it.

Radical modernism excludes both the ironic use of the past characteristic of

neoclassicism and the emotional continuity with the past characteristic of

expressionism. On the other hand, conservatism can be detected in works of both

Schoenberg and Stravinsky in the late 1930s and early '40s, following the emigration

of both these composers to the United States. Carter, although initially attracted by

the liberating daring of modernism, later grew impatient with its apparent lack of

discipline and its failure to engage the interest of the public. In the late 1 940s and

early 'SOs his early interest revived as he sought new means of musical expression.

However, he remained critical of the modernist aesthetic as it rose to prominence

again in the work of the post-war European avant-garde [see Edwards: 761.

It is difficult to define radical modernism except in negative terms, that is, as a

rejection of the past. This definition encapsulates the central historical problem of

modernism, namely its inability to sustain the momentum of its first radical impulse

without recourse to a reconciliation with traditional techniques and attitudes. The

history of music, in common with that of the other arts, does not conform to the

28
modernist ideal of a progressive "improvement", but instead follows a cyclic pattern,

in which periods of experimentation, reaction and consolidation rotate. In the course

of this century, extreme modernism has flourished during those periods when cultural

continuity with the past has been called into question most acutely; the period

surrounding the first world war and the aftermath of the second world war. In each

case, the desire of modernist artists to create a "Brave New World", independent of all

but the most forward-looking elements of the past, remained frustrated. The limited

capabilities of performers and performance technology and the conservatism or apathy

of the public have been important factors in this frustration. However, the isolating

effect of the divorce from tradition on individual artists may also bear some

responsibility for the relative brevity of extreme modernist movements. Many

composers have found difficulty sustaining a radical outlook and have sought to re-

establish links with their predecessors and the public by reaching an accommodation

with tradition.

A characteristic symptom of the need to find a replacement for reassuring

traditional certainties is the urge to construct theories. The growth of the importance

of theory to composers this century is such that it has led one writer (himself a

composer and theorist) to state that he finds "the notion of a composer who is not also

a theorist incomprehensible" [Perle 1990: 23J. However, it is also symptomatic of

the condition of twentieth-century music that the ideas exposed in these theoretical

works may often be more revolutionary and interesting than the compositions

supposedly based upon them.

Although short-lived, the Italian Futurist movement was prototypical for later

manifestations of modernism, especially in its fierce iconoclasm and (often naive) faith

29
in technology. Various other European composers, among them the Czech Alois

Haba (1893-1973) and the Austrian Josef Hauer (1883-1959) (both classic examples of

composers obsessed with the need to create a new theory) experimented with

alternative methods of pitch-organization, paralleling those of the Second Viennese

School but not sharing their expressionist aesthetic. However, it was in the New

World, far from the source of European musical tradition, that the spirit of modernism

seemed to flourish most freely. An American musical historian accounts thus for this

phenomenon:

The geographical peculiarities of the United States - its enormous spaces

and varied terrain - and the pioneering spirit that these fostered have
tended to shape the country and its people in a quite different mold from
that of the Old World. Early settlers were forced to rely largely upon
their own ingenuity, making do with resources at hand in responding to
the harsh demands of an unfamiliar environment, and something of the
pragmatic invention and improvisation encouraged by this experience
has been preserved in the national character. In addition, social forms
have tended to develop in a less conventionalized and structured way
than in Europe. The resulting tradition of tolerance toward a broad
range of individual differences has been supported by the belief that one
should make of one's life whatever one wishes, even to the extent of
defying generally accepted norms. Though mythologized - and
trivialized - in the popular media to the point of meaninglessness, the
ideal of the "rugged American iiidividualist" has a concrete historical
basis in the special circumstances of the nation's development.
[Morgan 1991: 296]

This analysis explains the unashamedly experimental nature of much American

music and the relative reluctance of American composers to band together into schools

or movements. However, the "uncompromising individualism" [Morgan] of figures

such as Ives, Edgard Varèse (1883-1965), Harry Partch (1901-1974), Conlon

Nancarrow (1912-97) and John Cage (1912-92) should not blind us to the fact that

various personal and professional associations played an important role in the

development of a "tradition" of experimentalism in American music. It is interesting

30
in this respect to compare Carter's perceptions of the situation he lived through with

the conclusions drawn by a later scholar of the period:

The main difference [between European and American expressionism],

as always, is that the state of American musical life was so inchoate that
a revolutionary movement in this art would necessarily be less well
thought out, less focused, and more of an affair of individuals only
agreeing in a general way, hence less corrosive of the fundamental
aspects of what seemed to all a moribund musical tradition, since the
situation was not seen with any clarity - and for that reason tended to
dissipate itself in superficialities and absurdities, as so often happens
even today. [CEL 76J

The unfavourable comparison with European music is regarded as an example

of uninformed prejudice by David Nicholls in his assessment of American modernism:

There is a commonly held view, particularly among those concerned

primarily with European musical traditions, that the American
experimental movement developed accidentally, in isolation, and in a
naïve and undisciplined way. It further considers the composers
associated with experimentalism as amiable eccentrics, whose works
are far less interesting than the anecdotes about them. This view is
clearly wrong. The composers discussed here [Ives, Henry Cowell
(1897-1965), Charles Seeger (1886-1979), Carl Ruggles (1876-1971),
Ruth Crawford (1901-1953) and Cage], as representatives of
experimentalism's first half-century, had a clear sense of direction both
individually and collectively. Their music and ideas are rigorous and
highly disciplined. [Nicholls: 218]

Nicholls stresses the co-ordinating effect of the pedagogical and theoretical

work of Charles Seeger (a composer for whom Carter has little respect). In

particular, Nicholls views the ideas contained in Seeger's essay "On Dissonant

Counterpoint" (published in 1930) as laying the foundation for much of the

compositional practice of Cowell, Ruggles and Crawford. However the cohesion of

this group was short lived; Cowell eventually developed his own remarkably forward-

looking theory based on the mathematical relationship between pitch and rhythm and

on the synthesis of stylistic diversities; Ruggles, as Nicholls admits, "owes as much to

31
 late Romanticism as to modernism, and lies essentially to one side of the main thrust

of American experimentalism" [Nicholls: 31; Crawford and Seeger turned away from

experimental music in the later 1930s to devote themselves to ethnomusicological

investigation. The ideas of these composers left little mark on musical life in general,

swept as it was in the 1930s and '40s by a wave of conservatism and in the '50s and

'60s by a second wave of radicalism. Only in the wake of the revolutions of total

serialism and indeterminacy could their achievements be re-evaluated, although, as

Carter notes in the essay previously quoted, this re-evaluation was still the result of a

reaction to European developments [CEL 72].

The "recurring preoccupations" of the "ultramoderns" have been summarized as

follows by Nicholls:

1 extreme chromaticism of both melody and harmony;

2 tone-clusters and noise;
3 the use of new or unconventional instruments (both electronic and acoustic)
and/or of conventional instruments in an unusual way;
4 rhythmic complexity, both simultaneous and successive;
5 implied or actual polytempo and/or polymetre;
6 implied or actual spatial separation of groups of instruments;
7 independent organisation of the various parameters of a musical line or idea;
8 large-scale and/or small-scale structuring of form, using extra-musical devices
and processes, including numeration;
9 graphic notation and/or semi-improvised music;
10 works which are indeterminate of their performance. [Nicholls: 218]

Points 1, 4, 5, 6, 7 and 8 are also characteristic of Carter's own later music and thus

reveal an affinity with some of this group's ideas, although his avoidance of electronic

instruments, improvisation and indeterminacy embodies the conservative side of his

nature. 2 However, the characteristic of American experimentalism most troublesome

to Carter is its bold eclecticism. Ives, in particular, deliberately challenged

2
For Carter's views on improvisation see Edwards: 78-9, on aleatoric music see
Edwards: 97-& and on electronic instruments see CEL 222.

32
conventional notions of stylistic integrity by producing works in both traditional and

revolutionary manners and by mixing these manners within single works. Carter, like

Ives, wished to find musical analogues to represent the plurality of experience in

modern America. However, in Carter's eyes, Ives abrogated his responsibilities as a

creative artist too often by neglecting to bring his works to a performable state and by

allowing borrowed material to dominate his music to the extent that it detracted from

the communicative purpose of the work. The criticism contained in Carter's review

of the Concord Sonata ("...more often original than good...")["The Case of Mr.Ives"

(Modem Music, 16, 3 (Mar. 1939) also CEL 87-901 exemplifies the younger man's

attitude towards his elder. However, later re-evaluations reveal an underlying

sympathy with Ives's ideas and problems, even if Carter is ultimately ambivalent about

his former mentor's true stature:

One could say that Ives was unable completely to digest his experience
as an American and make it into a unified and meaningful musical
expression. The effort of remodelling the musical vocabulary to meet
his own personal vision, almost without encouragement or help, was too
great, and too often he had to let hymn tunes and patriotic songs stand
for his experience without comment. ["Shop Talk by an American
Composer" The Musical Quarterly, 46, 2 (Apr. 1960), also CEL 2141

Thinking about Ives has been particularly fruitful to me: about how he
calls into question matters of style, coherence, and even the integrity,
the "seriousness" of serious music - and especially thinking about the
whole question of his inclusion of popular songs and hymns, which has
been constantly perplexing. Sometimes, as in the Concord Sonata, his
music seems like the work of an extraordinarily accomplished and
skilled composer, particularly the "Emerson" movement, where all the
motivic material is so highly organized and so closely interconnected, as
are the harmonic materials. And then there are other pieces that seem
to wipe all this aside and do something else. I have the impression that
Ives must have known very much what he was doing and thus must
have had many different intentions as a composer - sometimes to write
pieces in a high style and at other times to write sort of angry vaudeville
pieces. [Edwards: 631

33
 Carter has also been a perceptive critic of that other characteristic of

modernism, its obsession with pseudo-scientific theories. His review of Joseph

Schillinger's The Schillmger System of Musical Composition correctly pinpoints "the

assumption that the "correspondence" between patterns of art and patterns of the

natural world can be mechanically translated from one to the other by the use of

geometry or numbers" as the "basic philosophic fallacy" of his theory. ["Fallacy of

the Mechanistic Approach", Modern Music, 23, 3 (Summer 1946), also CELZ 15-16]

However, he clearly found some of Schillinger's ideas suggestive, particularly the

notion of rhythmic "interference patterns" which he exploited in the First Quartet and

the Quartet Sonata. The exhaustive nature of Schillinger's classification of harmonic

resources relates his work to that of middle-European theorists such as Haba and to the

later development of pitch-class set theory in the 1960s and '70s. Carter has a strong

link with this tendency as can be seen in his compilation of a "Hannony Book" (a two

volume catalogue of all possible three- to six-note sets together with possibilities for

combination and succession), beginning in the early 1960s.

In opposition to the strain of radical modernism, the twentieth century has

experienced a persistent and powerful conservative tendency. This has manifested

itself in music perhaps more strongly than in any of the other arts. Several factors

have contributed to this tendency. The growth of historicism during the nineteenth

century made artists more aware of the inherent value of earlier works of art and

caused them to question the attitude that saw the history of art as a triumphant

"progress" from past to present. As historical awareness grew, so did the artist's

sense of being dominated by the past and unable to match its achievements, of being

born too late to express oneself unselfconsciously and with originality. Responses to

34
the burden of the past varied from outright rejection (the course of extreme

modernism) through ironic distortion (neoclassicism) and the reinterpretation of

classical ideas in modern terms (the path taken by Schoenberg, Bartók and many

others) to submission to the past and stylistic continuity with it (Strauss,

Rachmaninov). (See Straus 1990: 1-20 for a wide-ranging discussion of these issues.)

The establishment and standardization of the economic foundations of musical

performance and publication played its part in the creation of a "canon" of

acknowledged masterworks of the eighteenth and nineteenth centuries. Pieces whose

"greatness" had already been universally endorsed were safer investments for the

publisher and impresario, whereas the promotion of contemporary music became an

increasingly risky venture. As the market for cultural products expanded in the

industrial age, "progressive" artists found themselves increasingly marginalized,

addressing themselves to an informed elite who constituted an ever dwindling

proportion of the total audience. Conservatism thus has two main roots; (i) the needs

and anxieties of the artist; (ii) the expectations of the general public as influenced by

the economic strategies of music's marketers. For some, the reconciliation of these

factors was relatively straightforward, for others it was a source of continual difficulty.

For American composers of the inter-war years, the personal and economic

factors contributing to a conservative aesthetic were closely bound up with social and

political factors, which will be discussed in the remaining sections of the chapter.

The economics of American musical life, through an unfortunate irony, operate in

direct opposition to the radical experimentalism encouraged by its social structures.

The theme of the financial difficulties of the serious musician in the United States runs

35
 throughout Carter's writings.3

However, the unfavourable distribution of economic patronage was not the only

factor in the decline of the fortunes of the "ultramoderns". In "Expressionism and

American Music", Carter draws attention to the role played by influential interpreters

of modern music:

The two important rivals in presenting modern music to the large

musical public were Leopold Stokowski - an irrepressible experimenter,
in those days, who played Schoenberg, Varèse, and Ruggles, and was a
supporter of the more extreme "ultramodernists" - and Serge
Koussevitsky, also dedicated to the new, but really more interested in
the Franco-Russian schools and in launching the (then) younger
generation of American composers [Copland, Harris, Piston, Schumanj,
giving them the kind of enthusiastic support he had previously given to
young Russians in Europe.. in the end Koussevitsky's energy and
persistence won a larger audience for the new American neo-Classical,
folkioric, and populist school and adherents of other esthetics were more
and more bypassed and forgotten. tCEL 73-4]

3. Elitism and Populism

As the century progressed and the political practice of democracy grew and

spread (if somewhat erratically), attitudes towards the arts began to change. The

traditional view of artistic endeavour as the province of a cultured minority came to

seem inappropriate in a social structure based - at least in theory - on the needs of the

majority. Many artists responded to this challenge in the 1930s and early '40s by

See for example "The Composer's Viewpoint", National Music Council Bulletin, 7,
1 (Sept. 1946) also CELz 3-5, "The Composer's Choices", Radio Broadcast (c.1960) also
CEL 210-214, "Shop Talk by an American Composer", The Musical Quarterly, 46, 2
(Apr. 1960) also CEL 2 14-224, "The Milieu of the American Composer", Perspecth'es of
N,wMusic, 1, 1 (Fall 1962) reprinted in expanded form as "The European Roots of
American Musical Culture" in CEL 62-72, "The Orchestral Composer's Point of View",
from The Composer's Point of 1'e Essays on Tndeth-Century Music by Those Who
Vfrote It, ed.R.S.Hines. (1970: University of Oklahoma Press) also CEL 235-250.

36
attempting to deal more realistically with social issues in their work. However, style,

as much as subject matter, was an important factor in developing a "populist" idiom.

A simplification of style, resulting in a greater transparency of language, would

enable the straightforward communication of the "message" of the work.

Two economic and political phenomena of the 1930s motivated the development

of a populist movement in the arts; the severe economic depression and the rise of

fascism. In the face of these threats to the security and stability of eveiyday life, the

experimental aspect of much contemporary music seemed irrelevant at best and

dangerously alienating at worst. In place of the introspective, speculative aesthetic of

modernism, populism favoured alternative views of the arts; either as a vehicle for

socio-political comment, warning or exhortation to action, or as a form of

entertainment, conveying messages of consolation and offering escapist fantasies.

The characteristics of populism could be discerned throughout Europe and

America, forming part of the general reaction to the radical atmosphere of the 1920s.

Ironically, the social climates of quite different regimes favoured similar approaches.

In the Soviet Union, the reaction against modernism took the form of official

suppression of avant-garde works, which were censured as "formalist", and the

cultivation of "socialist realism", an essentially conservative idiom marrying traditional

means of expression with themes exalting the simple virtues of ordinary people. In

Nazi Germany, too, a quasi-populist culture was manufactured by the state as a means

of ideological control. The challenge facing artists in the western democracies was to

make their art socially relevant without falling into the predictable banalities of the

"official" art of totalitarian regimes.

In the United States, the principal figure in the popularization of serious music

37
 was Aaron Copland. Coplarid's initial popular success, on his return from Paris and

Boulanger in 1924, rested on the freshness and vigour of his incorporation of jazz

elements into his music. However, in later works, such as the Piano Variations

(1930) and the Short Symphony (1932-3), he began to develop a more abstract style

which critics felt to be "austere" and "esoteric". At the same time, inspired by the

example of European models such as Les Six, he formed a "Young Composers'

Group". Faced with indifference and hostility among public and performers alike,

Copland felt obliged to reconsider the relationship between artist and audience:

During these years I began to feel an increasing dissatisfaction with the

relations of the music-loving public and the living composer. The old
"special" public of the modern music concerts had fallen away, and the
conventional concert public continued apathetic or indifferent to
anything but the established classics. It seemed to me that we
composers were in danger of working in a vacuum. Moreover, an
entirely new public for music had grown up around the radio and
phonograph. It made no sense to ignore them and to continue writing
as if they did not exist. I felt that it was worth the effort to see if I
couldn't say what I had to say in the simplest possible terms.
[Copland, quoted in Berger: 26-7]

Copland found the basis for the stylistic simplification he sought close at hand

in the folk-music of Latin America, the Caribbean and his native America. Diatonic

melody and rhythmic energy and elasticity became the essential elements of the new

style. In practical terms, he aimed at a broader audience than the concert-going

minority by writing for films (Of Mice and Men (1939), Our Town (1940), N,rtb Star

(1943), The Red Pony (1948)), for schools (The Second Hurricane (1937), Outdoor

Oivrture (1938, band arrangement 1941)), for radio (Music for the Radio (Saga of the

Prairies) (1937)), for the theatre (Quiet City(1939) and most importantly, for the

ballet (Billy the Kid (1938), Rodeo (1942), Appalachian Spring (1943-4)).

Furthermore, Copland felt he had a duty to promote a wider understanding of

38
music in general and modem music in particular, through writing and lecturing, as is

demonstrated by his books What to Listen for in Music (1939) and Our New Music

(1941), his frequent contributions to the periodicals Modern Music and The Musical

Quarterly and his occupation of the post of lecturer to laymen at the New School for

Social Research from 1927 to 1937.

Copland, and members of the Young Composers Group associated with him,

had broadly left-wing sympathies and actively participated in artistic projects forming

part of the Works Progress Administration during the "New Deal" period. Even

members of the ultramodern school tried to accommodate social issues within an avant-

garde framework; Cowell, Seeger and Crawford were all members of New York's

Composers' Collective during the early 1930s.

When Carter returned to the United States in 1935, populism had effectively

ousted ultramodernism as the dominant characteristic of American music. Having

followed the example of Copland and others in pursuing studies in Europe, Carter

naturally found himself involved in the same circles as these composers. He began to

work as a reviewer for Modern Music in 1937. In the same year he became musical

director of Ballet Caravan under the aegis of Lincoln Kirstein. Kirstein commissioned

both Copland's Billy the Kid and Carter's Pocahontas (1939), which were premiered

on the same occasion.

At this time, Carter's views on the function of music in society corresponded

closely with the populist mainstream, although he maintained a certain degree of

sympathy with composers of a more radical outlook:

It's true that Varèse, whom I used to see occasionally, especially during
the time when he was rehearsing my chorus To Music with an amateur
group, seemed very melancholy during this period, which was turning

39
 toward new, more populist artistic aims, thus putting into question the
more experimental attitudes of the best artists of his generation. It was
easy for me to sympathize with both the old and the new of that time.
During my studies and after, so many disastrous human situations
resulting from the depression, from the Moscow trials, and from the
Nazi-Fascist dictatorships haunted me. It was hard not to feel that very
simple human needs were unmet and that the high art we knew seemed
cruelly remote from this. Surrounded by so much violence and so
much need, one couldn't help wondering whether such a thing as
advanced modern music with its elite audience wasn't just beside the
point. [Edwards: 59]

The only work of Carter's based on explicitly populist subject matter is

Poc,ahontas, which explores the consequences of an early confrontation between native

Americans and European settlers. However, according to his later reminiscences, the

composer was not entirely convinced of the ideological acceptability of the plot:

At that time the American past was being whitewashed, I suppose in a

desperate attempt to make the "melting pot" idea work. I myself had
misgivings about the "colonialist" aspect of the subject, particularly as I have
some Indian blood of my own, but hoped to make it a parable of cooperation.
[Edwards: 57]

Although the subject matter of Pocahontas had something in common with that of

ballets by Copland, its musical substance was varied and often complex. The

unfavourable popular and critical reception of the ballet was one of the factors

prompting Carter's adoption of a simpler musical style in the early 1940s. ["The

Composer's Choices", CEL 210-2141

This was far from being the only problem which Carter experienced in working

within a populist framework. By nature, Carter was ill-disposed towards the urge for

simplification enshrined within populist ideals. The same critical attitude which

caused him to view the work of early modernists as naive applied to the shallow

pretensions of early "American musical foildorists" and would-be popularizers:

Like all other music, that intended for the masses can be good or bad.

40
 Effectiveness in putting across a message is no criterion of artistic
value. Both Eisler and Blitzstein have shown that real musical
imagination and originality can be of great service to their political
points by adding character and incisiveness. But such music on the
recent TAC [Theatre Arts Committeej evening, gallery-funny pieces like
Henry Brant's Marx Brothers or Morton Gould's Sonatma or Child
Prothgy, or gallery-serious cantatas like Kleinsinger's I Hear America
Singing or Earl Robinson's fresher Ballad for Americans, begin to sound
thin and the attitude of the composers condescending. Their apparent
assumption is that the masses don't know anything about music and
never will. I wouldn't be a bit surprised if works like the Sessions
Quartet or the Harris Spnphony were to become more popular than
these self-conscious and restricted compositions ever will be.
["American Music in the New York Scene, 1940", Modern Music, 17, 2
(Jan.-Feb. 1940), also CEL 53]

It is noteworthy that the American composer for whom Carter expresses the

most consistent praise with the fewest reservations is Roger Sessions. Sessions

studied for a while with Boulanger, wrote in a style owing much to Stravinskyan

neoclassicism in the late 1920s and early 1930s and formed a close personal

association with Copland, including the "Copland-Sessions Concerts" of 1928-31.

However, Sessions was never deeply committed to the populist ideal, or to the

aesthetic of neoclassicism. Like Carter, he maintained an interest in the composers of

the Second Viennese School and came to stylistic maturity only after a long considered

appraisal of a variety of influences. Carter clearly drew inspiration from Sessions's

consistent and uncompromising sense of purpose and from his engagement with "the

most serious and important issue that has faced contemporary music.. .the task of

finding new forms for the new material." ["Current Chronicle: New York, 1959", The

Musical Quarterly, 45, 3 (July 1959), also "Roger Sessions: Violin Concerto" in CEL

175-1801

In his own works, too, Carter revealed his unease with the restrictive aspect of

populism. Whereas Copland, in his Piano Sonata (1939-41) and Violin Sonata (1942-

41
3), was able to wed the newly-forged musical language of the stage works to the

abstract and personal medium of chamber music, Carter's instrumental works of the

period (especially the Holiday Overture (1944) and the Piano Sonata (1945-6)),

introduce a greater degree of formal and contrapuntal complexity into the populist

idiom than is typical of it. Carter reports that Copland's first reaction to the Holiday

Overture was to criticize it as a "typical, complicated Carter score" [Edwards: 58]. It

would probably be fair to say that whereas Copland achieved his definitive

compositional "voice" through the revision of musical language accompanying his

espousal of populist values, Carter suppressed his natural inclinations in order to

conform to an ideological ideal.

Perhaps the most frustrating aspect of the populist ideal of mass communication

through serious music was that it proved to be unattainable. The conservatism of

American audiences was proof against even the most ingratiating of offerings of living

American composers. Carter's reflections on his earlier aspirations reveal a complete

rejection of populism and a certain resentment of the lack of appreciation for his

efforts demonstrated by contemporary audiences:

They were - and still are - in the position I was in as a little boy, when
it comes to modern music - they aren't able to distinguish veiy much
about any of these things; they just know it doesn't sound very much
like Brahms, and that's about all, as far as I can see. In fact, I
probably should have known better than to try writing works like my
First Symphony and Holiday Overture in a deliberately restricted idiom
- that is, in an effort to produce works that meant something to me as
music and yet might, I hoped, be understandable to the general musical
public I was trying to reach for a short period after writing Pocahontas.
I did this out of a natural desire to write something many people could
presumably enjoy easily at a time of social emergency, but I did so
without appreciating just how serious was the audience paralysis
engendered by this lack of interest in or familiarity with the new in any
of its artistic forms. Thus I wrote music which escaped the average
listener, despite what seemed to me its directness. [Edwards: 58-58]

42
 After the second world war, the aggressive elitism of a new generation of

European modernists established itself as the cutting edge of musical thought. In this

environment, the fortunes of composers such as Copland declined somewhat, although

he did make efforts to come "up to date" with the development of serialism.

However, for those like Sessions and Carter who placed a high value on the expression

of complex thought in art, the new Zeitgeist was a liberating one, even if they felt that

the Darmstadt generation was afflicted with the same naivety as the earlier

"ultramoderns"4. The culmination of Carter's rejection of populism came in 1950-1,

the period during which he retreated to the Sonora Desert in Arizona in order to work

on his First String Quartet:

I decided for once to write a work very interesting to myself, and so say
tohellwiththepublicandwiththeperformerstoo. Iwantedtowritea
work that carried out completely the various ideas I had at that time
about the form of music, about texture and harmony - about everything.
[Edwards: 351

4. CosmopolitaniSm alid Americanism

The desire to create a recognizably American music, independent of the

European models which had dominated their predecessors, obsessed American

composers of the early twentieth century, whether populist or modernist in outlook.

However, the means of achieving this were greatly disputed. Both schools of thought

recognized Ives as the first truly original American composer. Ives himself had had a

conventional musical education (that is, one based on mastering the language and

forms of nineteenth-century European music) at Yale under Horatio Parker. Although

' See Edwards: 76, 83 and 117-8.

43
Parker did not share Ives's interest in the experimental (which Ives owed essentially to

his father), the technique which Ives acquired through his studies was a crucial part of

his musical make up.

The debate which followed centred on whether American composers could still

learn from the Europeans, or whether they should strike out on their own and sever all

links with the Old World. Battle lines were clearly drawn between the graduates of

the "Boulangerie" - Copland, Sessions, Carter, Piston et al - and the "ultramoderns".

Sessions stood at one extreme of this opposition, having spent most of his life between

1926 and 1933 studying in Europe and before this, having "apprenticed" himself to the

Swiss emigré Ernest Bloch. At the other extreme stood figures such as Harry Partch,

who received no formal musical training, grew up and spent most of his life in remote

parts of the western U.S.A. and rejected almost all aspects of traditional European

music, including such basic characteristics as its instruments, its system of tuning and

its conventions of performance.

Carter, although clearly inclining more towards Sessions than to Partch, was

well aware of the need for American music to achieve maturity and self-recognition

through independence from European thinking [see "American Music in the New York

Scene, 1940" Modern Music, 17, 2 (Jan.-Feb. 1940), also CEL 48-53]. In an article

on the early American "nationalist", Henry Franklin Belknap Gilbert, he summarizes

the stages of development of an autonomous musical tradition:

At the time when Henry Franklin Belknap Gilbert was making his effort
to write, as he put it, "some American music", nationalism was the
subject of wide discussion by critics, musicians and composers,
including Gilbert himself. A general historical sequence of periods had
been formulated to cover our national musical evolution: first, foreign
domination, and imitation of non-native art music; second, collection of
and familiarization with indigenous folk songs and dances; third,

44
 invention of a style consistent with folk material though without using
actual quotations; and fourth, the musical millennium, emergence of the
national masterworks written by native composers with a large native
background and inheritance.. .this thesis gave folk songs a basic position
as the root from which each national music culture is to grow. All the
elements comprising the "manner": rhythm, melody, harmony, and
form, evolve from this germ and generate a style that is to be expressive
of our native kind of "matter". To put it another way, in a search for
a means of expressing the "matter" of our national consciousness, it was
assumed that composers would inevitably follow this historical pattern.
The two interrelated doctrines of historical stages, and the
antitheses of matter and manner, combined with our special brands of
individualism and of progress, have deeply influenced the thinking of
our contemporary composers, particularly those of nationalist intentions.
The conflict of opinion over which of the Ibur stages we are now in is
the basis of many present arguments; while the manner-matter problem
perpetuates itself in questions as to the "abstraction" in contemporary
music and its "neo-classicism" or "flee-romanticism". ["American
Figure, with Landscape", Modern Music, 20, 4 (May-June 1943), also
CEL 1341

Carter criticized exponents of self-conscious "folkiorism" in the same manner

as he did the naive radicalism of the "ultramoderns" 5. His own temporary adherence

to the neoclassical style may be regarded as an example of a borrowed "manner",

which he later abandoned because it could not express the full complexity of the

"matter" of his musical thought. The post-war era, which saw the development of the

mature styles of Carter and Sessions, perhaps represents the true beginning of the

"fourth stage" of American musical development. However, the existence of a

variety of conflicting ideologies, producing an exciting plurality of possibilities,

seemed to him to be the sign and guarantee of the musical health of the nation:

Nevertheless, important composers are already with us in America.

Ours is a varied musical scene: the music-makers, few as they are in
comparison with other artists, write in every kind of style, whether it
derives from Europe, is boldly original, extreme, conservative, crude,
or highly polished. They make up a complex, interesting picture, as

See Edwards: 31.

45
interesting and much more individual than many a European scene.
Their quality, competence, and seriousness is on a generally high level.
Indeed, the variety of current esthetic attitudes is proof that the question
of musical competence is no longer the problem it was. One might
almost say that American music was born when these differences began
to take convincing shape in works. ["American Music in the New York
Scene, 1940" Modern Mnsic, 17, 2 (Jan.-Feb. 1940), also CEL 491

46
CHAPTER 3: ANALYTICAL CONTEXT

The analytical context in which Carter's works of 1945-55 are placed is two-

dimensional. On one axis, we should examine "neoclassicism", the musical idiom

within which the substance of Carter's earlier works was formed. Along the other

axis lies the concept of "transitional" music, which may provide several parallel

models for analysing the processes of change undergone in Carter's oeuvre.

Examination of neoclassicism will involve assessing various analytical approaches to a

body of work composed in the period c.1920-45, which, while far from homogenous,

exhibits many recognizable common features of aesthetic and musical style.

Examination of "transitional" music, on the other hand, necessitates consideration of a

wider variety of musical styles, which are themselves, of course, in a state of flux and

therefore present the analyst with methodological problems. The present chapter

surveys some important issues and approaches within these two areas, in order to

arrive at provisional principles for the analysis of Carter's transitional music. The

chapter concludes with a brief critical resume of analytical work already undertaken on

this music.

1. Analysis of Transitional Music

It may be debated whether there really is any such thing as "transitional

music". It is relatively easy to view almost any work as a half-way house between

two others and thus the whole history of music is, in a sense, transitional. On the

other hand, the same work may exhibit some form of internal coherence which is sui

47
 generis, requiring no explanation or justification in terms other than its own. All this

is merely another expression of the familiar mathematical concept that a line may also

be regarded as a set of individual points. It is no longer reasonable to view

transitions in matters of artistic style as straight lines leading to pre-ordained goals, or

to regard individual works as interesting only from the point of view of those projected

goals. Therefore, as this study unfolds, although the "line" of Carter's development

must be borne in mind, considerable space will be devoted to consideration of each of

the "points".

The species of transition which is of principal interest here is that between

tonality and atonality which took place during the early part of the twentieth century.

This area has been a focus of considerable analytical interest in recent years1,

principally because of the quality of the music itself, but also because of the intensive

reworking and refining of analytical ideas required in response to the variety and

complexity of compositional procedures it displays. As Jonathan Dunsby observes,

early twentieth-century music still conveys an effect of contemporaneity,

confronting us with the mystery of what we are still learning to
assimilate. There are analysts who believe that thinking about the
music of this period is - apart from the special challenge of absorbing
the music of our own age - the most important analytical thinking to be
done.[. . .1 It can be argued that in modern Western music there is no
period less amenable to the idea of analytical models than this one; no
period in which the very criteria for understanding any one piece have
been more entailed in that particular piece; no period which was marked
by such diversity of compositional practice, to the extent that the
possibility of generalized analytical explanation may be doubted.
[Dunsby 1993: ix]

The principal difficulty facing the analyst of music of this period is one of

methodology. There are well-established ways of approaching tonal music and, to a

'See, for example, Samson 1977, Kerman, ed. 1990 and Dunsby, ed. 1994.

48
 certain extent, post-tonal music. Schenkerian techniques have proven to be so

successful in the analysis of music of the "common practice" period (Bach-to-Brahms)

that they have virtually become the established orthodoxy [Dunsby and Whittall 1988:

7-81. Although no such orthodoxy exists for the repertoire of "classic" atonal music

(pre-serial Schoenberg, Berg and Webern), pitch-class set theory - originally developed

as a means of elucidating dodecaphonic music - has asserted itself as a powerful

resource, which must be acknowledged by anyone attempting to make further inroads

into this complex material. However, there exists a large body of music by

composers as diverse as Liszt, Debussy, Strauss, Mahier, Scriabin, Ives and the early

Schoenberg, Berg and Webern, which may be analyzed - partially - from either point

of view, and which cannot, therefore, unequivocally be called either "tonal" or

"atonal". The necessity for an historical category of "transitional music" arises not

only from the overlap of repertoires examined by analysts of tonal and atonal music,

but also from the seeming impossibility of producing convincing and comprehensive

analyses of these works from a single theoretical standpoint. "The whole point of the

music of [thej transitional period", according to Dunsby and Whittall, "is that it cannot

be analyzed with reference to one type of pitch - structuring alone." 2 [Dunsby and

Whittall 1988: 113]

Extensions of Schenkerian methods "forwards" into the atonal domain have

been efficiently summarized by James Baker in his article "Schenkerian Analysis and

2
Recognition that a single theoretical basis may not provide a sufficiently broad basis
fr analysis is not confined to the field of transitional music. Analysts of tonal music
have attempted to overcome this problem by developing synthetic or pluralistic methods
(such as David Epstein's fusion of Schoenbergian and Schenkerian ideas in &yund
rJ)rpheus [Epstein 19791, Christopher Wintle's concept of structure and counter-structure
PWintle 1985], or Kofi Agawu's conjunction of Schenkerian and semiotic techniques
Agawu 1991]).

(LOiO)
 Post-Tonal Music" [in Beach 1983: 153-186?. Baker divides the analysts he examines

into two groups; firstly, those, such as Adele Katz, who after testing Schenker's

methods against contemporary music, reject its applicability and propose that new

methods are required, without suggesting what these might be; secondly, those who

claim to find examples of prolongation and structural hierarchy in post - tonal music,

either (a) by identifying tonal elements and progressions, or (b) by attempting to

demonstrate the prolongation of dissonant sonorities.

A complementary tendency, the endeavour to explain certain aspects of tonal

music through recourse to methods usually applied to the post-tonal repertoire, is most

strongly represented in the work of Allen Forte. Forte has attempted to interpret

aspects of works by Liszt, Mussorgsky, Debussy and even Chopin in terms of set

theory [Forte 1987, 1988 and 19901. An important concept here is the proposition

that pc sets may be "prolonged" as linear middleground motives, rather than, as some

have argued, functioning purely as surface motives.

Both approaches are vulnerable to the criticism that an accurate and sensitive

analysis cannot be produced by forcing the music to "obey" one set of theoretical

rules. The obvious reaction to recognition of this flaw is the adoption of a synthetic

or pluralist methodology, based on the interaction of tonal and atonal theory.

Something like a methodological consensus in the analysis of transitional music may

be emerging in the Yale University Press series Composers of the llsvntieth Century,

under the general editorship of Allen Forte. Forte's own article "Schoenberg's

Creative Evolution: The Path to Atonality" (MQ 64, ii, 1978) perhaps provided the

model and inspiration by examining the work of a single composer - an acknowledged

Also see Baker 1990 and 1993 for further thoughts on post-tonal voice-leading.

50
 master of late nineteenth-century extended tonality and the prime originator of "classic

atonality" - and showing, through analysis of a few key works in the decade 1900-

1910, the extension and eventual abandonment of prolongational voice-leading

techniques and their replacement by pc set relations as agents of structural coherence.

This is also the approach of James Baker in The Music of Alexander Scriabin (1986)

and Richard Parks in The Music of Claude Debussy (1989), the latter making extensive

use of Forte's later addition to his theory, the pc set genera [Forte 1988]. These

studies aim to portray the development of the composers' musical language as a

continuous logical process, through a smooth "modulation" from one analytical

orthodoxy (Schenker's) to another (Forte's). Early examples of the composer's work

which exhibit conventional tonal Ursatz forms are followed by those in which tonal

structural elements are implied but not actually present, and finally, by those in which

no tonal background can be discerned. At the same time, atonal pc sets grow in

importance from being mere foreground presences to being determinants of musical

structure.

Baker's thesis hinges on the dual significance of certain pitch - configurations

in Scriabin's music, most characteristically that in Table 3.1. This progression may

Table 3.1 Whole-tone invariance/dominant prolongation in Scriabin

f b(=ck) f
db g(=aI,) db
b f b
g d g

b ll V5

51
 be understood both as a dominant prolongation and as an example of whole-tone

invariance (i.e. as a tonal and/or an atonal progression). Apparently, for Scriabin,

the exploitation of the limited possibilities of transposition and complementation of the

whole-tone scale was a stepping-stone towards the more general application of these

techniques in the atonal universe. Baker attempts to show that despite the presence of

both tonal and atonal elements in Scriabin's music, the works are still examples of the

kind of organic, unified coherence typical of nineteenth-century tonal masterpieces:

In Scriabin's compositions dating from around 1903-1909, tonal and

atonal procedures - traditionally considered mutually exclusive - function
inseparably within integrated musical structures. Atonal procedures are
used only at foreground levels and overall coherence is guaranteed by
the underlying tonal Ur.satz [Baker 1986: 268)

Baker feels that for this reason, he is able to work within the repertoire of orthodox

Schenkerian techniques and that there is no need to invoke such concepts as

"extended" or "hovering" tonality or "contrapuntal - structural chords" [Baker 1986:

2691. In practice this is not strictly true; there are examples in Baker's analyses of

what, in all but name, are "contrapuntal-structural chords". 4 Furthermore, the

innovative techniques of Scriabin's transitional works which Baker identifies - namely

"(1) omitting explicit endpoints of conventional wifoldings at any level, including the

background, and (2) loosening traditional vertical structures governing counterpoint so

that lines which would ordinarily move in conjunction with traditional harmonic

progressions are instead allowed to move out of synchrony" [Baker 1986: 268] -

suspend and threaten to destroy any really audible sense of tonal function and produce

exactly the sense of "hovering" Baker rejects.

In Baker's analysis of Scriabin's Op. 5112, the structural 5th degree is supported in
the background graph by what Baker describes as a "linear 412", which is clearly of
contrapuntal origin. [Baker 1986: 56]

52
 It may be argued that Baker's attempt to synthesize tonal and atonal elements

into a unified whole is misguided, not merely because it plays down any sense of

tension between the two, but because atonal elements are resistant to the kind of

unifying coherence characteristic of tonal music. Two contrasting approaches to

transitional music are now examined in an attempt to find alternatives to the Schenker-

Forte synthesis.

The issues of the relationship between foreground and background, and the

validity of extensions of prolongational concepts, are addressed by Derrick Puffett in

his analysis of the Fugue from Tippett' s Second String Quartet [Puffett 1986]. Unlike

Baker, Puffett is reluctant to try to bind tonal and non-tonal elements into a unity. He

begins by constructing a voice-leading graph with three structural levels, but finds that

the means of elaborating the contents of one level to create the next do not correspond

consistently with tonal prolongational techniques and therefore the "analysis breaks

down" [Puffett 1986: 2571. In certain passages, according to Puffett,

there is a discontinuity between foreground and background which

cannot be bridged by an ever-increasing liberality in applying pre-
existing concepts of prolongation. If such passages were to be
reconciled with the more remote levels of structure, new concepts of
prolongation would have to be found to do it. The line between
"finding" a "concept" and cooking an analysis is as narrow as the
challenge in Tippett to existing theory is wide. [:258]

Like Katz, Puffett tests analytical orthodoxy (Saizer's rather than Schenker's in

this case) against musical fact and finds it wanting. However, he draws a positive

conclusion from this finding; the "value" of the voice - leading graph is that it can

establish "the exact nature of Tippett's tonal language in that it demonstrates, through

its failure as much as through its success, the extent to which that language is tonal

and the manner in which it is so" [:2581. Although his analysis is necessarily

53
incomplete, Puffett raises a crucial point, which challenges Baker's view that

transitional music may still be organically unified through the relationship of

hierarchical levels. Instead, he suggests that a disjunction exists between levels of

structure, so that while "long-term harmonic movement can be represented in terms of

a Fundamental Structure, [iti cannot be reconciled with all the movement at the

foreground. This creates a discrepancy between harmonic and contrapuntal

considerations, which, because it is structural, goes deeper than.. .essentially stylistic

form and content questions..." [:2601

It may, of course, be argued that what Puffett describes as a "Fundamental

Structure" cannot be regarded as such if its unifying power is not transmitted through

all levels. However, the idea that a work which alludes almost constantly to tonal

harmony, without actually being tonal, can be interpreted in terms of a disjunction

between its structural elements is a striking and valuable one.

Puffett's reluctance to extend prolongational concepts beyond a certain limit is

echoed in articles on Mussorgsky by Michael Russ [Russ 1990, 19961. In contrast to

Puffett's single method and Baker's attempt at synthesis, Russ employs a freely

eclectic methodology. His graphic and symbolic analysis contains elements from a

variety of analytical vocabularies; Schenkerian foreground prolongations,

Schoenbergian tonal regions and Fortean pc sets, since "prolongation operates over

relatively short spans in this piece, its large scale role being assumed by association,

symmetries and motivic and fixed - pitch connections." [Russ 1990: 481 According

to Russ,

this methodological eclecticism is justifiable in that no single analytical

approach can do justice to Mussorgsky. It also stems from a belief that
it is better, in transitional music, to bring together methods which, to

54
 the extent to which they are used, are applied strictly, than to adapt
methods like Schenker's to new kinds of music; a multi-method
approach may throw progressive and conservative elements into clearer
relief. [Russ 1990: 49]

Russ's emphasis on "conjoining" rather than "synthesizing" elements represents

an abandonment of strict organicism and perhaps fulfils Dunsby and Whittall's wish

for a "useful acknowledgement of the symbiotic presence of distinct rather than

synthesized features" in transitional music [Dunsby and Whittall 1988: 113]. The risk

entailed in such a method is that the results may appear incoherent. Russ skillfully

avoids this but also produces a parallel analysis along unadulterated Schenkerian lines

(a Ia Puffett), which "complements and challenges" his first reading. [Russ 1990: 48J

The pluralism of Russ's approach is thus multiplied since it offers two alternative

views of Mussorgsky, one illustrating the importance of "realist" anti-organic elements

in his music, the other showing how "purely musical" elements do in fact "create a

higher unity" [Russ 1990: 61J.

To summarize the foregoing discussion, it may be said of transitional music:

(a) that it may be observed to be organized on different hierarchical levels as tonal

music is;

(b) that these levels may not be continuous in themselves or consistently and

organically related to one another;

(c) that it may exploit the ambiguity arising from differing interpretations of certain

pitch-configurations (as tonal hannonies, motives or pitch-class sets);

(d) that it may be appropriate (if not an inescapable necessity) to adapt techniques from

complementary analytical methods in order to express these ambiguities and

disjunctions;

55
(e) that the practices of individual composers and the unique context of individual

works are as important as the a priori concepts of a given theory in performing an

analysis.

2. Analysis of Neoclassical Music

While there are certain similarities between the stylistic transitions in Carter's

music and in that of early twentieth-century composers like Schoenberg and Scriabin,

there is an obvious difference in their starting-points. The musical language of

Schoenberg's early maturity was late Romanticism, an idiom which still maintained a

direct link with the harmonic principles operating at the time of tonality's birth some

three-hundred years previously. That of Carter's early maturity was neoclassicism,

which had evolved only since the First World War and which was already at a remove

from tonal convention. Carter's stylistic transition was therefore from one post-tonal

idiom to another and is perhaps more closely comparable with Stravinsky's late

conversion to serialism and Tippett's change of direction with Kianzg Priam. The

following discussion will attempt to establish the key questions in the analysis of

neoclassical music.

Probably the most important issue in the analysis of neoclassical music is that

of the degree to which the adoption of the materials of tonality entails the continuing

presence of tonal functions and hierarchies. Harmony and voice-leading often become

separated from one another in analyses of the neoclassical style, since they appear to

have differing relaionships with their counterparts in the "common practice" era.

Attempts at explanation of the harmonic usages of Stravinsky and others have

56
 thrown up a collection of terms such as "pandiatonicism", "polychords" and

"bitonality", which describe the symptoms of neoclassical harmony rather than

diagnose the underlying condition. The issue is touched on by Carl Dahihaus in his

article wToijjty: structure or process?" [in Dahlhaus 1987: 62-80]. Dahihaus states

that in twentieth century music, the concept of harmonic process has given way to that

of harmonic system. (This is the result of a shift of emphasis in a dialectical

relationship rather than of a complete overthrow of one principle by another.)

Harmonic process, by which DahThaus means the idea "that sonorities do not simply

exist side by side, linked by melody or voice-leading, but have their own inherent

momentum" [:66], began to take root in western music in the twelfth century and

achieved its final flowering in the tonal music of c.1600-1900. Harmony as system,

which consists principally of the composer's "freedom to posit axioms" without

recourse to justification by appeal to "nature" or "history", re-entered western music in

the twentieth century, most strikingly with Schoenberg's serial method.S Stravinsky's

"freedom to posit axioms" included the development of a technique of layering or

stratification of tonal areas. However, Dahihaus suggests that this resulted not in the

complete overthrow of tonal function, but in a tension between the dialectical poles of

process (tonal function) and system (layering or stratification):

if we wish to do justice to the complex techniques of superimposition in

Le Sacre du prrntemps or to the ironically ingenuous ones in The Rtike 's
Progress, we must not fail to appreciate that Stravinsky's neutralisation
of tonal functional chords should not be understood as a given, so to
speak self-contained fact which we have to accept for what it is, but as a
procedure which can be comprehended - though not always without
some effort. When a tonic is, so to speak crippled functionally by
having a dominant placed on top of it, the listener should become aware

5Dahlhaus regards Schoenberg's own appeals to nature and history for justification of
t1e twelve-note serial method as irrelevant.

57
 that this is the result of a conflict and not just a mute ....... the inner
tension of what we call Stravinsky's static harmony is due to its
suppressed dynamic element... [: 69]

It is the acceptance of complex superimpositions as "mute fact" that is a major

weakness in some of Felix Saizer's analyses of music by Stravinsky and Copland in

Structural Hearing [Salzer 1962]. Salzer employs the term "polychord" to describe

pitch formations created through the superimposition of triadic elements, though he

claims not to find it "accurate or indicative" because "the chord built on the bass will

always be the stronger one, and it is the bass and its chord which will determine the

chord grammatical status of the whole chord cluster. Polychord would thus be an

adequate denomination only if both [or all three] chords would amalgamate on equal

terms, which seems hardly possible." [: 192-1931 Put so bluntly, this equation of

ba±sith-root function seems crudely inadequate, although it might be rendered more

acceptable by the modification that a bass note will tend to be heard as a root in direct

proportion to the degree to which the chord above resembles a third-based structure.

Salzer's examples 416 and 472 [Salzer: Vol.2, 182-7, 234-7] show dissonant

polychordal sonorities acting as "tonics" in Copland's Piano Sonata (1942) and

Stravinsky' Symphony In Three Moiements (1945), the former based around a

superimposition of tonic and dominant triads of B flat minor, the latter on a

superimposition of triads of G major and D flat major. The bass notes of these

sonorities are regarded by Saizer as roots and the chords themselves as stable, thus, in

the case of the Copland, "a fusion of tonic and dominant chords has taken place",

while in the Stravinsky, there is a "prolongation of the polychord on G with the D flat

chord as a secondary chord of fusion" [Saizer: volume 1, 1941.

Saizer's next theoretical postulate is, however, even more problematic and is

58
the source of a long-standing controversy. This is the possibility of "the contrapuntal

prolongation of dissonant chords, especially of polychords."[: 193] The problem with

Saizer's analyses of the two works mentioned above is that they interpret harmonies as

prolonged merely because they are recurrent. The Copland example revolves around

exactly the kind of superimposition of tonic and dominant which Dahihaus refers to

with regard to Stravinsky (see above). Rather than explain this as a source of tonal

tension, Saizer regards it as a self-contained fact, which can be "composed out" by

means of motions into inner voices. [:194 and Ex.416].

Analyses such as these have prompted a critical response from, among others,

Joseph N. Straus [Straus 1982a, 1982b and 1987]. Straus asserts that a theory is

needed to account for surface style and deep structural organization and that this will

entail "a systematic and coherent view of pitch organization at all levels" [1982a].

Straus rejects the application of the Schenkerian concept of prolongation to

Stravinsky's music, but he retains the theoretical division of pitch relations into

harmony and voice-leading, by formulating his theory as a response to two questions;

if much twentieth-century music is organized around functional tone-centres, but these

are not tonal in the traditional sense, then "(1) what is the nature of the tone centres?,

and (2) what is the means of progression between the tone centres?" [1982a].

Straus's answers - respectively his concepts of "tonal axis" [1982b] and "pattern

completion" [1982a] - are synthetic elements grafted on to an organic system of

hierarchical levels. In this respect, Straus is no different from Salzer, although he is

at least clearer in his exposition of new theoretical principles.

Straus's "tonal axis" is defined by three conditions: (1) it is formed from

overlapping major or minor triads (such as e-g-b-d or e -g-bt, -d); (2) it must function

59
as a referential sonority (i.e. as a prominent foreground harmony); (3) it must embody

a conffict or polarity between its constituent triads, this polarity being the principal

determinant of the work's structure. Straus thus identifies the collection c-e-g-b as the

"tonal axis" of Stravinsky's Symphony in C, embodying a conflict between C major

and its mediant e minor. This finding is corroborated by the work of other writers on

Stravinsky [Whittall 1988: 61-641, although Straus's interpretation minimises another

important triadic relationship, that of the dominant - gbd - with tonic and mediant.

While the "tonal axis" may be a useful, if somewhat rigidly defined, concept,

"pattern-completion" is more problematic. An unordered collection or set, "usually a

tetrachord", is established as a melodic and harmonic "norm" in the foreground. The

presence of all but one member of the "pattern" in contextually prominent roles creates

the psychological expectation of the arrival of the missing member to "complete" it.

The pitches of the pattern are thus "associated" across stretches of intervening material

(rather than "prolonged" in the strict Schenkerian sense), forming a level of structure

akin to a tonal middleground. The resulting relationship between foreground and

middleground thus resembles the "motivic parallelism" observed by Schenker in many

tonal works. However, while in tonal theory, motivic parallelism may be expected to

fulfil only part of the structural role of deeper levels, in Straus's theory of post-tonal

music it constitutes the whole of the middleground, thus apparently providing post-

tonal compositions with a high degree of organic unity.

The concept of "association" rather than prolongation is potentially a useful one

and is amplified in Straus's later work [Straus 19871. However, in his work on

Stravinsky, he does not make clear his criteria for the selection of particular pitches as

carriers of greater structural weight than others and this inevitably weakens some of

60
 his analytical observations. Furthermore, the idea that a psychological expectation of

a particular pitch may be created is undermined when there may be more than one way

to complete the pattern. 6 Whatever the value of the idea of pattern completion, it is

doubtful that it can really be equated with voice-leading. It is noticeable that Straus

nearly always identifies pitch-class sets which resemble tonal scale-segments and which

therefore suggest a quasi-tonal goal-directed motion or a "gap-fill" construction, but

the concepts of completion (which l)ahlhaus would regard as a manifestation of

harmonic system) and progression (harmonic process) are fundamentally dissimilar.

Finally, the polemical stance adopted by Straus appears unreasonable. In dismissing

attempts to find tonal or octatonic structures in Stravinsky's music as "fallacious"

[1982b], he denies the possibility of the fruitful interaction of these elements with those

of his own theory. By rejecting the validity of a pluralist approach to Stravinsky (the

pluralist composer par excellence), Straus leaves himself open to the charge of

misunderstanding the true nature of the post-tonal music he defends.

The work of Pieter van den Toom [van den Toorn 1983, 1987, 1995, 1997] on

Stravinsky's music offers some useful concepts for analysis of Carter's earlier works.

Van den Toorn's work is based around the identification of the octatonic scale as a

major component of Stravinsky's musical language. Phenomena such as the

"polychord" g-b-dl' -f-al', which SaLzer takes as the underlying prolonged sonority in

the opening of Stravinsky's Symphony in Three Mowments, can thus be shown to

6
For example, in Stravinsky's Symphonies of Wnd Instruments, Straus identifies the
tetrachord [0,1,3,51 as the "pattern" [Straus 1982a1. Between rehearsal figures 6 and 8,
the second flute line revolves around the pitches dl', dl' and e, allegedly creating the
xpectation of the e4 which arrives at figure 8. There is no reason in Straus's theory
vhy b' would not fulfil the function of "pattern completer" equally well. (A similar
j,roblem exists with the pattern completion in the third flute part in this passage.)

61
belong to a single octatonic collection rather than being formed from the juxtaposition

of triads. While the octatonic scale itself is of little importance in Carter's music, the

relationship between the "closed" symmetrical structure of octatonicism and the "open"

structure of functional tonality, which van den Toom explores, is of great significance.

In discussing the relationship between octatonic and diatonic elements, van den

Toorn notes that the "dominant seventh" collection [0,4,7,101 and the "major/minor

third unit" [0,3,4,7], both important quasi-tonal components in Stravinsky's

neoclassical style, are also both subsets of the octatonic collection. However, a V'-I

cadence is not possible within a single octatonic collection [van den Toorn 1983: 330].

Thus,

"tonic resolution" often surfaces as a kind of terminating convenience in

contexts either octatonic or octatonic-diatonic in conception, so that,
apart from qualifying "impurities" in the concluding simultaneity itself,
the "resolution" may seem incidental to pitch organization generally.
[:3301

An important example of the above occurs at the end of the Symphonies of

'Mnds, where the pervasive octatonic sonority g-b-d-f/bb -(1-f-at) eventually "resolves"

on to c-e-g-b-d. This progression involves a re-interpretation of the octatonic

sonority; during the course of the work, it appears as a static, referential sonority, but

the final cadence invests it with a dynamic "dominant" quality which van den Toorn

describes as a "retroactive function" [:337]. Van den Toorn's approach thus allows

for the possibility that tonal function may play an intermittent role, rather than being

either ever present in distorted form (Saizer) or entirely absent (Straus).

A subtle analysis of the elusive and multivalent qualities of Stravinsky's

references to tonal sonorities and functions (in his serial as well as his "Neoclassical"

62
 works) can be found in Whittall 1982. In the Anthem of 1962, Whittall discovers

"various embedding? of tonal progressions within one another, "the effect of [which)

is to disperse local tonalities into general atonality" [Whittall 1982: 11). This

technique parallels that found in the first movement of the earlier Serenade in A, in

which the illusion of a tonal Fundamental Structure is created through emphasis on the

pitches a, e and a, but in which these pitches tend not to be heard "as members of a

single, integrated, unifying process, rather than as the representation of one process

enclosing a second process."[Whittall 1982: 131

Although the tones A and E are basic to its structure, it is difficult to

think of them as in a fifth-relation of anything approaching the
traditional diatonic kind. Rather, they are contrasted, polar
alternatives. More radically still, they may be the principal, centripetal
agents of the breakdown of what might have been a unified, goal-
directed structure, following traditional 'laws of continuity' and 'rules of
continuation'. [Whittall 1982: 131

The other principal figure of Neoclassicism, Paul Hindemith, has also been the

subject of analytical enquiry which may be drawn upon in this debate. David

Neumeyer's The Mnsic of Paul Hindemith [Neumeyer 1986]- another in the Yale series

- evolves an analytical method based to a large degree upon that composer's theoretical

pedagogy. Although intended expressly as a tool for the "description" of the music's

technical features and not as a "prescription" for its conect interpretation, Neumeyer's

explanation of his method clearly suggests a parallel with the views on post-tonal

music of Puffett and Joseph Straus:

It is.. .not a reductive theory as that term is usually employed nowadays,

despite its five hierarchic stages. The stages are used to separate the
different processes or components of the music and not to demonstrate a
closed system of structural levels. Only Stage ifi can properly be
called integrative, and it is the one most open to individual

' Also see Pople 1996, for amplification and critique of Whittall's ideas.

63
 interpretation. ...It is the essence of Hindemith's pedagogy and
compositional practice that he permits different levels of definition for
all the musical elements. In effect, any one of the elements has its own
continuum of definition from greatest clarity at one end to obscurity or
even randomness at the other, and the resulting network of definition
levels is the most important factor in determining the technical character
of a piece. [:49-50]

In other words, in place of the Schenkerian/Salzerian system of organically related

structural levels, Neumeyer posits a looser network of levels, whose relationship is not

governed by any a priori laws or forms but is flexible and variable from work to

work. ("There is no fixed relationship of the primary elements but any number of

possibilities for their "balanced cooperation."" [:49, footnote 1.1) This model has

much in common with Puffet's notion of the disjunction between structural levels in

Tippett, and Dunsby and Whittall's emphasis on symbiosis rather than synthesis in

post-tonal music.

The analysis of neoclassical music thus involves consideration of the following;

(a) the juxtaposition of triadic elements and the extent to which the sonorities thus

formed may retain a tonal function;

(b) the interaction of the "open" tonal system and "closed" symmetric systems;

(c) voice-leading through tonal "prolongation" or through non-tonal "association".

3. Analysis of Carter's Music

Before the task of outlining a suitable analytical method or methods for Carter's

transitional music is undertaken, the work of other analysts of Carter's music will be

reviewed. The first attempt at comprehensive exegesis of Carter's work was made by

David Schiff in 1983 in The Music of Elliott Carter. Schiffs study is the product of a

64
close personal association with the composer and this is both its strength and its

weakness. As an introductory account, it is exemplary, combining enthusiasm and

clarity of exposition. However, Schiff rarely extends far beyond the information

supplied by the composer with regard to overall form and basic musical materials.

Nor does he concern himself deeply with analytical methods and their applicability to

Carter's music, making only brief reference to pc set theory. Thus, despite some

essential insights into Carter's central ideas and techniques, Schiff does not achieve the

necessary independence from the source of his information for analytical detachment.

With the making available of Carter's manuscript scores and sketches the way

became open for more independent analytical work. Jonathan Bernard's article

"Spatial Sets in Recent Music of Effiott Carter" [M4, 1983] was one of the first

products of this availability. Bernard's approach is based primarily on the application

of set theory to Carter's mature music, modified by his work on Varèse which

emphasizes the spatial aspect of harmony [Bernard 1981]. (The article deals with the

way in which complex textures are built up from the combination and duplication of

intervals and sets.) Pitch, rather than pitch-class, is stressed, which is a significant

departure from orthodox set theoTy, and furthermore, the nexus set/set-complex

principle of Forte's theory is not brought into play, as it conificts with the principles of

organization of harmonic vocabulary outlined in Carter's own notes. Bernard

manages to give more detail and substance to "flesh out" the outlines suggested by

Schiff. However, this is still far from a complete view of Carter's musical language.

The concentration on spatial harmonic matters treats Carter's music in a rather one-

dimensional manner since it excludes the temporal, developmental aspect which is so

important to Carter. Furthermore, as Bernard acknowledges, the task of analysing a

65
 whole movement by Carter in the bar-to-bar detail assayed in the five to ten bar

segments tackled in the article, would be almost prohibitively demanding [Bernard

1983: 32J. Bernard's subsequent published work on Carter has similarly concentrated

on single aspects of the composer's technique and approach [Bernard 1988, 1995,

1996].

David Harvey's doctoral dissertation The Later Music of Elliott Carter: A Study

in Music Theory and Analysis [Harvey 19891, appeared almost in fulfilment of

Bernard's challenge. Harvey's thesis is an attempt at a comprehensive description of

Carter's musical language as it is embodied in three major works (the Second String

Quartet, Double Concerto and Concerto for Orchestra). Analyses of these works

follow several introductory chapters which place Carter in the context of American

composition and theory, outline his background, influences and early career, and

develop an analytical method suitable to various types of atonal music. The latter is

formed from a consideration of Fortean set theory together with criticisms and

modifications of it by such writers as William Benjamin, Christopher Hasty, Jonathan

Bernard and Joseph Straus. The product is a method which employs Fortean notation

to describe significant foreground components and also "background sets", which may

appear to dominate a passage through the repetition and contextual reinforcement of

particular pitches, and attempts to describe pitch-continuity in terms of intervallic

"process" and symmetry. While acknowledging Bernard's approach as one among

several which have helped to refine pc set theory into a more "context-sensitive"

analytical method [Harvey 1989: 41-6], the latter's "architectural" approach is

criticized as portraying Carter's music as rather static, and since Harvey's avowed

intention is to produce analyses which account for "perceived continuity in the domain

66
 of pitch" [:69], he takes equal note of analysts such as Benjamin and Hasty whose

bias is towards "process" rather than "structure".

Bernard's review of Harvey's work [Bernard 1990] pinpoints certain slightly

problematic features in what is otherwise a "usefully stimulating" publication. Bernard

quibbles with Harvey's description of Carter as an "empiricist", suggesting that a

comparison with the American composer-theorists such as Babbitt and Martino should

not denigrate Carter as "lacking in rigour". A more serious criticism is that Harvey's

criteria for the segmentation of the scores and the selection of particular pitches and

configurations for middleground or background priority are not always explicitly clear.

A wide-ranging methodological eclecticism would seem to be a natural

approach to analysis of the music of Elliott Carter, given the plethora of musical

influences - from Ives, to Mozart, to African drumming, to Stravinsky - that he

acknowledges. Harvey brings a suitably broad range of analytical ideas to bear on

Carter's later works. However, his one extended analysis of one of Carter's

transitional works, the Cello Sonata (1948), is disappointing. After exposing some of

the work's generative motivic cells, 8 Harvey attempts a graphic representation of the

first movement in terms of the "structural progression of focal pitches." The results,

however, are sketchy and inconsistent - at some points showing inappropriate quasi-

tonal "voice-leadings" and at others simply juxtaposing elements without explaining

their supposed relationship. (See Bernard 1990: 348 for further criticism.) Bernard

[:354] warns against the assumption that structural levels operate in Carter's music in

the manner "construed by Schenkerian theory and those theories of 'transitional' or

Harvey bases his analysis on the premises set out in Schiff 1983 [: 137].
}Iowever, an alternative explanation of the movement's basic pitch materials can be found
in Kies 1984 (see below, Chap.5, pp. 125-6).

67
post-tonal music based, however generally or partially, on Schenkerian models." It

remains to be seen whether the works of 1945-55 respond to analysis on these bases.

To summarize, the following are the main points which must be understood by

the analyst of Carter's transitional music:

1. Carter has left some useful clues, in the form of sketches and writings, about his

compositional methods, but these can only inform the beginning of an analytical

approach, they cannot tell the whole story.

2. All writers are in agreement that the dense complexity of Carter's scores makes

detailed analysis of large-scale works a daunting prospect. The prevailing image is of

a near-impregnable fortress being assailed by a battery of unwieldy siege-engines.

For an analysis to be effective, it must ultimately render its subject more readily

comprehensible. Inevitably, this entails simplification of the work's complexities, but

on the way to this final view, the complexities must be fully explored and response to

the work rendered even more complex. In the case of Carter's music this represents

a considerable challenge in terms of stamina.

3. The analyst must recognize that at the beginning of the period under consideration,

Carter was writing within the established idiom of neoclassicism, and take account of

methods currently available to approach this idiom, but must also be aware that Carter

may have (consciously or otherwise) subverted that idiom, given his later radical

departure from it.

68
4. Analysts should also accept that the transitional nature of this music necessitates

bringing a variety of methods to bear it, while recognizing that these methods should

be exploited only as far as they yield meaningful results that do not distort the facts

represented in the compositions themselves, and also being aware that the use of

different methods may yield complementary or contradictory results.

69
PART TWO: ELEMENTS OF CARTER'S MUSICAL LANGUAGE iN THE
TRANSITIONAL PERIOD.

70
 CHAPTER 4: KEYS, MODES, FIELDS

1. Diatonic fields

In common with the musical languages of his principal models at that time,

Stravinsky and Copland, Carter's was characterized, in the early 1940s, by a

prevailing diatonicism. However, diatonic pitch-content must not be taken as an

outward sign of conventional tonal harmonic structure, and we should therefore be

waiy of describing his music as being in particular "keys", as this term carries with it

the assumption of an organic hierarchy of elements with pre-ordained functions.

Accounts of this music which full to address the fundamental differences between the

key-relationships of tonal music and the varied pitch-structures of post-tonal music risk

seriously misrepresenting both.' As David Harvey observes with regard to the

second movement of the Cello Sonata, "'key" is usually only definable as a property

of the scalar relationships of the pitches present at any given moment, rather than as a

function of harmonic relation." [Harvey 1989: 341 In other words, we have

diatonicism without a tonic, or any degree identities. It will be useful, in discussing

this music, to develop a terminology which, while acknowledging both its prevailing

diatonicism and its tendency to be organised around "tone-centres" (analogous but not

identical to "tonics"), does not bind these two into the indissoluble unity that the term

"key" implies.

The diatonicism of some of Carter's music, when taken in conjunction with its

Witness Wilfrid Mellers's description of bb.33-41 of the first movement of Carter's

Cello Sonata: "In this passage the basic tonality is B major, which is, however,
telescoped with D major and 0 sharp major, which becomes enharmonically A flat."
[Mellers: 107]

71
avoidance of functional tonal relationships, may lead to descriptions of it in terms of

the ancient church modes. Although modal writing was a characteristic of some

neoclassical music, it is not really appropriate to describe Carter's works in these

terms. The use of "modes" still implies triadic harmony and a (loose) hierarchy

based on the "final" degree, neither of which could be said to be continuously in

evidence.

Consideration of a passage from Carl DahThaus's essay "Tonality: structure or

process'?" [in Dahihaus 1987: 62-721 will help to elucidate the nature of diatonic post-

tonal music such as Carter's. Dahihaus uses two principal criteria in order to

distinguish between various types of musical language or "system"; they may be

"open" or "closed" on one hand, and "centred" or "non-centred" on the other:

The chromatic scale, if understood according to the principle of

distance, as in dodecaphony, that is, without distinguishing between
diatonic and chromatic semitones, is a closed complex - as is the
pentatonic scale, which has no semitones at all. On the other hand, the
overtone series, which can in theory be extended into infinity, is an
example of an open system - as is major-minor tonality, whose ability to
accommodate remote harmonies is unlimited.

Open systems are and must be centred, since, in a situation where there
are no limits, the grouping around a central point seems to be the only
way of ensuring systematic coherence. Similarly, closed systems may
indeed also contain a tonic or principal note, yet having a centre is not
necessary, as dodecaphony shows; and where a centre is prominent in a
closed system - as in a mode in non-semitonal pentatonic music - it has
the appearance of being a secondary structure or something that has
been super-imposed. A centre is not essential in closed systems,
though it is in open systems. A tonic in pentatonic music, where it
exists, is, so to speak, an additional determining factor, in the overtone
series and the complexes derived from it, by contrast, it constitutes the
basis of tonal coherence. [Dahlhaus 1987: 67-81

Dahihaus's description of major-minor tonality as an open system requires

qualification. The equal temperament system of tuning, adopted since the eighteenth

72
century, forced tonality into a closed, cyclic structure, consisting of twenty-four major

and minor keys. One can only accept tonality as a genuinely open system if one

refers exclusively to the just-temperament system of tuning which adheres strictly to'

the properties of the overtone series. Nevertheless, if one temporarily accepts a

single tonic as the centre of the equal temperament system (as, for example, in

Schoenberg's "Chart of the Regions" [Schoenberg, 1969: 20], a network of functional

relations is set up which enables distinctions to be drawn between apparently identical

keys in terms of their "relative", rather than "absolute" values. For example, if we

take C major as tonic, the key of A major may appear as (a) the dominant of the

supertonic, (b) the parallel major of the relative minor, (c) the subdominant of the

mediant major, or in a large number of other possible relationships to the tonic. if

one accepts that all these keys are different entities because they are defined by

different relationships to the tonic, then Dahihaus's view of tonality as an open system

may be endorsed. With the removal of the notion of a tonic, the cyclic equal

temperament system collapses in on itself and, as Dahihaus states, the only relationship

between pitches which remains is that based on "distances" between them, measured in

semitones [Dahihaus: 64].

The harmonic organization of Carter's later music is a clear example of a

"closed system"; chords are categorized by their interval content (by counting distances

in sernitones) and distributed into separate "vocabularies", whose interaction, although

rich and varied, is governed by reference to a pre-compositional scheme. Dahthaus' s

assertion that a centre is not essential in closed systems is borne out by Carter's own

attitude to the question of the importance of "absolute pitch":

73
 The actual notion of "absolute pitch" is not significant in my pieces.
The pitches are chosen registrally as a matter of instrumental
practicality. In fact, I frequently transpose parts of my pieces up and
down, when I compose them, to try and decide in just which register
they would sound most characteristic, given the instruments that are
playing them. [Edwards: I 10]

However, in Carter's earlier music, a variety of approaches to "centredness"

are explored and the relationship between open and closed systems is very much a

subject for investigation. We encounter passages of apparent pentatonic or modal

simplicity, passages which exploit the ambiguity of one or two "false relations" or

enharmonic changes, passages which give the appearance of "bitonal" organization and

passages which display the kind of chromatic saturation normally associated with

"atonality" -

For the purposes of describing the pitch-content of the predominantly diatonic

parts of Carter's music without recourse to the terms "key", "scale" or "mode", we

will adopt the term "diatonic field". To describe a passage as being in the "diatonic

field" of C, means only that its pitches are diatonic to the scale of C major, not that C

is necessarily perceived as the tonic upon which all other pitches are structurally

dependent. Italic notation will be used to identify diatonic fields. Thus Cis "the

diatonic field of C". (In set-theoretic terms, this is equivalent to set class 7-35 with

the specific membership [11,0,2,4,5,7,9].)

The adequacy of the description given above may, however, be compromised

by the possibility that a passage may suggest a "minor", rather than a "major",

orientation. This point illustrates precisely the difficulties which may arise in

separating diatonic content from tonal function. if, in a passage whose pitch-class

content suggests C, the note a is given contextual prominence, an "a minor

74
 orientation" will naturally be heard. This does not, in itself, invalidate the

classification of the passage as being in C, since, as explained above, this only

describes the content of a set, rather than establishing a hierarchy among its members.

However, should the passage in question mimic certain conventions of the traditional

minor mode - specifically, the variability of the seventh degree of the scale (the

"sharpened leading note") - the classification of its diatonic orientation must be

amended to account for this. Thus, for example, a passage in the diatonic field a will

have an identical pitch-class content to one in C, except that in the former, the pitches

g and g# will be interchangeable. (In set theoretic terms, 8-26 [l,8,9,ll,O,2,4,5J.Ø

This phenomenon will be observed occasionally in Carter's music, but more usually,

other means of complicating the sense of diatonic orientation will be encountered.

Carter often employs what may be described as composite diatonic fields.

These are fields which are derived from two major scales adjacent in the cycle of

fifths whose pc content therefore overlaps by six elements. Thus the diatonic field

C/G contains eight pitch-classes, including the "false relation" V -f. (This is

equivalent to the set class 8-23 with the specific membership [4,5,6,7,9,1l,O,2J.)

Occasionally, three adjacent major scales may be used; CIGID contains nine pitch-

classes, including two "false relations" f# -ft1 and c# - cii (equivalent to set class 9-9

[11,0,1,2,4,5,6,7,9].)

Conversely, diatonic orientation may remain ambiguous because there are fewer

pitch-classes present in a particular passage than the seven constituting a diatonic scale

or field. In this case, the presence or absence of the tritone relationship is the crucial

factor. The collection (c,d,e,f,g,a) (set class 6-32 [0,2,4,5,7,9]) may be a partial

representation of either F or C. This will be represented thus: (F/C). A still smaller

75
collection, the set 5-35 [0,2,4,7,9], contains neither tritone nor semitone relations and

may therefore be representative of one of three diatonic fields; for example (c,d,e,g,a)

may be described as (F/C/G).

An obvious parallel exists between the relationships among diatonic fields and

those among tonal tonic and dominant key areas, the difference being that in a

composite diatonic field, "tonic" and "dominant" are equal partners, the focal or

central pitch-class(es) being determined by contextual factors rather than through

inherent structures. Motion between diatonic fields also has a parallel with

modulation in functional tonality (especially in Schoenberg's sense of motion between

regions), in that certain pitches may play a pivotal role by being open to

reinterpretation in different harmonic contexts. Motion may be smooth - to a field

whose pc content has a maximum overlap with the preceding one - or abrupt - to a

field which has relatively few pcs in common with the original. In the case of a

smooth "drift", this may be effected by the introduction of a single new "false

relation". In the case of abrupt "shifts", the enharmonic re-spelling of particular

pitches is often involved. For example, the harmonic reinterpretation of leading notes

as tonics in their own right (or vice iersa) is an important part of the harmonic process

of the Piano Sonata (a#, the leading note in B major becomes bk', the tonic of BI'

major).

The equation of various forms of diatonic field with pitch-class sets which has

been introduced above may serve as a point of contact with Allen Forte's theory of pc

set "genera" [Forte 1988]. Forte's theory seeks to establish "families" of pc sets, the

members of which are related through holding certain "progenitor" subsets (trichords)

in common. The sets which have been identified as significant in the classification of

76
 Carter's musical language in the early part of his career - 5-35, 6-32, 7-35, 8-23, 8-26

and 9-9 - are all members of either Genus 11 (the "diatonic") or Genus 12 (the "dia-

tonal"), which together form "Supragenus N". 5/7-35 and 6-32 are members of

both. Only 6-32 is a member of any other Genera (Genus 7, the "chroma-dia", and

Genus 10, the "atonal-tonal"). In this respect, these sets represent a relatively tightly-

knit group - almost a "closed system" - and thus perhaps demonstrate that Carter's

interest in working with clearly delimited pitch vocabularies predated his systematic

classification of chords in the "Harmony Book". The role of the pc set genera in the

classification of "chords" and smaller segments of the musical surface will be

examined in Chapter 5.

Three passages from Carter's earlier works will provide brief examples of his

handling of these diatonic fields; the song The Rose Family (1942) offers the earliest

and most straightforward of these examples, while two passages from the second

movement of the Piano Sonata (1945) demonstrate similar techniques used in a more

elaborate manner, taking us, in one case, to the edge of what can be explained in

terms of diatonic fields.

(a) The Rose Family

Robert Frost's miniature is a satire on the mutability and fragility of referential

meaning in modern poetry, in which the traditional concept of metaphor - one object

standing for another - may be developed to absurd extremes. 2 The text is reproduced

here;

2Gertrude Stein is the principal target of this satire, according to Schiff [1983: 92].

77
 The rose is a rose,
And was always a rose.
But the theory now goes
That the apple's a rose,
And the pear is, and so's
The plum, I suppose.
The dear only knows
What will next prove a rose.
You, of course, are a rose -
But were always a rose.

The certainty expressed in the opening pair of lines is undermined by the

subsequent "theory" that other objects may be symbolically or metaphorically regarded

as "roses". These are listed in the order "apple...pear...plum", exhibiting a

progression which has its own logic, but which takes us far from the literal concept of

a rose. The next pair of lines represent the opposite pole from the certainty of the

first pair; almost anything may conceivably represent, or be represented by, a rose.

Meaning is therefore undermined and devalued. The final pair of lines contain an

ironic twist; they return to certainty (with an echo of the word "always"), but admit

the validity of metaphor within certain "understood" conventions (here the courtly

compliment of a poetic lover).

Although the song (in keeping with the poem) is slight, its harmonic technique

and language succinctly demonstrate many of the salient features of Carter's style.

The poem's transformation from the apparent certainty of literal meaning to the

uncertain world of symbolic meaning finds an analogue in Carter's music in the play

between the closed system of pentatonic harmony and the rather more extended

harmonic sphere of the cycle of fifths. Example 4.1 shows the score of the song with

an analysis of its diatonic fields.

Bars 1-7, comprising the piano introduction and the first pair of lines, reflect

78
the meaning of the words by using rhythmic and harmonic ostinato. The naivety of

the poem's sense here is conveyed by the useofa pentatonic scale ona flat -ak, bk,

c, el, f. Although al' is emphasized by certain contextual factors (its position in

the extreme bass, its metrical position of the first beat of each bar and the

"arpeggiation" of the at triad in the vocal part), others detract from its status as

"tonic" (the absence of leading note g and subdominant dli). The harmonic material

of this section may, in fact, be defined as a continuous segment of the cycle of fifths:

al' -e -bL -f-c. Significantly, this set (5-35) is the largest such segment which can be

drawn from the cycle of fifths which contains no semitone relationships between its

members. The absence of semitones makes the harmonic texture free of tension in

the conventional tonal sense, but nevertheless leaves a 1eling of ambiguity, because of

the latent expectation that this segment of the cycle of fifths will be extended by two

elements, in order to make up a complete scale or diatonic field. Thus, Carter is able

to play on the paradoxical nature of pentatonicism within a tonal context; it may

appear, on one hand, a "closed system", and, on the other, as an incomplete

manifestation of a diatonic field.

Carter does indeed extend the cycle of fifths segment, bringing about a gradual,

but increasingly rapid shift of diatonic orientation, which mirrors the progression in the

poem away from the literal presence of the rose, towards the surreal symbol of "plum

as rose". The first stage in this process is marked by the introduction of dl' in the

phrase "But the theory now goes" (bb.8-lO), expanding the pc field from 5-35 to

hexachord 6-32 and thus focusing the diatonic orientation from (Db lAb IRk) to (Db lAb),

However, the presence of the seventh member of the field - g or gI' - is withheld

until bar 11. Here, g is introduced in the bass, but c is absent for the first time in

79
the song and is displaced in bar 12 by c (respelled enharmonically as in the

piano). This is quickly followed by the introduction of e11 , a and d4 in bars 12-13.

(The voice part continues to employ flat notation.)

The appearance of dI (emphasized by its extreme bass register) represents the

furthest pole from the opening A' and thus corresponds with the poetic "mutation"

from rose to plum. The disorientation expressed in the poem at this point ("The dear

only knows...") is reflected by the reappearance of the opening material transposed a

semitone higher. (Semitonal relationships such as this play a vital role in the

organisation of much of Carter's music of this period.) The stability of bars 14-17,

which express the field (DINE), is short-lived, however, as gl and c displace g# and

c# in bars 18-19. This would appear to suggest that the flatward drift is continuing.

However, c# and g# are reinstated in bars 20-21 and the next few bars provoke a

"crisis" in terms of diatonic orientation; in place of the gradual movement through the

cycle of fifths, a more abrupt motion appears, resulting in the close juxtaposition of

"false relations", for example d4 -d# in bar 22. In bars 24-5, there is a rapid

sharpward motion, bringing a#, e# and b# (=c) in quick succession and leading to the

song's second climax in bar 25 on a chord which expresses the diatonic field (E. LB.).

From this point to bar 32, there is a gradual, almost imperceptible, return to

the diatonic orientation of the song's opening. The d}1 of 25-6 disappears in 27-30

and is displaced by d in 30-31. During the final part of the song ("But were always

a rose") neither d nor dL are present, nor are g or gk, and thus the (D1'/Ab/Eb) field

of the opening is restored.

80
 (b) Piano Sonata, Second Movement, bb.1-26.

This passage (see Ex.4.2) reveals several striking similarities with the song

described above; the harmonic texture is almost exclusively diatonic, although "pure"

triads are avoided; there is a large-scale motion between contextually determined tone-

centres a semitone apart; this motion is effected through a gradual "drift" through the

cycle of fifths.

The internal formal organization of the passage may be understood in relatively

conventional terms as "exposition" (1-15), "development" (16-20) and "climax and

restatement" (21-26), the latter taking the form of an explicit recall of the opening

texture and rhythm (c.f. bb.1-2, 25-26). The climax may be regarded as a more

oblique reference to preceding material, retaining the melodic contour, but altering the

actual intervals (see Ex.4.3).3

A Saizerian interpretation of this passage would divide it into two phases; a

"prolongation" of (d,b) as a referential, or tonic, sonority extending from b. 1 to b. 15,

with the bass moving from d in its "obligatoiy register" to the b of an "inner voice" in

b.15; the second phase (b.16-26) would be regarded as a passage of "prolonged

motion", in which the structural d gives way to an el ' neighbour note, the structural

significance of both pitch classes being indicated by their presence in no fewer than

five octave positions. (This analysis would continue by demonstrating the

prolongation of eL' through bars 25-51 and its "resolution" with the return of d in bars

This technique of interval expansion or contraction is a common device in Carter's

transitional music (it occurs in the Cello Sonata, Quartet Sonata and First Quartet) and
demonstrates an interest in finding ways of relating thematic material other than through
nore obvious motivic/intervallic procedures. It will be more thoroughly explored in
chapter 7, pp. 180-184.

81
51-52; see pp.84-S of this chapter.) The question of whether we may speak of

"prolongation" in this music must wait until a discussion of harmony and voice-leading

is undertaken. However, the relationship of tone-centres a semitone apart is of

undeniable importance in the work (the first movement plays on the relationships of b

with a#/b1 and b with c) and an "association" of d and eI may be accepted with few

qualms.

The harmonic material of most of the passage bears an oblique relation to

tonality; as in The Rose Family, its almost pure diatonicism is set against an avoidance

of triads and of functional tonal progressions, particularly those involving motion by

semitones. The resulting harmonic texture thus seems more "modal" than tonal and

thismodalityisreflectedinthewayinwhichthenotedisusedasatone-centre. A

further reference to Dahihaus' s notion of "open" and "closed" systems should elucidate

the function of the pitch d in bars 1-15. These bars inhabit a harmonic structure

which is "almost closed". Tension is generated between the overlapping diatonic

fields of C and F and between the static, "anti-progressive" characteristics of modality

and the dynamic, processive nature of tonality. The harmonic structure is "almost

closed" in that it admits one ambiguity, the alternation of b and br'. The reiteration of

d as a melodic starting and end point has some of the characteristics of a "secondary

structure" or "an additional determining fuctor", giving the music a "dorian modal"

quality, but it takes on greater significance through its opposition to g. Example

4.4(a) illustrates the relationship between the modes on d and g, which hinges on the

exchange of the semitone b-c for a-b'. Example 4.4(b) is an abstract representation

of the pitch-relationships within the upper voice melody of bars 1-15. A symmetrical

structure is set up, with d and g as focal points or poles, each decorated by upper and

82
lower neighbours. To fill the space between g and d, however, the symmetry must

be broken, and one of two diatonic alternatives chosen. The music thus contains a

destabilizing equivocal element which counteracts the otherwise static and self-

contained modal structure.

From bar 15 to the end of the passage a stronger sense of "directed motion"

emerges. The jostling for position of bli and b generates a process of "modal

modulation", which is not so much the establishment of a succession of keys, as the

gradual displacement of elements of the current mode by those of the mode transposed

by a perfect fifth, producing a sharpward "drift". (F is displaced by f.# and c by c# in

b.16,gbyg#anddbyd#inb.l8etc.) Anenharmonicchangetakesplaceinbars

20-21 so that c## becomes d and e# becomes f.

From this point, the harmony not only begins a reverse flatward drift, but

becomes increasingly dissonant. In terms of diatonic orientation, bars 21-6 take us to

the composite field el' /a, in which the predominantly flat notation is contradicted by

the "raised leading notes" g11 and d11. The climax and resolution of the passage are

brought about through the introduction and subsequent elimination of harmony which is

not explicable in purely diatonic terms. This represents an intensification of the

process in The Rose Family, where no single sonority is actually non-diatonic in itself.

In the harmonic vocabulary of the passage from the Piano Sonata, the most obvious

distinction lies between those vertical sonorities which are subsets of the diatonic scale

(7-35) and those which are not. The former make up the greater part of the passage

by far. The latter, only three different sonorities, are confined to the final section of

the passage (bb.23-4). The relatively greater dissonance of these three sonorities may

be explained by the fact that they contain semitone relationships which cannot be

83
 interpreted diatonically. The succession of vertical sonorities in bars 21-6 - single

note, [0,3] dyad ("minor third"), 3-11 ("minor triad"), 4-26 ("minor seventh"), 3-10

("diminished triad"), 4-3 [0,1,3,4], 4-18 [0,1,4,7], 4-5 [0,1,2,6], 3-5 [0,1,5] (no

diatonic "equivalents"), octave - clearly represents a pattern of intensification and

relaxation suggestive of traditional harmonic process. (A more thorough investigation

of Carter's use of particular "chords" or "sets" will be undertaken in Chapter 5.)

The musical language of this passage thus represents an advance in terms of

complexity over that of the earlier song. The principal reasons for this are (i) the

immediate introduction of an element of ambiguity (the "false relation" b-bk') and (ii)

the appearance, at the climax of the passage, of non-diatonic sonorities (which can

nevertheless be explained as the result of stepwise voice-leading).

(c) Piano Sonata, Second Movement, bb.27-52

The passage immediately succeeding the one described above offers contrasting

thematic material and texture, but is nevertheless bound to it by the overall motion

back from eE to d (see Ex.4.5). The harmonic process involved in conjunction with

this motion is similar to that employed in the opening paragraph of the movement, but

the assault on pure diatonicism is carried to further extremes. This, combined with a

progressive breakdown of rhythmic, melodic and textural stability, produces a passage

whose violent disruption of the norms established at the beginning of the movement is

almost expressionistic in effect.

The passage begins stably in the diatonic field G which persists from bar 26

to 29. During the next two bars, there is a gentle sharpward "drift" of the kind

84
 familiar from the passages described above; the introduction of ct in bar 30 and of dt1

and gLj in bar 31 take us through the cycle of fifths to the field (Eb /Bb). In bars

32-33, Gb is reestablished through the reintroduction of g, d' and c. This

process of departure and return reinforces the local stability of G.

This diatonic field is then sustained for another eight bars, during which time,

the music takes on a greater feeling of conventional regularity than has previously been

experienced in the movement; a new rhythmic ostinato figure emerges in the outer

voices of a three-part polyphonic texture, the middle part developing a song-like

melodic line. The only harmonic "foreign body" which occurs during bars 33-40 is

thee(?I in thebassofbar37. Thecontextin which thispitchoccursallowsusto

interpret it contrapuntally, that is as a passing note in the "bass voice" between the

pitches eli (in b.36) and d (in b.38). The latter are both members of the prevailing

diatonic field and therefore apparently more "stable" than the intervening pitch, which

thus appears dependent upon them. The association of these three pitches through

similarity of registral position and metrical placement naturally creates the impression

of a descending linear progression.

The precedent set by the appearance of such contrapuntal phenomena is used to

undermine the diatonic stability of the harmony of the passage from bar 41 onward.

Here, we have a sense of "bifurcation" of harmonic orientation, with one part of the

texture remaining within the prevailing diatonic field and the other drawing us away.

Thiseffectisbroughtaboutby theappearanceofthepitchbb inthelefthandinbar

41 which is placed in direct opposition to the b s which appear higher in the texture.

Thus while the upper part of the texture remains within (3k during bars 4 1-5, the

lowest voice persists with the downward linear motion begun in bars 36-40, apparently

85
taking us into the field Ck. However, although the bI of bar 41 at first appears to

be a passing note, identical in function to the el of bar 37, its subsequent

enharmonic respelling as a in bars 42-4 suggests a more radical challenge to the

prevailing diatonic field. This challenge materialises in bar 45, where the descending

linear motion in the bass arrives on gt1 , thus suggesting a diatonic field of (Bi / F) in

the left hand. Meanwhile, the right hand part moves towards C with the

introduction of the pitch fl in bar 45. Thus at this point, there is a radical separation

between the harmonic fields of the two parts of the texture, with only one pitch-class

(bIt) in common. (The position of the bi in the texture at this point - in octaves in

the "middle register" - combined with its metrical stress - it appears on the downbeats

of bars 43-47 - emphasizes its role as a central pivot.)

The matter of diatonic orientation is considerably complicated in the succeeding

bars (46-51). The reinterpretation of fi' as el and the exchange of & and

between the hands in bar 46 suggests the field i but the f# in the extreme bass which

continues the linear descent in that register (bt - a - g - ff) obviously conflicts with

this. Pivotal shifting between distantly related fields accelerates during the next four

bars so that by bars 48-9 only three to four consecutive notes remain within a single

field and by the end of bar 49 sometimes only two consecutive notes do so. Clearly,

this represents the point beyond which the concept of harmonic orientation according

to diatonic fields ceases to be appropriate. (See Chap.6, pp.157-9 and Ex.6.13 for an

alternative analysis of this passage and the preceding one in terms of voice-leading.)

86
2. Chromatic complementation

The "atonality" of Carter's later music is a result not only of the use of pc sets

and intervals divorced from their tonal functions, but also of the constant circulation of

all twelve pcs of the chromatic scale. Repetition, rhythmic stress, extremes of

register and other factors may give prominence to particular pitches, but these

nevertheless appear against a background of the total chromatic field, rather than as

the nodes of a network of tonal relationships.

The gradual accumulation of the 12-note aggregate as a means of organizing

pitch in an atonal work has been observed in pre-serial works of Schoenberg and

Webern. Similarly, in Caner's works, before the development of a method of

ordering the twelve notes into all-interval "tonics", the principle of chromatic

complementation is occasionally used systematically.

Example 4.6 from the first movement of the Piano Sonata (bb.83-5), illustrates

this principle at work. The analysis shows that, despite the obvious centring on c, the

full chromatic complement is used. The upper line shows the gradual process of

aggregation in the main melodic line. This does not achieve chromatic "completion"

until the penultimate note of the "answering phrase", g; thus, perhaps, the melodic

closure is reinforced by the notion of chromatic completion. The lower line

represents a similar process, taking in the total pitch-content of melody and subsidiary

"hannonic support". This achieves chromatic completion with the a concluding the

first melodic phrase; thus the bi-partite phrase-structure is underlined by the handling

of the total chromatic field. (See Ex.6.9(a) for an alternative analytical approach to

this passage, examining set structure and symmetry.)

87
 This technique of withholding one pitch-class for rhetorical effect or structural

purpose is given further illustration in Example 4.7, which shows the opening six bars

of the Cello Sonata. Here, chromatic aggregation is not complete until the cello's

first note, thus highlighting its entry. Further, more sophisticated, examples of this

technique can be found in the first movement of the Quartet Sonata (see Exx.5. 16 and

10.8).

88
 CHAFItR 5: CHORDS AND/OR SETS

The previous chapter has dealt with one aspect of the large-scale harmonic

organization of Carter's music. The present chapter will examine the harmonic fabric

in greater detail and will attempt to trace the changes in the types of and relations

between harmonic building-blocks used by Carter during the period 1945-55. It will

be useful to refer to the composer's own pronouncements on the subject of harmony

and to compare these with attempts by other writers to interpret his ideas and apply

them to analysis of the music. Much of this chapter will therefore consist of an

exploration of the relationship between pitch-class set theory and Carter's

compositional methods.

The title of this chapter is intended to express something of the semantic

difficulty involved in reconciling the terminology of the composer with that of his

analysts. The relatively conservative mode of expression Carter adopts when dealing

with technical aspects of his music contrasts with the complex vocabulary developed

by theorists in order to describe music of such extreme sophistication adequately.

The situation is succinctly encapsulated in the following exchange [Bernard 1990(a)]:

JB Even though "chord" is usually specific in its meaning...

EC . . .of simultaneity...
JR . ..whereas you clearly mean by the word "chord"...
EC Yeah, I know, I know...
JB . . . the linear presentation of notes as well.
EC I'm aware of that. "Set" is a word for tennis, while arpeggiation is a
more familiar term in my vocabulary

The problem is that the word "chord" is no longer entirely apt as a description

of the basic harmonic building-block of much twentieth-century music. Schoenberg's

dictum that "the two-or-more-dimensional space in which musical ideas are presented

89
 is a unit" ["Composition with Twelve Tones (1)" (1941), in Schoenberg 1975: 2201,

forms an unspoken background to the conversation quoted above. Harmony and

melody are, according to this viewpoint, to be regarded as indivisible; a "motive" may

be projected either vertically or horizontally and thus the distinction between "motive"

and "chord" is dissolved. This basic premise is one of the fundamentals of pitch-class

set theory.

In Carter's case, however, his attachment to the word "chord" is not merely the

result of an "old-fashioned" approach to terminology; it does in fact strongly reflect his

compositional practice, which assigns primacy to the vertical dimension.

Combinations of pitches are categorized and tried out in various successions and

spacings as chords in the so-called "Harmony Book", which Carter began compiling in

the 1 960s.

The particular registration and spacing of these chords is as important a part of

their identity as their intervallic content 1 . Most recent analysts of Carter's later music

have been sensitive to this fact and have attempted to adapt set theory by incorporating

some means of considering specific pitch rather than pitch-class alone, which may

appear too abstract a concept to be genuinely useful. Thus, for example, Jonathan

Bernard:

In the case of Carter's music [... it seems quite possible that individual modes
of presentation are inseparable from pitch-class set equivalence. In other
words, it becomes necessary to define analysis primarily in spatial terms, in
which the identity of a pitch collection is a function of its actual intervallic

"Spacing in chords is a very important matter in my work, since I distinguish

between an interval and its inversion and often use them for very different musical effects.
Even in chords of three notes, spacings become as differentiating as mirror inversions -
maybe more. After all, a four-note chord has 24 different spacings of intervals if none of
the intervals exceeds an octave - 48 if you count those of its mirror inversion separately,
of course." [Carter, quoted in Bernard 1990(a): 201]

90
 configuration. There is plenty of evidence, both in his writings and in his
sketches, that Carter composes with spatial criteria in mind. [.. .1 It is clear that
the literal size of the interval is crucial, since inversional equivalence is also
explicitly excluded. [Bernard 1983: 7]

Harvey, too, argues for a "more generous definition of interval class [...] as an

absolute measure of "distance" in musical "space", defined in size but not necessarily

in terms of precise pitch identity" [Harvey 1989: 209-10].

Taking due note of these observations, it is nevertheless undeniable that

Carter's classification of three- to six-note chords is remarkably similar to Forte's list

of prime forms of pc sets in The Structure of Atonal Music [Forte 1973: l7918112.

As David Schiff states, Carter and Forte "agree on basic definitions and on the number

of chords" [Schiff 1983: 324]. Furthermore, Carter's interest in the relationships

between chords demonstrates a fmiliarity with the concepts of inclusion and

complementation:

This has become for me a whole new field of thought, involving such questions
as what two-note groups are contained in three-note groups and so forth. This
way of working allows me to make all sorts of harmonic identities by adding
and subtracting notes and so produce a whole gamut of harmonic colours all
related to each other. [Edwards: 1081

In short, in Carter's later works, the selection of abstractly defined pitch-class sets is

an indispensable step in the compositional process, after which the physical

characteristics of spacing and instrumentation and association with rhythmic elements

may be assigned. However, the question of whether matters of pitch are actually

determined before those of rhythm and texture is difficult to establish; the evidence of

Carter's writings often suggests the contrary to be the case:

2
Since the works to be examined in this study all pre-date the compilation of Carter's
larmony Book and since Forte's numbering system is more generally known, the latter
'ill be adopted throughout this study. See Appendix 1 for a cross-referencing of Carter's
umbering of chords with Forte's list of prime forms.

91
 [...]the first stage of conception was the general dramatic plan [... J The next
stage was that of working out this over-all plan concretely: and determining the
specific rhythmic detail-patterns of the basic material, and how they could be
variously combined and interrelated, and how they could acquire pitch.
[Edwards: 104-51

The writings of Schiff, Bernard and the composer himself [Schiff 1983: 61-9,

Bernard 1983: 5-11], promote the view that Carter's harmonic practice has progressed

gradually towards a "deliberately global method" [Bernard 1983: 61; one in which

maximum harmonic variety is achieved, but not at the expense of coherence, since all

the harmonic material employed is related to a governing chord or chords. The range

of possible subsets and other relatives of the governing chords is divided into

"families", whose members are related through similarities of interval content.

Norms of spacing and registration are established as reference points, around which

patterns of departure and return may evolve; these harmonic norms are perceptually

reinforced through their association with distinctive features from other parameters,

such as rhythm and instrumentation. The size of the governing chords, and of the

members of the intervallic or chordal vocabularies formed from them, grew during the

middle part of Carter's career; the first two String Quartets (1951 and 1959) and the

Double Concerto (1961) use the all-interval tetrachords - the latter two works assigning

intervallic vocabularies to different instruments or groups -, the Piano Concerto

(1965) uses two twelve-note governing chords and divides the twelve possible three-

note chords between soloist and orchestra, while the Concerto for Orchestra (1969)

deals in repertoires of three-, four-, five- and seven-note chords. In more recent

large-scale works, twelve-note all-interval chords have been Carter's favoured

harmonic resource, but the principles evolved in the 1960s have continued to form the

basis of his harmonic practice.

92
 The period under consideration here is that leading up to Carter's discovery of

the potential of the all-interval tetrachords in the early 1950s. During this period, we

may observe the composer exploring the possibilities of creating distinctive chordal

vocabularies through the varied combination of intervals. This process of exploration

extends throughout the period, encompassing the largely diatonic language of the Piano

Sonata and the highly chromatic language of the Quartet Sonata alike. Certain kinds

of chord or set, it will become apparent, proved more useful to Carter than others and

were used extensively in more than one work.

1. A note on pitch-class set genera

As previously suggested [above, p.76], talk of "families" of chords suggests a

possible link between Carter's harmonic practice and Allen Forte's concept of pc set

"genera" [Forte 19881. These "genera" are in fact families of sets derived from

common "progenitors"; they are formed from groups of sets of cardinality 4 to 9

which have a particular trichord or pair of trichords as common subsets. The twelve

harmonic genera which result from this division are not equal in size or mutually

exclusive in terms of membership, nor is any twentieth-century work or section of a

work likely to contain only members of one genus. Rather, through a process of

segmentation of the musical surface and a statistical analysis of the generic affiliations

of the resulting sets, a genus or small group of genera may be identified as

predominant. What the theory intends to show is that the distinctive harmonic

languages of particular works, composers or groups of composers can be categorized

systematically in terms of their relation to the complete range of possibilities inherent

93
in the twelve-note universe. The changes in harmonic language during the late

nineteenth and early twentieth centuries can thus be ascribed to a gradual working

through the twelve genera from the familiar "diatonic" and "tonal" genera (these are

Forte's informal descriptive titles), via various forms of chromaticism represented in

the "diminished", "whole-tone" and "augmented" genera, among others, to the

"atonal". Certain pc sets, being members of more than one harmonic genus, are

capable of reinterpretation in different contexts and may thus be seen as agents of

transition. For example, 4-27, the "dominantlhalf-diminished seventh",

which is a hallmark of late 19th-century experimental music, connects Genus 2,

the whole-tone genus, with Genus 3 the "diminished" genus, and, at the far end
of the spectrum, joins the traditional "dia-tonal" genus, Genus 12. In fact,
from a historical vantage,[. . .this corresponds] to the transplantation of the
familiar half-diminished seventh (in particular) from its native diatonic clime to
the exotic habitat of the diminished and whole-tone genera. [Forte 1988: 2041

Carter's harmonic practice, which seeks deliberately to divide the total

chromatic universe into separate but related spheres, would seem to provide suitable

musical material upon which to test the theory of genera, and indeed, Forte includes a

study of the first movement of the Concerto for Orchestra [: 249-2521 in an historical

survey extending from Chopin to Stockhausen. Forte's finding that "the main generic

components of the harmonic species of the movement are Genus 4 and Genus 8 0 [:

2491 agrees closely with the composer's published hand-written sketch of the work's

basic constituents [Carter 1973: 54-55]. Genus 4 consists of sets containing trichord

3-12, while Genus 8 consists of sets containing both 3-3 and 3-4. These three

trichords are precisely those listed by Carter as "characteristic" for the movement.

Furthermore, Forte's table of prominent sets and their principal generic affiliations

lists only those trichords, tetrachords and pentachords which appear on Carter's chart

94
 for the movement. This suggests that Forte has sought only to analyze the

characteristics of the pre-compositional materials rather than those of the movement

itself. As Forte himself states, the fact that "we have the composer's own catalogue

of harmonic materials [...1 by no means confirms that the issue of detailed

segmentation is totally solved" [:2491. Since lack of space prevents Forte from

providing any musical examples in support of his "table of results", this issue is not

addressed, let alone resolved.

The observations made with regard to the movement's harmonic materials are

intended to demonstrate that the generic structure is "more 'unified' than might appear

at first" [:249], since of the five genera represented, three are bound into a

"Supragenus" while another has a "uniform relation" to each of these three. This

uniformity is emphasized by Forte's practice of reducing the multiplicity of relations

between sets and genera to a statistically significant few. According to Forte, the

material of this movement demonstrates "a very rich harmonic panorama" which has

to be "brought into focus" through the operation of various rules of genus formation in

order to be interpreted effectively [: 2501 . Forte makes no comment on whether

"uniformity" or "richness" is the more important or characteristic quality in this music,

but it may be presumed that a balance between the two is desirable and is a

prerequisite for satisfying musical creation.

It is not my intention to portray the development of Carter's harmonic language

as a steady progress through the pc set genera, as this might suggest an inappropriately

teleological view. Furthermore, the trichordal basis of Forte's system of genera does

not always fit comfortably with Carter's earlier harmonic practice, which often takes

95
 the combination of dyads as its starting point 3 . This observation does not invalidate

the applicability of Forte's theory, but it suggests an important change of emphasis; the

network of relationships created through dyadic combinations is quite different from

that created through trichordal combinations.

Not all the interrelationships among sets described in pc set theory, including

the system of genera, can be shown to be significant in the compositions under

analysis here. However, recourse will occasionally be made to this theoretical

apparatus in order to clarify certain relationships which are exploited in Carter's

compositions.

2. "Key-chords"

In response to enquiries about the basis of his harmonic language, Carter has

often cited his use of particular chords or sets as governing sonorities. The latter,

according to the composer, play a central role in two ways;

(i) they may be established as norms through repetition and other means of

contextual emphasis. Thus, a loosely hierarchic system may be created, in

which the "key-chord" becomes a quasi-tonic, upon which other harmonies are

dependent;

(ii) they may act as a source of harmonic materials. Their constituent intervals

and subsets may be extracted and recombined to form a distinctive harmonic

repertoire or vocabulary.

"[. ..1in all my pieces written before the Piano Concerto the pitch-behaviour of the
sub-voices constituting the separate contributing characters was almost entirely linear or
built up of two-note intervals [...]" [Edwards: 107]

96
 The second of these two functions seems to have been the one which came to play the

defining role in the constitution of such "key-chords". Often, in earlier works, the

distinctive sonority of a chord may have recommended itself to the composer as a

suitable harmonic focus or reference point; in the transitional music, however, the

richness and cohesiveness of the intervallic relationships to be drawn from the "key-

chord" are paramount. The discovery of the potential of the "all-interval" tetrachords

4-Z15 [0,1,4,6] and 4-Z29 [0,1,3,71 was clearly a crucial step in the development of

this method of harmonic organization, as is revealed in the following passages:

In all my works from the Cello Sonata up through the Double Concerto I used
specific chords mainly as unifying factors in the musical rhetoric - that is, as
frequently recurring central sounds from which the different pitch material of
the pieces was derived. For example, my First String Quartet is based on an
"all-interval" four-note chord, which is used constantly, both vertically and
occasionally as a motive to join all the intervals of the work into a
characteristic sound whose presence is felt "through" all the very different
kinds of linear intervallic writing. This chord functions as a harmonic "frame"
for the work [...] which makes all the events and details of a piece of music
feel as if they belong together and constitute a convincing and unified musical
continuity. [Edwards: 106-7]

Q: Does your music ha e any harmonic plan?

A: A chord, a vertical group of pitches either simultaneously sounded or
arpeggiated, like a motif, is a combination to be more or less clearly
remembered and related to previous and future chords heard in the same work.
Whether a composer is conscious of it or not, a field of operation with its
principles of motion and of interaction is stated or suggested at the beginning of
any work [...1 In my First String Quartet, I did use a "key" four-note chord,
one of the two four-note groups, that joins all the two-note intervals into pairs,
thus allowing for the total range of interval qualities that can still be referred
back to a basic chord-sound. This chord is not used at every moment in the
work but occurs frequently enough, especially in important places, to function,
I hope, as a formative factor. It is presented in various kinds of part-writing
and interval combination, the number of notes is increased and diminished in it
[...] ["Shop Talk by an American Composer", MQ 46/u (1960), also CEL
2 18-91

Although Carter himself cites the Cello Sonata as the first to employ such "key-

chords" or "source-sets", David Schiff states that the idea began to evolve several

97
years earlier, but was applied with "increasing rigour" in the Piano Sonata and Cello

Sonata [Schiff 1983: 641.

An important note of warning should be sounded here. The "key-chord"

concept outlined by Carter in response to these questions, and developed by Schiff, by

no means provides the basis for a complete explanation of the composer's harmonic

practice, but is rather in the nature of a convenient "pat" answer, convincingly

technical but sufficiently simple to be understood by the student or dilettante analyst.

Those analysts who have undertaken a more detailed examination of Carter's work

than is offered in Schiffs book, have inevitably found the "key-chord" explanation to

be too simplistic or only a partial truth.

Several examples may be cited in support of this statement. Jonathan Bernard,

in a discussion of the first movement of the Cello Sonata, finds that Schiff s

description of its harmonic materials is inadequate:

[...] the first part of the opening Cello phrase can be interpreted as overlapping
instances of pc sets 5-11 [0,2,3,4,71, 4-14 [0,2,3,71 and 4-17 [0,3,4,7], with 4-
14 and 4-17 each expressed once as literal subsets of 5-11. [Bernard 1988:
172] [...] My analysis differs significantly from Schiff's. Both 4-14 and 4-17
are included in his tabulation of four-note subsets of a controlling six-note set
(6-Z43 [0,1,2,5,6,8]), but 5-11 apparently does not figure in his scheme, since
it is not a subset of 6-Z43. [:200, footnote 181

A doctoral dissertation by C.R.Kies [Kies 1984], which concludes with a study of the

first movement of the Cello Sonata, also disagrees with Schiff's interpretation, in that

it identifies 6-20 [0,1,4,5,8,9] as the governing sonority, rather than 6-Z43. David

Harvey's interpretation of the same work's harmonic materials, although heavily

dependent upon Schiff's, views the opening tetrachord, 4-7, as the governing sonority.

In fact, it will be seen that the techniques connected with "key-chords" underwent

considerable changes during the course of the Cello Sonata, so that this work may

98
indeed be regarded as the linchpin of the "transitional period".

The examples given above may appear to be mere quibbles that do not

challenge the fundamental idea of a "key-chord"; taken in conjunction, however, they

demonstrate the difficulties that often arise in attempting to reconcile the composer's

statements with a detailed examination of his actual practice. A more fundamental

criticism of Carter's "key-chord" explanation appears in Canton Gamer's review of

Flavved Words and Stubborn Sounds [Gamer 1973]. One symptom of the "lack of

specificity and analytical inadequacy" [: 1541 which he finds disappointing in the book,

is that "the harmonic unity which admittedly exists in Carter's music can often be

more adequately explained in other terms than his own"[: 154]. In particular, he

questions whether the composer's alleged use of the all-interval tetrachord 4-Z15 as a

"harmonic frame" in the First Quartet is really an accurate reflection of his harmonic

technique. Gamer's approach to an understanding of Carter's technique in this work

depends upon an appreciation of the richness and diversity of the intervallic content of

the sets used in the opening cello solo, rather than upon the unifying power of the all-

interval tetrachord. Segmenting the work's opening ten bars into nine four or five-

note sets, he notes that "the six unique pitch structures of size 2 (that is, the six

interval classes), the twelve unique pitch structures of size 3, and the twenty-nine of

size 4 are each contained either within one or another of the given sets or within the

unions of adjacent sets of which certain phrases of the solo are comprised", and draws

from this the conclusion that "every vertical sonority played by any one of the four

instruments anywhere in the entire quartet can be heard as an instance of one of the

pitch structures contained within the sets of the first 10 measures"[: 154]. This

observation is interesting in that it shows Carter developing a musical language which

99
embraces the maximum variety of intervallic combinations, but it does not explain how

be draws coherence from that variety. Gamer's argument, if taken to its logical

conclusion, would imply that the opening of the First Quartet could be related to any

other piece of music, since it contains every possible 2, 3 or 4-note set. Clearly,

some sets must be privileged above others if coherence is to be achieved, the all-

interval tetrachords being chief among these privileged sets because of their power to

integrate various possibilities of intervallic combination into a single sonority. Such

"key-chords" should therefore not be regarded as all-powerful generators, but as

occasional points of focus.

The following sections will survey the changing character of the basic harmonic

building blocks used by Carter and their relationship with the idea of a "key-chord"

over the transitional period.

3. Diatonic Sets

As previously described, the harmonic language of Carter's earlier music is

characterized by diatonicism, but tends to eschew simple triadic harmonies. The

hierarchic relationship between consonance and dissonance, which governs harmonic

relationships in tonal music, has, therefore, a severely attenuated role in Carter's

music. There are chords and pitches whose function is analogous to that of the tonic

and the tonic chord in tonal music, and there are similar "quasi-dominant" chords and

pitches, but the effect of both these types of harmony is often dependent to a large

degree upon their intervallic resemblance to tonal triads. Beyond recognizing these

quasi-tonics and dominants, it is extremely difficult to find a consistent relationship

100
between the kinds of harmony employed by Carter and the network of triads and scale-

steps of tonal music.

However, this is not to suggest that Carter's earlier music admits no hierarchic

distinction between harmonies akin to that between consonance and dissonance. Two

basic observations may be made with regard to Carter's harmonic practice. The first

is that harmonic density plays a similar role in Carter's music to that which it plays in

tonal music; three and four note chords tend to be the norm, with five and six note

chords reserved for moments of harmonic tension or complexity. This distinction,

although infonnal and flexible, operates throughout the earlier part of Carter's career.

The second observation is that the intervals of semitone and tritone (interval classes 1

and 6) are treated as more dissonant and less stable than the others, while interval

classes 3,4 and 5- those which make up the traditional triad - are treated as the most
stable. Interval class 2 occupies an intermediate position, frequently being accepted

as part of a stable harmonic entity.

These two observations may be partially explained and subsumed by

consideration of another, more penetrating distinction; that which may be drawn

between diatonic and non-diatonic harmonies. In Carter's earlier music it is the

former which act as the harmonic norms, while the latter play a role analogous to that

of dissonances in tonal music. "Diatonic chords", in this context, may simply be

taken to mean those which are formed from subsets of 7-35 [0,1,3,5,6,8, 101, the set

representing the major scale or complete diatonic field. The total number of sets of

cardinality 3 to 6 which fall into this category is 35, clearly far exceeding the number

of chord-type.s normally used in tonal music.

Among these 35, further subdivisions may be made, enabling us to draw

101
 distinctions between varying degrees of consonance and dissonance. It will be

recalled from the previous chapter that 7-35 may be created from a continuous

segment of the cycle of fifths, as may its subsets 6-32 10,2,4,5,7,91 and 5-35

[0,2,4,7,9]. These two sets may be used to create further "filters", separating sets

with strong diatonic implications (i.e. a high incidence of interval class 5 in their

interval vectors) from those with relatively weaker diatonic allegiance. The subsets of

6-32 form an inner core of 16 and those which are also subsets of 5-35, a smaller

nucleus of 7. Appendix 2 lays out these relationships in tabular form. Three sub-

categories of diatonic set are thus created: "Type A", which are subsets of 5-35;

"Type B", subsets of 6-32, but not of 5-35; and "Type C", subsets of 7-35, but not of

6-32. An examination of the intervallic content of the sets in each of these sub-

categories shows further internal consistencies. Type A sets contain only interval

classes 2, 3, 4 and 5, thus, semitonal or tritone relationships are not possible among

this type. Type B sets add interval class 1 to the intervallic repertoire, while type C

sets also include a single tritone relationship. It will also be noted that for type C sets

of cardinality 4 and greater, two semitonal relationships may exist, thus creating the

possibility of a greater degree of dissonance than is available among the type A and B

diatonic sets (see sets 4-8, 5-Z12, 5-20, 6-Z25 and 6-Z26).

An examination of some of Carter's earlier works will show that diatonic sets

of types A and B are treated as normal and stable (with precedence given to the

former, especially sets 3-7, 3-9, 4-23 and 4-26), while type C sets are often used to

create a degree of harmonic ambiguity and to act as a bridge to non-diatonic

harmonies.

102
(a) The Rose Family

Returning to the song The Rose Family, one may observe that every individual

vertical sonority can be found in the table of diatonic sets shown in Appendix 2 (see

Ex.5. 1). Indeed, the harmonic make-up of the song is almost exclusively drawn from

types A and B, suggesting that the harmonic language of this work is diatonic to an

unusually strong degree. (There are only two instances of type C sets, both occurring

only once and lasting for a single quaver, 4-Z29 b. 12, 5th quaver, 4-8 b.21, 4th

quaver.) Carter exercises a control over the types of chord used in this song which

matches the careful management of diatonic fields. During the first seven bars of the

song, since all the pitches used fall within the field (D/A./E4) (or set class 5-35,

with pitch membership [al' ,bt' ,c,dk ,et' }), all the harmonies used are of type A. In

fact, for the first four bars, the only harmony denser than a dyad is set 3-9 [0,2,71.

The latter, which maximizes interval class 5 and only contains one other interval class,

is strongly representative of the cycle of fifths and may therefore be regarded as the

"seed" of the harmonic language and process of the song, since these depend upon

motion through this cycle. The relatively low harmonic density and reliance on type

A diatonic sets is clearly a musical analogue of the "naive" mood of the poem's first

two lines. The opening figure in the bass, two rising fifths (at' -e -bt'), is an instance

of set 3-9 but also has an allusion to tonal function, as it suggests overlapping

arpeggiations of the tonic and dominant triads of A flat major.

As the song unfolds, Carter employs harmonies of progressively greater density

and complexity. The first vertical tetrachord to be heard (4-22) occurs in bar 5, third

quaver. The next such harmony, occurring on the third quaver of bar 8, is 4-14 [bk,

103
 c,dk ,fJ. This is the first type B set to be heard in the song, and its occurrence

during the phrase "But the theory now goes" mirrors the text's move away from the

naivety of the opening. From this point until the second quaver of bar 11, the total

pitch field is represented by 6-32 (Db lAb) and, therefore, a mixture of type A and B

sets may be used. However, Carter still maintains the relative purity of type A, only

employing one type B set, trichord 3-4, on the fifth quaver of bar 9. Nevertheless, a

greater variety of trichords are employed. This passage also sees a far greater

occurrence of tetrachordal harmony, as set 4-22 is heard three times during bars 9-10.

Bars 11-13 lead to the song's first climax through a rapid flatward motion

through diatonic fields. The climax itself, on the word "plum", is the fourth of a

succession of tetrachords (4-Z29, 4-22, 4-26, 4-14), the first such sequence in the

song. This progression is formed from the arpeggiation of an A major triad in the

piano right hand, with the uppermost part (e-a-c#) moving in symmetrical contrary

motion with the bass line in the left hand (d# - b-f#-dt). The final chord, 4-14, is

constructed so as to suggest a superimposition of tonic and dominant triads of D

major, thus linking this climactic moment with the song's opening.

Similar processes of thickening of the texture and movement from type A to

type B sets are present in the second part of the song. Here, five-note chords are

employed as the climactic sonorities and the relationship between the three type A/B

pentachords 5-35, 5-23 and 5-27 is highlighted. Type A harmonies alone are

employed from the second quaver of bar 14 to the third quaver of bar 19, with one

appearance of 4-22, in bar 17, being the sole harmony denser than a trichord. From

bar 19 onwards, as the diatonic orientation becomes more complex, so does the nature

of individual harmonies. Type B sets 3-4, 4-11 and 4-14 are introduced in bars 19-

104


20, while in bars 21-22 an almost unbroken series of tetrachords leads to the first

pentachord of the song, 5-27, on the first beat of bar 23. The succeeding bars in turn

lead to the climax of the whole song - a vertical statement of 5-23, supporting the

words "You, of course" (bb.25-6). From this point, the three pentachords mentioned

above form the entire harmonic material; after the chordal statement of 5-23, the piano

part presents 5-35 (bb.27-30) and 5-27 (bb.30-31) in a similar 1shion. The

remainder of the song is built entirely from the pitch field represented by 5-35 as it

appeared at the opening.

The presentation of these three chordally stated pentachords is worth examining

in greater detail for the information it gives about Carter's style and technique. Each

of the chords is spaced in such a way as to emphasize its quasi-tonal characteristics;

they are constructed as "stacks" of perfect fifths and/or thirds, so that the lowest note

may be interpreted as the root of a triadic formation. However, their succession

relies less upon the concept of harmonic "progression", found in tonal music, than

upon properties of permutation and invariance. Between them, the three pentachords

exhaust the range of possible different 5-note subsets of 6-32. The similarity

relationships between the three sets are displayed in the table below:

Table 5.1 Similarity relations between pentachords appearing in The Rose Family

5-23 [132130]

5-27 [122230] R2R.,, 5-27 [1222301

5-35 [032140] R2R

105
 The table shows that the R1., relationship4 holds between all three sets. In the passage

in question, this relationship is strongly represented in that common four-note subsets

are held at the same pitch-class level between each pair of pentachords; as 5-23 gives

way to 5-35 in bars 26-27, set 4-23 (bt ,c,e1 ,f) is present in both chords with d in

the first exchanged for a in the second; in bar 30, the change from 5-35 to 5-27 is

executed by exchanging b for di' while set 4-26 (e ,f,at' ,c) remains invariant. The

exchange of pitch-classes which takes place during this sequence of chords effects a

gradual change in diatonic orientation from (AI'/Eb/Bb) in the first chord, to (Db/Ab

/E) in the second, and (Gb/Db/Ab) in the third. The pitch classes which link each

pair of chords (at', b, c, el' and 1) themselves make up the same form of 5-35 as

that which forms the central chord of this sequence of three and which opens and

closes the song. This form of 5-35 may thus be regarded as the "key-chord" of the

song, since it is both the source of most of its harmonies and the origin and end-point

of its harmonic trajectory.

(b) Dust of Snow

The song Dust of Snow (1942), another of Carter's Frost settings, demonstrates

a similar carefully controlled use of diatonicism (see Ex.5.2). The texture is mostly

spare, with only two independent parts, but changes to a four or five note chordal

texture in the central contrasting section (bb.23-32). As in The Rose Family, type A

and B sets predominate; however, while The Rose Family employs a gradual motion

See Forte 1973: 46-60 for an explanation of similarity relations between pc sets.

106
between diatonic fields, Dust of Snow relies on a more abrupt contrast between

harmonic blocks.

The opening section (bb. 1-18), which set the poem's first four lines, stays

exclusively in the field E. The sparing use of the pitch a (which is not introduced

until bar 6, and then again only in bars 14-16) leaves a slight ambiguity so that a field

of (E(B) might be inferred at the opening. Most of the harmonic material of this

section is therefore of types A and B. The total pitch collection of the vocal line

makes up 5-35, thus forming a central core of pitches around which the piano may

elaborate. This relationship parallels the rhythmic one, in which the voice has a

cantus firmus -like line in minims and semibreves, around which the piano has more

rapid figuration in quavers.

This song makes more frequent use of type B sets, and its harmonic language is

therefore rather more dissonant than that of The Rose Family. The first harmony

heard is the type B tetrachord 4-14. This "key-chord" is presented in such a way as

to emphasize its triadic components; the lower perfect fifth c-b and the upper minor

third d#-f# create the effect of a superimposition of triads of E and B major. This

effect is enhanced on the reappearance of the piano's opening gesture in bar 11, where

the lower perfect fifth has an added g#, forming set 5-27. The latter is also the final

chord of the song.

The contrasting central section primarily inhabits the diatonic field (B!' IF), a

tritone away from that of the opening and closing sections. This contrast extends to

the nature of the harmonies used; the "change of mood" of the text is reflected in the

turn towards the "softer" dissonance of type A sets. Set 4-26 (f,g,b ,d) is sustained

throughout in the piano part, while the bass oscillates between f and c in a manner

107
 which mimics the tonic-dominant relationship, producing an alternation of sets 4-26

and 5-35.

The central section is approached via a brief transition (bb.19-22), during

which harmonic instability is created through rapid fluctuations of diatonic field and

the introduction of type C sets. The interval of a perfect fifth plays a pivotal role in

this transition. This interval is prominent in the opening sonority 4-14 and is

frequently sustained in the remainder of the opening section (see bb.6, 14 and 17-18).

The transitional bars make use of pairs of perfect fifths to produce symmetrical

tetrachords, namely 4-26, 4-20 and 4-8. The first of these is significant in this

context since it may be formed either from a pair of perfect fifths or from a pair of

minor thirds, the two intervals which are most prominent in the key-chord 4-14 as it

appears at the opening of the songS A comparison of the pitch-class and interval-class

content of the two sets reveals that they are in the close relation R2R.,:

4-14 [0,2,3,7] [1111201

4-26 [0,3,5,8] [012120]

Since 4-26 is also a subset of the final chord 5-27, it may be regarded as a subsidiary

key-chord for the song.

In bar 19, the pitch b is a pivot between two successive forms of 4-26 (b d f# a

- c# e g# b), which together form 7-35, with the introduction of di effecting a move

to the diatonic field A. Once again, the perfect fifth is prominent at the top of the

texture (d-a). This is contradicted in the following bar by the reintroduction of d#

and a return to the motive of the opening bar (via a "passing TM form of 4-20).

However, the dyad c#-b of the latter is replaced in bars 20 and 21 by set 4-8, which is

presented as two perfect fifths, g#-d# and a-e. In bar 22, the dyad g-d acts as a pivot

108
 between two forms of set 4-26; while the upper perfect fifth is sustained in the right

hand, the pitches of the perfect fourth f-br' in the left hand move outwards

symmetrically to the perfect fifth e-bI. This progression encapsulates the overall

harmonic contrast in the song between the fields (E(B) and (BI'/F) and is repeated in

bars 27-28.

(c) Pastomi for viola, clarinet or cor anglais and piano

The Pastoral is a much more extended and ambitious work than either of the

brief songs described previously. Its harmonic language is based on similar premises,

but is developed to a higher degree of complexity and sophistication, as an analysis of

the opening paragraph (bb.1-45) will show (see Ex.5.3). Carter's use of sets as

motives rather than purely as chords will begin to be examined, as will the

contrapuntal interaction of sets presented linearly.

The prevailing field during these bars is A and the focal pitch is a, but, this

field is not established immediately. By withholding the pitch d until bar 11, Carter

restricts the harmonic field of the opening to (NE), or set 6-32. Type A and B sets

are therefore available. The type B set 4-10 forms the opening sonority and is

presented as two minor thirds (f#-a and g#-b). These two thirds are joined by another

(a-c#) in bars 2-5, forming a total pitch combination of set 5-23. The f#-a dyad is

sustained at the top of the texture in bars 1-6 and is prominent until bar 10, while the

g#-b dyad is inverted to form a sixth in bar 2 and is then presented melodically in bar

3, forming the beginning of the work's first melodic motive. The introduction of the

pitch e in bar 3 widens the total pitch field to 6-32. The viola's first phrase (bars 8-

109
12) forms 5-35, thus acting as a purer diatonic centre within the wider field suggested

by the piano.

Within this field, there are certain motivic consistencies. Set 4-23 appears in

two prominent forms. Its appearances are (a) as a melodic line of 4 quavers (b-c#-e-

f#), which is gradually isolated from the g# which first precedes it in bar 3 (see bb.3,

5, 6-7, 7-8, 9, 10-11 etc.) and (b) as a pair of dyads, the perfect fifth c-b followed by

the minor third f#-a (bb.6, 8-9, 10). The e and b are heard as neighbour notes to the

f# and a. These two forms present the same set in different aspects; (a) as two

major seconds separated by a minor third and (b) either as a perfect fifth and a minor

third presented harmonically, or as two major seconds presented melodically in

contrary motion. This kind of exploitation of the different intervallic patterns inherent

in a set is a hallmark of Carter's technique. In bars 15-16, the melodic form of 4-23

is transposed and the second interval expanded, so that the two major seconds are

separated by a perfect fourth (d-e-a-b) (piano right hand, imitated by viola).

The next harmonic turning point occurs in bar 11, where the pitch d is

introduced (piano, left hand). This not only widens the diatonic field to A, but

introduces a type C harmony for the first time. Set 4-13 is sounded in the piano,

comprising a perfect fifth d-a and a minor third g#-b. These two dyads are related to

the pair forming 4-23 as it appears in the right hand of the piano in bars 9-10 (perfect

fifth c-b and minor third f#-a). In conjunction with the f# in the viola, this forms set

5-25. The> marking draws attention to this more complex sonority, which signals

the beginning of a move away from the straightforward diatonicism of the opening.

Bar 14 marks the beginning of a long transitional passage, leading eventually to

the establishment of a new tempo, quaver = 160, and diatonic field, C#, at bar 45.

110
During this transition new pitches and sets are introduced which disrupt the diatonic

field of A and break away from the diatonic vocabulary laid out in Appendix 2.

These new sets are the product of a divergence between the diatonic fields of two or

more independent contrapuntal lines. Carter thus sets up a polarity between diatonic

norms and chromatic disturbances which parallels the traditional distinction between

consonance and dissonance.

In bar 16, the piano's b# and dl suggest in themselves a move to C/but are

irreconcilable with the viola line, which still maintains A. In the following bar, bli in

the piano contradicts the field of CI', and, in conjunction with a#, suggests the field B.

The clash of b# against c# in the piano and the dLj of the viola (b. 16, second quaver)

creates a momentary instance of 3-1 [0,1,21, the chromatic trichord, which is, of

course, completely alien to the diatonic language, as is set 3-3, which immediately

follows it [b#,c#,e], although both sets arise quite logically from voice-leading

motions. Another form of 3-1 appears in the following bar, where the al-b of the

piano clashes with the at of the viola.

The climax of the piano's phrase is a 6-note type C chord in bar 18, 6-Z25,

which suggests a diatonic field of D, the latter being maintained (with the exception of

bars 20-21) until bar 25. This chord, despite its dissonance, is spaced in such a way

that perfect fifths and minor thirds are prominent, thus affirming its link with the

simpler diatonic harmonies such as 4-23 and 4-13 of the earlier part of the work. The

articulation of the chord in bar 18 recalls the very opening. The lower minor third

c#-e brings an association with the earlier f#-a and g#-b, while the perfect fifth g-d at

the top of the chord is sustained, thus forming an association with the sustained dyad

f#-a of the opening chord.

111
 Later in this transitional passage, there is further conflict between contrapuntal

lines which, although diatonic in themselves, produce non-diatonic sets in combination.

Interval classes 2, 3 and 5 continue to play an important role and set classes 3-7 and 3-

9 are particularly prominent in this process.

In bars 30-35, the harmonic language reaches the maximum of dissonance and

complexity thus far in the work. While the viola line in bars 30-35 maintains a

diatonic field of D, the piano begins to diverge from this with rapid sharpward shifts.

In the right hand of the piano in bar 33, with the sudden intrusion of a#, e# and b#, a

form of set 5-23 arises, suggesting the field C#, while the viola and piano left hand

maintain D. Several forms of set 3-7 are presented here; the piano has d-e-g (left

hand) in bar 31 and e-c#-f# (left hand) and d#-b#-e# (right hand, lower line) in bar 33.

The juxtaposition of the diatonic fields CMind D produces a series of verticals (4-13,

4-11, 4-6, 4-8, 4-19) which shows a progression away from the simpler diatonic

sonorities towards more complex type C and non-diatonic dissonances. The same can

be seen in bar 34, where the appearance of 4-9 on the third quaver as two perfect

fourths separated by a tritone (b-eIe#-a/^), is emblematic of the conjunction of two

semitonally related diatonic fields. This is further exemplified by the juxtaposition of

two semitonally related forms of 3-9 in the right hand of the piano in bars 34-35 (d#-

e#-a#/ d-e-a).

In bars 36-42, the diatonic field oscillates between CI and FIB before settling

on the former in bars 44-45. In bars 36-44, the interval of a perfect fifth plays a

focal role; in particular, at the top of the texture, the fifths g#-d# and e-b alternate in

chords 4-14 and 4-Z29 (these two sets being in the close relation R2R) each of which

contains a strong triadic component (3-11) with an "added" note between the third and

112
fifth of the chord, producing a strong allusion to the triads of C# and E. At the end

of bar 42, both of these fifths appear together as a form of set class 4-20.

4. Non-diatonic harmonic principles.

While diatothcism is the norm for most of Carter's works of the early 1940s,

many passages may be found which demonstrate other means of harmonic

organization. During the earlier part of the period, the sonorities used often have a

resemblance to conventional triadic patterns, and may appear to be alterations of them.

At a further stage of removal from conventional harmony lie those sets which may be

formed from the recombination of the "triadic" intervals - the major and minor third

and the perfect fourth/fifth (interval classes 3,4 and 5) - in unconventional patterns.

This process of intervallic combination leads, in the later works of the period, to the

establishment of repertoires or families of sets related through similarities of interval

and pc content. Among these families of sets, symmetrical pitch structures are often

privileged, since they offer an alternative to conventional means of achieving tonal

focus. An important effect of considering harmony as the product of the combination

of intervals is that "motives" begin to replace "chords" as the basic building units of

harmony. It is important to note, however, that there are some works and passages

where the harmonic organization is apparently resistant to schematic analysis, usually

because the primary organizational function has been assigned to a parameter other

than pitch. (See, for example, the discussion of the first of the Eight Etudes and a

Fantasy, p.123 and Ex.5.12.)

113
(a) Triadlic alteration and superimposition.

Mention has already been made of Carter's use of sets which resemble tonal

triads altered by the addition of "extra" pitches, or "polychords" formed from the

superimposition of two triads. In general, such chords are reserved for moments of

important harmonic focus, such as beginnings, cadences and climaxes, and are often

used as "key-chords" in the earlier works because of their association with the

traditional function of tonal triads. Altered and superimposed triads as a category

cross the boundary between diatonic and non-diatonic harmony, on one hand evoking

familiar tonal phenomena, on the other, challenging conventional tonal hearing. Such

examples of "bitonality" and "wrong-note harmony" link Carter strongly with other

neoclassical composers and are among the features of his music which gradually

disappeared as he began to find a more individual style.

The simplest such chords are those which may be formed through the

superimposition of elements from tonic and dominant triads. These chords are, of

course, members of the family of diatonic sets laid out in Appendix 2, but they are

discussed here because of their link with non-diatonic sets. Several examples may be

found. The use of 4-14 and 5-27, resembling the superimposition of triads of E major

and B major at the opening and close of Dust of Snow, has already been remarked.

Similar use of 5-27 and its subsets is made in the Piano Sonata (see Ex.5.4); 5-27,

formed as a superimposition of triads of B and F# majors, is the closing sonority of

the second movement; this chord and its subset 4-14 are also prominent near the

beginning of the first movement (4-14 appearing in bb.20 and 25, 5-27 in b.30).

Such fusions of tonic and dominant tend to function in a manner similar to a

114
conventional tonic by confirming the diatonic field of a passage, but nevertheless

contain a degree of tension and ambiguity. After the Piano Sonata, Carter's works

contain few of these types of sonority, which suggest tonal models so strongly.

More complex superimpositions involve elements from triads which cannot be

reconciled within a single diatonic field. These were frequently used by Carter in the

early 1940s as a means of encapsulating tonal conifict or ambiguity. One of the most

dramatic examples is the appearance of set 5-21 [0,1,4,5,8] at the climax of the first

movement of the Piano Sonata (bb.246, 249-251, 262-263) (see Ex.5.5). This set is

formed from the superimposition of triads of G1 major and Bt' major and expresses

the pivotal role of the pitch b in the work. It conflates a triad of BL' in its own

right as a tonic, with a triad (enharmonically respelled) of the dominant of B major,

the tonic key of the Sonata, in which bIaI is merely the leading note. The pivotal

nature of the climactic chord is clearly revealed in bars 262-263, where first a B

triad and then a (fl' triad are allowed to emerge from the complete five-note chord.

The octave doublings of the bi' and f obviously reinforce the traditional feel of the

texture here.

A similar relationship, more gently expressed, occurs at the end of the song

Voyage (1943) (see Ex.5.6). The final chord here, 4-19 [0,1,4,8], is constructed

from the superimposition of a B major seventh and a triad of El' major. Once more,

certain pitches are pivotal; d#/e and a/t/b may appear as the third and seventh of

thechordonBorastherootand fifthoftheEk chordandthe senseoftonalcentreis

delicately balanced between the two keys.

Another special category of harmony employed by Carter requires separate

consideration. This may be informally described as "seventh chords" and includes

115


both diatonic and non-diatonic harmonies. The former include 4-26 [0,3,5,8] - the

"minor seventh" chord - and 4-20 [0,1,5,8] - the "major seventh" chord. These two

sets are both remarkable for having a symmetrical interval class structure which

includes a pair of ic5s, a fact which will become important in the discussion of

Carter's later music. The non-diatonic sets include 4-18 [0,1,4,7] - which may be

presented as a "diminished triad plus major seventh" - and 4-19 [0,1,4,8] - which may

be presented as a "minor triad plus major seventh" or "augmented triad plus major

seventh". The "dominantlhalf-diminished seventh" chord 4-27 [0,2,5,8] is

comparatively rarely used by Carter. An important relative of 4-18, 4-19 and 4-20,

which deserves consideration here is 4-17 [0,3,4,71 - which may be presented as a

triad with major and minor third. This set, like 4-20 and 4-26, has a symmetrical

interval structure, and like 4-18 and 4-19 may appear as a non-diatonic alteration of a

conventional triadic model. The relationships between these sets are summarized in

Table 5.2:

Table 5.2 Similarity relations between "quasi-seventh" chords

4-17 [0,3,4,7] [1022101

4-18 [0,1,4,7] [102111]
4-19 [0,1,4,8] [101310]
4-20 [0,1,5,8] [101220]
4-26 [0,3,5,81 (0121201

4-17
4-18 R2R 4-18
4- 19 R2R. ____________ 4-19
4-20 R1R1, ____________ R2R. 4-20
4-26 ______ ______ ______

116
 The variety in Carter's use of these chords is exemplified by examining

passages from works spanning the entire period and crossing the stylistic divide. The

song Warble for IJia c-time (1943) matches the unfettered joyousness of the text with a

harmonic language that draws its richness and vibrancy from its original use of

diatonic and triadic elements. The central slow section (bb.90-127) provides an

excellent example of this (see Ex.5.7); the majority of sonorities are formed from the

combination of triadic components, often moving in parallel. Sets 4-20 and 4-26 - the

"major" and "minor seventh" chords - are particularly prominent. Stepwise melodic

motion predominates, with motion in parallel perfect fifths and fourths, thirds and

complete triads pervading the texture.

A rather more ironic use of such triadic sonorities can be found in the Piano

Sonata's first movement (see Ex.5.8). Between bars 109 and 116, the texture

contains two main elements, a sustained triad held in the middle range of the piano,

and fragmentary melodic gestures played in octaves. The first of these elements

parodies the traditional harmonic device of gradually changing from one triad to

another by chromatically altering one note at a time. The pitch b is pivotal in the

move between triads of G major and B major, however, the two intervening steps are

not tonal triads but are both examples of the non-diatonic atonal set 4-19.

A similar passage from the first movement of the Quartet Sonata (bb.9-16)

shows this technique used more extensively and in a harmonic context far removed

from the extended diatonicism of the previous two examples (see Ex.5.9). It will be

noted that sets 4-17, 4-18, 4-19 and 4-20 are prominent among the sustained

sonorities. (A more thorough discussion of this passage will be undertaken in Chapter

10.)

117
(b) Combination of intervals.

The quasi-triadic sonorities described above are really a special case within the

more general principle of Carter's harmonic practice, which is the creation of larger

hannonic units from the combination of individual intervals. The simplest and, at this

stage, the most important example of this technique is the creation of tetrachords from

pairs of dyads. An examination of the various charts of harmonic materials drawn by

David Schiff from the composer's sketches to illustrate his discussions of the works

shows the extent of this preoccupation, which culminated in the discovery of the all-

interval tetrachords. (See in particular, the charts for the Cello Sonata and the

Quartet Sonata, [Schiff 1983: 137 and 1651.) Appendix 3 takes these charts as a

starting point for a complete classification of tetrachords according to their dyadic

content and will be an important point of reference for the following discussion. The

appendix is set out in two parts; the first takes each possible combination of the six

interval-classes and lists all the tetrachords which may be formed therefrom, together

with their interval vectors and a diagrammatic representation of their interval-class and

pitch-class structure and content. This part of the appendix begins with all the

symmetrical tetrachords, which is to say, all those which may be formed from pairs of

identical interval classes. The second part of the appendix summarizes the first part

by listing each tetrachord and showing the two or three different ways in which it may

be constructed from pairs of dyads. The first part of the appendix shows that there

are twenty different "families" of tetrachords, varying from two to five in

membership. The second part of the appendix shows that the membership of these

families overlaps to a considerable extent, since each tetrachord belongs to either two

118
or three families. Without going into too much theoretical exegesis, a few important

observations may be made. The first is that the two all-interval tetrachords 4-Z15 and

4-Z29 are obviously distinctive in that they may be formed in three different ways,

each of which involves a pair of unique interval classes; less obvious is the fact that

these two tetrachords are the only ones which may be formed by pairing interval

classes 3 and 6. Other observations to be made regarding these two sets will lead to

an understanding of the underlying symmetry of the network. 4-Z15 pairs interval

class 1 with ic2, and ic4 with ic5, whereas 4-Z29 pairs id with ic4, and ic2 with ic5.

Interval classes 1 and 5 thus stand in a reciprocal relation, as do ics2 and 4, and ics3

and 6, all of which can be seen throughout the network of tetrachords. For example,

there are exactly the same number of sets (4) in the family formed from id +ic2 as

there are in that formed from ic2+ic5; a similar balance exists between the numbers in

the families id +ic3 and ic3+ic5 (5), id +ic4 and ic4+ic5 (4), and id +ic6 and

ic5+ic6 (2).

This symmetry extends to the relationships between the individual members of

these families; if, for example, the members of the families id +ic3 and ic3+ic5 are

compared, it will be observed that the interval vectors for one family may be obtained

by exchanging the id and ic5 entries in the interval vectors of the other group (a

process identical to the so-called "MS mapping" [Rahn 1980: 153-5]). Where these

are identical, obviously the same set will be represented in both families, otherwise

each set will be balanced by an R 1 relative (see Table 5.3):

119
Table 5.3 Comparison of interval vectors for tetrachords formed from id +ic3 and
ic3 + ic5

id +ic3 ic3+ic5

set interval vector set interval vector

4-1 1321000] 4-23 [0210301

4-4 [211110] 4-14 [111120]
4-13 [1120111 4-13 [112011]
4-18 [102111] 4-18 [102111]
4-20 [101220] 4-7 [2012101

The same relationship holds between each of the pairs of families which hold one dyad

in common and exchange id for ic5. Interestingly, it also holds between the groups

icl+icl and ic5+ic5.

There are three tetrachords which are multiply symmetrical, which is to say

that they may only be formed from pairs of similar dyads; these are 4-9 (icl+icl,

ic5+ic5 or ic6+ic6), 4-25 (ic2+ic2, ic4+ic4 or ic6+ic6) and 4-28 (ic3+ic3 or

ic6+ic6). Between them, these tetrachords exhaust the possibilities for ic6+ic6 and

partition the remainder of the pairs of similar dyads symmetrically, according to the

reciprocal relation mentioned above.

It should be stressed that, at this stage in his development, Carter appears not

to have recognized the symmetrical structure of this network, or, if he did, it was

probably not important to him. As several other writers have pointed out, Carter is

not the kind of composer for whom the systematic exploitation of theoretical

possibilities is the main creative stimulus. The exposition of these theoretical

possibilities therefore serves only as a background to the particular focus of Carter's

interest.

In the earlier part of the period 1945-55, the sets which proved most useful to

120
Carter were found among those which combined the "triadic" interval classes, ics 3, 4

and 5 (i.e those ics which may be found in the interval vector of the "major/minor"

triad 3-11 [0,3,7]). Much of the "quasi-diatonic", "quasi-triadic" and "quasi-tonal"

music of this period can be explained with reference to the tetrachordal families

formed by ic3 ^ ic3 and ic5 + ic5 in particular.

(i) Piano Sonata, second movement, bb.1-26

This passage, which has already been examined with regard to the handling of

diatonic fields, also offers an excellent example of Carter's manipulation of tetrachords

as harmonic motives. The earlier analysis of the passage dwelt principally on diatonic

orientation and was able only to classify individual vertical sonorities as diatonic or

non-diatonic (see Chap.4, pp.81-3). However, equipped with a means of classifying

sets or chords according to their intervallic content, a rather more penetrating

examination may be undertaken.

Ex.5. 10 approaches the passage in two parallel ways. Firstly, on the upper

pair of staves, each vertical sonority is shown, as in Exx.5.1-5.3, and each set of 3 to

5 members is identified. Secondly, on the lower pair of staves, certain motivically

significant tetrachords are highlighted. The upper staves show a situation already

familiar to us: the trichords 3-7 and 3-11 are by far the most numerous sonorities, with

4-26, which contains both these sets, being the most frequently occurring tetrachord.

These are all diatonic type A sets and may be regarded as harmonic norms for the

passage; they act as stable sonorities at the beginnings and ends of phrases, while the

type B and C sets represent slightly greater dissonances. As has been remarked

121
before, the climactic point of the passage in bars 21-5 sees the only non-diatonic

sonorities brought into play. The lower pair of staves in Ex.5. 10 show occurrences

of the three principal pitch motives used to weave the fabric of this passage. All are

symmetrical and may be constructed from interval classes 3 and/or 5, thus forming a

tightly knit group. Carter uses these sets not merely as chords, but as motivic units;

motive x presents set 4-10 in the form ic3+ic3 (in this case, as two minor third or

major sixth dyads connected by parallel motion through major seconds); motive y

presents set 4-23 in the forms ic3 + ic5 (y1) (a minor third, moving outward

symmetrically to a perfect fifth) and ic5 + ic5 (y 2) (parallel perfect fourths or fifths,

moving through a major second); motive z presents set 4-26 in the form ic3 + ic3 (two

minor thirds moving through a perfect fourth or fifth). In each case, the individual

ic3 or ic5 dyads may act as pivots between overlapping forms of these sets. For

example, at the end of bar 19, the dyad g#-b in the left hand may be associated with

the preceding dyad f#-c#, forming y', as well as forming two statements of motive x,

one vertical (with the dyad a#-c# in the right hand), the other horizontal (with a

similar dyad an octave lower in the left hand at the beginning of bar 20). The

association of these dyads into patterns which mimic conventional voice-leading

through stepwise connection, especially in parallel, is so strong that it survives the

registral dislocation in bars 2 1-26. Other symmetrical tetrachords which pair ic5s,

particularly 4-8 and 4-20, also play a part in the harmonic organization of this passage.

One factor which complicates an analysis of this passage according to

intervallic combinations is the presence of the equivocal "false relation" bfb. The

substitution of b for b into motive y 1 shown in Example 5.10 produces a form of the

type C set 4-16, which in these circumstances may be regarded as a distortion of the

122
"pure" type A 4-23; this happens in bars 9-10 and 13-14. The association of set 4-16

with the pitch class bb as an agent of harmonic tension is made near the beginning of

the passage; at the end of bar 3, the first appearance of b! also brings the first vertical

sonority denser than a trichord - set 4-16.

(ii) Cello Sonata, second movement, bb.91-112

This passage makes an interesting comparison with the preceding one (see

Ex.5. 11). The basic principle of harmonic construction is similar, in that it is built

from sets which contain interval classes 3 and 5, frequently using these dyads as pivots

between tetrachords. However, the sense of diatonic orientation fluctuates much more

rapidly here and the vocabulary of sets used is much wider, including several non-

diatonic sets. Although the range of sets used is apparently diverse, it will be

observed that the great majority fall into the families of tetrachords formed from

ic3+ic3, ic3+ic5 and ic5+ic5 (see Appendix 3). This statement applies equally to

sets arising from linear and vertical combination. The context of this passage within

the movement and the origins of its harmonic vocabulary will be explored in more

detail in Chapter 8.

(iii) Eight Etudes and a Fantasy, Etude I

This work is a collection of brief pieces, each focusing on a single aspect of

compositional technique, whose concentrated development here led to the liberation

123
and enrichment of Carter's musical language in the immediately succeeding pieces, the

First String Quartet and the Quartet Sonata. The first etude is described by Schiff as

a study in the use of musical space [Schiff 1983: 1431. Its harmonic language is

typical of the later Carter in that there is no reference to a tonic or to particular

diatonic fields, and a wide range of sets, both diatonic and non-diatonic,

(predominantly trichords and tetrachords) is used in both the vertical and horizontal

dimensions (see Ex.5.12).

However, the harmonic language of this work is perhaps less unified than that

of the passages previously examined. The probable reason for this is that the

exploration of various kinds of texture and spacing is more important here than that of

a particular harmonic vocabulary. Nevertheless, certain characteristics may be

observed which lend cohesion to the piece. Combinations of interval classes 3, 4 and

5 predominate. This is particularly clear towards the end (from bar 15 onward),

where much of the texture is constructed as pairs of dyads. Set 4-12 [O,2,3,6J,

spaced as ic3+ic4, becomes a temporary point of focus in bars 15-16; this is an

important junction in the work's form, as a condensed recapitulation of the opening

gesture occurs in bar 16.

(c) The Kh(6-20) subcomplex.

During the later part of the transitional period, Carter attempted to find means

by which to achieve coherence within the liberated harmonic world of intervallic

combination. One of these, as has been recognized and will be discussed in the

succeeding section, was the use of the all-interval tetrachords as formative or

124
integrative factors. Another important and distinctive harmonic resource, which has

been less widely recognized, is the group of sets which may be described, using

Fortean terminology, as belonging to the Kh subcomplex about the hexachord 6-20.

In general, Forte's concepts of set complexes and nexus sets play little part in the

technical vocabulary of this study. However, the Fortean description of the

membership of the group, and of the relations between them, is apt in this case, since

they correspond closely with the relationships exploited by Carter.

Set 6-20 [0,1,4,5,8,91 is one of the "all-combinatorial" hexachords; in other

words, it may be mapped on to itself or its complement by subjecting it to

transposition, or to retrograde, inversion or retrograde inversion combined with

transposition [Babbitt 19551. One result of this property is that there are only four

different possible transpositions of this hexachord; only 6-35, the "whole-tone scale",

has fewer. Two further consequences of this property are (i) that the set has multiple

internal symmetries, and (ii) that it has a relatively small number of different subsets.

The Kh subcomplex centring on 6-20 has only nine members. (Again, among the

hexachords, only 6-35 has fewer.) The practical consequences of these properties are

respectively (i) the symmetry of 6-20 itself means that it may be exploited as a striking

referential sonority, and (ii) the restricted number of members of Kh(6-20) creates the

possibility of using a small and tightly unified group of sets as a basic vocabulary.

Thus 6-20 would appear to possess the properties appropriate for a "key-chord"

described in section 2 of this chapter. Indeed, it appears to play this role in two

major works of the period, the first movements of the Cello Sonata and the Quartet

Sonata.

Before the role of 6-20 in these works is discussed, the properties of its subsets

125
 will be examined. The interval vector for 6-20 [3036301 is unique among hexachords

in that it contains two zero entries - those for interval classes 2 and 6. The resulting

preponderance of thirds and fourths/fifths creates a harmonic vocabulary in which the

quasi-triadic or "seventh" sets, described in section 4 (a) of this chapter, are the norm.

6-20 has only one pentachordal subset, 5-21[0,1,4,5,8], which is therefore identified

very closely with the parent set. Of the four tetrachordal subsets, three, 4-7 [0,1,4,51

4-17 [0,3,4,7], and 4-20 [0,1,5,81, are themselves symmetrical and form a very tightly

knit group, since they form a "transitive tuple", with the similarity relation R1R,

holding between each pair5. The fourth tetrachord, 4-19 [0,1,4,81, is in the relation

R2R1, with each of the other tetrachords. The trichordal subsets are 3-3 [0,1,41, 3-4

[0,1,51, 3-11 [0,3,71 (the "major/minor triad") and 3-12 [0,4,8] (the "augmented

triad"). 3-3, 3-4 and 3-11 are all in the relation R1 R, with one another, while 3-12 is

in the relation R with each of the other trichords. Thus the members of the Kh(6-20)

subcomplex form a distinctive and highly unified harmonic vocabulary.

(i) Cello Sonata, first movement.

As stated in section 2 of this chapter, differing opinions have been expressed

regarding the harmonic constitution of this work. Schiff describes the set formed by

the pitches of the first bar [0,1,2,5,6,816, or 6-Z43, as containing the "basic harmonic

material for all four movements" [Schiff 1983: 1361. He states that this set is

See Forte 1973: 52-56 for an explanation and discussion of transitive set relations.
6 Wrongly printed as [0,1,2,5,7,81 in the first edition of Schiff's book. [Schiff 1983:
137]
126
 immediately repeated in bars 3-4 in transposed form, as though to confirm its

importance. 7 However, the aui in bar 3 does not belong to this set, a fact which

suggests that the reappearance of 6-Z43 here may not be as important as those of

other, smaller sets. (The ab forms part of a vertical statement of 4-17, for example.)

Moreover, 6-Z43 appears to play no further part in the work, leaving its status as

"key-chord" or "source-set" in some doubt.

More significant are the teirachords which Schiff derives from the opening bar

[Schiff 1983: 137, "Chart 9"] (see Ex.5.13). These break down into two groups

according to the pairings of dyads described above in section 4(b); those formed from

ic4+ic4 and those formed from ic3+ic5. The link between the two groups is 4-7

[0,1,4,51, the opening chord of the movement, which may be formed in either way.

An inspection of Schiff's "Chart 98 yields two interesting points; (i) the ic4+ic4

tetrachords which Schiff lists are 4-7, 4-17 and 4-20, the three members of the

transitive tuple described in section 4(c), while the ic3+ic5 tetrachords include 4-18,

which, as discussed in section 4(a), is closely related to sets 4-17, 4-19 and 4-20; (ii)

of the six different tetrachords listed by Schiff, only four are subsets of 6-Z43. Taken

in conjunction, these observations suggest that 6-20 might act as a more convincing

"controlling set", if one is to be found, than 6-Z43.

Kies goes as far as to ascribe the entire pitch material of the movement to the

manipulation of two complementary forms of 6-20, which he terms H 1 [11,0,3,4,7,81

' Harvey repeats this observation, and illustrates it in his Example 3.10 [Harvey: 34].

The sets and relations described here are reproduced by Harvey in his Example
3.12. The latter contains two errors; the interval vector of 4-13 appears as [110211]
correct form [112011]) and the prime form of 4-18 appears as [0,1,6,9] (correct form
[0,1,4,7]).

127
and H2 [1,2,5,6,9,10] [Kies: 63]. Although Kies does not use Fortean terminology,

he is clearly aware of the symmetries and transitive similarity relations which exist

between members of the Kh(6-20) subcomplex, as is shown in his Figures 11 and 12

[Kies: 66, 75] and in the appendix to his study. However, difficulties arise when

Kies attempts to maintain the separation between H 1 and H2 , as certain passages can

only be explained by describing them as using a mixture of elements from both

hexachords. This implies a kind of conflict between the hexachords which is in

practice very difficult to detect, since the sense of a tonal centre fluctuates almost

constantly. Furthermore, Kies finds he is obliged to consider sets which are not

subsets of 6-20, or else to place pitches in parentheses in his analysis if they do not

belong to the prevailing hexachord.

Clearly, an account of the movement's harmonic resources should acknowledge

the importance of 6-20 without forcing the music into a straighijacket by attempting to

show that it is continually present (something the composer himself disclaims - see the

quotation on p.97). The proper function of 6-20 is to act as a point of focus, which

crystallizes the intervallic relationships existing among the primary constituents of the

movement's harmonic vocabulary. Nowhere is this shown more clearly than at the

end of the movement (see Ex.5. 14). Here the texture becomes "frozen" into isolated

chords, thus focusing the listener's attention on the harmonic atoms from which the

music is constructed - the interval classes 1, 3, 4 and 5. The resulting vertical

sonorities may therefore all be related to 6-20 and, on the last beat of bar 127, the

hexachord duly appears, its status as a goal emphasized by its symmetrical spacing.

The emergence of 6-20 as a focal sonority in the first movement of the Cello

Sonata is one example of this work's "transitional" nature. The first movement,

128
according to Schiff [Schiff 1983: 1361, was the last to be composed, and therefore

represents the furthest extreme in a process of harmonic concentration which can be

traced through the work. The harmonic language of the second movement (the first

to be written) seems much less focused by comparison; 6-20 plays little or no part in

its harmonic make-up, which is instead dominated by a wide range of tetrachords

combining interval classes 3 and 5. Only later in the composition of the work did

Carter discover the integrating power of the nucleus of three tetrachords, 4-7, 4-17 and

4-20 and of their common superset 6-20. However, once this had been discovered,

he was able to return to it in a later composition and exploit it even more consistently.

(ii) Quartet Sonata, first movement.

Carter's use of 6-20 and its subsets in this work exploits their symmetrical

properties to the utmost. A full discussion of the harmonic structure of this

movement will be reserved for Chapter 10. However, a few salient moments will be

examined here in order to illustrate Carter's more systematic use of Kh(6-20) members

in this later work. The appearances of 6-20 in the movement are few, but are of

crucial importance. The first forms part of the initial splashing" gesture of the

harpsichord (see Ex.5. 15). This widens symmetrically about an axis of d-e and

takes the form of tetrachord 4-17 followed by 6-20, the former partitioned into

ic3+ic3, the latter into two forms of 4-20. This symmetrical division of the

hexachord between the two hands is adopted as the characteristic spacing, and it recurs

in bar 24 and in the final bar (both of these examples in the harpsichord part).

The passage in bars 23-4 is an expanded inversion of the opening gesture and is

129
the harpsichord's most extended if rnarcato statement (see Ex.5.16). It is framed by

two complementaiy statements of 6-20 [0,1,4,5,8,9] and [2,3,6,7,10,111, the former

partitioned into two trichords 3-3 and the latter into two 4-20s. These two framing

statements of 6-20 share a common axis of symmetry about e-f, although the

intervening material departs from strict symmetry.

In the final bar of the movement, 6-20 appears as the goal of an extended

process of harmonic expansion across the range of the harpsichord, beginning in bar

30. The twofold 4-20 spacing is retained, but with the player's hands at opposite

extremes of the keyboard rather than overlapping in the centre. (The right hand f and

left hand b mark the harpsichord's pitch extremes for the entire movement.) As

Example 5.17 shows, the whole of the passage from bar 30 to the end of the

movement may be seen as a huge expansion of the opening symmetrical "wedge"

gesture.

6-20 may thus be regarded as a true "key-chord" in this movement since it

appears as a starting point and as a goal on both the large and the small scales. These

appearances may be seen as strictly symmetrical pillars, between which the looser

symmetries of the intervening material hang.

(d) The all-interval tetrachords.

Carter's use of the all-interval tetrachords is one of the best-known aspects of

his harmonic technique. However, as has been suggested above (see the references to

[Gamer 1973 ]), detailed discussion of the composer's actual practice has been

relatively sparse. Schiff briefly draws attention to the use of 4-Z15 in the opening

130
cello solo of the First Quartet;

Its strong structural possibilities can be seen in the cello solo which opens the
First Quartet; the chord is stated five times in different inversions and shapes.
Note that in this passage and throughout the Quartet Carter uses chords closely
relatedto(0, l,4,6)-(0,2,5,8)inbar2,(O, 1,3,7)inbar6,(0,3,4,7)
in bar 7 differ from the source chord by only one pitch. [Schiff 1983: 641

Schiffs analysis of this passage makes an interesting comparison with that by Gamer,

the former attempting to show the unity of its harmonic language, the latter its

diversity. Schiff's remarks regarding related sets are suggestive and require further

exploration. However, there is one set which he appears to rule out - the other all-

interval tetrachord 4-Z29 [0,1,3,7] - which he states "Carter avoided at first because of

its strongly tonal feeling" [: 64 ], although his own analysis shows that it is prominently

presented in bar 6 (third crotchet). In fact, it will be seen that 4-Z29 plays an

important subsidiary role in the Quartet.

The most explicit expressions of 4-Z15 come towards the end of the work. In

the climactic passage at bar 430 and in bars 438-443, a ten-note chord is sounded,

leaving the pitches a#, c, d# and e exposed. After this point, the tetrachord appears

more frequently as a chordal sonority - see for example, bars 449-451, 455 and 477

(Ex.5. 18).

However, there are less obvious ways in which 4-Z15 influences the harmonic

language of the work. For example, it may be combined with other important related

sets to form larger supersets. By way of illustration, we may examine the second

violin part at the opening of the first movement "prope? (bb.22-30; see Ex.5.19). A

straightforward segmentation of the passage reveals that 4-18 is a particularly

prominent set, usually articulated as a trichord (3-5 or 3-10) plus a melodic semitone.

The importance of 4-18 here may be traced back to the second violin's first entry

131
(b. 12), during the cello cadenza, where the combination of the two instruments forms

this set. 4-Z29 is also prominent as the opening motive of the line (also recalling the

first entry). The succeeding passage develops the relationship between 4-18 and the

two all-interval tetrachords. It will be noted that 4-18 differs from each of the latter

by only one pitch (4-Z15 [0,1,4,6], 4-18 [0,1,4,71, 4-Z29 [0,1,3,7]), and may thus

form a close association with both.

In the first bar of the passage, the second violin states the octad 8-6

[0,1,2,3,5,6,7,8], articulated as 4-Z29 followed by 4-12. However, alternative

segmentations reveal that 8-6 may be partitioned into two forms of 4-18 and,

furthermore, that it contains both of the all-interval tetrachords no fewer than four

times each. This set may therefore be regarded as a particularly rich source of

harmonic material, saturated with forms of significant sets. In the bars that follow,

the second violin part remains under the influence of this group of three tetrachords

and their supersets. In bars 22-26, the pitches form 6-Z19 [0,1,3,4,7,8], which

contains 4-18 twice and 4-Z29 once. During bars 27-30, as the part becomes denser,

moving in vertical dyads and triads, successive pairs or trios of chords form pentads or

hexachords, each of which contains 4-18 or either or both of the all-interval

tetrachords or combinations of them:

b.27 6-Z50 (2 x 4-Z15, 2 x 4-Z29, 4-18)

5-20 (4-Z29)
6-Z36 (4-Z29)
b.28 5-31 (4-18)
6-16 (4-Z15, 4-Z29)
6-Z13 (2 x 4-Z15, 2 x 4-18, 2 x 4-Z29)
b.29 5-32 (4-Z15, 4-18)
6-Z26 (2 x 4-Z29)

Thus, although the tetrachords 4-Z15, 4-18 and 4-Z29 are not always clearly

132
prominent on the surface of the musical texture, they play a pervasive role in the

formation of the harmonic language. A comparison of the second violin part with that

of the cello during part of the same passage (bb.22-26) shows that similar processes

are at work in the shaping of this line.

Whereas the previous example has shown Carter's use of the all-interval

tetrachords in creating individual parts or lines, examples taken from much later in the

work will show how all four parts are harmonically coordinated. Once again,

attention will be focused on the relationship between 4-Zl5, 4-18 and 4-Z29. 4-18 is

prominent in Part ifi (Variations) of the Quartet, as part of a thematic idea; the

"chorale" motive [Schiff 1983: 161] consists of two inversionally related forms of this

set, sharing two pitch-classes. The varied repetitions of this motive make use of

subsets and supersets of 4-18 to produce effects of harmonic relaxation and

intensification respectively (Ex.5.20). In the richly chordal passage from bars 392-

402 of Part Ill, the relationship between the "chorale chord" 4-18 and the two all-

interval tetrachords is vividly exploited. The texture is extremely dense, varying from

four to eight parts, but coherence is achieved through the almost constant presence of

4-Z15, which runs through the texture as a core sonority. Example 5.21 demonstrates

this; each vertical sonority in the passage is given its set label and an indication of the

number of times 4-18 and the two all-interval tetrachords are contained within them.

Particular attention should be given to the three appearances of 5-32 [0,1,4,6,9] which

contains both 4-Z15 and 4-18. However, even more remarkable is Carter's control of

the larger sonorities (septads and octads) which contain high multiples of all three

tetrachords. Significantly, the richest of these, 8-28, is reserved for the climax of the

passage; being multiply symmetrical, it contains all three tetrachords eight times over.

133
 CHAPTER 6: VOICE-LEADING AND LINEAR MOTION.

The preceding chapters have dealt with the local associations of pitches into

fields and sets on the surface of Carter's compositions. Analysis of this music by

these means reveals something about the harmonic vocabulary employed by the

composer, but little about the syntax of musical construction. Sets and fields are

"democratic" entities, whose members are indistinguishable from one another in terms

of importance, but in practice, we are clearly able to hear that "some notes are more

equal than others", since Carter's transitional music, in common with much other post-

tonal music, creates hierarchical relationships among pitches through the adaptation of

tonal procedures.

This is not to say that post-tonal music exhibits the same degree of structural

unity as tonal music. The structural levels which, according to Schenkerian theory,

govern tonal music, are indivisible unities, held in relationship through a similarly

indivisible hierarchy, in which every element is referable to one of greater structural

importance and ultimately to the Ursatz In post-tonal music, however, such

hierarchies may be loose or fragmentary, since the a priori elements and relationships

which exist for tonal music are absent and cannot be replaced by any other than those

determined within the context of the individual work. In a certain body of post-tonal

music, which is sometimes described as exhibiting "extended tonality" Schoenberg,

1969: 76], the harmonic and contrapuntal relationships employed bear some

resemblance to those of tonality. These may function in a way which is genuinely

parallel to tonal practice, or they may mimic the appearance of tonal relationships (or

parody them) while actually obeying other, non-tonal, principles such as symmetry or

134
 complementation.

This chapter will examine various ways in which Carter shapes musical

structures and will proceed from the small to the large scale, but first, a preparatory

discussion of terminology will undertaken. The terms "foreground", "middleground"

and "background" are used here informally as an indication of scale, rather than in the

strict Schenkerian sense. The terms "prolongation" and "association" must also be

examined, since they encapsulate some of the principal differences between concepts of

tonal and post-tonal structure. "Prolongation" refers to the process in tonal music

whereby the content of a lower structural level is created through the linear elaboration

of a harmonic entity on a higher level. "Association" - a term coined by Joseph

Straus [Straus, 1987] in order to explain the workings of post-tonal music - is simply

the mental connection of pitches which axe perceived to be structurally important and

which, taken in conjunction, may be regarded as forming a unity on a higher level.

Prolongation would appear to be the stronger principle, association the weaker, but

more adaptable. In Straus's theory, patterns of association are in effect the

"composing-out" of motivic pitch-class sets, although they may resemble traditional

tonal prolongations:

In many twentieth-century works musical motions over large spans of time

seem to follow traditional tonal patterns, like the descending perfect fifth.
Usually, however, their customary meaning is effectively neutralized. They
evoke traditional practices but derive a more potent meaning from the specific
post-tonal context in which they occur.[. . .J Large-scale motivic statements
frequently provide coherence over large musical spans. The tones that make
up a motive can be associated in a variety of ways, by register, timbre,
metrical placement, dynamics, instrumentation, articulation, or shared value in
any musical domain. [Straus, 1990 169-701

While acknowledging the usefulness of the concept of association, a degree of

caution should be exercised in its application. An analysis which attempts to show a

135
 continuous, unified middleground or background, consisting of associated pitches, may

be as dubiously grounded as one which attempts to show atonal "prolongation".

When the criteria for assigning structural weight to particular pitches are

fundamentally subjective, the danger exists that facts will be distorted to fit the theory;

specifically, the discontinuous and fragmentary nature of post-tonal structures may be

misrepresented by a theory which attempts to "reveal" connections and identities.

Another important factor is the simple issue of scale in relation to aural

perception. A small-scale motion which resembles a tonal prolongation at the

foreground level may be accepted as such, but if repeated and expanded at the

middleground level may be downgraded to an association, while at the background,

even an association may not be perceptually valid. The importance of factors other

than pitch in determining patterns of association (see the quotation from Straus, p.135),

may be interpreted as a sign that, in the long run, specific pitch relationships are

bound to lose their efficacy as the principal means of ordering musical structure in

post-tonal music. Carter's evolution of rhythmic, temporal, intervallic and timbral

schemes of relationship in his later work would appear to be a practical

acknowledgement of this fact.'

1. Foreground.

The foreground is the level at which the diversity of tonal and post-tonal

'See [Bernard, 1990(b): 352]; "The real issue here is whether structure in Carter's music
beyond the level of local, note-to-note considerations can in fact be grasped by means of an
analytical method based upon assumptions of priority for pitch-class relationships.[. ..1
Relegating all other domains [...] to the status of the 'presentational' seems rather too blithe
a dismissal of Carter's claim that such matters are every bit as important as pitch class."

136
 techniques employed by Carter is most clearly revealed. In common with other

composers working within the language of extended tonality, Carter's use of or

reference to tonal techniques has several guises. He may make use of genuinely

functional tonal relationships, he may alter these relationships so that their

conventional function is still discernible but their surface appearance is unconventional,

or he may shape non-tonal elements into patterns which allude to tonal ones. This

section will focus on several harmonic and contrapuntal phenomena which have their

origin in tonality, but which are extended or adapted in Carter's transitional phase; (a)

the tonic-dominant relationship, (b) the neighbour note, (c) linear progression, (d)

arpeggiation, (e) octave doubling and octave transfer.

(a) The tonic-dominant relationship.

The relationship between the scale steps of tonic and dominant and their

respective triads is the most fundamental in tonal harmony. For music to be

perceived as tonal, we must be aware not only of a central tonic which governs

harmony and counterpoint, but also of a structural counter-pole - the dominant - whose

existence at first challenges the supremacy of the tonic, but ultimately affirms it by

acting as a "leading-note sonority" which must proceed (resolve) to the tonic. In the

music of extended tonality, we may see a split between these two dominant functions

of structural counter-pole and leading-note sonority, so that each may be taken by a

different pitch element.

One of the simplest and most characteristic uses of the dominant in tonal music

is in alternation with the tonic at the opening of a work The creation of a I-V-I-V...

137
pattern serves to prolong the tonic and thus establish the tonality with the greatest

economy. Adaptations of and allusions to this technique are frequent in Carter's

earlier music (see Ex.6. 1). Example 6.1(a) is taken from bars 4-7 of The Rose

Family. Here, the entry of the voice suggests an alternation of scale degrees 3 and 2.

The motion of the bass ostinato also suggests an alternation or amalgamation of tonic

and dominant over a tonic pedal, but since this is not coordinated with the vocal part,

the correspondence with the conventional tonal model is not straightforward.

A similar situation is presented in Ex.6.1(b), taken from bars 5-8 of the

Pastoral. Here it will be noticed that the pitch f# forms part of the quasi-tonic

sonority. The dyad e-b, whose pitches act as neighbours to f# and a, plays the role

of quasi-dominant and the pitch a persists as a tonic pedal in the bass. Once again, a

straightforward tonal reading is complicated by contrapuntal factors; the moving part

in the piano left hand begins with g#-b, which in itself suggests the dominant, but is so

frequently sounded with the fA'-a "tonic" that it is practically amalgamated with it.

Ex. 6.1(c) is the first two and a half bars of Voyage. As in Ex. 6.1(a), the

upper part suggests an alternation of scale degrees 3 and 2, here supported by motion

in parallel tenths in the bass. However, the bass, rather than moving from tonic to

leading-note and back, moves in exact parallel, descending a major second to the

flattened leading-note, atj. This produces a "false relation" with the a# of the inner

part, thus creating an ambiguity with regard to the mode of the dominant harmony

which is implied. This alternating pattern is extended in bars 38-44 and 93-105.

Ex.6. 1(d) shows a pattern of alternation arising from motion in the bass. The

bass moves from c to f, suggesting V-I in F major, while the upper parts maintain an

unchanging f-bL -d-g. While the chord over the c may be interpreted as a "dominant

138
eleventh", that over the f is less convincingly interpreted as a tonic, unless the upper

voices are regarded as suspended. Since the latter never resolve by step, this

interpretation is relatively weak. This is, therefore, an interesting example of a

progression which alludes to tonal function but avoids purely triadic elements.

Example 6.2 illustrates the use of the dominant in its "leading" or cadential

function. It must be said that the unambiguous use of the perfect cadence V-I in

Carter's music is extremely rare, since emphatic closure is usually avoided.

Ex.6.2(a) shows the end of the Pastoral, where an A major tonic is asserted. The

effect of a dominant is dependent on the e in the viola at the bottom of the texture in

bar 275, enhanced by its flattened upper neighbour ftj. In bar 276, the sense of

dominant function is tenuously present as the pitches of an "e dominant seventh" are

dispersed through the bar in a way which makes them marginally more prominent than

the other pitches.

Ex.6.2(b) shows passages from the first movement of the Piano Sonata, a work

which, unlike most other Carter, is characterized by grand cadential gestures. The

passages in question are taken from the "second subject" and the approach to it (bb78-

83 and 101-102). The final beat of bar 78 moving to bar 79 gives the effect of a

conventional, if extremely rapid, 3-2-1 over I-il-V-I harmony in c minor. Bars 81-82

emphasize the lower neighbour, b, harmonizing this with a G major triad (temporarily

obscured by a flourish making use of a tonally unrelated form of set 4-23) before

returning to c in bar 83. Midway through the course of the second subject, this

cadential emphasis of c is repeated with parodic effect (bb.101-2). The bass motion

g-c is "normal", but the chromatic flourish of bar 81 is greatly expanded in bar 101,

forming an eleven note collection, which mocks the tonal conventionality of the bass

139
 progression. This kind of progression, where motion through a fourth or fifth in the

bass might suggest a tonal cadence, while the upper voices militate against such an

interpretation or at least render it ambiguous, becomes common later in Carter's

output. Ex.6.2(c) shows several instances from the first movement of the Cello

Sonata2.

Allied to the use of the dominant in the perfect cadence is its use as a

"divider", a point of temporary rest or a structural interruption such as may be found

at the end of the antecedent phrase of a period in tonal music. The functiofi of the

dominant in such a case is both to act as a point of arrival and to create the

expectation of continuation. Example 6.3 illustrates Carter's use of the dominant in

these situations. Ex.6.3(a), from the second movement of the Piano Sonata, is a

variant of a standard means of dominant preparation, the augmented sixth chord. The

pitch ei is emphasized here not only through its appearance in five octaves, but also

through the "leading-note" effect of its chromatic neighbours f and d[1. The factors

which blur a clear tonal reading of the progression are Ci) the fact that e' is sustained

against its neighbour notes in bars 24-5 and (ii) that the voice-leading suggested by the

figured bass is not completely realized and is subject to extremes of registral transfer.

The emphasis on eL as V in a' minor is clearly felt nevertheless, though Carter

avoids the obvious by leading us towards Gk or CL' in the following bars (27-29).

The emphasis on eb as a dominant here also has the effect of strengthening the

2 Harvey [: 36] draws attention to the first of these (b. 1), which approximates a
perfect cadence in B major, also noting the "arpeggiation" of a B major triad in the bass.
However, the appearance in octaves of the pitch d# may alternatively be explained as the
result of an extension of the intervallic relationships presented in the first chord, 4-7.
The upper dyad, a#-e#, spans a perfect fifth; adding another perfect fifth below a#
produces d#. The lower dyad, f#-a, outhnes a minor third, which, if added to in a
similar manner also produces d#.

140
 connection with the d of the opening bars of the movement (see Chap.4, p.81). The
'4.
latter may be tentatively interpreted as scale degree 5 in G major, with the progression

d-c
b-a "unfolded" below it. The dominant quality of d remains only latent, since an

unequivocal G triad is avoided.

Ex.6.3(b) (Cello Sonata, second movement, bars 35-44) shows a passage which

hovers around the tonality of A' major/minor. The first two bars of the passage

appear to outline, through unfolding and voice-exchange, a I-V progression in

minor. The "answering phrase" initiated by the cello begins with an implied perfect

cadence and changes to the major mode. The bass imitates a familiar tonal pattern, a

sequence of root progression through falling thirds I-V1-W, before being diverted back

to V via V/V in bars 39-40. This interpretation is, of course, dependent on supplying

"implied notes" to the otherwise sketchy voice-leading. It is more difficult to

incorporate the next two bars into such a scheme. The abrupt shift of the "root" to gl

followed by dl' and the complex relationship of the piano right hand part to these roots

in bar 41 make a tonal interpretation in At' major problematic, while the intrusion of

eIj and b11 in the cello in bar 42 are positively disruptive. Nevertheless, when the g-

b' dyad is asserted again in bars 42-3, it functions as an incomplete dominant

sonority, being strongly associated with similar occurrences in bars 37 and 40.

(b) The Neighbour Note.

The neighbour note is perhaps the simplest form of melodic prolongation

available in tonal music and it is therefore not surprising that it should be "borrowed"

141
 or alluded to frequently in post-tonal music. The only problem in applying the term

neighbour-note outside a tonal context is that it presupposes a hierarchic relationship in

which the neighbour note is dependent upon the prolonged note. This situation does

not invariably exist in post-tonal music, where both notes may be equally important

members of a pitch-class set or motive.

Example 6.4 illustrates several uses of the neighbour note in Carte?s earlier

music (also see Ex.6.l(a) (b) and (c)). The first of these, Ex.6.4(a), is relatively

straightforward; the neighbour notes may be identified as such because they do not

belong to the prevailing harmony and are resolved by returning to adjacent notes which

are. Set 4-26 (f-g-b' -d) is sustained in the piano, while the g and d are arpeggiated

in the vocal part and prolonged by use of the neighbour notes a and e. These notes

produce an effect of greater dissonance than those surrounding them, since they form

semitonal clashes with the pitches b(' and f of the sustained harmony.

Ex. 6.4(b) shows the opening of the Piano Sonata (bb. 1-7), which makes

complex and expressive use of the neighbour note. Several neighbour relationships

are exploited here; (i) the motion from b to a# underpins the passage (this is

emblematic of the tonal progress of the movement as a whole), (ii) the pitches of the

first chord in bar 2 (g#-c#-e#) are initially perceived as neighbours to the implied B

major harmony produced by the overtones of the five octave b, (iii) the "resolution" of

this chord produces further complication, since a c is introduced instead of the

expected b; this has the effect of retrospectively altering the perception of the g#-c#-e#

chord so that it appears the more stable sonority, while the c-d#-f# chord appears part

of an incomplete dominant seventh, composed of neighbour notes to its predecessor,

(iv) the introduction of dL1 in bar 3 produces a further ambiguity, since it might be

142
heard either as a neighbour to the c, or as an alteration of the d#, producing a modal

"mixture", (v) the e# and b# in bar 5 are clearly neighbours to f# and c# in the

following bar. As a whole, this passage also exhibits the "unfolding" of a non-tonal

sonority, set 4-26, which may be regarded as one of the work's "key-chords". The

pitches d#-f#-a#-c# are deployed in two pairs of dyads, first in minor thirds (bb.2-4),

then in perfect fifths (bb.6-7).

Ex.6.4(c) demonstrates the use of neighbour notes in a context that is further

removed from conventional tonality (Cello Sonata, third movement, bb. 12-21 and 29-

39). The pitch content of these passages is strongly influenced by the Kh(6-20) sets

described in the previous chapter. These sets are deployed so as to maximize their

resemblance to tonal patterns. The particular form of 6-20 used at the opening of the

first passage is [d-eI -f#/g' -g-bI' -b/c'], which contains major and minor triads of (I,

E and B. Within this symmetrical set structure, the pitches d and et' should be

regarded as having equal value. However, the prominent repetition of the pitch d

makes it a point of focus, while the pitch e' is frequently used in a manner which

mimics the conventional neighbour note, the most important difference being that the

"neighbour note" e is often sounded against the "prolonged note" d.

Having discussed a few instances of Carter's use of the tonic-dominant

relationship and the neighbour note, we will examine a particularly characteristic

feature of his style which involves the conjunction of these two phenomena (see

Ex.6.5). In each of these passages, there are two tonic-dominant axes, which are

separated by a semitone. In some cases, one of these axes appears to be primary,

while the other stands in a "leading-note" or "Neapolitan" relation. In other cases,

the four pitches appear locked together in tense symmetry. The first example, taken

143
from the final bars of the first movement of the Piano Sonata, encapsulates the tonal

argument of the work as a whole, which hinges on the harmonic interpretation of the

pitch al/b. (see Chap.5, pp.114-5). Here, the stark juxtaposition of pitches from the

tonic and dominant chords of B major against their neighbour chords F major and B!

major makes it difficult to decide which is the true tonic axis and which the neighbour

axis. The final settling on f and bL is only a temporary solution in the context of the

work as a whole, as may be deduced from the sudden drop in dynamic and the

marking Pià tranquillo.

Examples 6.5 (b), (c) and (d) are all taken from the first movement of the

Cello Sonata. David Harvey attempts to show that the movement has a unified

middleground structure, in which focal pitches are nsustainedw through neighbour-note

and fifth relationships [Harvey: 361. While reserving judgement on the validity of

this view, we can see that at the foreground such relationships axe much in evidence.

Ex.6.5(b) is the least complex, being taken from a passage for solo cello. The

resemblances to tonal features such as arpeggiation, octave transfer and neighbour note

motion require little explanation. Ex.6.5(c) is typical of several brief sections in the

large paragraph from bar 19 to bar 67; the bass, in itself, suggests the kind of axial

relationship previously described, but in the context of the three-part contrapuntal

texture, it is impossible to ascribe a genuine root-function to this bass line. The third

of these passages, Ex.6.5(d), is the most complex in its allusion to tonal procedures

(compare Ex.5. 14). The pitches e and b appear frequently in the bass, which,

together with the e-g dyad sustained by the cello from bar 123 until the end of the

movement, suggests an e minor tonality. The use of tetrachords from the Kh(6-20)

group, with their strong triadic associations, contributes to the pseudo-tonal feeling;

144
several of the sonorities employed might be construed as tonic or dominant chords in e

minor with added neighbour notes, the pitches a# and c frequently appearing in

conjunction with b. The alternative axis to e-b is d#-a#, most strongly presented

across the bar line of 125-6. As in the example from the Piano Sonata, the pitch d#

appearsinthispassagebothasthethirdofthedominantchordandasarootinitsown

right, but the superinipositions of tonic and dominant and their neighbours are much

more difficult to "disentangle" in the later work, and it may be reasonable to suggest

that the fifth-relationships in play here are primarily the outcome of the interplay of

intervals and sets rather than of tonal 'progressions".

(c) linear progression.

The concepts of linear progression in tonal and post-tonal music, while

superficially similar, are fundamentally different in structural implication. In tonal

music, a linear progression connects two notes which have a "genuine relationship'

[Schenker, 1979: 74 1, which is to say that, at a deeper structural level, they may be

regarded as belonging to a single harmony, which is then "unfolded" and "composed

out" by means of the linear progression. Thus, the only intervals which can be

composed out through linear progression are those which may form part of a triadic

harmony - the third, fourth, fifth, sixth and octave. In post-tonal music, there is no a

priori relationship between the notes which begin and end a linear motion; instead,

borrowing a phrase from a description of Hindeinith's music, "the voiceleading is spun

out between [...1 two points like the cables of a suspension bridge" [Neumeyer. 65].

Thus a linear progression at the foreground level in post-tonal music does not

145
 necessarily have implications for deeper structural levels, although it may do so.

Linear motion is an aid to coherence in post-tonal music, since it is the most

aurally obvious method of leading from point to point. This is especially true in

contrapuntal music, much of Carter's being a case in point. The effect of a linear

progression is often enhanced by doubling it with a parallel motion - a device much

favoured by Carter - or occasionally, by combining it with a contrary motion.

Linearity is such a powerfully simple constructive principle that it served Carter

through the whole of the period under examination. Example 6.6 illustrates this with

a few characteristic passages.

Examples 6.6(a), (b) and (c) have in common a string of parallel triadic

intervals (perfect fifths or major thirds), forming long, sometimes meandering, chains

of linear motion. The example from The Rose Family shows descending linear

motion in parallel fifths in bars 22-24, which preserves a diatonic field of D, Ibliowed

by ascending parallel triads in bars 25-6, which disorientate the sense of diatonic field.

The use of parallel fifths here, in the approach to the main climax, further intensifies

the importance of this interval in the song (parallel fifths have already been introduced

in bars 20-21, piano right hand).

Ex.6.6(b) demonstrates a much more thoroughgoing use of parallel linear

motion in thirds, fourths, fifths and complete triads, which makes up most of the

middle section of Warble for lilac-time. Bars 93-4 make use of an interesting

variation of the voice-leading of bars 90-92; the upper pair of lines are inverted,

forming a perfect fourth instead of a fifth and the order of the upper dyads is reversed.

The streams of parallel motion seem to find no cadential resting place, an effect

which is probably designed to reflect the words of the text "Thou, soul, unloosened

146
the restlessness after I know not what;ICome, let us lag here no longer, let us be up

and away!"

Ex.6.6(c) is from Carter's most extended exercise in parallel linear motion, the

song Voyage, which sustains parallel motion in major tenths virtually throughout its

122 bars. Against this, the middle part of the piano has a tortuous line derived from

a single three-note motive (reflecting the "infinite consanguinity" of the poem's

opening lines), while the voice has a rhythmically freer, declamatory part The

constant motion in parallel results in a harmonic texture which resists definition in

relation to a single tonic and which never resolves its tension but comes to rest on a

chord combining elements of B and E majors (see Chap.5, p.1 15). The rise and fall

of the parallel lines generates the song's large-scale form. The opening paragraph

(bb. 1-52) is constructed in three "arches", each longer and reaching higher than the

previous one; (i) bb.1-1O, b-d# rising to f-a in bb.6-8 and falling back down to a-c#;

(ii) bb.1O-21, rising to g-b in b.20, falling to b-d#; (iii) bb.21-52, rising to a. -c in

bb.28-31, falling to an alternation of a-c# and b-d# in bb.38-47 and to g-b in bb.47-49.

The middle section, marked "Slightly Faster" (53-92) may be divided into two; (i)

bb.53-78, where the parallel motion receives further doubling, is distorted by octave

displacement and is temporarily abandoned; (ii) bb.79-92, a return to the rising pattern

of the opening, which is extended so that it covers two octaves. The final section

alternates b-d# and a-c# in bb.93-105, before finally subsiding towards etr -g in bb. 113-

4. This dyad is balanced against the b-d# of the opening in the brief coda.

The Piano Sonata contains many passages which exploit parallel linear motion,

though never to the extreme represented by Voyage. Ex.6.6(d) shows a passage from

the opening of the second movement, familiar from discussion of diatonic fields and

147
pitch-class set motives (compare Exx.4.2 and 5.10). Although the parts frequently re-

align themselves and may be displaced by octave transfer, patterns of parallel octaves,

perfect fifths and minor thirds are clearly audible, with the latter playing a particularly

important role, as is shown in the subsidiary diagram. Examples 6.6(e) and (t) show

two related passages from the first movement. In the first (163-6), an upper dyad of

a perfect fifth gradually ascends through the same interval. The voice-leading below

this is less consistent, but tends to create a minor sixth below the lower note of the

dyad. The second passage (232-243) expands this idea, maintaining the upper perfect

fifth at the beginning, but eventually dropping this and combining the topmost line with

other patterns of ascent and descent and some "pedal" notes. The final two stages of

this ascent - from d# to e# and from e# to f# - are more widely separated, thus

bringing the question of "prolongation or association" into play. The octave f#s at the

end of the progression are a strongly asserted goal, and the whole of the preceding

passage may be regarded as an extended "dominant preparation".

Passages of strict parallel or contrary motion are less common in Carter's later

music. Examples 6.6(g) and (h) give isolated instances from the first movement of

the Cello Sonata and Part ifi of the First Quartet. More typical are Examples 6.6(i)

and (k) (Cello Sonata, first movement and Quartet Sonata, second movement), where

one predominantly linear part receives support from others which approximate a linear

shape.

Another aspect of linear writing which is a highly characteristic feature of

Carter's style during this period, is his use of rapid scalic figuration, often irregularly

accented in order to give a sense of rhythmic freedom. Examples can be found in the

outer sections of Warble for JJJac-thne, the first movement of the Piano Sonata, the

148
third and fourth movements of the Cello Sonata, the second and eighth of the Fight

Etudes and a Fantasy and in various places throughout the First Quartet and the

Quartet Sonata (see Ex.6.7). The later examples avoid allegiance to particular keys

or modes, mixing major and minor seconds freely. This kind of scalic writing all but

disappears in Carter's later music, in which his creation of distinct intervallic

vocabularies leads him away from such relatively conventional and harmonically

neutral patterns.

(d) Arpeggiation.

As with linear progression, there are fundamental differences between genuine

arpeggiation in tonal music and allusion to it in post-tonal music, which result from

post-tonal music's lack of a priori elements and rules. While arpeggiation of a tonal

triad can be recognized as such because understanding of the concept of a triad is a

given element of tonal theory, post-tonal music has no equivalent to the triad and thus

no norms for its prolongation. If the members of a non-tonal pitch-class set are

presented horizontally, it is not possible to argue that this represents an "arpeggiation"

if we cannot recognize this set as a vertical harmonic entity.

In post-tonal music there are, principally, two ways in which arpeggiation may

be emulated. The first is by simply presenting the pitches of a tonal triad

successively; in these cases, reference must be made to the harmonic context in which

this "arpeggiation" appears in order to ascertain whether it may genuinely be perceived

as such. The second may more accurately be described as "intervallic process", to

borrow a term from Harvey [Harvey: 551; this involves the multiple replication of a

149
motion through a particular interval and direction. This is, in one sense, akin to the

principle of linear motion in post-tonal music; no pre-determined relationship need

exist between the end points of the motion; what binds them together is the consistency

of direction and size of step of the motion itself. From another perspective,

intervallic process may be interpreted as the "composing-out" of a pitch-class set,

which will usually have a symmetrical interval structure.

Simple examples of quasi-tonal arpeggiation can be seen in the opening solo

phrases of The Rose Family (voice, bb.3-7) and the Pastoral (Viola, bb.8-12).

Carter's exploitation of the rather more complex interplay between different forms of

arpeggiation is shown in Example 6.8. Ex.6.8(a) is taken from Warble for lilac-time,

the passage beginning "The robin where he hops, bright-eyed, brown-breasted" (bb.53-

56). The piano (right hand) plays on the relationship between the arpeggiation of

tonal triads [0,3,7] and of forms of set 3-4 [0,1,5]. Ex.6.8(b) is taken from the

closing pages of the first movement of the Piano Sonata and is characteristic of the

scorrewle passages of the movement as a whole. This exhibits a similar interplay

between [0,3,7] triads and set 3-9 10,2,7], which appears either as two perfect fourths

or perfect fifths.

Example 6.9 shows several passages in which an "intervallic process" akin to

arpeggiation is used to carry out a larger scale melodic motion. At (a) (Piano Sonata,

first movement), the centrality of the pitch c is established by its appearance as a bass

pedal, and as the starting point and goal of the melodic lines. The principal melodic

lines themselves are created from overlapping or adjacent statements of set 4-23

[0,2,5,7], as the second stave of the analysis shows. These lines are constructed so

that they consist only of the intervals present in 4-23, the major second, minor third

150
and perfect fourth or fifth. The underlying symmetry of the passage grows from the

SUbdiViSiOn of the octave by the interval of a minor third, this interval being

"prepared" by the c-e" dyad on which the music comes to rest in bar 80. The first

phrase rises from c to a and the "answering phrase" rises from e' to c as the upper

voice completes its octave progression, moving from a to c. This implies a

symmetrical division of the octave, c-eL' -gLi -a-c, an interpretation which is supported

by the shaping of the melodic lines. The structural importance of the minor third is

underlined by the detail of the parallel motion in the left hand in bar 83. In Ex.6.9(b)

(Cello Sonata, second movement), a similar intervallic process occurs, making use of

cycles of minor thirds, this time in parallel perfect fifths.

Ex.6.9(c) shows the opening of the third movement of the Cello Sonata. The

opening sonority is set 4-14 [0,2,3,7], in a form which pairs interval classes 3 and 5.

These two interval classes play an important role in the progression that follows. The

two instruments take the outer pitches of this chord, c#/cli' and a, as their starting

points and, over the course of bars 1-4, outline a symmetrical progression, moving

outward by semitone:

dl'

The internal structure of this progression also shows symmetry operating in different

ways, with the two instruments producing contrasting patterns from similar pitch

material. The cello part subdivides ic5 into ic2+ic3 in a repeating pattern, while the

piano part rearranges a straightforward ic5 cycle by reversing the position of the poles

dl' and c, thus creating a symmetrical pattern of ics 4 and 5:

151
 Table 6.1 Intervallic structure of Cello Sonata, ifi, bb. 1-4

5 5 5
r
cello g_j cd f bb

piano dl' f_bUic

45554

Rearranging the sequence of pitches, we can see that both instruments use

transpositions of the same hexachord, 6-32 [0,2,4,5,7,9], which may be represented as

a chain of perfect fourths separated by a minor third:

55555
_
cello adg c

piano f bl el al dl'

Carter's use of quasi-scalic figures in the later music of the period has its

counterpart in the frequent appearance of rapidly moving chains of thirds (and

occasionally fourths) in patterns which mimic the conventional arpeggio, but which

avoid tonal definition. Ex. 6.10 illustrates this with passages from the Cello Sonata

(second movement), First Quartet (Part ifi) and Quartet Sonata (first and third

movements). As in the case of the scale-type figuration shown in Ex.6.7, this kind of

material gives way after c. 1955 to patterns whose intervallic content is more rigorously

controlled.

152
 (e) Octaves.

The aural equivalence of pitches separated by an octave, or multiples of an

octave, is one of the most fundamental concepts in music, whether tonal or atonal.

The practice of octave doubling in order to reinforce the prominence of a pitch or a

melodic line became engrained early in the development of tonality. Its power was

such that it was one of the features (along with tonal triads and minor dissonances)

which Schoenberg recommended should be avoided in serial composition, in order not

to privilege one note above others. The use of octave doubling is thus one of the

features by which we may distinguish tonal from atonal music. "Coupling", or the

structural connection of pitches an octave apart, is a closely allied phenomenon - the

composing out of these octave relationships.

In "extended tonality", composers may make liberal use of octave doubling and

coupling, either as a simple continuation of common tonal practice, or in order to give

structural emphasis to certain pitches in the absence of clearly defined tonal

hierarchies. In !ct, this is one of the features which distinguishes music of Carter's

early and transitional periods from that of his later period (post-1959). Examples of

octave doubling from Carter's earlier music are almost too widespread to occasion

comment, save that it is a particular feature of his piano writing, which exploits the

instrument's potential for effects of resonance. Octave doubling becomes more

notable as it becomes rarer. It is already less frequent in the Cello Sonata than in the

Piano Sonata, and within the later work, the first and fourth movements (the last two

to be written) contain fewer octaves than the second and third. This is partly

explained by the transition during the course of the work from a primarily harmonic

153
conception to one which explores leaner contrapuntal textures, but it is also the result

of a less emphatic reference to tonality. In the First Quartet, octave doubling, while

not banished entirely, is comparatively rare, being reserved for climactic or assertive

passages (e.g. part ifi, bb. 105-113 and bb.431-437), while in the Quartet Sonata, they

are still less frequent, being used, in most cases to highlight some of the interesting

timbral combinations possible within the ensemble (see first movement, bb.45-50).

A similar course may be traced through the transitional period by examining

Carter's use of octave coupling. The use of "local" melodic octaves is an important

feature of the musical language of the earlier works, especially the Piano Sonata (see

Ex.6. 11(a)). The Cello Sonata exhibits far fewer of these local octave transfers,

which are confined to moments of structural emphasis (Ex.6. 1 1(bXc)).After the

composition of this work, such use of the octave becomes as rare as its harmonic use,

being confined mainly to passages which exploit changes in spacing and timbre (see

Ex.6. 11(d), (e) - Eight Etudes and a Fantasy, no. I, and the Quartet Sonata, second

movement). Other examples of the association of pitches an octave apart can be

found in the First Quartet and the Quartet Sonata, but emphasis on particular pitch-

classes here is transient in effect (see Ex.6. 11(1), (g)).

Carter's use of "progressions" which span an octave is a more complex issue.

In Examples 6.6(c) and 6.9(a) we have already seen passages in which structurally

important pitches (or harmonies) an octave apart are connected by means of consistent

voice-leading techniques, whether linear progression (as in bars 79-93 of Voyage) or

intervallic process (as in bars 83-85 of the Piano Sonata, first movement). The

questions of the scale of these progressions and their harmonic content must be

considered here. In Ex.6.6(c), while an "association" between the harmonies of bars

154
79 and 93 may be acknowledged, it is difficult to regard the intervening passage as a

"prolongation" of the b-d# dyad, since the harmony wanders so far from the implied B

major of this dyad and is, in any case, resistant to meaningful interpretation in terms

of root progression. In Ex.6.8(a), the shorter span of the passage, the continuing

presence of the pitch c and the consistency of the symmetrical division of the octave

contribute to a much securer sense of prolongation, despite the chromaticism of the

pitch material.

These octave progressions have an enclosed and stable harmonic character

which is not typical of Carter's later music. Indeed, the most elaborate example of

this kind of progression may be regarded as parodic or ironic in effect. This is the

statement of the fugue subject from the second movement of the Piano Sonata (bb. 103-

112, see Ex.6. 12). The initial presentation of this theme is a fuintly comic moment,

self-consciously proclaiming the beginning of a fugue with mock-academic seriousness.

Its slower and stricter tempo, apparent metrical regularity and diatonic and tonal

clarity contrast with the nervy fragments of the passage from which it emerges. The

length of the subject is another contributory factor to its character; a two octave ascent

and descent is achieved through a rather labyrinthine series of overlapping

arpeggiations and descents. (In fact, this passage demonstrates the use of all the

features described in this chapter; fifth-relationships, neighbour notes, linear motion

and arpeggiation). The effect is compounded by the literal repetition of the whole

subject as a "real answer in the dominant" in bars 112-121. However, the rhythm of

the subject and its harmonic relationship with its countersubject conspire to undermine

its pretended conservatism. The suggestion of hemiola and the irregular accentuation

in bars 106 and 108 produce a subtly shifting sense of metre, which contradicts the

155
surface regularity of the 6/8 metre. In tonal terms, the subject may be analyzed along

the Saizerian lines suggested in Ex.6.12, that is as an example of prolongation through

means of slightly extended tonal voice-leading. However, the putative harmonic

structure set out here is flatly contradicted by the combination of "answer" and

"countersubject" in bars 112-121.

2. Middleground and Background.

It could be said that in Carter's transitional works there is not "a

middleground", but that several middleground elements exist. These may form

patterns of association having variable internal consistency and structural strength. If

this is so, then the concept of a background is at even greater risk. Even for works

which begin and end "in the same key", such as the Pastoral and Warble for Lilac-

time, the deduction of a meaningful background is a dubious enterprise. This is so

not only for the reason that the succession of tonal areas visited may be resistant to a

reductive analysis, which attempts to subordinate some and privilege others, but also

because the kinds of harmonic language employed may vary so much within a work,

as has been shown in chapters 4 and 5. It is therefore impossible to form an

impression of the "background" of a work without consideration of issues such as

pitch-fields and set vocabulary, factors which are not easily represented in the

conventional Schenkerian graph, which shows only harmonic and voice-leading

relationships. Indeed, in some of the later works of the period, such as part ifi of the

First Quartet and the third movement of the Quartet Sonata, it is rhythmic and metrical

rather than harmonic factors which govern large-scale form. A fuller discussion of

156
"middlegrounds" and "background? will be reserved for the analytical chapters on

complete movements which conclude this study. For the present, some important

general issues will be discussed through the use of a familiar passage.

In principle, the relationships discussed at the foreground level may all play a

part in the deeper structural levels. In practice, functional tonal relationships and

prolongations are rare, association and allusion being more usual, and the number of

different types of structural connection which can be genuinely heard is reduced. The

strongest factors in determining the structural importance of pitches are generally

register and metrical placement, a fact which is most clearly seen in the construction

of large-scale neighbour-note and linear motions. The second movement of the Piano

Sonata (bb.l-52) will once again furnish a suitable example (Ex.6.13). Previous

paragraphs have discussed the relationship between the five-octave d in bar 1 and the

five-octave e in bar 24. The association formed through metrical placement and

register is reinforced by the allusion to tonal function, both the d and the e' acting as

quasi-dominants. The recurrence of the d-b dyad in bar 52 forms another association

with the first bar and may be taken as the end point of a large-scale neighbour-note

motion.

Within this paragraph, prolongation and association are both brought into play.

During bars 1 to 15, the continuing presence of d in the upper voice, frequently in

conjunction with its neighbour note e, may lead us to regard this pitch as being

prolonged during this passage. The d in the bass, recurring in bars 3, 6 and 13, has a

similar relationship with its lower neighbour c, although the much more extreme

changes of register in the bass disrupt a sense of continuous prolongation. Between

bars 14 and 26, the development of clearly audible linear relationships carries the ear

157
from the prolonged d to the er', thus creating a "progression" whose continuity

appears logical. The first half of the paragraph (1-26) thus exhibits a strong

resemblance to tonal procedures of prolongation and progression.

In the balancing section (26-52), similar methods are brought into play. E'

persists in the bass and middle register in bars 26-32, while the diatonic orientation

shifts around this pivot. Between bars 33 and 36, in the highest and middle registers,

e' and its neighbour d' alternate in an ostinato in association with gi', a pattern

which suggests Gt' major, with dI as scale degree and el' as its neighbour note. In

bars 36-41, a linear progression develops in the bass, descending from e' to bL I,
while the e' /dE alternation continues in the upper voices.

In the remainder of the passage, stability and logical continuity are

progressively disrupted as the music is made to bear greater expressive weight. The

clue to this process is given in the expression marks; molto intensarnente in bars 42-3

leads to precipitoso in bar 48 and rubato in bar 49. Rhythmic values decrease from

quavers to triplets in bars 45-6 and semiquavers in bars 48-51, while the underlying

pulse is disrupted, the crotchet motion of bar 47 giving way to dotted quavers and

quavers in bars 4 8-9, syncopation in bar 50 and irregular grouping in bar 51. These

metrical and rhythmic disruptions are matched by the disintegration of linear logic.

The linear motion in the bass of bars 36-41 is continued, but is subjected to

displacements of rhythm and register. Having arrived at bk in bar 41, this pitch is

re-spelled as a and persists from bar 42 to 44. From this point, the descent

continues, moving through g in bar 45, f# in bars 46-7, f in bar 48, e in bar 49, e' in

bar 50 and d in bar 52. However, the audibility of this "progression" is challenged

by the gradual loss, beginning in bar 46, of the distinctive rhythmic profile of this idea

158
and the transfer of the f, e and e to a higher octave. If the bass line" offers a

thread of continuity, the upper voices militate against this, with apparent arpeggiations

giving way to semitonal motions, some of which are displaced through octave transfer,

the last of these semitonal motions being el' -d across the bar line of 51-52. Thus the

continuity of the middleground "progression" from ei to d in the upper voice is lost

and is only recaptured by the allusion to it in the final detail of the foreground of the

passage. Bars 26-52 thus represent a mixture of prolongational and associational

elements. The linear motion in the bass might be interpreted as an octave transfer of

the pitch eI and thus suggest the prolongation of this pitch throughout the passage, but

the absence of such a continuity in the other voices means that we can only infer an

association. Perhaps what is most characteristic of Carter's style in this passage is the

dynamic relationship which exists between principles of prolongation and association;

our expectations of one mode of continuity are challenged by his resort to another.

159
 CHAFFER 7: THLMES, MOTIVES, ROWS.

1. Introduction

The variable relationships which exist between keys and fields and between

chords and sets have parallels in Carter's approach to thematic material. One might

speak of a transition to "athematicism" corresponding with that to "atonality", but the

term "athematic", originally an abusive expression', is of dubious usefulness,

suggesting as it does the absence of meaningful content. The question of what could

or could not be "thematic" in music underwent intense scrutiny in the early years of

the century in the works of the Second Viennese School and their contemporaries,

leaving their successors a much wider array of possibilities than they themselves had

inherited. It would, therefore, be more accurate to state that Carter's view of what

might constitute thematic material underwent considerable changes during the

transitional period. Identification and analysis of thematic material is inextricably

linked with consideration of its treatment and, therefore, with issues relating to musical

form. Thus, although the focus of this chapter will be on content, it is inescapable

that much of it will be devoted to Carter's approach to form, and hence it will

(together with the later parts of Chapter 6) mediate between the consideration of small-

scale features in Part 2 and the analyses of complete movements in Part 3.

Carter's approach to thematicism is multi-faceted. On one hand, we may see

patterns of statement, repetition and variation, which have their origin in nineteenth

century music. On the other, there are movements built primarily from processes of

See [Samson 1993: 1851

160
 continual transformation which are more akin to the experimental "stream of

consciousness" prose of Joyce or the Surrealists. Passages of literal or transposed

recapitulation are occasionally encountered, as are transformative or synthesizing

passages, which attempt to demonstrate or "explain" thematic resemblances, and

passages in which an idea previously only tentatively proffered is allowed to blossom

and dominate. (Schiff calls these "epiphanies", borrowing Joyce's term [Schiff 1983:

37-8].) Melodic lines which unfold in a seemingly spontaneous, unstructured manner

coexist with quasi-serial passages whose pitch-sequences exactly reproduce those heard

elsewhere in the work. Passages of apparently neutral thematic significance may be

found, where rhythm or texture are the focus of interest; in certain extreme cases (the

third and seventh of the Eibt Etudes and a Fantas, the musical material is pared

down to a single chord or note so that Kianglarbenmelodie is the focus of the piece.

There are "themes", "motives", "tone-rows" and - perhaps the most forward-looking

aspect of all in the context of Carter's later development- passages where individual

intervals appear to take on thematic significance2.

The context in which this multiplicity was engendered was shaped by the work

of earlier twentieth-century composers, particularly Schoenberg. "Theme" is as much

an aesthetic concept as a technical one, so consideration of it ought to begin with a

brief examination of what it meant to Carter's forebears. Schoenberg sought to make

his music "thematic" through and through in the sense that to be "thematic" is to carry

2
See Whittall's reference to "interval-class thematicism" in Carter's later works
.#hittall, 19871

161
 expressive meaning3 . As Dahihaus notes, "in Schoenberg, the terms 'theme', 'basic

shape' and 'idea' tend to overlap" [Dahlhaus 1987: 128], and are impossible to define

rigidly, since, in the composer's writings, the significance of these categories extend

"from the initial motif of a movement, through the method of mediating between the

various shapes, to the form as a whole" [: 129]. The imprecision of this terminology

is, in Dahihaus's view, inevitable, given the great variety of compositional contexts in

which it is put to use. He does, however, give in passing a useful summary of the

possible interpretations of the term "basic shape":

[Its] meaning can vary between an actual theme defined in all its
parameters, an abstract interval structure and a still more abstract basic
pattern reaching back behind the intervals to mere outline and
expression. [: 1321

Furthermore, Dahihaus clarifies the fact that probably the most significant of these

meanings for Schoenberg was the intervals or complexes of intervals, which provided

the "true substance" or "content" of the music [: 130-13 1]. What is clear, is that in

analyses of the works he most admired (his own as well as those of his predecessors4)

and in composition, Schoenberg was driven to demonstrate the operation of "musical

This interpretation relies on Dahihaus's reading of the Wagnerian concept of

"unending melody" and Schoenberg's concept of "musical prose", as expressed in his
essays "Issues in Composition" [Dahihaus 1980: 40-78] and "The Obbligato Recitative"
[Dahlhaus 1987: 144-148]. To summarize, Dahihaus defines Wagner's use of the term
"melody" as indicating that which is musically eloquent and meaningful, as opposed to
formulaic and inexpressive, in other words, that which is thematic, rather than non-
thematic. "Unending melody" is thus music which maintains an unbroken communicative
purpose through the continual deployment of recognizable motives. Schoenberg's own
interest in "musical prose", in which asymmetrical phrasing arises from the excision or
elision of the repetitions and "padding" normally employed to create symmetrical phrases,
is clearly descended from Wagner's theory and practice.

See, for example, the analyses of his own works in "Composition with Twelve
Tones" [Schoenberg 1975: 214-245] and of music by Brahms in "Brahms the Progressive"
[:429-4411.

162
 logic" which derived all material from an underlying basic shape through the process

of "developing variation". Put at its most direct:

Whatever happens in a piece of music is nothing but the endless

reshaping of a basic shape. Or in other words, there is nothing in a
piece of music but what comes from the theme, springs from it and can
be traced back to it; to put it still more severely, nothing but the theme
itself. [Schoenberg 1975: 290]

The technical consequences of this view were of paramount importance for the

development of modern music. In his desire to pursue the Brahmsian principle of

motivic unity and the Wagnerian one of "unending melody" to their logical

conclusions, Schoenberg, in his Expressionist works, created a musical style in which

redundant repetitions and cadential and linking passages were banished, and even

"accompaniments" were thematic. An extreme point in this idiom is the fifth of the

Opus 16 Pieces for Orchestra, Das Obligate Rezitativ (1909), in which a single

melodic line and its accompanying counterpoints are generated from ever-evolving

motivic shapes without recapitulation or interruption. The development of serialism,

which Schoenberg felt to be the logical outcome of this tendency, guaranteed

underlying unity through the continuous presence of the "genetic fingerprint" of the

series, while enabling the creation of a satisfactory degree of surface variety.5

The approach of the mature Carter to musical form and content would be

unthinkable without the example of the Expressionist music of Schoenberg and his

pupils. There are three fundamental ways in which this influence can be seen in

Paradoxically, however, Schoenberg's "discovery" of the serial principle after 1912

oincided with his decision to recover and recreate the earlier formal principles of
hematic statement, development and restatement, especially as manifested in sonata form,
.n aim which, in retrospect, aligns him with the neoclassical tendency in European music.
ihus, the works of the inter-war years display a complex interaction between traditional
nd modern approaches to thematicism.

163
Carter's music; (i) in the practice of creating a pre-compositional resource (be it a

tone-row or a vocabulary of chords and/or rhythms), from which thematic material can

be drawn during actual composition; (ii) in the desire to make all parts of the musical

fabric continuously "thematic" in the aesthetic sense of "significant" and the technical

sense of "derived from a common source", which is manifested in Carter's marked

preference for polyphonic textures in which all parts employ thoroughgoing motivic or

intervallic working; (iii) in the treatment of thematic material as a fluid substance,

subject to continuous change and relying more on "developing variation" than on literal

repetition.

However, as has already been noted in Chapter 2, Carter's development

towards this "neo-Expressionist" outlook was a matter of re-discovering his earlier

musical preoccupations after a period of self-imposed subjection to the aesthetic of

neoclassicism. In its most transparent guise, the neoclassical style involved a return,

at least on the surface, to traditional manners of thematic presentation (for example,

division of texture into "tune and accompaniment") and formal differentiation (for

example, structural division into "themes", "bridge-passages", "episodes" etc.). In

other words, it involved a recreation of the traditional distinction between the

characters and functions of "thematic" and "non-thematic" music. Although rarely

employing such obvious methods, Carter's works of the early 1940s lean rather more

towards this mode of articulation than do his later works. The first movement of the

Piano Sonata has a close relationship with sonata form, while many other works

employ variants of ternary or rondo structure and are to some extent comprehensible

in terms of the structural functions mentioned above.

A further important issue to consider is the gradual weakening of the semantic

164
 bond which makes "theme" synonymous with "melody". In tonal music, according to

a leading figure in the cognitive psychology of music,

Melody.. .represents the level of the greatest differentiation..., the level

at which our evaluative and critical faculties are most immediately
engaged. It is the aspect of music which is nearest to the 'surface',
and that which, for most listeners, most immediately characterizes the
music. [Sloboda: 52]

This statement may be used to support Dahihaus's theory that, in an effort to

make their music more explicitly communicative, nineteenth-century composers made

the melodic dimension the principal carrier of expression. "Theme" thus usurped

tonality as the determining factor of musical form. [Dahihaus 1980: 78]

Furthermore, as the function of harmony in post-Wagnerian music tended towards

increasingly enhanced expressive power at the local level, it became divorced from its

former bond with melody and became motivic in its own right 6. Melody no longer

contained inherent tonal implications. Instead, ever more elaborate and sophisticated

harmonies could be created from the contrapuntal combination of melodic motives7,

while these harmonies in turn could be "arpeggiated" into melodic forms. This

situation gave rise to the Schoenbergian concept of the motive, which could appear,

and carry meaning, both vertically and horizontally. The thematic function of "key-

chords" in Carter's music will already be apparent from discussion in Chapter 5, and

this function is particularly clear when the harmony is associated with a specific

register, spacing and instrumental scoring. The use of individual harmonic intervals

as thematic entities, a common feature in Carter's later music, is also occasionally

6
Dahihaus cites the Wagnerian leitmotif of "Day" in Tristan and the "Mystic Chord"
Parsifal. [Dahihaus 1980: 78]

Dahihaus gives the "Tristan" chord as an example of a harmony created through the
eraction of the melodic motives of "yearning" and "suffering". [Dahlhaus 1979: 611

165
anticipated, for example in the prominence of the minor third in many passages in the

Piano Sonata (see Chap.8).

As well as this harmonic dimension to thematicism, the early part of the

twentieth century also saw the occasional elevation to thematic status of other musical

features, particularly rhythm and texture. Rhythm may be seen to have an enhanced

function in works by Stravinsky, such as lie Sacre du Printemps (1913) and Les Abces

(1914-17), as well as acting in a motivic fashion (as Hauptrhythmus) in many works

by Berg, but it is really in the output of Carter's experimental American predecessors

that it takes on a primaiy constructive and thematic role, the most obvious examples of

this being works written for percussion ensemble, such as Cage's First Construction in

Metal (1939). The enterprise of "total senalism" in the post-war period represents

another step in the attempt to equalize the importance of all musical "parameters", the

domain of rhythm being treated as a repertoire of durational values (see Babbitt's

Three Coinposftions for Piano (1948), Messiaen's Mode de valeurs et d'intensitds

(1949) and Boulez's Structures Ia (1952)). Carter eschews this serial approach but

uses rhythm thematically and systematically in the sense that distinctive rhythmic

features may be assigned to different parts of the musical texture. These features may

be specific metronomic pulses, particular polyrhythmic combinations or other ways of

behaving, such as acceleration and deceleration ISchiff 1983: 24-341.

The importance of texture in post-tonal music can hardly be overestimated.

Between the extreme pointillism of Cage's Music of Changes (1951) and the vast,

dense clusters of Ligeti's Atmospheres (1961), not to mention the huge array of

electronically produced or modified sounds, a range of textural possibilities has been

opened up in the twentieth century which defies description. Carter's mature works

166
make wide and dramatic use of these possibilities. (See, for example, the Concerto

for Orchestra or A Symphony of Three Orchestras.) However, along with this

explosion of resources came a desire to promote texture to the level of a form-bearing

medium. Debussy is generally credited with taking the first steps in this direction in

works such as the sets of Est1vnpes (1903) and Images for Piano (1905, 1907-8) and

Orchestra (1906-11), in which the distinctive instrumental colours employed are at

least as important as the melody and harmony they are associated with. Schoenberg's

well-known prophecy (made in 1911 in his Harmonielehre) of a Klangfarbenmelodie

with a logic of its own [Schoenberg 1978: 421-21, is another example of early

recognition of the formal and thematic potential of timbre and texture. The third of

his Op. 16 orchestral pieces, Farben, served as a model for many later composers in its

drastic reduction of melodic, harmonic and rhythmic material, which enables the ear to

focus on nuances of tone-colour. Carter's sensitivity to the potential of timbre and

texture is clear from his writings [see Edwards: 67-76, 98-103], and from his

compositions from the Piano Sonata onwards, though it is equally clear that he regards

texture as one element among the many that contribute to the unfolding of musical

form:

I find that I am rapidly bored with music that is entirely "textural" in its
construction [...] The "textural" effect by itself ceases to be surprising
after the first hearing, because it is immediately clear that it does not
contribute to any but a very primitive and simple-minded point-to-point
continuity of sound moments... [Edwards: 76] What began to interest
me was the possibility of a texture in which, say, massive vertical
sounds would be entirely composed of simultaneous elements having a
direct and individual horizontal relation to the whole progress or history
of the piece [: 1001

In the discussion that follows, the terms "exposition", "development" and

"recapitulation" are used to map out the issues involved in examining thematicism.

167
 However, the invocation of this traditional vocabulary is intended to suggest starting

points rather than to dictate a traditional approach to the subject.

2. Exposition

Paradoxically, given its status as the initiator of the thematic process, this is

perhaps the most intractable area to investigate. Implicit in a discussion of techniques

of thematic exposition is the assumption that it is possible to discuss a theme as an

entity, separately from its manner of presentation, and yet our recognition of the theme

depends upon this presentation. Thematic exposition in the "Common Practice"

period depends upon making the theme distinct from its surrounding context (a) by

differentiating it from the accompanying texture and (b) by creating a degree of

structural closure and separation from preceding and/or succeeding material. Thus

one expects a degree of melodic continuity in a "theme", coupled with a clear sense of

metrically organized phrasing and a harmonic structure articulated through cadences.

All of these features came under attack during the Romantic and early Modern

periods; continuity became more fragmentary, melodic and harmonic closure was

rendered ambiguous, leaving thematic statements open ended, and "quadratic phrasing"

was abandoned in favour of "musical prose". To adopt Schoenberg's terminology,

the forms of presentation of a musical idea - the "period" [Schoenberg 1970: 25] and

"sentence" [: 20] - were broken down, leaving the content - the "motives" [: 8-91 - to

be presented in an apparently direct way, ungoverned by any previously existing

notions of balance and symmetry. Even more radically, the nature of the content

might actually change: Schoenberg's motive, a "rhythmicized succession of notes" [:

168
9], could thus be replaced by a more abstract idea, a pitch-cell or a timbral

combination for example, thus making constructs such as the sentence and period

redundant, based as they are upon outmoded concepts of phrase rhythm and tonal

harmony.

The examples which follow demonstrate the variety of Carter's approaches to

thematic exposition, from the relatively conservative to the relatively radical. Carter

tends to avoid obvious repetitions, sequences and symmetries, even in his neoclassical

phase. However, a number of thematic presentations occurring in earlier works

reveal the vestiges of traditional methods of phrase-construction.

(a) The Rose Family, bb.3-15

The opening vocal melody of The Rose Family (Ex.7. 1) approaches the

sentence type of thematic construction, in that the opening two-bar phrase - marked

"a" - is immediately repeated, and then subjected to development, which alters some

aspects of its pitch and rhythmic content while maintaining its basic contours. The

underlying sequential nature of this development can be seen in the line of pitches

traced by taking the downbeats of each phrase from bar 8 onwards (e -<IL' -ci' -b b).

As this sequence proceeds, it accelerates, condensing the phrase lengths in the manner

described by Schoenberg as "liquidation"[: 581, leading to a climax produced by

extending the initial arpeggio motive ("x"). The final drop of an octave, being

unrelated motivically to the rest of the passage, may be regarded as the "melodic

residue" [: 63].

169
(b) Piano Sonata, second movement, bb.1-26

The regularity of metre and simplicity of motivic content and treatment in the

foregoing example, while not entirely typical, provide a useful model for examining

more complex thematic statements, for example, those at the beginning of each

movement of the Piano Sonata. The second movement (see Ex.7.2) begins with a

paragraph of twenty-six bars which may be regarded as an extended sentence-type

structure. The analysis of the passage breaks it down into three broad sections, the

first (b.1-15) basically expository, the second (bb.16-20) developmental, being

characterized by an overall ascent and dynamic crescendo, and the climactic final

section, which consists of a severely distorted version of the opening material,

followed by a falling octave "residue" reminiscent of the previous example.

Within this macro-structure, the phrase structure is rather more complex.

Three basic phrase types are identified, a, (3 and 'y. a is the most clearly defined of

the three, consisting as it does of a metrically regular repetition of a simple

combination of motives. However, after its initial appearance, which creates the

expectation of continued regularity, it plays no further part until the very end, where

the resumption of this regularity is a clear indicator of closure. Phrase types (3 and y

are much more fluid and unstable, in terms of metre, length and motivic content.

The basic contour of the first (3 phrase is clearly repeated in bars 13-15 at the end of

the expository section, thus giving these bars the function of a mini-recapitulation.

The intervening phrases, however, subject this shape to considerable transformation.

Phrase i32 maintains the overall fall and rise, but the length of the phrase is extended

from 9 to 12 crotchets and the pitch content is more varied as new transpositions of

170
motives (c), (d) and (e) are combined. By ending as it began, with the rising major

second of motive (c), it suggests instability and invites continuation. Phrase j33 is

drastically condensed and more urgent, being essentially a reworking of the pitches of

the right hand part in the last two bars of phrase 2. The shedding of motive (d),

which had begun the previous phrases, gives phrase j32 a rising overall contour, thus

making it the most distantly related to the original phrase (II.

The -y phrases introduce new motives (the rising major third (g) and the falling

perfect fifth (hD, making them significantly different from the a and /3 phrases,

although they continue to make use of recognizable motives from these phrases. The

evolutionary nature of this part of the passage may be seen from the radical

transformation of the motivic elements of yl in 'y2. In effect, the continuity of

texture (an homogenous four-voiced polyphony), of pitch (the prominent b and f#) and

of the process of shift of diatonic field extending through yl and 'y2 are as important,

if not more so, than the motivic content in linking these two phrases.

The climax of the passage in bars 2 1-26 is a good example of Carter's ability

to suggest formal balance through the recall of motivic elements without compromising

the dynamic, developmental nature of his style through redundant recapitulations.

The overall contour of j35 clearly relates the phrase to its predecessors, but the

individual motivic components have undergone intervallic distortions; (c) has shrunk

from a major to a minor second, and (e) has changed from a fulling major sixth and

rising perfect fifth to a falling minor seventh and rising perfect fourth. in summary,

while clearly-defined formal divisions and functions can be discerned within this

thematic statement, the tendency in Carter's themes towards evolutionary growth and

transformation rather than symmetrical balance tends to blur this articulated structure

171
into one continuous process of change. Two more examples will demonstrate more

complex and radical manifestations of this tendency.

(c) Piano Sonata, first movement, bb.1-32

The opening of the first movement of the Piano Sonata (see Ex.7.3) presents a

dense complex of motives, some of which are identified primarily by texture and

rhythm rather than pitch. Attempts to create a phrase structure based on balanced

repetitions are eventually swept away by a stream of continual transformations. The

motivic relationships themselves are worth commenting on; although one of the

obvious characteristics of the passage is the number of different types of thematic

material it encompasses, the variety of connections and transformations between them

are remarkable. The principal motivic ideas are (a) the rising octave first heard in the

bass (and mdeed the actual harmonic sonority of the octave), (b) the rising semitone,

(c) the pattern of rising perfect fourths or fifths, (d) the falling major second and (e)

the rhythm I' J. which links motives (a) and (d). In addition, there are two other

more complex ideas which are labelled (x) and (y). (x) is the composite motive

created by combining motive (b) with the inversion of (a) in the upper melodic line of

bars 2-4, and (y) the motive appearing in the bass in bars 20-22, which is adumbrated

throughout the preceding passage, being derived from motive (c) through its initial

rising perfect fourth and semiquaver anacrusis.

These materials are initially presented in such a way as to suggest the opening

of a Schoenbergian sentence; there is a "tonic form" of a phrase, al, which consists

essentially of motives (a) and (x), followed by what resembles a "dominant form",

172
a2. Processes of transformation are already at work in the second phrase; motive (a)

is drastically condensed to create the rhythmic profile (e), motive (x) appears in

rhythmic diminution and a balancing statement of motive (c), which seemed to appear

almost parenthetically in cr1, is missing.

The phrases labelled /3, which follow, are characterized by the use of motive

(c) at the beginning and the prominent use of motive (d), particularly in conjunction

with the rhythm of (e). The first, /31, employs a greatly augmented version of motive

(c), making it more akin to the expressive style of the phrase immediately preceding it.

/32 is basically a condensation of /31, returning motive (c) to its scurrying semiquaver

pace and giving more emphasis to motive (d) through the stark octave scoring.

However, the third of these phrases, /33, embarks on a completely unexpected

development, using the semiquaver motion of motive (c) to extend the phrase into a

welter of rapidly mutating variants of the motives presented so far. Thus, motive (c),

which had initially appeared almost ornamental, has become a much more dynamic

force. The rapid melodic ascent in bars 14-15 is followed by a more gradual descent

over the next four bars, abruptly truncated by an accented chord and a sudden

simplification of the texture, allowing a statement of motive (y) to be heard clearly in

the bass. The process of liquidation of motives arising from the development of (c) is

thus interrupted by the forceful presentation of yet another important theme. This

itself is then liquidated into the dyad a#-c#.

The continuation further compounds the complexity of thematic relationships.

There is a return to the maestoso tempo and another variant of phrase a, suggesting

that a further paragraph, balancing bars 1-13, is about to begin. However, this

appearance of the maestoso is much condensed and is still penetrated by additional

173
developments of the scorrevole ideas. In phrase a3 the contour of motive (x) is

altered to emphasize the dyad a#-c#, thus forming a link with the end of the previous

passage. There is also a "codetta" to the phrase in bars 27-29 which attempts to

integrate the florid scorre wile development of motive (c) with the macstoso motives (d)

and (x). In bar 30, another a phrase begins, with motives (a) and (x) presented

simultaneously. This time, however, the notes of motive (x) are drawn out and made

to combine harmonically; note also the reappearance of the melodic semitone (b) in the

left hand part in bars 30-31. This phrase has an ambiguous effect; the rhythmic

augmentation suggesting closure, while the harmonic density and dissonance suggests

the need for resolution. The answer comes in the form of an interruption again, with

the scorrevole rhythm and texture presented in a starker, clearer form than in bars 15-

19, this leading into the main body of the movement. The balance between the

relative importance of the macstoso and scorre wile material has thus been completely

reversed.

The key ideas this crucial passage are the variety and fluidity of thematic

material and the ambiguity of phrase structure used to present it. In the following

example, the thematic material is even more volatile, so that it is virtually impossible

to speak of a "phrase structure" at all.

(d) Quartet Sonata, second movement, bb.69-99

The opening of the Quartet Sonata's second movement (see Ex.7.4) is an

intriguing study in modes of continuity and discontinuity. No two phrases are entirely

alike in terms of thematic material, texture, length or rhythmic character, thus giving

174
the passage an almost improvisatory character. The thematic material varies from the

rhythmic and timbral inflection of a single pitch to long sequences of pitches, while the

texture grows progressively more complex from the single note to polyphony and the

rhythmic character proceeds from the virtually ametric to regular pulses.

An attempt has been made to label the phrases in the manner used in the

previous examples, but as the motivic material proliferates - especially after the fourth

phrase - it becomes almost meaningless ta try and differentiate degrees of similarity

and difference between the phrases; each one is a new beginning. Indeed, the choice

of end-point for the extract is almost arbitrary, since the process of thematic mutation

continues throughout the movement.

The movement begins in a manner similar to no. VII of the Eight Etudes and a

Fantasy, with a passage of Kianglarbenmelodie on the note g. The dynamic swell in

the flute part in the first bar may be regarded not merely as a "colouring" effect but as

a symbol or generator of the dynamic tension of the whole passage, and indeed the

whole movement. The syncopated rhythmic pattern of attacks in the following bar

( I ci. ) created by the entries of the oboe and cello becomes a motive in its own
right when it is taken up and extended by the harpsichord in the next bar. The

registration of the harpsichord further reinforces this motivic significance, with the

"more intense" manual H taking over the oboe's role, while the "more relaxed"

manual I emulates the fragile sound of the cello harmonic. This combination of ideas,

namely the sequence of timbres - flute, oboe, cello harmonic - and the syncopation,

recurs in bars 81-82. As well as creating a thematic motive, the idea of

Klangfarbenmelodie suggests the basic formal processes of the passage, which are, on

the one hand, the passing of ideas between instruments, and on the other, the splitting

175
of the ensemble into fragments, just as the opening note is split into different timbres.

This can be seen on the small scale in bars 72-3, where the new motive of the rising

semitone in the harpsichord is echoed by oboe and flute, with successive notes actually

being played by different instruments. On the larger scale, a process of drawing apart

into two groups can be observed, as the statements of the harpsichord and the

remaining trio become increasingly independent of one another from bar 75 onward.

The process of thematic transformation begins almost at once; the single

sustained note in bar 69 is turned into a repeated note in the next two bars arid then

into a rising semitone in bars 71-72, the latter motive retaining the syncopated rhythm

of the previous bars. A rising minor third is then appended to the rising semitone

(see flute and oboe, bb.72-73) to form a three note motive. The notes of this motive

are subjected to rhythmic variation and permutation of their order in bars 74-76 in the

right hand of the harpsichord part and the oboe and cello. At the same time, in

transposed form and in triplet crotchet rhythm, it forms the basis of the left hand part

of the harpsichord in bar 74. The latter is inverted in bars 76-77, and this inverted

form of the motive is subjected to rhythmic augmentation into minims in the oboe part

in bars 78-80. Meanwhile, the rising minor third, initially an "appendage", begins to

take on an independent existence. It is used in inversion to create a new melodic

shape in the oboe in bar 75 and the pattern of two descending minor thirds (c#-a#-g),

formed as the oboe passes the line to the cello, is then rhythmically augmented and

"stretched" in the right hand of the harpsichord (c#-a#-e#) in bars 76-78. The last

interval of the latter phrase is stretched further as it is echoed in the left hand part in

bars 77-78 (d#-g#). The melodic pitches at the end of this last harpsichord passage

are then transformed into a harmony, with small changes of register and chromatic

176
alteration of one pitch, as the other instruments re-enter in bar 78.

Further transformations of the minor third motive many be seen from bar 80

onwards. The flute has two ascending minor thirds in bars 80-81, inverting and

augmenting the pattern of bars 75-76. In bar 82 a significant new motive is heard in

the oboe, consisting of a rising minor third followed by a rising major seventh, the

rhythm of three equal quavers being derived from the harpsichord and flute motives of

bar 75. This motive is then developed in the right hand of the harpsichord part in

bars 83-86, at first in crotchet rhythm and then in syncopated minims, before being

taken up by the oboe again in bar 86, after which point it is loosely imitated by the

other instruments, its rhythm being reduced to semiquavers and the size of its

constituent intervals again being altered. At the end of bar 88, the minor third (in

particular the pitches e and g, which have been prominent at the beginning of each

phrase since bar 80) appears as an harmonic interval. Again, this is subjected to a

brief Klanglàrbenmelodie treatment, as it passes from flute and oboe, to double-

stopped cello and then becomes the uppermost interval of the dense chordal flourish of

the harpsichord's new marcato, maestoso motive.

The obvious conclusion to be drawn from the above narrative is that the

process being described is a continual evolutionary development, in which thematic

material never arrives at a "definitive" form but is always in a state of becoming

something new. Traditional methods of exposition (particularly the Schoenbergian

sentence) usually include developmental elements, but they play a subordinate role to

the establishment of definitive forms of the principal themes. It is therefore clear that

we have crossed the boundary into the next category to be addressed in this chapter.

177
3. Development

Carter's ingenuity in the matter of thematic transformation and "disguise" is

one of the most fascinating aspects of the transitional music. Apart from the

traditional devices of transposition, inversion and rhythmic augmentation and

diminution, Carter developed a significant number of other techniques, including

intervallic alteration, polyrhythmic combination and changes of speed.

(a) Interval expansion and contraction.

This technique involves retaining the rhythmic outline and basic melodic

contour of a motive, while altering the actual size of the melodic intervals used, and is

a relatively traditional idea, related to Schoenberg's concept of developing variation.

An interesting feature of interval alteration is the effect it has upon the pc set structure

of motives. Numerous examples can be found throughout the transitional period, with

the first String Quartet containing the most thoroughgoing exploitation of this

technique. Both movements of the Piano Sonata, the second and fourth of the Cello

Sonata, Part ifi of the String Quartet and the first movement of the Quartet Sonata will

provide examples.

(i) Piano Sonata, first movement

The principal melodic idea of the "second subject" in this movement (bb.83-4)

is dominated by the intervals of the major second and perfect fourth, emphasizing its

178
 diatonic nature. As the theme progresses, it is developed through the alteration of

these characteristic intervals, coupled with rhythmic changes; the major seconds

contract to semitones and the perfect fourths to major thirds in bars 85-6 and 87-8 (see

Ex.7.5). This produces a more angular and chromatic profile than the original. In

both of the subsequent reappearances of this idea, its intervals are subjected to similar

alterations. Bars 156-60 reproduce the pattern of bars 84-5, which starts with the

original form of the phrase and continues with the altered version, producing an

increase in tension. At the climax of the development section, there is a further

recollection of this idea (bb.205-13) in condensed form and subjected to extreme

rhythmic distortion as the texture fragments.

(ii) Piano Sonata, second movement

Example 7.6 illustrates various forms of interval alteration from the second

movement of the Piano Sonata. Ex.7.6(a) shows the generation of the fugue subject

from the fragmentaiy motives of the misterioso passage (bb.76-103). This is a

relatively straightforward example; the opening interval of a rising minor third (b.76),

which has been prominent in the preceding section (see bb.70-73), becomes a perfect

fifth when the fugue subject itself is announced.

In Ex.7.6(b), the effect of increased harmonic tension is similar to that in the

first movement. The example compares the melodic motive of bars 1-2 with its

altered forms in bars 362-3 and 381-2. In the original form, the two falling minor

thirds are separated by a major second, forming a diatonic set, 4-10. In bars 362-3,

the major second contracts to a semitone so that the set formed is the non-diatonic 4-3.

179
 This alteration not only produces the harsher dissonance of a minor ninth, but the

semitones between b and a# and between d and c# create a leading-note effect which

suggests b as a tonal centre, conflicting with the g or d of the opening. In 381-2, the

level of dissonance introduced in 362-3 is maintained and developed. The pattern of

falling intervals is complicated by the tonal disjunction between the bass and upper

voices; the g} in the bass conflicts with the f# in the upper voice in bar 381 so that

these form intervals of a major and a minor third with the d#. The following bar

presents a major third and perfect fourth. This alteration produces an effect of

tonicizing the note f#.

Ex.7.6(c) illustrates the varied career of the phrase that immediately follows the

one just discussed (bb.3-5). Towards the end of the first paragraph of the movement,

this idea is "stretched" both intervallically and rhythmically, so that the resemblance of

bars 21-3 to bars 3-5 is less obvious. Later in the movement (from b.362), this

phrase is subjected to a variety of transformations as it is used to link the work's main

melodic motives and passes through various harmonic contexts.

(lii) Cello Sonata, second and fourth movements

Example 7.7 illustrates the various shapes taken by the motive first heard in

bars 61-3 of the second movement. The opening falling and rising semitone is

maintained, but the falling minor third is expanded to a major third or augmented

fourth, or is inverted. The motive itself may be extended sequentially or, in bars

204-7, may be used in contrapuntal combination with itself in altered forms.

Example 7.8 shows a brief passage of motivic development in the fourth

180
 movement. Here, both intervals and rhythmic values are augmented, while the

concluding rising semitone of the motive remains unchanged.

(iv) First String Quartet, Part Ill

Interval expansion becomes particularly important towards the end of this work,

where various thematic ideas are drawn together to prepare for the cyclic return of the

opening material (see Exx.7.9 and 7.10). At bar 463, the first violin has a faster

version of the theme first heard at bar 281. The rhythm of this new version recalls

that of the "Threes and Twos" 8 idea (see first violin, bb.13-16). The viola then

expands the intervals of this idea from seconds and thirds to fifths (b.465), from which

the next step is to transform the idea into the theme first fully stated in bars 431-6. In

bars 469-72, the cello phrase is related to the preceding material by the final interval,

a rising minor third. Example 7.10 shows how the two halves of this phrase are

isolated and their intervals expanded before the original cello phrase is shown to be an

intervallic contraction of the phrase which opened the work (see viola, b.481 and

violin I, b.483).

(v) Quartet Sonata, first movement

Example 7.11 shows some of the many transformations of the wedge-shaped

gesture which begins the movement in the harpsichord part and which dominates the

opening paragraph. As can be seen, the degree of variation in pitch is quite wide, the

8
Schiff's term [Schiff 1983: 1611

181
 contour and rhythmic features being the factors which maintain the identity of the

motive.

(b) Polyrhythinic combination.

Carter's technique of polyrhythm has been discussed thoroughly elsewhere

[Bernard 1988]. During the transitional period, he experimented with a particular

variation of this, deriving from Joseph Schillinger's theory of rhythmic "interference

patterns" 9. Independent melodic lines moving in simple pulse ratios such as 2:3 or

3:5 are played within the same pitch range, producing a composite rhythmic and

melodic contour. The First Quartet, Part ifi and Quartet Sonata, third movement

both exploit this device.

(i) first String Quartet, Part ifi

Carter uses polyrhythmic combination in this work to transform one idea into

another or to suggest thematic relationships. Much play is made from the

combination of melodic lines moving in the pulse ratios 3:5, 4:5 and 3:4:5. The

composite melodic contours formed from these combinations occur elsewhere as single

melodic lines; thus one thematic idea may be shown to be the result of a "cross-

fertilization" of two or three others. Some of the most important instances of this

technique are shown in Example 7.12. At bar 82, the violins combine in the pulse

See Joseph Schillinger, The Schillinger System of Musical Composition (New York:
Carl Fisher, 1946), Carter, "Fallacy of the Mechanistic Approach [The Schillinger
System]", Modern Music 25,3 (Summer 1946) and CEL: 15-16 and [Harvey: 27].

182
ratio 5:3, the second violin having a variant of the "Turn" idea (which plays a part in

most of these polyrhythmic combinations) and the first violin another recurrent motive.

The composite contour produced forms part of the "Gigue" idea, heard as a single

melodic line in the first violin in bars 78-80 and in the viola in 248-9. At bar 205,

the violins combine again, now in the ratio 4:5. Each has a line of rising thirds,

derived from previously heard material (see violin I, b. 141, cello b. 190 etc.). The

composite contour is the "Five against Four" idea as it appears in bar 181-4 (violin I).

The full combination of 3:4:5 is heard at bar 289. The second violin line is once

more derived from the "Turn" idea, while the composite contour is a chain of rising

and falling thirds which echoes the "Threes and Twos" idea and has already been

heard as a single melodic line (see violin I, bs.263-5, cello, bs.266-7 and viola,

bs.267-9).

(ii) Quartet Sonata, third movement

The Quartet Sonata demonstrates a simpler use of the technique of rhythmic

interference patterns (see Ex.7. 13). In the final movement, the characteristic dotted

rhythm of the opening is reproduced by the composite contour of the flute and oboe

parts in bars 332-45.

(c) Acceleration and Deceleration.

The presentation of themes at different speeds is by no means a new idea, but

in Carter's hands, it is developed into a formal principle in its own right. While in

183
later works, such as the Piano Concerto, particular metronomic speeds are associated

with "abstract" pitch material (the twelve possible trichords, in the case of the Piano

Concerto), in earlier works, the presentation of a distinctive "theme" at different

speeds draws the listener's attention to the actual process of tempo modulation.

(i) First String Quartet, Part III

The major transformational technique of Part m of the Quartet is acceleration;

the main thematic ideas undergo drastic alterations of character as a result of the

gradually increasing tempo brought about by the almost constant process of "metric

modulation". Acceleration is used not merely as a transforming device, but also as a

structural principle. Table 7.1 shows the successive appearances of what Schiff terms

the "Passacaglia" idea; on each occurrence it is played at a faster speed than the

previous one. The idea is thus gradually transformed from a pulse-pattern at the

beginning, to a quasi-melodic phrase in bar 231, to an impassioned flourish in bar 404

and finally to a flickering tremolo in bar 456 (see Ex.7. 14). This process assumes

great structural significance as it extends over almost the whole of Part ifi and

embraces such a wide variety of speeds.

Table 7.2 shows an analogous pattern, working on a smaller scale, but also

extremely important. At bar 199 the cello begins unobtrusively to state a new idea; a

series of twenty notes meandering up and down in scalic patterns and articulated in

even pulses (see Ex.7. 15). On each successive appearance, this idea proceeds at a

faster speed and with greater prominence until at bar 334, it transforms into the

rushing scalic figures of the coda to Part II, which is then partially recapitulated (Part

184
ifi, b.334 = Part II, b.170). The accelerative process of Part ifi is thus wedded to

the cyclic process of the whole work.

Table 7.1 Acceleration of "Passacaglia" idea in String Quartet no.1, Part ifi

Bar Instrument Pulse unit MM speed

2 cello 24
42 cello 25.6
70 cello i. L 48
115 viola, cello i 71 57.6
140 violin 1 60
179 viola, cello 1 72
215 viola dl 90
231 violin 2 Ii 120
309 violin 2 dl 182.25
345 violin 2 r 216
382 violin 2, viola, cello 288
404 violin 1 300
408 violin 1, violin 2 384
452 violin 1, viola 504
456 viola 1008

(ii) Vañalions for O,rhestm

The processes described above are further developed in Carter's Variations for

Orchestra (1953-5) [see Schiff, 1983: 177-91. In this case, two "themes" take part in

complementary processes of acceleration and deceleration. The first to be heard

(b.26) is a twelve-note "row" whose seven appearances are at successively slower

speeds, ranging between MM 405 and MM 24. The other (first heard beginning in

b.42), is a quasi-scalic idea similar to that found in the Quartet. It too is played

seven times, gradually accelerating from MM 18 to MM 640. These appearances —

185
 Table 7.2 Acceleration of scalic "row" in String Quartet no.1, Part ifi

Bar Instrument Pulse unit MM speed

199 cello I 45
221 violin 1 r 77 60
245 cello 90
283 viola I'. 108
300 violin 1 162
304 cello r
'-.3 243
307 viola F
L.)
364.5
324 cello f
I-'
486
328 violin 1 648

which lead Schiff to describe the ideas as "ritornelli" - are spaced over the entire

length of the work, often overlapping so that "the criss-crossed trajectories of the

ritornelli outline the large formal shape of the work, which moves from extreme

opposition to neutralization to renewed contrast." [: 178] It is interesting to note that

the scalic idea consistently moves in equal pulse units, much as that in the Quartet,

while the twelve-note idea, whose intervallic content is more varied, is subjected to

rhythmic alterations, particularly a form of "stretching" of certain durations (see

Ex.7. 16).

(d) Quasi-senal procedures.

The examples in the preceding section, particularly the last, are closely linked

with those appearing below. Under examination here are those passages in Carter's

music which rely upon the recurrence of quite lengthy sequences of pitches akin to the

linear presentation of a series or note-row. Such passages hover between

conventional notions of thematic recapitulation and a "primitive" serial approach to

186
thematicism, which treats the row as a kind of theme, whose identity may be modified

through rhythmic manipulation, but is nevertheless discernible through preservation of

the sequence of intervals or pitches. The term "quasi-serial" in this context thus

applies to those techniques which preserve the intervals of a pitch row but alter the

rhythmic presentation. The classic serial operations of transposition, inversion and

retrograde are included here, but Carter's "rows" are not necessarily twelve notes in

length nor do they exhaust the total chromatic range. It should also be stressed that

Carter's use of quasi-serial techniques is confined almost exclusively to the horizontal

dimension and is never an important organizing force at more than the local level. In

other words, his rows are presented linearly rather than harmonically and they are

usually just one thread in the midst of a texture whose harmonic characteristics are

governed by the forces described in previous chapters.

(i) Piano Sonata, second movement, bb.104-329

The contradiction between the "implicit" harmony and voice-leading of Carter's

fugue subject and those of its contrapuntal context has been discussed above (see

pp.155-6 and Ex.6.12). Some of the harmonic organization of the fugue may in fact

be explained in quasi-serial terms. The subject itself is treated as a "line" of pitch

classes whose recurrence guarantees coherence in a complex contrapuntal texture.

Carter condenses this "row" by treating only the first and last few notes of the subject.

This is then subjected to various rhythmic variations. Some of the resulting

transformations are set out in Example 7.17. Bars 169-189 demonstrate the

combination of these quasi-rows in a way which suggests something of both the

187
 conventional idea of stretto and the motivic saturation of serial music (see Ex.7. 18).

Elsewhere in the fugue, as has been pointed out by David Schiff [Schiff 1983:

129-1301, Carter employs this quasi-serial technique to deliberately ironic effect.

Between bars 209 and 226, a 21-element pitch class row is combined with a 40-

element rhythmic row derived from the rhythm of the fugue subject. Thus the quirky

white-note episode a Ia Copland turns out to be the product of a sophisticated rhythmic

device relating to the combination of "color" and "talea" in the medieval isorhythmic

motet (see Ex.7.19).

(ii) Cello Sonata, first movement

Carter's adaptation of fligal technique in the Piano Sonata looks forward to a

recurrent feature of the works of the later 1940s and early 1950s. Example 7.20

illustrates the appearances of a "thematic row" in the first movement of the Cello

Sonata. The final statement has an obvious effect of recapitulation, while the third

statement, which is the only one to be transposed, suffers the greatest rhythmic

alteration and is further disguised by the insertion of "extra" pitches.

(iii) Quartet Sonata, third movement

Example 7.21 demonstrates an interesting variation on this technique from the

final movement of the Quartet Sonata. Between bars 349 and 364, there is a partial

recapitulation of material from the beginning of the movement (bb.205-216). The

melodic line is subjected to transposition, changes of instrumentation, insertion of new

188
passages and, in 354-7, rhythmic alteration.

The examples from the First String Quartet (Exx.7.14-15) and the Variations

for Orchestra (Ex.7. 16) employ a similar technique. However, from the Double

Concerto (1961) onwards, Carter abandoned this fundamentally empirical device as he

evolved a more systematic approach to pitch-organisation.

4. Recapitulation.

The proportion of a sonata-type work accounted for by exact repetition may be

taken as a simple index of its progressiveness or conservatism. Forms which, in the

classical era, were dependent upon thematic recapitulation gave way in the nineteenth

century to those dependent upon thematic transformation. The principle of cyclic

form, in which the main thematic material returns at various points in transformed

guises, arose in conjunction with this development. Schiffs observation that

"Carter's forms depend on patterns of motion rather than the repetition of fixed

events" [Schiff 1983: 50] is true of his later music, but the traditional concepts of

thematic transformation and recall inform much of Carter's work during the

transitional period, although in his hands, transformations are often more radical and

recollections more fleeting than in nineteenth-century works.

The issue of transposition in relation to recapitulation highlights an important

aspect of Carter's stylistic transition. In most of the earlier works, Carter

recapitulates material at its original pitch in a manner which affirms his allegiance to

conventional notions of tonality and thematicism. Some early works make significant

use of transposition in their pattern of recapitulation. The Rose Family recapitulates

189
its opening material at t' after the first climax (bb. 14-17) to symbolize the semantic

"distance" from the opening. In the first movement of the Piano Sonata, the "second

subject" is recapitulated at t6 (compare bb. 102-108 with 265-270) and is therefore

based on f# rather than the original c, thus bringing this material within the orbit of

the tonic, b, in an interesting reworking of the conventional notion of sonata form

recapitulation. The final movement of the Cello Sonata recalls material from the end

of the first movement at t' (IV, 166-171 = I, 117-122). This has the effect of

shifting the prominent semitonal clashes of a# against b of the original to b and c in

the transposed version. Since the movement is eventually to come to rest on an

octave c, this would appear to be an adaptation of the idea of resolution through

recapitulation in the home key, despite the fact that, in this work, the sense of key is

severely challenged. In some of the later music, transposition becomes rather less

significant, even arbitrary, since notions of pitch-centricity, let alone key, are

weakened at the local level and abandoned on the larger scale. At the end of the first

String Quartet (Part ifi, bb.483-), the most likely reason for the transposition of the

opening material at t7 is to facilitate its performance on the violin (the first appearance

of this material was on the cello).

Recapitulation fulfils the same rhetorical and formal purposes in Carter's works

as in those of earlier composers. The most basic use of recapitulation is to create

balance after a passage of contrasting material within a clear-cut sectional form.

Simple ABA forms can be found throughout Carter's earlier output. Clear examples

are provided by Dust of Snow and Warble for Lilac-time, in each of which, there is a

contrasting middle section, differentiated from the outer sections by tempo and texture.

Some larger scale movements demonstrate similar basic patterns, albeit carried through

190
with far greater complexity. These include the second movement of the Piano Sonata

and the first two movements of the Cello Sonata. Each of these examples contains

some important degree of modification. The recapitulation in Dust of Snow is very

brief, that in Warble begins at t'° before returning to the original pitch and then veers

into new developments of the highly plastic main material. That in the first

movement of the Cello Sonata is subject to a change of register, while those in the

second movement of the Piano Sonata and the second movement of the Cello Sonata

are considerably reworked, with some passages being condensed, while others are

expanded.

A second approach to recapitulation is to treat it as a triumphant return of

material in a climactic release of tension. This approach is most closely associated

with sonata form, especially in the later nineteenth century; the end of the development

section is often fused with the beginning of the recapitulation, so that the tonal and

thematic processes of the development are oriented towards the goal of thematic

return. Such an approach can be seen in the first movement of the Piano Sonata

(bb.243-251). This passage exhibits a rhetoric which belongs to the late Romantic

tradition and which Carter was later to eschew; the build-up of tension produced by a

steadily rising linear process in fast tempo leads to a massive chordal climax in which

the sense of pulse is arrested; the recapitulated material dwells on repeated gestures

and is subjected to registral displacements which emphasize the "dramatic" gesture of

the octave leaps present in the original theme. Since this is Carter's only real

engagement with sonata form in his chamber music, this is a somewhat isolated

example, but it is characteristic of the work as a whole, which relies more than most

of Carter's subsequent works on repetition of large sections.

191
 Carter's most characteristic use of recapitulation is in a kind of ritornello form,

in which each reappearance of the main material serves as a springboard for new

development. Such is the case in Voyage and the Pastoral. In Voyage, the main

idea recurs at the same pitch throughout, whereas the Pastoral employs a variety of

transpositions and transformations. The last two movements of the Cello Sonata and

all movements of the Quartet Sonata and the first String Quartet employ developments

of this kind of form, which tends towards continuous process rather than sectional

juxtaposition. The second movement of the Quartet Sonata is built from a collage-

type sequence of relatively brief ideas, some of which recur in very similar form while

others are continually altered.

Finally in this section, Carter's handling of cyclic form will be examined.

Carter often attempts to create the impression that the work is returning to its origins,

creating a circular pattern (see Schiff 1983: 47-48). This is sometimes combined with

the idea or illusion that we have never left the origins' 0. The Piano Sonata, Cello

Sonata and first String Quartet all employ some variant of this device. In the former,

there is a coda which synthesizes elements of themes from the second movement with

those from the maestoso passages of the first movement and gradually liquidates their

characteristic features. The Cello Sonata and String Quartet share a process of

heightening tension (principally carried out through acceleration), culminating in the

recall of earlier material, at first forcefully, but gradually liquidating and fading.

10
See Carter's reference to the influence of Cocteau's film Le Sang d'un poete.
[CEL 233] and also Schiffs mention of Joyce's Finnegans Wake [Schiff 1983: 47].

192
PART THREE: ANALYTICAL STUDIES

193
CHAPTER 8: PIANO SONATA, FIRST MOVEMENT

The Piano Sonata is widely regarded as the first of Carter's works to deserve

inclusion in his mature oeuvre, despite its basically neoclassical idiom. It is certainly

clear that the composer himself thinks of it as a turning point:

I [...] have been interested in pursuing the possibilities of dramatic contrast and
interplay offered by the individual character of instruments and have attempted
in all my works, at least since my Piano Sonata, to exploit these possibilities in
the most vivid ways I could imagine. [Edwards: 681

By 1945... I became more concerned with the formation of musical ideas, of

types and qualities of continuity, and with the fascinating possibilities of
musical flow and change. The opening of my Piano Sonata of 1946 will
illustrate what I mean. It is the first passage in my works that is not primarily
thematic. Its central idea comes from the total sound of the piano writing.
Notice particularly the variety and flexibility of rhythm, the frequent changes of
character, the oppositions of register, of manners of playing, and of slow and
fast. All of these were to become increasingly important. [Radio Broadcast
"The Composer's Choices" c.1960 and CEL 213]

Other commentators on the work have elaborated on the Sonata's pivotal role,

the prevailing argument being that while it adheres to the harmonic and rhythmic

language of Carter's earlier period, some aspects of its form and the treatment of the

instrument adumbrate the composer's later practices. Schiff states that the Sonata

"looks in two directions" [Schiff 1983: 123]. Harvey maintains a similar view: on

one hand, the Sonata contains for the first time in Carter's music "the direct

association of the musical materials of a work with instrumental characteristics and the

performing situation", and on the other, it represents "Carter's last extended essay in

musical Americanism" [Harvey: 19].

The obvious "backward-looking" features of the work are its allegiance (at least

on the surface) to traditional tonal forms (sonata in the first movement, fugue in the

194
second), its thematicism (including much use of cyclic quotation) and the retention of

key-signatures (although these do not indicate a conventional approach to tonality).

The most obvious "forward-looking" element is the extreme rhythmic fluidity of the

scorrewle material, which prompted Carter, under the influence of his editor Kurt

Stone, to abandon time signatures in the first movement. However, there are aspects

of Carter's technique in this piece which are connected more fundamentally with his

later music. These include the formalized tempo relationship between the maestoso

and scorrewle of the first movement (sharing a common pulse of minim = 66), the

derivation of harmonic materials from the basic acoustic properties of the instrument

and the exploitation of the relationship between musical materials of opposing

character.

This chapter will begin with a consideration of the influences on Carter's

Sonata, particularly its place in relation to the "American" idiom, before proceeding to

an analysis of the first movement's structure and language in the light of the issues

addressed in Chapters 4 to 7.

1. Populism and other influences

The style and technique of the Sonata seem at times to invoke Copland and
Debussy. The opening phrases of both movements suggest Copland's Piano
Sonata which predates Carter's by four years. Carter also cites, in passing,
Copland's Piano Variations: these echoes I believe were intended as a homage.
Carter's Sonata is written in the tradition of Copland's piano music I...] but it
transforms Copland's gestures into a recognizably new and personal idiom with
a greater variety of moods and textures, and a greater contrapuntal density.
[Schiff 1983: 124]

Schiffs observations on the relationship between the Piano Sonatas of Copland

and Carter raise an important question; the extent to which Carter on the one hand

195
 adopts the musical language of American populism as a natural means of expression

and, on the other, challenges and undermines it from within. Carter's uneasy

relationship with Populism has been discussed earlier (see pp.36-43). Evidently there

was a conflict between his personal belief in the artistic value of sophistication and

complexity and a politically motivated commitment to simplicity and immediacy of

communication. This conflict can be seen to have consequences for the musical

substance of the Sonata. The elements of the musical language which Carter inherited

in the 1940s have their origin in Copland's style. However, the uses to which they

are put in Carter's Sonata occasionally transform them in ways which conflict with the

aesthetics of populism and neoclassicism. A comparison between the Piano Sonatas of

these two composers will show the debt of Carter to Copland and also the degree to

which Carter turned the populist idiom of the older composer to his own ends.

For both composers, the genre of chamber music presented an opportunity to

develop the newly created musical idiom on an intimate rather than public scale.

However, the similarity in motivation ends here. Copland's Piano Sonata and Violin

Sonata (1942-3) are relatively isolated chamber works within a phase dominated by

works aimed at the larger audience, stretching from El Salon Mexico (1936) to the

Third Symphony (1946). For Copland, the scaling down of musical language from

the concert works and ballets to the Sonatas was successfully achieved, producing

works which are among his most perfectly balanced and lucid. Carter, however, was

drawn to chamber genres largely through dissatisfaction with the state of American

orchestral music and its audience. Chamber music offered the possibility of writing

works of greater complexity, whose performance would not be compromised by the

difficulty of rehearsing larger groups and whose reception would not be doomed from

196
the outset by the stubborn conservatism of the public. Between 1945 and 1953 he

wrote only one orchestral work, the ballet The Minotaur (1947).

The Sonatas by Carter and Copland share many elements of musical language,

both general and specific; a small selection will be discussed in order to highlight the

similarities and differences of approach. Harmonically, both works employ the

unconventional diatonicism that is a hallmark of neoclassicism. The characteristic

features are:

(1) the creation of harmonic tension from the potential for "false relations"
within the system of diatonic fields described in Chapter 4;

(2) the avoidance of conventional tonal relations and pure triads except at points
of closure, where they are referential rather than functional;

(3) the employment of certain "Stravinskian" pitch configurations such as the

superimposition of tonic and dominant triads or other juxtaposed triads;

(4) the use of consonant intervals - especially perfect fourths and minor thirds -
moving in parallel in stepwise motion in a manner which parodies conventional
tonal voice-leading.

Rhythmically, similar ideas can be detected in the two works. Regularity and

symmetry are played off against their opposites. Beats, bars and phrases are expanded

or contracted to produce an ambiguous or fluctuating sense of metre. Ostinatos are

set up only to provide a foil to the prevailing sense of fluidity and elasticity.

Formally, there are some interesting similarities. Both composers employ

cyclic themes and cross references. Copland's final movement grows from the germ

of material first heard in the "trio" of the scherzo and culminates in a return of the

main material of the opening movement. The coda of Carter's Sonata mingles

material from both movements. A structural similarity can be discerned between the

first movement of Copland' s work and the second of Carter's. In each, moderately

197
paced outer sections flank a faster central section, the final section being a condensed

recapitulation of the first. The outer sections of both movements contrast

predominantly chordal main material with more lyrical subsidiary material.

Probably the single most important factor which distinguishes Carter's work

from Copland's is the approach to formal continuity. Copland, attaching himself

very closely to Stravinskian neoclassicism, favours the juxtaposition of sharply

contrasted blocks within which harmonic motion is relatively static. Carter, in

contrast, proceeds through sweeping evolutionary arcs, during the course of which

musical matter is continuously transformed. It would be wrong to describe either of

these approaches as more radical or genuinely "modern", since both have auspicious

precedents dating back to the works of the acknowledged masters of early twentieth

century music. It would also be wrong to ascribe one manner of composition

exclusively to either composer. What the comparison does show is Carter's

impatience with music in which, in his words, "first you do this for a while, and then

you do that" and his desire to "mix up this and that" [CEL 229].

In general, the differences between Copland's Sonata and Carter's from this

perspective are those of degree. Carter's work is more contrapuntal, rhythmically

more fluid, more varied in texture and harmony and therefore more complex and less

accessible to "popular" understanding. A comparison at this level suggests merely

that Carter sought to enrich the idiom in which he had chosen to express himself.

However, other factors suggest a more ambivalent attitude towards American

populism, one which viewed it as flawed by naivety and therefore susceptible to ironic

treatment which undermines its implicit aim of straightforward communication. It is

in the area of thematic typology and its relation to the elements of musical vocabulary

198
that this ironic strain is most clearly revealed, particularly in the form of parody.

The thematic typology of American populist music derives to a certain extent

from native American music, whether jazz or popular marches and hymn tunes. The

cowboy tunes, Latin American dances and Quaker hymns which find their way into

Copland's music have no place in Carter's. (The example of Ives, whose borrowing

of popular material appeared to Carter to be uncritical and simplistic, probably

dissuaded him from literal quotation of this kind.) However, many Copland

fingerprints can be detected in Carter' music. Copland's style, particularly during the

populist phase, depended on the cultivation of instantly recognizable thematic "types,

almost comparable with the system of rhetorical figures of a much earlier period.

Some of the principal rhetorical modes of Copland have closely corresponding

counterparts in Carter's music of this period and are perhaps the most obvious (and

superficial) illustrations of the influence of the older composer on the younger. One

characteristic "type", for example, is the nervous, fragmentary style, in a fast tempo

but metrically irregular and/or using rubato. (See Ex. 8.1, the opening of the

second movement of Copland's Sonata and bb.76-103 of the second movement of

Carter's.)

A relatively straightforward example of Carter's adaptation of Copland's

rhetoric may be obtained by comparing the triumphant recapitulations of the opening

material in the first movements of the two Piano Sonatas (Ex.8.2 Copland 1:206,

Carter 1:243). Both passages employ massive octave displacements of melodic

material which was previously restricted in range, combined with harmonic

intensification. These passages both exemplify a type of grandiose declamation which

is a common rhetorical type of American populism, but it will be noticed that the

199
degree of transformation in the Carter is more extreme.

Within this same rhetorical type, another comparison shows Carter in

subversive mood. The "second subject" material in the first movements of both

Sonatas is subjected to similar apotheosizing treatment shortly after its first statement.

In both works, this takes the form of massive downward-sweeping gestures, gradually

expanding in length and mainly employing stepwise motion (see Ex.8.3(a) Copland

1:123 and (b) Carter 1:102.) In the Carter example, however, an ironic tone is

introduced by prefacing this passage with an overblown parody of a perfect cadence,

whose extreme chromaticism creates an exaggerated contrast with the "fervently" pure

diatonicism of the succeeding passage. Despite its surface similarity to the Copland

example, this passage is far more complex contrapuntally, since it takes the form of a

three part canon, with the third part imitating by inversion (and later doubled at the

third). A further irony arises from the cultivation of such learned contrapuntal

sophistication in a passage whose static modal harmony and meandering melodic

motion almost create an effect of aimlessness.

The frame of reference for Carter's Sonata is of course much wider than that

suggested by its conjunction with Copland's in this section. The conflict between b

and b flat, which is pursued throughout the work suggest a point of contact with

Beethoven's Hainmerkiavier Sonata or Liszt's B minor Sonata. The two movement

form, incorporation of material of widely varying tempo and character within single

movements and the use of fugue also suggest a Beethovenian connection (see the

Sonatas Op.lO9, 110 and ill). The context of American piano music in general

should also be considered; in particular, the influence of tonal and rhythmic elements

in the First Sonata of Roger Sessions (1927-30) and textural and contrapuntal elements

200
 in Ives's Concord Sonata (1911-12) on Carter's work should be remarked. This

breadth of reference is another factor which distinguishes Carter's Sonata from

Copland's; while Copland may be more "himself' in his work than Carter is in his,

the greater scope of the latter's imagination is clearly in evidence here and is at odds

with the self-contained clarity of the former.

2. "Sonata Form".

The reinterpretation of sonata and other conventional forms in modern terms

was a crucial item on the compositional agendas of all those who wished to give their

work a legitimizing link with the music of the past, whether or not their aesthetic was

an explicitly neoclassical one. The subject has been discussed at length by J.Straus in

Remaking the Past [Straus 1990: 96-1321. The context of that book is the supposed

"anxiety" of early twentieth century composers in relation to their forebears; its thesis

is that in order to assert their individuality, it was necessary for the later composers to

employ strategies of creative "misreading" of traditional materials, techniques and

structures.

Carter's approach to the sonata principle in this movement is an ambivalent

one, drawing on modernist ideas as well as on nineteenth-century practice. Avoiding

the question of whether this exhibits any "anxiety" on Carter's part, it is nevertheless

clear that two of Straus's observations are applicable in some degree to this movement:

that in order to engage in a struggle with tradition, composers may (1) "subvert the

form's traditional impetus towards reconciliation" (as in the first movement of

Stravinsky's Symphony in C) or (2) "set in motion musical forces that strain against the

201
boundaries of the form" (as in Bartk's Piano Sonata). The latter, in particular,

chimes with Schiff's assertion that in Carter's hands "sonata is no longer a 'form', but

a rushing stream echoing across a vast landscape" [Schiff 1983: 125 ] . At first glance,

the potential for a dynamic conflict of opposites, whether tonal or thematic, offered by

sonata form would appear to make it an attractive vehicle for a composer of Carter's

disposition, but the first movement of the Piano Sonata is the only example of a

"sonata form" movement among his sonatas. Speculation, based on knowledge of his

later development, regarding the reasons for the rarity of Carterian sonata form leads

to two conclusions; (i) that Carter's gradual abandonment of tonality and thematicism

rendered the sonata principle an inappropriate form-building idea and (ii) that the basic

expository procedure of sonata (and many other traditional forms), which is to present

opposed ideas successiely is superseded in the later Carter by techniques of

simultaneous presentation.

The points of contact between conventional sonata form and Carter's are clear;

there is a contrast of thematic material, characterized by opposing expressive

characters, tonal centres and tempi; there are clearly demarcated passages of

exposition, development and recapitulation, following the expected sequence of events;

and there is an organic relationship between the outwardly contrasting thematic

material. The table below summarizes the internal structure of the movement:

EXPOSITION DEVELOPMENT RECAP. CODA

lo 'L I

The divisions on the table may be justified straightforwardly. The junction

202
 between exposition and development is marked by a double bar, a return to the

opening maestoso tempo and a ghostly reappearance of the opening theme. That

between development and recapitulation is more difficult to detect aurally because of

the continuity of tempo and figuration, but the restoration of the key signature of five

sharps at the start of bar 223 and the almost exact repeat of bars 15-18 in bars 222-226

are clear indications of the composer's intention. The coda begins with a resumption

of the scorrewle tempo of the main body of the movement.

Carter's formal innovations lie in the relationships between the tempi of the

thematic materials, in the movement's internal proportions and in the approach to key-

structure. The principal tonal and thematic material of the movement combines slow

(maestoso) and fast (scorremle) music, related through a common pulse. The

maestoso opening of the Sonata initially sounds like a recreation of the classical slow

introduction, but it becomes apparent that this is not to be the case. It reappears at

the opening of the development (bb. 123-133) and at the climax of the recapitulation

(bb.252-264), the latter creating a significant reordering of the presentation of the

movement's main themes when compared with the exposition. The reappearance of

material from a slow introduction at crucial stages during the course of a movement

has many precedents (viz. Beethoven String Quartet Op. 127, Schubert "Great" C

major Symphony, Brahms Symphony no.1); what is unusual here is the close

interpenetration of the two types of material, the scorrewle appearing to grow

seamlessly out of the maestoso and melt back into it.'

The subsidiary material of the movement (which fulfils many of the

'See Chap.7, pp.172-174 for an illustration of this effect in the first thirty-two bars of
movement.

203
conventional functions of a "second subject") unfolds at a moderately paced meno

mosso ( J = c.72). The very imprecision of the metronome mark sets it apart from

the calculated interplay of opposites of the main material, as do the frequent rubato

and ritard./a tempo instructions (see bb.9land 109). There is no attempt to relate this

tempo to those of the maestoso or scorrevole systematically. Instead, the nieno mosso

tempo is confined to two passages (bb.83-122 and 265-270) which are clearly cut off

from the surrounding music by silence or sustained notes, the absence of audible pulse

being the antithesis of the technique of metric modulation Carter was later to employ.

On the face of it, this approach would appear to be a conservative throwback to the

nineteenth-century practice of placing the "feminine" second subject in a more

leisurely tempo, but there are connections between the main and subsidiary material

which suggest a less conventional attitude. Most importantly, the meno mosso

continues the metrical flexibility of the other sections, but with a quaver pulse and

predominantly scalic melodic shapes taking the place of the semiquavers and

arpeggiated shapes of the scorre pvie. Furthermore, there are thematic connections;

the septuplet gruppetti of bars 81 and 297-9 in the maestoso/scorrevolc tempo are

clearly meant to have the same effect as those in bars 108, 112 and 114-5 of the meno

mosso and are presumably to be played at the same speed.

The internal proportions of the movement are not those of the conventional

model, but nevertheless follow a plan which is not entirely unprecedented. Expressed

as percentages of the total performance time of the movement, the durations of its

constituent sections are: exposition 48%, development 26%, recapitulation 17%, coda

9%. Thus, almost half the duration of the movement is given over to exposition,

after which there is a far-ranging, but concentrated, development leading directly into

204
a drastically curtailed recapitulation and a brief coda. As each major section takes

only just over half the playing time of the previous one, the effect is of progressive

concentration of argument and acceleration towards a final goal. This approach is

essentially dynamic and developmental, the antithesis of the static, symmetrical model

suggested by Straus as one of the main modernist misreadings of sonata form [Straus

1990: 132]. The scorre vole material, with its upward thrusting gestures and tendency

to proceed through rising sequences, is the main agent of this dynamism. The meno

inosso, on the other hand, achieves its definitive melodic shape in downward-drifting

phrases (see b.102), before allowing itself to "disintegrate" into meandering and

registrally disconnected phrases and fragments in bars 109-121. The relatively swift

progress of the development is due to its continuous adherence to the scorre vole tempo

and manner. This in turn necessitates the drastic reduction of this material in the

recapitulation in order to avoid redundant repetition. The dynamic process of the

development is allowed to spill over into the recapitulation, so that tension is only

dissipated in the wake of a massive, apotheosis-type restatement of some of the

exposition's thematic material (bb.252-264). After this, there is a very brief recall of

part of the "second subject" material before the coda restores the scorrevole tempo and

the dynamic urgency of the development. While Straus's reference to "musical forces

that strain against the boundaries of the form" and Schiff's to a "rushing stream"

capture the exhilaration of the movement, it should be noted that Carter's approach is

not unique; twentieth century sonata-forms as diverse as the first movements of

Shostakovich' s Fifth Symphony and Ravel's Sona tine employ some variant of this plan.

Harmonically, the movement does not exploit the conventional tonal tension

between tonic and dominant, or tonic and relative minor, but employs a variant of this

205
 idea, with the tonic b, being challenged by its own leading-note a# (respelled as b'

when it asserts its independence from b) during the "first subject" material, while the

"second subject" takes c as a tonal centre. Tone centres a semitone apart thus take

the place of more conventional tonal relations. However, the classical resolution of

tonal tension through the ultimate triumph of the tonic is avoided. While the coda of

the movement includes passages which emphasize b as a melodic and dynamic high

point (bb.288-9 and 293), these are not underpinned by metrical or harmonic closure.

In the closing bars, the tension between b and a# remains unresolved, either through

the use of ambiguous complex harmonies (such as the implied superimposition of tonic

and dominant in b.301) or through the avoidance of harmony altogether (the texture of

the final three bars is reduced to bare octaves). The last two sustained notes, f and bb,

clearly pose a tonal question, implying as they do a usurpation of the expected tonic

resolution.

The use of a non-dominant counter-pole to the tonic is by no means

unprecedented in the late nineteenth and early twentieth centuries, nor is a sense of

tonal ambiguity. However, coupled with the conflict between tonal centres, there is a

conifict between different kinds of harmonic context (diatonic/chromatic,

consonantldissonant), engendered through Carter's manipulation of diatonic fields. In

this respect, his work shows a kinship with works by Debussy, such as "Reflets dans

l'eau", in which key-relationships are less significant than the contrast between

diatonic, whole-tone and pentatonic fields. In order to explore this area more

thoroughly, the harmonic building blocks of the movement must be analyzed.

206
3. Harmonic inateriajs: stnictural consequences

In descriptions of the musical language of the Piano Sonata, much has been

made of the derivation of musical material from the nature of the instrument itself.

Both Schiff and Harvey take as their starting points the interaction between acoustic

and tonal phenomena in the work's harmonic organization. Schiff describes it in

terms of "two extremes and a mediator" fSchiff 1983: 126J, the "extremes" being the

natural resonance of the overtone series and the cancellation of this effect of resonance

by chromatic, semitonal dissonance. The "mediator" is the cycle of fifths, which in

its entirety constitutes a twelve-note set, but when presented partially, generates

diatonic and pentatonic scales. Schiff illustrates his theory by writing out the cycle of

fifths as two overlapping diatonic sets:

B major

I I
E B F# C# G# D# A#=B F C G D A

B major

The argument is summarized with the statement that "the pitches B and A# and the

tonalities of B major and B flat major are the major elements of harmonic organization

in the Sonata." [Schiff 1983: 126]

Harvey, in his brief discussion of the work, develops Schiff's line of thought

and states that "both the characteristic harmony and the elaborate figuration of the first

movement grow from the overtone series of the work's first pitches and their conifict

with fifth-cycles based on the same pitches b and a#, thus developing the texture from

207
 the Particular resonance of the instrument itself." [Harvey: 1 9J

These accounts, while containing some truth, require more careful

qualification. 2 Essentially, the harmonic language of the Sonata derives from

alternative interpretations of a central "Key-chord" or source set. This is presented in

its clearest form not at the opening of the Sonata, but in two passages whose unique,

translucent sonority makes them stand out from the dense textures of the rest of the

work. (See Ex.8.4(a) I, 123-8 and (b) H, 388-92.) At these two points, the thematic

material of the work is literaily generated using the acoustical properties of the

instrument. In Ex.8.4(a), the silent depression of the c 4 and e keys allows these

strings to vibrate sympathetically, generating harmonics when other keys are sharply

struck. In the case of the first sonority, the harmonic et' 6 appears so strongly because

it is present within the audible harmonic spectra of both lower notes, being the third

harmonic of a fundamental at' and the seventh of a fundamental f. The c6 is similarly

well supported, being the third harmonic of a fundamental f and the fifth of a

fundamental at'. The composite sonority (f,ab ,c,eb) may be regarded as the "key-

chord" of the work. Having thus been generated acoustically, it is exploited by

Carter for its intervallic content and possible tonal implications.

In set-theoretic terms, this sonority is the symmetrical tetrachord 4-26. It can

2
Both writers seem to imply that the harmonic series can actually be reproduced on
the modern piano, but since the piano is fixed in equal-temperament tuning, it can only
produce an approximation of the harmonic series, and in practice it is virtually impossible
to detect harmonics other than the first few - the compound octave, perfect fifth, major
third and minor seventh above the fundamental. Schiff's statement (leaving aside its
inappropriate reference to "tonalities") implies that the harmonic series stands in a similar
relation to the cycle of fifths as the chromatic and diatonic scales do; in other words, that
one is equivalent to, or a subset of, the other. This is not, in fact, the case, since the
cycle of fifths in equal temperament is based on an adjustment of the simple frequency
ratios of the harmonic series.

208
be partitioned into complementary pairs of intervals (ic3 and 5) as may be seen by

examining the interval vector [012120]. (See Appendix 3.) This symmetrical pairing

is made explicit in the opening bars of the Sonata (refer back to Ex.6.4(b)). The

single major second (ic2) in the tetrachord is also exploited as a motive in bars 10 and

13. The minor third (ic3) also takes on a highly important thematic role (see

especially bb. 171-1 80 and 197-200 in the development).

Returning to Ex.8.4(a) and taking in the total pitch-content of bs.123-8, it

will be observed that it consists of two alternating versions of 4-26, pivoting around a

shared minor third dyad. The symmetrical structure has thus been extended to form

the diatonic hexachord 6-32, as in Example 8.5. 6-32 is a segment of the cycle of

fifths and of the diatonic scale; both segments are symmetrical structures. In fact the

symmetry of 6-32 can be expressed in a variety of ways, there being three axes of

symmetry (the dyads b-a#, d#-f# and c#-gff). One of these symmetrical arrangements

partitions the collection into two overlapping statements of tetrachord 4-23. (See

Ex.8.5(b).) This tetrachord shares with 4-26 a symmetrical interval structure

including ic5 + ic5. Like 4-26, 4-23 has a pervasive thematic significance, especially

in the form of the arpeggio figure in fourths and fifths which it takes in bar 3 (see

Ex. 8.6).

The interaction between the symmetrical and the diatonic implications of these

sets is an important feature of the work's harmonic organization, and one which

influences our perception of its "tonality" or "atonality". Example 8.7 illustrates how

the set 4-23 is used to generate passages of quite different harmonic character, which

nevertheless share some textural features and a common pitch-centre. (Also see

Exx.4.6 and 6.9(a).) In Ex.8.7(a) the beginning of the "second subject" (bb.83-4) is

209


shown. As was illustrated in Ex.6.9(a), the melodic line is formed from adjacent

linear statements of 4-23, creating a texture in which chromatic saturation is counter-

balanced by the privileging of c as a tonal centre through registral placement and

symmetrical intervallic processes. Ex.8.7(b) shows the second phase of the second

subject (bb. 102-4). The same initial transposition of 4-23 110,0,3,51 results from the

canonic imitation of the falling major second, but here, it is the diatonic rather than the

symmetrical aspect of 4-23 which is emphasized, since the canonic lines are extended

to form segments of the diatonic scale set 7-35. C is again felt as the tonal centre,

because of its emphatic use as a pedal and as the temporal and registral starting point

of the highest melodic voice. This tonicizing of c within a pure diatonic field of A,%

produces a feeling of "Phrygian" modality.

Further elements of the work's harmonic structure may be introduced by

referring again to Ex.8.4. Comparing the total pitch content of Ex.8.4(a) and

Ex.8.4(b), it will be seen that the two passages share a common subset, the pitches a,

e' and b (or g#, d# and a#). The harmonic content of the later passage may be

obtained by "pivoting" that of the earlier passage around the pitch d#/e':

b------------ ------a#
g f#
e------------------
c b
a ' -------
f

This could be seen as a realization of the potential suggested in Schiff's diagram

(reproduced on p.192), in which the cycle of fifths is partitioned into two overlapping

diatonic scales of B major and B flat major, but in the actual context of the work,

there is a more complex relationship between diatonic fields and tonal centres to

210
consider. The pivoting around a central pitch produces the effect of a reorientation of

diatonic field within the cycle of fifths, so that while the earlier passage suggests a

field of Ak/Es, the later one suggests B/F#. However, the situation is complicated

by a contextual sense of tonal bias which unbalances the pure symmetry of this

scheme. In Ex.8.4(a), the dyad c-et' represents the tonal centre of c which has been

in operation for most of the preceding passage (the "second subject" of the first

movement). However, in the music following Ex.8.4(b), the dyad d#-f# comes

eventually to be absorbed into a composite tonic/dominant chord based on b (the final

b-d#-f#-a#-c# sonority of the sonata). Thus the shift of diatonic field between the two

passages, combined with the asymmetrical shift of tone centre produces quite different

scalic structures:

Ex. 8.4(a)

a e b f c g
f# c# g# d# a#

Ex. 8.4(b)

Ex.8.4(a) c [d/d] e' f g a bI'

Ex.8.4(b) b c# d# [e/e#] f# g# a#

The idea of harmonic pivoting around a fixed pitch, which can be reinterpreted

enharmonically, is a crucial one in the sonata. In particular, the reinterpretation of d#

as e during the course of the exposition of the first movement is an important adjunct

to the opposition of tonal centres a semitone apart. This pitch-class forms the third of

a triad, under which b and c alternate as roots. Example 8.8 offers a summary of this

process.

211
 It is also possible to regard Ex.8.8 as a "background" manifestation of set 3-3

[0,1,4]. The large-scale progress of the exposition would thus appear to be related to

the details of the local level, since this set an its relatives are the principal agents in

the chromatic assault on the prevailing diatonic context.

The analysis which follows focuses on a passage from the movement which

exemplifies the harmonic, linear, tonal and thematic processes outlined thus far,

putting them into the formal context of the movement as a whole.

4. Exposition, bb.1-82

This passage, which comprises the exposition of the "first subject" material (the

maestoso and scorrewile) and its subsequent development and "crisis", leading into the

second subject, is undoubtedly the most deftly controlled handling of oppositional

tensions in Carter's early work up to the Cello Sonata. The basic oppositions are:

(i) between the slow pace, low to middle register, sustained tone and

predominantly chordal texture of the maestoso and the rapidity, volatility, high register

and predominantly linear writing of the scorrei.vle. While the fundamental linear

processes of the niaestoso can be heard in operation most clearly in the lowest register,

this is only intermittently present in the scorrewile, which concentrates similar

processes in the uppermost voice;

(ii) between the harmonic clarity of passages which adhere to a particular

diatonic field, and the ambiguity and tension created through chromaticism and non-

diatonic dissonance;

(iii) between the tonal centre of b and those of its semitonal neighbours a# and

212
c (oppositions which entail much play on enharmonic pivots, especially d#/dp , f##Igj

a#/bt and b#1c4);

(iv) between regular and irregular rhythmic grouping;

(v) between different modes of continuity (unbroken process versus abrupt

change);

(vi) between clearly identifiable themes and more neutral material embodying

harmonic and linear processes.

(a) Bars 1-32

This opening section exposes virtually all the oppositions described above.

Perhaps the most fundamental overall tension is that between b and a# as tonal centres,

an opposition which subsumes that between the different tempi. The relationship

between these two pitches is played out over several registers in a manner which

suggests comparison with Schenkerian structural levels. The initial five-octave b

(spread over bar 1 and the beginning of bar 2) sets out the registral space in which the

ensuing maestoso is to evolve. In bars 24-25, this opening gesture is repeated

transposed down a semitone, but with the added complication of g# appearing in the

second and fifth octaves. While the association between these two points is clear, it

would not be appropriate to describe this relationship as a prolonged neighbour note

motion since so much of the intervening scorrewle material, especially in the lower

register, is deliberately discontinuous and tonally ambiguous. However, within the

maestoso passages themselves (bb. 1-14 and 24-32), the harmonic and linear jostling

between the two pitches is presented in a more sustained fashion. Tension arises

213
 from the increasing ambiguity of their relative structural status; which is the structural

tone and which the neighbour note?

(i) Maestosix bb.1-14

The status of the initial b as a tonic is immediately challenged by the harmony

superimposed above it. As was illustrated in Chapter 6 (p.1 37 and Ex.6.4(b)), the

pitches of the first chord in bar 2 may be interpreted as neighbours to an implied B

major chord. Combined with the pitch b, they produce a "dominant seventh" type

sonority, the conventional resolution of which would involve the b descending a

semitone to a#. The actual progression to a more extreme dissonance on the last

minim of bar 2 violates this expectation and further complicates the sense of tonal

function. The c 4 in the left hand is ambiguous in itself, since it suggests two

possible aural interpretations. It could be regarded enharmonically as b#. This would

entail hearing the two chords of bar 2 (excluding the octave b) as a I-V7 neighbour

chord progression in C# major, requiring resolution by raising the b# a semitone to c#.

(Notice that the semitonal link from the c# to the putative b# has been displaced by an

octave.) Alternatively, the ci1 might act as a "Neapolitan" neighbour note to b,

needing to be absorbed back into the tonic by descending a semitone. Carter denies

both these possibilities by proceeding to a "b minor chord" as the cI rises to dl1.

There is also a sense in which this d11 4 is a displaced semitonal descent from the d#5 of

the previous chord, moving in parallel with the falling octave f#, and echoing the c# 5-

c 4 relationship of the previous progression. The direction of these displaced linear

connections is reversed as bar 3 proceeds to bar 4; the semitonal descent of the b 3 and

214
 d4 is thrust up an octave to a#4 and c#5, thus bringing the latter dyad into a register in

which the aural connection with the d#çf#5 of bars 2-3 is clear.

Such a complex matrix of semitonal neighbour relationships - actual, displaced,

implied and often contradictory - is emblematic of the harmonic language of the work

as a whole, but, in order to gain a balanced view of this subject, the interaction

between such quasi-tonal relationships and pitch-class sets and relations must be

investigated. The important thematic role of Type A diatonic sets 4-23 and 4-26 has

been made clear, but the role of non-diatonic sets must also be pointed out, especially

those related to the atonal trichord 3-3 mentioned above (p.217). The first non-

diatonic sonority heard in the work is that containing the "troublesome" ct in bar 2: 4-

18 [O,1,4,7]. While this appears at first as a B major triad infiltrated by a chromatic

"foreign body", it should also be regarded as an embodiment of some of the

movement's important intervallic relationships: it contains two ic3s (as does 4-26),

may be partitioned into ic3+ic5 (like 4-23) and contains the all-important semitone.

It will reappear at moments of harmonic tension as a referential sonority.

In bars 4 to 7 (first minim), registral and harmonic factors give all the

ascendancy. All is presented in four octaves, the middle two of which are sustained

throughout. The rhythmic figure of bar 4 represents a drastic contraction of the

work's opening rising octave. In place of the extreme harmonic ambiguity of bars 1-

3, this passage presents a stable diatonic field of CI, until the f## in the middle voice

of the left hand part takes it into GI. The level of dissonance is reduced, with fewer

semitonal clashes employed, although the linear use of semitones is still an important

The importance of this set in Carter's work in general has already been pointed out
Chapter 5 (see p.116).

215
feature. The only tonal uncertainty is whether a# is to be heard as the leading note of

B major or as the dominant of D# major. The final harmonic twist of this phrase (the

motion from f# to f##) appears to tip the balance in favour of the latter. This possible

interpretation of a/f as V of d# is latent in the progression used in bars 10 and 13, and

is alluded to later in the movement, but is not realized until the last page of the sonata.

(See bb.395-8 of the second movement.) The comparatively straightforward harmonic

character of this phrase when compared with bars 1-4 is underlined by its temporal

contraction of the events of the first phrase from eleven minim beats to six.

B is asserted strongly again in bar 7, with a repetition of the rising octave

figure of bar 4. Now the conflict between b and a# is presented explicitly in the

vertical dimension. and b3 are sustained in the left hand in bars 7-10, supplanting

the a#2 and a#3 of the preceding bars, and the aggressive rising octave is repeated in

bar 9. Against this, the right hand unfolds a melodic line which emphasizes the

octave from alt3 to a#4. The level of dissonance is correspondingly higher, the initial

sonority of this phrase being another non-diatonic relative of 3-3, 4-19 [0,1,4,81.

This dissonance is gradually softened, however, first by the rising motion of f/f/f to g#

in bars 7-8, which is repeated in bar 9, and then by the relinquishing of b in the lower

registers in favour of alt again in bar 10. At the end of the phrase, the use of the

diatonic and relatively consonant 4-26 rooted on a# produces a clear sense of harmonic

relaxation. The augmented repeat of bars 10-11 in bars 12-14 clearly confirms the

arrival of a/f as the local tonal centre, especially through the textural devices of

reducing to bare octaves in bar 12 and recapturing the extreme bass register, which

has not been heard since the first bar.

216
 (Ii) Legato scoirevole bb.14-23

The opening of the scorrewle passage (bb. 14-15) fulfils several functions:

(1) it begins with a repetition and extension of the rising semiquaver figure of

bars 3 and 11, revealing and developing the potential of this brief motive to generate

musical ideas independently of the maestoso material which originally surrounded it.

(2) it presents a third repetition of the a#-a#-g# motive heard in bars 10 and 12-

13, and then uses this as a springboard for further motivic development.

(3) it adumbrates the left hand motive of bars 20-22.

(4) it opens up the higher registers of the keyboard, up to b 6, continuing the

interplay of octaves of the preceding section.

(5) it presents, for the first time in the work, the diatonic field of B, which,

combined with the brief appearance of b1 in bar 15, swings the b/a# argument

temporarily back in the favour of b.

(6) it replaces the regular minim pulse and simple duple/triple alternations of

pulse groupings of the inaestoso with an intensely irregular pattern of groups of

semiquavers.

The types of tonal ambiguity - caused by chromatic alterations, false relations

and enharmonic changes - employed in the scorre wile are similar to those of the

maestoso, but because of the much faster tempo, they barely have time to register.

This, coupled with the frequent disjunctions of line in the left hand part in bars 15-19,

makes a passage such as this harder to hear in terms of harmonic functions and linear

processes than is the first niaestoso. Those processes which can be clearly discerned

are:

217
 (1) a gradual descent in the right hand part in bars 16-19, in two stages; (i)

beginning on e6 and falling twice through ic3 to a# 5; (ii) falling in a linear pattern

through ic2 from a#5 to g#5 and f#5 in bar 17 and thence to e4 in bar 18 and d11 5 in

bar 19. This d may be regarded as transferred down an octave in bar 19 and then

slips down a semitone to c#4 in bar 20. Since the latter note is transferred up an

octave in bar 22, we may retrospectively understand a linear connection from dip in

bar 19 to c#5 in bar 22-3.

(2) a gradual corruption of the diatonic field of B through the introduction of

chromatic notes in the left hand. These tend to occur in pairs a perfect fourth or fifth

apart: b# and e# in bar 17, gij and cL1 in bars 17-18, a11 and d11 in bar 18. The latter

two pairs have the effect of seeming to place the left hand part into the field of C. In

each case, the chromatic note generates a strong semitonal dissonance, the strongest of

which is the chromatic cluster 3-1 occurring at the beginning of bar 19, where d11, c#

and b# clash.

Both processes seem to point to the pitch c# as a resolution; the spelling of the

left hand dyad in bar 19 suggests a symmetrical semitonal motion outwards. The

"expected" c# comes in bar 20, but only in the upper register and only after an abrupt

interruption by a chord which reminds us once again of the b/a# opposition, a rather

"Stravinskian" superimposition of an F# major triad over a bass b. Straightforward

linear and tonal connections are thus challenged again.

Under the sustained c#4 in bars 20-22, a "new" motive is announced in octaves.

In fact, it proves to be organically related to material already heard. Its shape has

been anticipated in the right hand part of bar 15, while the rising fourth aLj -dij clearly

echoes the appearances of this pair of pitches in bars 18 and 19 and the rising whole-

218
tone scale from d11 to a# is an inversion of the principal right hand line of bars 17-19.

The sustained a#4-c#5 dyad of bars 22-23 forms a local association with the same dyads

one and two octaves lower, heard at the beginning of bar 22, and with that an octave

higher heard at the beginning of bar 20. This interplay of octaves, combined with the

explicit recall of the same dyad as that used in bar 4, effect a transition back to the

maestoso.

(iii) A teinw, maectoso bb.24-32

The second macstoso passage brings the b/a# opposition back into focus and

recalls the opening thematic material, both altered by the effect of the intervening

scorrewle, and, in several cases, condensed into harmonic form. During bars 24-27,

the note b is made to appear subordinate to a//, since the latter is sustained in the

middle registers, while there is a displaced neighbour note motion (c#6-b4-c#5) above

this, doubled in parallel thirds (a#5-g#4-a#4). As mentioned above (p.213),, the five

octave a# in bars 24-5 is coloured by g#1 and g#4, this vertical association of g# and a#

being a condensed reminder of the melodic motive of bars 11 and 13. The g# 4 is

itself made to fall a major second in bar 25 and then to rise in a scalic fashion which

parallels the middle voice linear movements of bars 2-10.

The main appearances of the note b in this second macstoso are associated with

reminders of the "Stravinskian" tonic/dominant chord of bar 20, a form of set 4-14

[0,2,3,71. Thus, b is not able to assert itself independently of a#. At the end of bar

25, the same pitch-classes are present in as in bar 20, but the recall is much more

explicit in bar 30, with a repetition of the exact pitches of that bar, the addition of a

219
d# completing the "tonic triad" part of the chord and enlarging the set to 5-27

[0,1,3,5,8]. (This set embraces the important 4-26, as is made clear by the omission

of b from the sonority after the first beat of bar 30.)

Similarly, the note a# appears locked into certain registral and harmonic

patterns, namely its combination with c# in registers which recall its appearances in

bars 20 and 22-3. The a#-c# dyad is presented vertically in bars 25, 27, 29 (left

hand) and 30 and horizontally in bars 28-9.

Although the passage begins with a resumption of diatonic stability in the field

(B/F,#), the same tendency towards chromatic obfuscation seen earlier begins to cloud

the texture, as the scorrew,le material insinuates itself briefly (bb.27-9). This

tendency is seen first in bar 27 with the addition of a d neighbour to the c# of the

semiquaver motive of bar 26. In contradiction of this, the rising chains of perfect

fourths and fifths in bars 27-9 (note again the use of set 4-23) generate the pitches d#,

e# and b#, thrusting the diatonic field sharpwards to C#.

The restoration of b(1 in bar 30 immediately cancels this out, but the progress

from bar 30 to 31 introduces the most complex set of thematic, intervallic and linear

relationships so far. The a#5-c#6 dyad of bar 30 is associated with the e#5 -g#5 of bar

31, producing a form of set 4-26 at t 7 with respect to the parallel passage in bars 2-4.

The f#5 and c#6 are transferred down an octave from bar 30 to 31 in a manner which

clearly recalls bar 6. The notes of the d# 3-f#3 dyad in the left hand proceed to their

chromatic neighbours d1 and gl in a fashion characteristic of so much of the work's

voice-leading thus far and clash against the f#4 and c#5 of the right hand, producing a

form of set 4-8 [0,1,5,61, whose possible partitioning of id + id or ic5 + ic5 symbolize

the intervallic tensions of the music. The left hand d 3 falls to b2 at the end of bar 31,

220
bringing to mmd once more the thematic importance of ic3, but, ironically, this last b

of the passage is unable to establish itself as a tonic here. Despite its position as

"bass note", the sonority above it is so dissonant and self-contradictory in terms of

diatonic orientation that any sense of tonal function is cancelled out.

(b) &onrivJe bb.32.-82

The remainder of the "first subject" will be examined in a less exhaustive

manner, focusing on processes that stretch across the whole passage. The following

topics will be examined: register and linear motion, tonality and diatonic field, and

thematic processes.

(i) Register

Essentially, the scorrevole is constructed as a series of rising waves, each one

cresting with a more powerful climax than the last. Thus, despite the continuing

importance of thematic, harmonic and tonal developments, it is the element of register

which is the most potent shaper of musical form. This foreshadows Carter's approach

in some later works, for example, the Adagio section of the First Quartet.

These "waves" are: (i) bars 32-5, (ii) 36-43, (iii) 44-50, (iv) 51-64 and (v) 65-

79. There is a clear progression in terms of length and complexity from section to

section and also a greater sense of dynamic force at successive climaxes. The first

wave achieves a c#7, played f at the end of a crescendo (b.35). The second does not

reach so high, but has greater sonorous weight, as it climaxes with a three octave b#,

221
 accented at the end of a crescendo from f (b.42).

The third wave is more complex, since the registral, dynamic and tonal

climaxes are slightly displaced from one another. The actual melodic high point is

the d#7 of bar 49, but the dynamic climax (an accented /7), marked by the rising

semitone from d#6 to e6 on the down beat of bar 50, is displaced down an octave.

The avoidance of e7 here makes its appearance at the subsequent climax (b.63) all the

more telling. However, the progress from d# to e is not the end of right hand's

rushing scalic line, whose momentum carries it through to another, more important,

semitonal motion, that from a#5 to b5 (again transferred down an octave). This

melodic climax is the first of the series to be combined in counterpoint with

thematically significant material in the left hand part, which here plays a distorted

version of the a#-a#-g# motive of bars 12-13. (The tonal and harmonic significance

of this will be discussed below.) The fourth climax is prepared by an extension of the

material used in the build-up to the third (compare bb.47-50 and 58-62), taking place

through a five bar continuous crescendo to an accented sif(b.63). Here, for the first

time, the extreme bass and treble registers are combined, as e 7 and e6 in the right hand

coincide briefly with the sustained c 1 and c2 of the left.

The last of these five waves is rather different in construction from the first

four. The established pattern of the previous waves has been a gradual ascent to a

combined registral and dynamic climax, after which there is a rapid descent, but in

bars 65-79, the centre of gravity shifts emphatically to the bottom of the final descent.

The melodic high point of the last wave is the e 7 of bar 74, but this only appears at a

dynamic level of p, after which, there is a gradual crescendo, tornando a! Tempo I and

an increasingly precipitous descent to the accented if triple octave c in the extreme

222
bass register in bar 79.

(ii) Linear motion

An essential characteristic of the scorrevole material is its fluent and constant

metamorphosis between different types of scalic and arpeggiated figuration which form

the content of the textural waves described above. On the local level, these are used

to create simple quasi-tonal voice-leading patterns, for example, the a#-b#-a#

neighbour note figure of bars 33-34, or the octave coupling, subdivided by

arpeggiation through the fifth, of the right hand part in bars 36-39. As was discussed

in Chapter 6, however, attempts to discern similar processes operating on a

"middleground" level are fraught with difficulties unless one accepts the model of an

associational, rather than prolongational middleground. Bars 40-44 offer an

interesting case: a linear process of rising semitones can be heard, linking the notes at

the beginning of each bar - b -b -b#-c# - despite the abrupt disjunctions of register

and diatonic orientation. However, it is questionable whether this progression

represents an "unfolding" of the minor third a#-c# (which was of such importance in

bb.25-30) since the beginning and end of the progression do not appear harmonically

related.

Middleground linear associations in this passage are most obviously formed

through repetitions, allied with contextual factors such as register, harmony and

rhythm. Associations between pitches in the uppermost register are simple enough to

construe. In the lower and middle registers, however, the sense of line is frequently

challenged by registral fragmentation and tonal ambiguity. For example, in bar 51,

223
the d -c#5 melodic cell in the upper voice appears to present the former as the

neighbour of the latter. Alihought the d "resolves locally to c#, a "middleground"

quasi-sequential pattern over the next few bars suggests that the d5 is the initial pitch of

an ascending linear progression. This rises a step to e in bar 52, and thence through

f#, g# and a# to b5 in bar 56, although this overall rising line involves much retracing

of steps. Similar processes unfold in the lower registers, but with progressively less

clarity the lower we look. The c#5 rises half way through bar 52 to d#5, where it

remains until bar 56. Beneath this, there is a great degree of ambiguity: at the bottom

of the right hand part, it is difficult to establish whether the d 4 and f#4 in bar 51 are

part of the same or different contrapuntal voices. The presence of the pitches e4 4 and

g#4 in the corresponding places in the next bar seems to indicate the latter, but the

figuration soon changes, so that the impression of a progression in parallel major

thirds gives way to an irregular rotation of the pitches f# 4 , g#4 and a#4. In the left

hand, the d#4 in bar 51 conificts with the dtj 4 in the right hand, but again, it is difficult

to hear it unequivocally either as a neighbour note to the dL 4 or as a displaced

neighbour to the e#3. The d#4 is repeated twice in bars 51-2 and then seems to rise to

e4, after which, this register is abandoned until it is recalled by the e 4 of bar 54 and

the d#4 of bars 55-6. At the bottom of the left hand part in bars 51-2 enharmonic

notation is used to underline a similar effect of ambiguity of line. The notation e#3,

g(1 3 in bar 51 implies an incompatibility of diatonic field between these two pitches,

whose main function seems to be as destabilizing chromatic neighbours to the f# in

higher registers. However, in bar 52, f9, gt1 3 and alj are assembled into an
ascending line, establishing some coherence in this register, but still conflicting

strongly with the content of higher registers in terms of diatonic field. This register

224
then becomes the locus for a gradually descending process, leading through g#3 in bars

52-3, g 3 in 53-4 and f 3 in 54-5 to e3 and d#3 in 55-6, although the staccato

articulation and frequent interruptions and octave displacements militate against a

straightforward linear interpretation.

To summarize the above description, it will be seen that the texture of the

scorrew)Ie is difficult to reduce to a simpler voice-leading model with a consistent

number of parts and relationships between these parts. A principal voice may be

detected, usually in the highest register, but the rest of the texture consists of isolated

groups of notes which may form themselves into scalic clusters or fragmentary lines,

often conificting rhythmically and harmonically with one another and with the principal

voice.

In the later stages of the passage (from b.51), as the harmonic context becomes

more chromatic and dissonant, the linear process in the uppermost voice becomes the

leading agent of coherence. The most important aspect of this line is the play on the

relationship between d and d#/e, exploiting the uneasy tensions between neighbour

notes which was generated in the movement's first phrase. The progress from d 5 to

b5 in bars 51-6 has already been charted. From thence, a clear continuation of this

line can be discerned, with the principal pitches highlighted through registral extremity

and octave doubling. The line leads through c#6 in bar 56 and d#6 in bars 57-8, this

movement being reproduced an octave higher in bars 59-60. In bar 61, the d#7 is

reinterpreted as et' 7 as there is a shift in diatonic field, and this pitch is brought into

direct conflict with d 516 in the left hand. This conflict continues into bar 62, where

the dissonance between these two pitches is most explicitly presented, the right hand d4

6fl clashing with the left hand d#516. The pressure of this dissonance is relieved by the

225
climactic e11 in the right hand in bar 63. In the rapid collapse of this registral wave

(bb.63-4), d is absent and the ascendancy of e is short lived as it is drawn back to its

lower neighbour d# (see e 4-d#4, left hand, bb.63-4, and d#5-e5-d#5, right hand, b.64).

Within the last of the five registral waves, the enharmonic reinterpretation of

d# as el' in the uppermost textural line continues to be of central importance. Once

the issue is resolved in favour of e, there is a descent, akin to a Schenkerian

Fundamental Line, through d to c. D#/e is established as the focal note in bars 66-

70 through repetition, accent and metrical placement, though it lacks consonant

support. Once again, there is much interplay of neighbour notes; at first, in the

rather static passage from bars 66-68, c#5 is frequently used as lower neighbour to d#5.

Below this, other neighbour relations are heard, mimicking the upper voice (c#4-b3

and a#3 -g#3 in the right hand, while the left hand has the same an octave lower).

There are also more dissonant semitonal neighbour relations, between d 4 and c#4 in

the right hand and b#2 and c#3 in the left, which disturb the diatonic stability of the

passage. With the replacement of d#5 by e' in bar 69, the upper neighbour ft15

comes into play. The insistent repetition of the figure e' 5-f 5 in bars 69-70 is

underlined by its simultaneous inversion c 5-b and leads to the continuation of the

upward movement by the establishment of g 5 in bars 70-71. From this point, the

upward reaching movement becomes more dynamic. G 5 is established as a temporary

point of focus through its repetition at the beginning of bars 71, 72 and 73, each time

supported by c5 and d 4 or dl' . Each repetition is the starting point for an upward

rising motion, reaching progressively higher (a 5 b.71, bI 5 b.72, c6 b.73). In bars

73-4 there is a much more rapid ascent to e , the first time this register has been

sounded since bars 59-63, but as has already been stated, this brief melodic high-point

226
does not serve as the climax of this section. From here the process of descent begins.

The e 7 is immediately transferred down two octaves and reaches d 5 in bar 75.

Octave displacement of the d takes place in both directions in bar 76, but in the Tempo

I (bb.77-8) there is an accelerating drive downwards, each successive descending

figure in the right hand being more compressed (7 semiquavers, then 5, 4 and 3) and

reaching lower until c3 is achieved in bar 79. In the last group of semiquavers in bar

78 there is an encapsulation of the overall descent e 4-c of bars 74-79.

(iii) Tonality and diatonic field

Tonally and harmonically, the overarching processes of bars 32-82 are the

destabilization of diatonic fields on the sharp side of the cycle of fifths, associated with

a tonal centre of b, and their replacement by flat-side diatonic fields, associated with c.

The pivotal role of d#/e' in these processes will already be evident from the preceding

discussion, although other enharmonic relationships, for example f##/g, b#Ick and

a#fb', are also crucial.

Within the first section of the scorreu1e (from bar 33 until the climax in bar

50) there is no non-diatonic dissonance, and so the sense of diatonic orientation, while

not static, is always clear. The section begins with a passage unequivocally in the

diatonic field C# (bb.33-6), resolving the extreme ambiguity of the previous chord,

and then shifts gently to G# in bars 37-9. The frequent repetition of a# and d# tends

to make these notes sound like dominant and tonic, a relationship made especially clear

in the sequentially rising melody of bars 36-9. The fourth repetition of this sequential

phrase in bar 39 leads to the first abrupt shift of field in the passage. The expected

227
e# and a# are respelled as 11 and b, leading to a brief appearance of the field F.

The simple device of repeating this passage up a semitone in bar 41 restores the field

to Fl, while the b# in bar 42 suggests that C# has been confirmed as the goal of this

section.

From this point until bar 50, there is a gradual tendency to shift flatwards.

The gently unwinding semiquavers of bar 43 take us imperceptibly through GMmnd Cl

to F/and in bar 44 the material from bar 36 is repeated transposed down a tone, so

that the music continues in this field. F,revails until the introduction of e in bar 48

causes a shift to B. Although the field of bars 44-50 remains quite stable, the sense

of tonal focus within this passage is ambiguous; f#, g#, c# or d# could all be heard as

tonally significant. Eventually, however, with the shift to the field B and the shaping

of the right hand line into an ascending scalic pattern from bar 48 onwards, the note b

begins to exert an attractive force but, as discussed above, the attainment of this

melodic goal in bar 50 is undermined by other factors. Notably, at this point, the

pitches g 314 are introduced in the left hand, contradicting the prevailing diatonic field

and, through their combination with d# and a#, suggesting a rival tonal centre of d#/eb.

Bars 51-64 are the most tonally unsettled thus far. The g introduced so

forcefully in bar 50 brings in its wake many other chromatic contradictions of the field

B; gI', aLj and f j are all heard in bars 50-55, creating a high level of dissonance.

These accidentals are briefly purged from the texture in bars 56-7, which return to the

field B/Fl. In bars 58-9 there is a return of the material of bars 48-50, including the

prominent g4 (now appearing an octave lower), later respelled as f## taking the music

into the field of G# in bar 59. From bar 60, the diatonic field of the left hand part

228
plunges dramatically flatwards to C, while the right hand part, having attained d#7 in

bar 60, adopts the enharmonic notation e' in bar 61 in order to try and accommodate

with the left hand, thus briefly producing a field of c. As the passage moves towards

the stark confrontation of the pitches d11 and d#/e in bars 61-2, the actual tonal status

of these two notes becomes uncertain. However, in bar 62 g1 returns once more,

this time stretching down across four octaves, and acts as a quasi-dominant to cLi,

which appears for the first time as a harmonic goal in bar 63. The "C major"

sonority of the downbeat of bar 63 is extremely short-lived as the diatonic field of the

right hand part reverts to B, while the left hand Ctj is recast as b# in bar 64.

Bars 65-68 present a situation similar to that of 5 1-55; the harmonic field is

quite static, but full of semitonal tensions, the basic field of B being contradicted by

the insistent presence of d11 4 and b#2. An abrupt change of field is brought about in

bar 68, through the pivotal exchange of d# for e, and is confirmed by the change of

key signature to four flats in bar 71. Although Ab is the prevailing field here, the

frequent incursion of a% and dIj in bars 70-73 unsettle the diatonic stability of the

passage. This destabilization is exacerbated in bars 74-6 as the field is subjected to

flatward changes (ct' 3 and ft 314 appearing in the left hand part) at the same time as the

sharpward alterations become increasingly prominent (dI,516 appearing in the right

hand and g in the left). in the Tempo I (bb.77-79), diatonic stability is almost

completely dissolved at first as the diatonic field plunges kaleidoscopically through the

flatward tending cycle of fifths from F to Pb.

The appearance of ckj in bar 79 as "terra firma" after this chromatic landslide

is almost arbitrary in effect. Although the sonority of the extreme bass register has

already been associated with the quasi-cadential bass g-c in bars 62-3, any connection

229
between that point and bars 78-9 must be regarded as a psychological association

rather than a tonal prolongation. The paradoxical relationship between a weighty

cadential gesture and a chromatically saturated harmonic context which Carter creates

here continues to be a theme of the "second subject" (see previous discussions of

bb. 83-4 and 101-2). Bars 79-82 encapsulate the tonal process of the whole of the

preceding section in several ways. Although they could be read as "I-V in c minor",

the spacing and voicing of the harmonies emphasize other linear and harmonic

relationships. The emphasis on the note b within the "G major" chord and the use of

the three lowest registers in bars 79 and 81 for the notes c and b recalls the

importance of the semitonal neighbour-note relationship between b and a# in the

work's opening page. The absence of g from the "c minor" chord of bar 80 throws

emphasis on e", thus underlining the importance of this note as a pivot in the diatonic

orientation of the preceding passage. Finally, the septuplet flourish in bar 81, which

superimposes two diatonically incompatible ideas (the cycle of fifths segment 4-23 in

the right hand and a G major triad in the left), is emblematic of much of what has

gone before: the clash of chromatic neighbours (bk' -bt, dl' -dIi i -e", at -glj, the first of

these relationships being of greatest significance), the emergence of diatonic stability

out of chromatic chaos, and the thematic importance of set 4-23.

(iv) Thematic processes

The scorre vole between bars 32 and 82 is typical of much of the Sonata in that

its thematic material is subject to continual development of melodic and rhythmic

shape to the point that its characteristics are liquidated into a freely evolving stream.

230
The most important thematic ideas are (i) the repeated note and falling major second,

with its associated lombardic rhythm, first heard in bars 12-13; (ii) an upward

thrusting motive consisting of an anacrusic perfect fourth followed by a whole-tone

scale segment, sometimes concluding with a minor third; (This idea has been heard in

its "definitive" form in bars 20-22, but was presented embryonically in continuous

semiquavers near the opening of the first scorrewle in bar 15. Both versions are

subsequently used.) (iii) the pentatonic motive first heard in bars 36-9, which has

prominent rising fourths in common with (ii) and thiling major seconds in common

with (i). Although (iii) is the first new idea heard in the passage and is ostensibly its

"main theme", the other two ideas play a crucial role at climactic moments. It will

be observed as a general point that all three ideas lose their distinctive shape and are

absorbed into the stream of linear and rhythmic invention as the passage progresses.

The succession of "waves" described previously meshes closely with Carter's

treatment of thematic material. The brief first wave (bb.32-5) begins with a reference

to (ii) as it appeared in bar 15 and links this with the descending major second of (i)

(a#6-g#6 in b.34). Towards the end of bar 35, the irregular rhythmic grouping of

three and five semiquavers anticipates (iii).

Idea (iii) functions as the starting point for the second and third waves. At the

beginning of the second wave, it is presented in a simple two-part counterpoint against

running scales in the left hand. Its boldly confident character is enhanced by the

conventional tonal device of repeating its opening five-note motive transposed

alternately up a fifth and a fourth. However, the rising optimism of this gesture is

somewhat undermined by the gradual diminuendo, which eventually leads to the piano

return of (ii) in the disoriented diatonic field of F in bar 40, a return which flows

231
naturally from the fact that both (ii) and (iii) begin with the same interval. The

immediate repetition of (ii) in a more characteristic field (B) and manner of articulation

(Ibrte inarcato) restores the outgoing character of the passage and leads to the climactic

b#3141 of bar 42, after which, a long tail of semiquavers is formed in bar 43 through

liquidating ideas (ii) and (iii); the first three notes of this "tail" are the concluding

notes of (ii) and this overlaps with notes 3, 4 and 5 of (iii), which is then repeated in

descending sequence, culminating eventually in a simple b major arpeggio.

The third wave begins in a similar fashion to the second, but idea (iii) is almost

immediately subjected to development of its rhythm and contour, leading in bars 46-7

to a repeated downward arpeggio. Under this, attention is drawn towards the left

hand's scalic semiquavers, which begin to form a rising pattern. In the lowest voice

of the texture in bars 46-7 (stems down, left hand), there is a suggestion of the shape

of idea (ii) (d#-g#-a#-b). In bars 48-50, the scalic figures break through to the top of

the texture, pushing the wave on to its climax. The chordal material beneath it in the

left hand makes use of the rhythm and pentatonic content of idea (iii) in bar 49 and

caps this in bar 50 with a fortissirno distortion of idea (i), its rhythmic shape made

irregular and its descending interval stretched to an augmented second.

In the remainder of the passage, thematic identity is eroded by the practically

continuous stream of semiquavers. In the fourth "wave" (bb.51-64), the themes

struggle to reassert themselves but are abandoned at the climactic point, while in the

final "wave" (bb.65-82) there is only the most fleeting reference to them. The fourth

wave begins with a two note motive in the uppermost voice (the descending minor

second d5-c#5). The thematic significance of this motive is at first enigmatic, but it is

developed during bars 52-5 by expanding the interval to a minor third and

232
incorporating other notes, so that by bar 56, it has recognizably become a new version

of idea (iii). This point coincides with the attainment of the dynamic level nzfand the

restoration of the field B. However, this point of arrival is swiftly passed over as

there is a return in bar 58 to the thematic and textural situation of bars 48-50; this

time, the right hand semiquavers push up to a greater registral extreme, while the left

hand unfolds a melodic line in marcato octaves which amalgamates ideas (iii) (bb.58-

9), (i) (b.60) and (ii) (bb.60-61). The last of these is the most forceful statement of

this idea thus far and coincides with the critical harmonic clash between dL and d#Ie.,

but the actual climax in bars 62-3 consists only of the rushing semiquavers over a

cadential bass.

The beginning of the final wave (bb.65-70) is similar in texture and content to

that of the fourth. Intervallic fragments of thematic material are animated by

irregular rhythmic groupings within the perpetuum mobile of semiquavers, the most

prominent intervals being the minor third, perfect fifth and major second. In the

Meno mosso, more continuous melodic lines begin to evolve, but there is no reference

to the main thematic material, apart from the use of some of the rhythmic features of

idea (iii) to animate the repeated dlj in the right hand in bars 75-6.

(c) Summary

This passage clearly shows the interaction between traditional and radical

approaches to form-building in Carter's transitional work. The tonal process of

sonata exposition, in which the stability of the tonic is undermined, leading to its

replacement by the dominant, has an analogy in the replacement of b as tonal centre

233
by c and the reinterpretation of the major third d# as the minor third ek. However,

the stability of b is called into doubt from the outset of the movement, so that a further-

level of complexity, involving the relationship between b and all, is created.

Furthermore, the means by which tonal centres are established are more tenuous than

in conventional tonal harmony. The reservation of the extreme bass register for

pitches of focal importance is Carter's most obvious and effective technique in this

passage. Linear patterns play some part in the connection of these pitches, although

"middleground" associations over longer stretches are relatively weak and relationships

between neighbour notes are more often exploited for their potential to undermine

rather than confirm the pull of a tonic. In the absence of fully functional key and

voice-leading relationships, the looser principles of diatonic orientation operate; pitches

will tend to seem more stable points of focus if any dissonance in the surrounding

harmonic context can still be accounted for as diatonic. The thematic process of

statement, development and recurrence can be observed, but is challenged by Carter's

tendency towards continuous evolution. Above all, the form-defining role of texture

is apparent here, foreshadowing the composer's later development.

234
CHAPTER 9: SONATA FOR CELLO AM) PIANO, SECOND MOVEMENT

1. Genre and Character

Carter's Cello Sonata is a crucial work in the development of the composer's

mature style. It ushers in a phase in which conventional approaches to musical time

and space are rejected in favour of more experimental strategies. David Schiffs

study of Carter's works revealed that the movements of the Sonata were conceived in

the order II, ifi, IV, I and suggested that the work "evolved as it was composed"

[Schiff: 142]. The first movement has therefore received the greatest attention, being

most representative of the musical language of the later Carter. However, for the

purposes of illuminating the transitional character of Carter's musical language during

the late 1940s, the second movement may prove an equally rewarding subject for

study, representing as it does the point from which Carter was to launch his change of

direction.

The movement's status as a significant "end-point" is symbolized by the often-

remarked fact that it is the last of Carter's works to bear a key-signature. The

relationship between the expectations arising from the employment of key-signatures

and the actual harmonic language of the piece is, therefore, an obvious starting point

for discussion. Certain aspects of the movement's form and genre are also

representative of the earlier, more conventional Carter; the ABA'(+ coda) formal

outline and the scherzando character of the main body of the movement (indicated by

the tempo marking %'fce, molto leggiero, the lean texture, the frequent use of

pizzicato and the incorporation of jazz-derived syncopation) place the movement firmly

235
in the generic category of "scherzo and trio".

Conversely, other features of the work's harmonic language and formal

processes indicate the new directions in which Carter was beginning to move; the

avoidance of conventional tonal sonorities and progressions, the "motivicization" of

harmony and the continual evolution of musical material during the course of the

movement adumbrate later developments in the composer's musical language.

The basic formal outlines of the movement are easily identified; double bar

lines, combined with changes of metre, tempo and key signature are the outward signs

of a ternary structure;

A: 1-57 (5,
2/2, J =84)
B: 57-129 (1I, 6/8,1. =112)
At (+coda): 130-213 (2, 2/2, =84) I
As is characteristic of Carter's music, the sections are not closed and separate, but run

into one another; bars 51-56 constitute a transition, as do bars 123-130, the latter

beginning with a repetition of the material of bar 1. The A and B sections are

contrasted in texture and character, but in the final section of the movement, elements

of the more violent and sinister B section affect the light, playful tone of the material

of the A section. In the coda (which may be regarded as beginning in bar 186, after

the movement's main dynamic climax), material from both sections is brought into

close conjunction and motivic connections and similarities which have been hinted at

during the course of the movement are made explicit. It would be wrong, however,

to regard this as a straightforward synthesis or resolution; the expressive character of

the middle section is so forceful that it tends to dominate and undermine the

scherzando tone of the first section of the movement. The sense in

236
which the issues of the movement remain unresolved is connected with the cyclic

process of the Sonata as a whole; the impassioned tone of the B section of the

movement is carried over into the succeeding slow movement, the connection in terms

of tempo and motive between bars 203-4 of the second movement and bar 1 of the

third being a concrete sign of this relationship. Similarly, the fast moving scalic

figuration of the more agitated passages of the slow movement gives rise to the

material of the turbulent finale. A sense of resolution is only achieved at the end of

the latter by a return to the enigmatic, emotionally "neutral" material of the opening

movement.

As can be seen by reference to the ABA' formal outhne of the second

movement, the proportions of the ternary structure are not balanced and symmetrical,

but are dictated by patterns of growth and development. Two approximately equal

sections are followed by a much longer third. This final section is far from being a

straightforward recapitulation; very little material from the A section is restated

literally, some is restated with minor modifications and much that was originally only

briefly stated receives extensive new treatment. Some material does not reappear at

all, resulting in an unusually high degree of "redundancy". (For example, the

passages from 15-21, 27-35 and 45-56 in A have no equivalents in A'.) Thus in

terms of both genre and structure, the movement evinces a degree of tension between

the convention of "scherzo and trio" and Carter's more fluid approach to form.

A further aspect of the movement which requires consideration is the

relationship between the two instruments. The opening movement sees the first

crystallization of Carter's technique of counterpointing opposing instrumental

"characters", Its texture "is concerned with mediation between relative dissociation

237
and relative association" [Whittall 1988: 221]. The second movement, although less

sharply focused, also explores a range of possible relationships between the

instruments. Musical material is shared between the instruments almost equally; for

example, the movement begins with the piano imitating the cello's opening gesture

(although the contrast between the capabilities of the instruments is emphasized in this

case as the piano transforms the cello's single pitches and dyads into dyads and

trichords.) In particular, Carter exploits the timbral similarity between the sharp

staccato attack of the piano in the low and middle registers and the pizzicato of the

cello. This is first heard in the passage of intertwining two-part counterpoint in bars

3-10, but receives its most extended treatment in bars 136-140, where the two

instruments divide a single melodic line and the piano is actually marked "quasi pirz".

The high point of "association", however, is seen in bars 167-170, where the texture

consists of a single melodic line presented in three registers, the cello occupying the

middle register and the right hand of the piano rhythmically displaced by one quaver to

create a heterophonic effect. Maximum "dissociation" occurs when the instruments

revert to type, the cello playing "cantabile espressiw)" lyrical lines, while the piano

has more fragmentary, faster moving material (for example, in bb. 17-20). The final

stages of the movement also tend towards dissociation in that the instruments are heard

separately more frequently than in the rest of the movement; there are three "solo"

passages (for cello in bb.191-194, piano in bb.194-196 and cello again in bb.210-213),

the last three and a half bars being the longest unaccompanied stretch of the

movement. Bars 204-207 represent an uneasy balance of associative and dissociative

factors; the main thematic material of A (in the piano) and B (in the cello) is

superimposed, the contrasting thematic characters and instrumental sonorities being

238
offset by the attempt to accommodate them both into the same metrical, harmonic and

textural framework.

2. Form and Thematic Process

The following is an informal description of the second movement's thematic

material and its surfce continuity. This will be a necessary preliminary to the more

detailed analysis of its musical language. Charting the movement's progress through the

appearances of its main thematic material will help to demonstrate both the degree to

which conventional thematicism plays a controlling formal role and the degree to

which it is challenged by other factors. Any attempt to deduce a large-scale tonal

plan or background contrapuntal structure must be reserved until the complex issues of

harmony have been addressed.

As in most of Carter's music of this period, the musical material of this

movement evolves freely from a handful of brief motives which are instantly

recognizable through distinctive melodic and/or rhythmic features. Some of these

remain almost unchanged, while others may undergo extensive alteration.

A (bb.1-57)

Example 9.1 shows the main motives from which most of the material of the

section is derived. Motive (i) is sharply characterized by its initial wide downward

leap and syncopated rhythm. Its concise first presentation by the cello is imitated and

expanded by the piano in bars 2-3, providing the starting point for an opening

239
paragraph in which other important motives gradually evolve. In bars 15-16 two

further variants of (i) in the cello begin the next paragraph. The piano part of bars

23-34 may also be derived from (i) (this is most clearly seen in 25-27 and 30) although

as the section progresses, the resemblance to the original form fades. The motive's

status as an initiator of action is confirmed at other important points in the movement;

for example, a suggestion of (i) in the piano in 50-51 begins the transition to the B

section; its literal restatement in the cello in 123-4 heralds the beginning of the

retransition to At and the beginning of A itself takes the form of overlapping variants

of (i) in both instruments. Motive (ii) grows from the descending minor third in bar 3

(piano, 2nd and 3rd beats). This also has a prominent role as a punctuating device,

particularly as a climactic end-point (see bb.7, 10-11 and 35-36). A rearrangement of

the paired minor thirds of motive (ii) produces motive (iii) (piano right hand, b. 12).

This appears only briefly in A, but subsequently plays a dominant role in A1.

The remainder of the material of the A section may be divided into two basic

types of articulation, staccato and legato, the former tending to appear in continuous

quavers often in quasi-scalic or arpeggio patterns (for example bb.5-9; this is the type

of material in which the timbral similarity between the instruments may be exploited).

The latter first emerges in the piano in 12-13, but appears most characteristically in the

cello with the marking "cantabile espress." (for example, in 17-20, 22-24, 27-34).

The rapidly varied combination of these contrasting types of material contributes to the

scherzando character of the section, while the pervasive syncopation (deriving from

motive (i)) lends it a jazzy "American" character, which, according to the composer,

is intended to be parodic ICLL 2301.

Formally, A progresses from small fragments to longer, more continuous

240
 passages which tend to end with an abrupt breaking-off. In bars 1-14, for example,

a one-bar cello phrase is imitated and expanded by the piano in the next two bars, and

gradually grows towards a four bar phrase for both instruments in 11-14. This

opening paragraph is punctuated by three abrupt cut-offs (bb.7, 11 and 14) each one

more violent than its predecessor. A second paragraph begins in bar 15, during

which the legato material which first emerges in bars 12-13, tends to appear in longer

stretches and resemble conventional melodic structures more closely. The cello line

in bars 15-35 might be analyzed in terms of balanced, periodic phrasing as follows

(also see Ex.9.2):

15-24: "antecedent" (15-20) + "consequent" (21-24)

2(1+1)+2+2 2+2
27-35: "antecedent" (27-31) + "consequent" (32-35)
3+2 2.5+1.5

This scheme, of course, only expresses a partial truth. The sense in which

each of the "consequent" phrases is conclusive is severely compromised by the

typically abrupt endings (bb.24 and 35). Furthermore, the lack of obvious closural

attributes in the harmony, the evolutionary nature of the melodic growth and the

incompatible phrase structure of the piano part conflict with the nature and the

substance of this division.

The climax in bars 34-37 is the most forceful up to this point and introduces

the most fully developed statement of motive (ii) so far heard. From this point, the

tail of the motive (the descending third b -g) is incorporated into a new melodic idea

which comes close to a conventional sentence structure occupying, fourteen bars (see

Ex.9.3).

241
 The remainder of the section (bb.51-57) consists of a transition to B. Its

transitional character is indicated by insistent repetition, especially of the minor thirds

of motive (ii). The persistent rhythmic irregularity of A is now pressed into recurrent

patterns in the cello, first of three quavers and then of two (the increase of pace

hastening the progress towards the junction of the sections - see Ex.9.4).

B (bb.57-129)

A straightforward metric modulation, which maintains the speed of the quaver,

relates the tempi of the A and B sections through the ratio 3:4. However, the

rhythmic characteristics of the two sections are sharply contrasted; while the A section

relies on the essentially conventional device of syncopation within an alla breve metre,

the B section exploits a more complex technique - a polyrhythm of 3:5. (The more

conventional device of hemiola is also used.) Example 9.5 shows the section's main

material; (iv) quintuplet quaver figures, either on repeated notes (cello, bb.57-62, 106)

or as a tremolando (mostly on a minor third: piano, bb.64-66, 113-116 etc., but also

on perfect fifths and fourths: piano, bb.68-71, and minor ninths: piano, bb.76-80); (v)

an expressive legato melodic line, mainly moving in dotted crotchets, and deriving

from the Dies Irae - like motive stated in the piano in bb.61-62; (vi) a five-quaver

motive, first heard in bar 60 (piano RH), which undergoes various transformations and

generates much new material, both in this section and in the remainder of the

movement.

The character of B is in marked contrast to that of A. The scherzando tone

gives way to a rather more unstable temper. Extremes of register are explored (the

242
piano's rise into the "stratosphere" in 68-73 is counterbalanced by the cello's plunge

into the depths in 79-85) and there is a wide range of textural contrasts, from the bare

octaves of 74-75 to the dense 5-or 6-part harmony of 113-117. There are sharp mood

swings from the delicacy of 68-82 to the violence of 95-112. There is, in short, an

air of unpredictable fantasy which is far-removed from the nonchalant mock-popular

idiom of the previous section.

Whereas A shows a relatively conventional formal tendency towards greater

continuity and evolutionary growth, B tends towards dissolution or fragmentation of

texture and liquidation of thematic material. Twice during the course of the section,

the thread of continuity runs very thin; firstly, in bars 68-73, the texture becomes

extremely rarefied, as the piano line "evaporates" with a succession of oscillating

perfect fifths ascending to the highest register, while the cello line is gradually

fragmented, ending with an isolated c# 5 in 70-71; secondly, in bars 105-112, at the

opposite dynamic extreme, after a passage in which irregularly grouped chordal

flourishes have become dominant, all movement in the piano is arrested by a ibrtisskno

chord, while the cello has a declamatory solo line. In each case (at bb.74 and 113),

momentum is regained with the return of motives (v) and (vi) in conjunction.

A1 (bb.130-213)

The final section of the movement may be described as a "developing

restatement" of the first A section. Table 9.1 summarizes the course of the final

stages of the movement in terms of thematic derivation and/or cross reference.

Passages of literal recapitulation are rare; more frequently, ideas from the first section

243


Table 9.1: Relationship between material in A and A1.

Bar nos. Motivic Corresponding pasges in A:

in Al derivation Recapitulation Development
close altered

130-134 (i) 2-7

135-136 (ii) 10
136-140 (ii)/(vi) 8-9/93-94
141-146 (ii) 38-43
146-149 (iii) (piano) 41/12-14
(i) (cello) 1
149-150 (u) 42-43
150-151 (ii)/(vi) 93-94
151-155 (ii) 38-40/43-45
155-157 (ii) 42-43
157-166 (iii)
166-173 (iii) (piano, cello to 170) 13-14 (piano)
(ii)/(vi) (cello from 170) 11-12 (cello)
174-177 (ii) (piano) 11-12 (cello)
(iii) (cello) 13-14 (piano)
177-185 climactic chordal section; distantly derived from (1) and (ii)

Coda
186-190 2226
190-191 10-11/33-34
191-196 (vi) 93-4
197-203 (iv) + (v) 118-122
204-207 (v)(cello) 113-114
(i) (piano) 2-3
208 (ii)/(vi)
209-210 (ii) 7/10
210-2 13 (ii)/(vi)

are allowed to "have their head" in a succession of climaxes, culminating in the

movement's dynamic high-point at bar 185. The music which follows may be

regarded as a coda, including recollections of material from both sections, the

culmination of which is the simultaneous juxtaposition of forms of motives (i) and (v)

in 204-207.

It will be noticed from the table that motive (vi), which is first suggested in bar

244
60, but which appears in its definitive form in bars 94-5, plays an increasingly

prominent role during the later stages of the movement. Its continuous quaver motion

clearly relates it to the staccato quaver material of the A section. Motive (vi) is

recognized through its pitch contour and content, rather than through its articulation or

rhythm; in A1 the characteristic legato articulation of the B section is abandoned in

favour of the staccato of A, thus bringing about a synthesis of sorts. Successive

appearances of material derived from motive (vi) are longer. For example, the last

two occurrences (bb. 194-196 and 190-191) are 23 and 27 quavers long respectively

(compared with the 10 quavers in bb.93-4).

Bars 130-140 represent a developed restatement of material from bars 1-14 of

A, climaxing in the extended development of the pizzicato idea. The material of 15-

35 is then omitted. Only in 186-191, after the movement's main climax, is any of

this material restated. Instead, A 1 continues with material from the second half of A

(that shown in Ex.9.4). This appears in a form reasonably close to the original.

However, it is interrupted by a development of a complex of motive (i) and an.

extension of the material of Ex.9.4 and a suggestion of the rhythm and texture of

motive (iii) (bb. 146-9), climaxing in another dual motivic statement - (ii) with (vi)

(bb. 149-151).

The texture and motives of bars 146-9 recur and receive extensive development

in 157-173. This passage may be regarded as the "definitive" presentation of motive

(iii), which appears explicitly in 166-7, the high point thus far of cooperation between

the instruments. From 170, the instruments begin to diverge again, the piano

continuing with the material of the immediately preceding passage, the cello running

into arpeggio-like streams of quavers, stemming from the material of bars 11-12. In

245
 174-8 the roles reverse, the cello taking the legato material (with another appearance

of motive (iii) in 175-6) and the piano taking the staccato quavers. This leads into the

climax of the whole movement in 178-185, which, perhaps echoing the texture of 91-

112, is primarily chordal and disrupts the rhythmic regularity which had been

established in the previous section.

This climax represents the point of furthest remove from the identities of the

movement's main motives. This is immediately compensated for in bars 186-189 by a

passage of literal recapitulation (compare bb.22-25). The latter takes a new turn in

190 with another climactic statement of motive (ii) recalling that of 34-35. The

climax subsides rapidly and is succeeded by two further statements of motive (vi), first

for the cello (191-3) then for the piano one octave higher (194-196). The cello

statement leads to a low "pedal" dyad of c#-a, over which the piano statement is

heard. The latter leads, via a simple metric modulation, to a restatement of material

from B, the piano part from 199.5 to 203 being a transposition through a tritone of

bars 119-122. The previously discussed juxtaposition of motives (i) and (v) follows in

204-7. The final six bars of the movement revert to the metre of A and to the

characteristic texture of bars 3-8. The piano drops out after bar 210 leaving the

pizzicato cello to end the movement unaccompanied in true scherzando fashion. It

should be noted that the cello's final bar, outlining a Bb major triad, is the final

transformation of a motive which has evolved gradually during the closing stages of

the movement (see the cello part in 178 and 192).

246
3. Harmony, Tonality and Sets

The harmonic language of the second movement of the Cello Sonata receives

attention in the studies by Schiff and Harvey. Two important issues arise from a

reading of their work; the nature of Carter's harmonic vocabulary and the overall

control of tonal motion. These may, perhaps, be described as considerations of

"foreground" and "background" respectively, and interaction between these levels (if

indeed any such hierarchic organization can be deduced) opens up a third area for

investigation. Schiff's comments are quoted in full;

Harmonically, the second movement - the last Carter was to write with
a key signature - contrasts B major and B minor, an opposition stated
in the first two measures. Carter had extensively explored this polarity
in the Piano Sonata and Emblemc, here, however, the fluctuations
between keys are far more rapid than before and moments of diatonic
simplicity sound ironic rather than affirmative - Carter has described the
more tonal parts of the movement as parodies of "pop" music (see bars
37-40 for example). The contrasted tonalities of this movement tend
increasingly to combine into chromatic figurations that emphasize the
minor third, and in retrospect the listener may note that the apparent
tonal opposition is in fact derived from the six-note cell heard at the
beginning of the first movement. Minor thirds, often complemented by
perfect fourths to form the "key" four-note chord of the Sonata,
dominate both the harmonic and thematic materials of the movement.
(Schiff 1983: 139-1401

Harvey deals with the movement in a similar manner, describing it as "tonally

the most straightforward" of the four and comparable to the composer's earlier music

in its use of scale and mode. Harvey finds the "opposition of B and B" in the

movement links it with the Piano Sonata and the Wind Quintet, the fact that there are

notes in common between B major and Bb minor being exploited to produce an

"equivocal" harmonic texture in which "'key' is usually only definable as a property

of the scalar relationships of the pitches present at any given moment, rather than as a

247
 function of harmonic relation." [Harvey: 34]'

These assessments of Carter's harmonic practice require elaboration and

qualification. The opposition of the pitches B and B' and the keys of B major and B

minor is not the central issue that it is in the Piano Sonata. Harvey's assertion that

these two keys are juxtaposed at first successively (in bb. 1-2) and then simultaneously

(for example, in what Harvey describes as a "bitonal" passage in bb. 15-25) is suspect;

it ignores the presence of the pitches a (in b. 1) and ct' (in b.2), and it confuses the

enharmonic notation of bars 15-25 (the cello part being notated in sharps, the piano in

flats, but both sharing pitch-class content) with actual bitonality. Similarly dubious is

Schiff's statement that the "six-note cell" (6-Z43 [0,1,2,5,6,8]) heard at the beginning

of the first movement, is the source of the second movement's tonal opposition.

The basic harmonic building blocks of the foreground of Carter's music are

obviously more diverse than the triadic components of traditional tonality. However,

as already argued in Chapter 5 (see pp.1 17-21), in the music of the late 1940s, a clear

relationship between the two can be seen; Carter takes the basic intervals of the triad -

the third and fourth (or fifth) - and recombines them in more complex formations.

Schiff's "Chart 9" of "Harmonic Materials" [:137], shows various intervallic

combinations which "are heard melodically and harmonically throughout the work."

The six tetrachords which Schiff identifies are said to be "neither thematic [...1 nor

serial in implication or treatment"; instead, a "chromosomal" intervallic pattern can be

detected running throughout the work. This constitutes a rather limited group if it is

to be considered the basis of such a wide harmonic vocabulary. Harvey recognizes

The semantic problem of the word "key" may, of course, be bypassed by discussing
the harmonic context in terms of diatonic fields.

248
 this in his analysis, which adds three trichords (3-2, 3-3 and 3-4) to the core of

significant sets.

However, even this is not sufficiently broad to describe the harmonic

vocabulary of the second movement. Schiff's chart identifies four of the possible

combinations of ic3 + ic5. One of these, 4-7 [0,1,4,5] which opens the work, is

described as the "key-chord". Another of its intervallic properties - it may be formed

from a pair of major thirds (ic4 + ic4) - relates it to two other tetrachords which share

this property, 4-17 and 4-20. The latter two sets may also be formed from a pair of

minor thirds (ic3 + ic3) and a pair of perfect fourths (ic5 + ic5) respectively. This

suggests another possible point of departure for reconstructing Carter's set vocabulary

for the work. The fact that the tetrachordal pairing of ic3 + ic 5 which Schiff omits

from his chart - 4-23 - may also be formed from ic5 + ic5, strongly suggests that the

various symmetrical pairings of ic3 + ic3 and ic5 ^ ic5 should be considered as well

as the ic3 + ic5 combinations. The basis of an analysis of the movement's intervallic

vocabulary should be a comprehensive categorization of tetrachords formed from the

pairing of interval classes 3 and 5. Table 9.2 is abstracted from Appendix 3 and

provides such a categorization.

The contents of Table 9.2 represents a harmonic universe in which all interval

classes are present, but in which some (ic3 and ic5) have a privileged status. This

anticipates Carter's later, increasingly systematic, exploitation of intervallic

combinations and his desire to create characteristic interyallic and set vocabularies.

It will be noticed that two sets, 4-26 and 4-23, appear in two groups in Table 9.2, thus

reducing the total number of tetrachords from 15 to 13. The choice of 13 out of a

possible 29 tetrachords may seem rather unfocused in terms of creating a characteristic

249


Table 9.2. Tetrachordal vocabulary of Cello Sonata, second movement

ic3+ic3 ic5+ic5

Ic and pc set Interval ic and pc set interval

structure vector structure vector

34 56
II{ }1 3 1 4-3 12121001 51{ } j5 1
4-8 [2001211
01 01

35 06
1 3 1{ }1 3 1 4-10 11220101 151( }I 5 I 4-9 [2000221
02 71

37 51
I 3 { } lI 4-17 [1022101 15K )1 5 1 4-20 [101220]
04 08

38 57
ll{ }1 3 1 4-26 [0121201 151{ )1 5 1 4-23 [021030]
05 02

39 58
II{ } 1 3 1 4-28 [0040021 15K }1 5 1 4-26 [012120]
06 03

ic3 +ic5

icandpc set Interval

structure vector

45
1 3 1{ )Il 4-7 [2012101
10

36
1 3 K }l 5 l 4-13 [1120111
01

37
13K } 1 5 l 4-14 [111120]
02

40
3 K }1 5 I 4-18 [1021111
17

50
1 3 K }II 4-23 [0210301
27

250


set vocabulary, but two points should be borne in mind; (i) a form of hierarchy may

be demonstrated to operate among the sets, with greater importance being assigned to

the symmetrical tetrachords, especially to the multiply symmetrical 4-28; (ii) the

"transitional" nature of the work has already been noted - the comparatively "loose"

approach to harmonic material in this movement had been refined by the time Carter

came to compose the first movement, which selects from this large pool of sets the

smaller network represented by the Kh(6-20) subcomplex as its most important

constituents (see pp.124-8).

The trichordal subsets of these tetrachords also play an important part in

defining the vocabulary; the symmetrical tetrachords may be generated from pairs of

inversionally-related trichords as well as from dyads. Table 9.3 sets out the trichordal

subsets of the tetrachords of Table 9.2. It will be noted that 3-11 (the major or minor

triad) is the most frequently represented, being contained within 5 of the 13 tetrachords

and thus perhaps confirming the subliminal influence of the triad on the harmonic

language of the movement.

Table 9.3. Trichordal vocabulary of Cello Sonata, second

movement

tetrachord trichordal subsets

4-3 3-2 (0 1 3) 3-3 (0 1 4)

4-7 3-3 3-4(015)
4-8 3-4 3-5(016)
4-9 3-5
4-10 3-2 3-7(025)
4-13 3-2 3-5 3-7 3-10(036)
4-14 3-2 3-4 3-9(027) 3-11(037)
4-17 3-3 3-11
4-18 3-3 3-5 3-10 3-11
4-20 3-4 3-11
4-23 3-7 3-9
4-26 3-7 3-11
4-28 3-10

251
 A study of the similarity relations among tetrachords and trichords (Tables 9.4

and 9.5) reveals a few more facts pertinent to analysis of the movement. The

"special" status of the "diminished triad" 3-10 and the "diminished seventh" 4-28 is

shown by the fact that they exhibit the minimum number of similarity relations among

their respective groups. Those similarity relations which do hold pertain exclusively

to pitch-class rather than interval class. This is unsurprising since the interval vectors

of these two sets are minimally varied with respect to interval class; in other words

they maximize ic3 and ic6 at the expense of all others. These sets thus represent the

"purest" form of presentation of ic3, which may be regarded as the motto interval of

the movement. (Note also that the interval vector of 4-23, the set which maximizes

ic5, contains three other ics including one instance of ic3.) The sparing, but

structurally prominent uses of sets 3-10 and 4-28 will be noted during the analysis.

The maximum number of similarity relations among the trichords is shared by

the "scalic" 3-2 and the "triadic" 3-li which are in the relation R or R 1 R to all the

other trichords. This means that they are less distinctive and hence more flexible than

3-10. Seven out of the thirteen tetrachords contains either 3-2 or 3-11. One, 4-14,

contains both, and it is therefore not surprising that this set represents the maximum

for similarity relations among the tetrachords.

Carter's music is evolutionary, continuously developing new shapes from basic

intervallic constituents. A problem in applying p-c set theory to music such as this is

that it tends to treat music as static structure rather than as dynamic process. This

problem may be partially overcome be adopting segmentation strategies which reflect

252
Table 9.4. Similarity relations among tetrachords in Table 9.2.

4-3

4-7 Rp 4-7

4-8 Rp 4-8

4-9 R2 4-9

Rp

4-10 Rp 4-10

4-13 Rp Rp Rp R2 4-13

Rp

4-14 Rp Rp Rp Rp 4-14

4-17 Rp RI RI Rp 4-17

Rp

4-18 Rp Rp Rp Ri Rp R2 4-18

Rp Rp

4-20 Ri Rp R2 Ri 4-20

Rp Rp Rp

4-23 Rp Rp 4-23

4-26 Ri Rp Rp R2 Rp Rp Rp 4-26

Rp

4-28 Rp

253
Table 9.5. Similarity relations among tnchords in Table 9.3.

the fluidity of the musical surface. In particular, the technique of imbrication will be

extensively used since this can show how individual pitches, intervals and small sets

are reinterpreted as the music progresses. A pitch, dyad or set which forms part of

one grouping may act as a "pivot", combining with another element to form another

significant group. This free association of intervals is an important facet of Carter's

harmonic practice at the local level.

Bars 91-112 offer a relatively clear example of how this free association of

interval classes 3 and 5 operates in the movement (see Ex.5. 11 and p.123). Most of

the melodic and harmonic activity of the passage derives from the combination of these

interval classes in various permutations. This passage is perhaps exceptional in that it

presents the harmonic "roots" of the movement in their most concentrated form, as

254
 pairs of dyads combined harmonically or melodically. (The position of this passage -

almost exactly half way through the movement and at the climactic point of the B

section - may be significant from this perspective.) Immediately following it is a

passage of denser harmony (bb.l 13-122) whose 5 or 6 part texture may be broken

down into constituents which prove to be similar to those of Ex.5. 11 (see Ex.9.6).

A similarly density of motivic harmony arises from the combination of motives

(i) and (v) in 204-207 (see Ex.9.7). Here the manipulation of sets and intervals is

seen at its strictest. The relationship between the melodic and harmonic dimensions is

particularly interesting. The cello has a series of eight trichords, of which, numbers

5-7 are identical to numbers 1-3 while number 8 is a transposition of number 4.

Analysis of the intervallic content of these chords reveals a palindromic pattern which

uses only set classes 3-4 and 3-11. These trichords are so spaced as to create upper

dyads of ic3 or ic5, thus giving rise to several of the work's characteristic tetrachords

through the pairing of these dyads. The two halves of the chordal sequence are

connected by the "key chord" 4-7. Furthermore, the upper and lower "voices" of this
/
chord sequence spell out various forms of motive (v); the upper voice has set 3-3

followed by 3-5, the lower voice provides 3-2 and 3-3 in counterpoint to this, varying

the motive through inversion and permutation of note order. This passage thus

encapsulates the "developing variation" of motive (v) which has taken place during the

movement, its initial form of 3-2 being expanded to 3-3, 3-4 and 3-5 (see bb.61-63,

67-69 and 74-76). The piano, meanwhile, performs a similar act of integration, its

downward octave leap to the movement's "tonic" in bar 204 being a kind of resolution

of the major seventh which opens the movement; this is supported by a sustained

tetrachord 4-23, which is built form ic3+ic5, but is also diatonic to B flat major.

255
 Segmentation of the harmonic surface of the movement at any point will reveal

a similar make-up in terms of the sets described above. However, this establishes

only that Carter is working with a consistent vocabulary; it does not provide the

syntactic "rules" which enable us to perceive the music as meaningful. Syntax is the

area in which the transitional nature of the Sonata is most apparent, in that a single set

of "rules" will not suffice to explain how the music is organized. Ambiguity arises

from the possibility of deducing different principles of organization from the harmonic

materials themselves; the predominance of the "triadic" interval classes 3 and 5 gives

rise to a host of quasi-tonal pitch-structures. Reference back to the tetrachords of

Table 9.2 shows that a small majority (seven out of thirteen) are subsets of the diatonic

major scale (set 7-35). Even more suggestive is the fact that only one tetrachord (4-

9) cannot be derived from either the major or harmonic minor scale. This set

vocabulary thus possesses the potential to create a harmonic texture which is

predominantly diatonic and may therefore also contain allusions to functional tonality.

Tetrachords formed from the dyadic pairing ic3 + ic3 have a particularly

important role in this respect, since they can be made to resemble parallel triadic

motions. Example 9.8 shows some of the occurrences of these sets during the

movement. Tetrachords formed from ic5 + ic5 have a variety of tonal characteristics;

4-20 and 4-26, for example, are identical with the tonal "major seventh chord" and

"minor seventh chord" respectively; 4-23 may be presented linearly to form a "cycle

of fifths" progression; 4-8 and 4-9 may be similarly presented, suggesting a

"Neapolitan" pattern. These, and other possibilities, are illustrated in Example 9.9

However, allusions to particular tonalities and voice-leading patterns are rarely

unambiguous. Carter creates harmonic ambiguity through the tension between tonal

256
function and diatonicism. Tetrachords which might be tonally suggestive in

themselves are often placed within a harmonic context which undermines the sense of

a fixed tonality. This is the cause of what Harvey describes as the "paradoxical"

harmonic nature of the movement. In this movement, however, enharmonic changes

are much more frequent and rapid and have fewer implications in terms of tonal

function.

The progress of the A section in terms of diatonic fields is shown in Table 9.6.

One outstanding feature of the section in these terms, is the gradual emergence of AI

as a relatively stable environment in bars 27-50, the composite fields E.6/Ab and Al'

1Db being used extensively. This is coupled with some of the clearest references to

functional harmony in the movement in bars 37-40, producing a moment of apparent

stability (see Ex.6.3(b)).

Another important feature is the "opposition" between sharp and flat

orientation. The opening two bars present this opposition immediately; the first 5

pitches of the cello are restated enharmonically by the piano at the beginning of bar 2.

The cello dyad e-a, which introduces the "!lse relation" a#-alj , defines the diatonic

field of bar 1 as B/F. However, the continuation of bar 2 suggests Gl'or eb. The

Table 9.6. Succession of Diatonic Fields in A section.

Bar 1 2-8 9-11 12-14

E'E Gb(oc eS') Cello: Ci'- g
(includes brief movements to Aband C)) Piano: E-1-Bh-bb -g

Bar 15-16 17-21 21-22 23-26 27-31 31-35 36-37 38-40 41

Cello B C# E Fl Ab Ab - ab/db ab EbM4 Ab/f
Piano C) E*IGb Ck Gb/Cb Ab/Dd Ak and Ckclash ab Eb/Ab A$/Db

Bar 42-43 44-50 51-54 55-56

Cello ? Ak Db db?
Piano Eb/Ab Ak (touches of (7.'/Cb/Fb) unstable
 balance of similarity and difference is also seen in the spacing of the five shared

pitches; the cello's major seventh an-b and the minor seventh f#-g# are retained in the

piano as b ' -á and gi' -at', but the cello states these pitches melodically, the piano

harmonically.

For the remainder of bars 1-14, the music of both instruments falls within the

composite diatonic field e /ai /d and makes use of the "false relations" gi' -g(j, ck -c*

and f -ft1. The exception to this is the excursion to the sharp side in bars 8-1 1.

Here, the pivotal use of enharmonic relations is once more crucial. This passage,

which wheels rapidly through the relationships between relative arid parallel majors

and minors, thus creating a chain of minor thirds, contains the seed of an extremely

important structural idea. The chromaticism which challenges the fifths-based

diatonicism is often based on the "ic3 cycle", or set 4-28 or the diminished seventh, to

give it all its possible names. Thus the minor thirds and perfect fourths/fifths (ic3 and

ic5) which form the roots of the foreground hannonic language, also play a larger

structural role in the fragmentary middleground.

The fiat/sharp opposition recurs in bars 15-25, although it would be simplistic

to describe this passage, as Harvey does, as "bitonal". While the cello in 15-16

espouses B and then shifts to CMn 17-2, the piano has C in 15-16 and Dbmoving to

G in 17-21. Whether this simultaneous presentation of a field in its sharp and flat

guises is meant to aid reading or is a genuine expression of tonal conflict is a

debatable point. Similarly ambiguous is the recapitulation of bars 22-26 in bars 185-

189; during the former, the cello moves through FIB to B/F#while the piano maintains

Cb 1Gb; in the recapitulation, both instruments have the sharp notation. A possible

reason for this is that the flat notation of the former allows a smoother optical

258
transition to the i* field of 28-9, while the latter moves to Did/F.

The tension inherent in the cross-relations of composite diatonic fields rises to

the fore in bars 3 1-35. Gb and ci' are heard simultaneously with gI and cI in the

climactic bar 35. Here, fj and c are pivotal, representing Cb in relation to the

piano left hand, but being reinterpreted as e11 -bL by the piano right hand. The

uppermost line in 33-35 is a further manifestation of 3-10 and 4-28 (f-as -ci' followed

by d'-f-g-bI').

The relative diatonic stability of bars 37-51 has already been noted. However,

it is important to point out that moments of instability are once more provided through

the agency of set 4-28. Thus, in bars 42-3, the combination e-d' -bi' -g is heard in the

cello, the preceding e-b-e having disrupted the prevailing DIi/A4 field.

In 51-6, a passage of obviously transitional character, diatonic stability is

disrupted by the appearance of fourths from fields other than the prevailing Dl' (a-d in

b.52, c -fb -bi' in b.53). Set 4-28 and its complement, 8-28 the "octatonic scale",

are once more in evidence as the insistence on minor thirds becomes more intense (see

b.54, beat 4 - b.57).

The harmonic material of the B section is difficult to describe in terms of

diatonic fields, even composite, overlapping or conflicting fields. The harmonic

context is clearly of a different order from that of the A section, even if the basic

constituents, the trichords and tetrachords, remain identical. The increased

chromaticism of the B section is partially explained by the fact that "diminished"

harmony and octatonicism play an enhanced role, although it would be overstating the

case to say that octatonicism displaces diatonicism as the background harmonic

"wash".

259
 This change of harmonic substance is not the result of an arbitrary choice, but

is based on an alternative interpretation of the implications of the trichords and

tetrachords identified in Tables 9.2 and 9.3. It has been noted that most of the

tetrachords in Table 9.2 have diatonic associations. It should be further pointed out

that the larger sets forming diatonic fields, 7-35, 8-23 and 9-9, can all be represented

as segments of the cycle of fifths. Another way of putting this is that the interval

vector of each of these larger sets maximizes ic5; 7-35 [254361], 8-23 [465472], 9-9

[676683].

The octatonic set 8-28, by contrast, maximizes ic3, being made up from two

forms of 4-28, the "ic3 cycle". Its interval vector is [4484441. Further investigation

of the tetrachords in Table 9.2 reveals that 8 out of the 13 are subsets of the octatonic

scale 8-28 (4-3, 4-9, 4-10, 4-13, 4-17, 4-18, 4-26 and 4-28). The change from

background diatonicism to octatonicism thus represents a change of emphasis from

structures built on ic5 to those built on ic3.

Explicit occurrences of 8-28, 4-28 and 3-10 in the B section are frequent; the

succession of tremolando fifths in bars 70-73 outlines two forms of 4-28 moving in

parallel, combining to form 8-28 (see Ex.6.9(b)); the cello line in bars 79-81 and 83-

85, made up primarily from overlapping statements of sets 3-2, 3-3 and 3-7, also

creates forms of 8-28 (see Ex.9. 10). The metrical disposition of the former also

emphasises a segment of the octatonic scale.

Sets 4-28 and 3-10 also play important roles in the climactic passages of the B

section. During bars 90-112 pitch classes 0, 3, 6 and 9 are particularly prominent.

d#-f# (or ei gi) occurs frequently in bars 90-99. The long-sustained c-eu dyad in

the cello in 102-105 is matched by piano dyads f#-a in 106-109 and 110-112. The

260
upper voice in these two piano dyads rises from f# (b. 106) to a (b.110), while the

cello emphasises c in 107-108 and e' in 110. The "association" of these pitches

across passages of intervening music creates a "middleground" statement of set 4-28

(see Ex.9.11).

4. Conclusion

This movement encapsulates the important issues in understanding Carter's

transitional music; (1) the subversion of conventional forms or genres (the scherzo and

trio) and styles (jazz-influenced populism); (2) the development of a harmonic language

which is based on the recombination of the triadic interval classes 3 and 5; (3) the

tentative formation of a principle of contrast through switching emphasis from one

interval class to another and the consequent decay of tonality, however "extended", as

a large-scale organizing factor.

261
 CHAPTER 10: SONATA FOR FLUTE, OBOE, CELLO AND HARPSICHORD,

FIRST MOVEMENT.

1. Introduction

Carter's Quartet Sonata has been described as a relaxation after the strenuous

effort of composing the First String Quartet (1951). For Schiff, the Quartet

represents "a hard-fought victory" [Schiff 1983: 164] in the struggle to create a new

musical language, while the Sonata is the subsequent "spontaneous celebration". In

the later work, "having assimilated and mastered a new language, he was able to use it

in a quick, relaxed and improvisatoiy manner." Two stylistic features in particular -

the comparatively simpler rhythms and textures - distinguish the Sonata from the

String Quartet and justify this view of it as a relaxation. The motivation for this

simplification appears to have been both practical and aesthetic. On one hand, the

Harpsichord Quartet of New York, who commissioned the work, were not a twentieth-

century specialist ensemble and might have found technical complexities as great as

those of the First Quartet too challenging. On the other, according to Schiff, "he

wanted to show that his new approach to musical form was broader in implication than

the specific procedures used in the Quartet" [Schiff 1983: 165 ]. Carter therefore

reduced the role of polyrhythmic counterpoint and metric modulation in the Sonata in

order to reveal this "new approach" more clearly. The work is also the first in which

Carter employs a version of Schoenberg's Hauptstimtnc notation in order to highlight

the principal voice, a technique obviously intended to aid the performers, but which

also suggests a more hierarchical relationship between the sharply differentiated

262
instruments of the ensemble, compared with the "democratic" polyphony of the String

Quartet.

Wilfrid Mellers, who clearly regards the Sonata as welcome light relief in the

midst of the formidable intellectual strenuousness of Carter's output, praises it because

"one is aware that the processes, whatever their origin, have become quintessentially

human" [Mellers: 121]. The formal structures and processes referred to by Schiff

and Mellers are, in common with those of other works of Carter's in the period 1945-

55, generated by patterns of growth, change and decay, and by the interaction between

contrasted instrumental "behavioural types". In this case, "the somewhat acid timbre

of the harpsichord is amplified by the other instruments, and its typical gestures are

anticipated, echoed and modified in terms of their original characters" (Northcott:

22 3]. Carter's formal processes appear to grow naturally and logically from the

musical material, thus creating a balance between logical system and apparently

spontaneous flights of imagination. It is this, together with the ease and freedom with

which Carter manipulates a unified musical language, which gives the work the

particularly relaxed quality remarked on above. There is, however, no abandonment

of intellectual rigour in its construction.

The first movement, being relatively concise and less complex than the other

two, offers a convenient, though not unrepresentative, subject for an analysis of

Carter's "new approach to form" and his "new musical language."

2. Form: processes.

Carter's own comments on the Sonata indicate that its central inspiration was

263
the variety of timbres available on the Pleyel harpsichord, the instrument on which it

was originally played:

the whole range of musical expression, details of shape, phrasing,

rhythm and texture, as well as the large form [...] were all determined
and grew out of a desire to explore the many colorful possibilities of the
modem harpsichord, with the other three instruments serving as a frame
to set this off in best relief, and with their "musical behaviour"
conditioned by this aim. [Edwards: 69]

Analysis of the first movement will confirm that texture (including the manipulation of

timbre) is indeed the most significant parameter in the articulation of its form. While

the harmonic and motivic language of the movement will be examined for its typical

characteristics, it will be observed that harmony and motive have been relegated to a

supporting, rather than a determining, role.

Carter's sleeve notes for the 1969 Nonesuch recording state that the movement

begins "RIsoluto, with a splashing dramatic gesture whose subsiding ripples form the

rest of the movement." [CEL 231] Schiff expands on this comment by suggesting

that the "form of the movement is the attack-decay contour of the harpsichord in slow

motion" [Schiff 1983: 1661 and Bayan Northcott has similarly described the movement

as a "long diminuendo" [Northcott: 223]. This phrase succinctly encapsulates one

aspect of Carter's "new approach" - his creation of form through the use of continuous

processes. Although it is going a little too far to assert, as Schiff does, that "there

are no themes" in this movement [Schiff 1983: 166], passages of large-scale

recapitulation are non-existent, as are the principles of movement between tonal areas

and contrast between opposing thematic ideas so that the movement has little

relationship to any conventional musical form, and thus represents a further step away

from the traditional models which, albeit radically adapted, form the background to the I

264
movements analyzed in Chapters 8 and 9.

In fact, several processes, incorporating rhythmic, metrical and textural factors

as well as dynamics, contribute to the effect of a "long diminuendo. However, as is

typical in Carter's music, all of these processes are complicated by factors which

interrupt them or interact with them, thus making the progress of the movement rather

less predictable and mechanical than the notion of a "continuous process" might imply.

The principle of juxtaposed contrasts continues to be an important one, working here

in combination with the processes mentioned above to produce a structure in which

sectional opposition and continuity are finely balanced. Analysis of the movement's

form will begin with those processes which most clearly embody the "long

diminuendo", before proceeding to more complex issues.

(a) Dyniimics

Most obviously, the movement traces a pattern of dynamic decay, from if

molto marc., draminatico in bar 1 to a quieter level in the later stages. There is,

however, a distinction to be drawn between the constant dynamic levels of the

harpsichord, which can only be modified by changes of registration, and the much

more rapid flexibility of the other instruments, whose dynamic levels may be coloured

by crescendi and diminuendi, accents and forte-pianos.

Analysis of the dynamic and articulation markings, together with the

harpsichord's registrations, suggests a division of the movement into four sections:

A (1-8) B (9l52) C (l5-30) D (30-68)

In section A, the general dynamic level is very loud. All parts are marked

265
molto niarcato, with marcatissimo appearing in the oboe and cello parts in 7-8. The

harpsichord is initially marked if and begins with "Tutti" registration, subtracting the

16' stop after bar 1. The flute, oboe and cello parts are predominantly within the

range mtif and make frequent use of accents, sudden dynamic swells and fp

markings.

Section B is characterized by an immediate contrast. The harpsichord is

marked p throughout, although an effect of gradually increasing volume is created by

the movement of the principal rhythmic part (constant staccato quavers) from the 8'

stop of the upper manual (bb.9-11) to the 4' and 16' stops of the lower (bb.1 1-14), to

which the 8' stop is added in bar 14. The flute, oboe and cello are marked p or jp,

apart from a brief burst of mf at the beginning of bar 14.

Section C brings a return to the dynamic level of A. The harpsichord is once

more marked f, increasing to if from bar 21 with the reintroduction of "Tutti"

registration. For the other instruments, this is the most variable section, with

dynamics ranging from pp to if As in section A, their level is predominantly f or if

and marcato, and there are even more pronounced dynamic swells and subito p or mf

markings. At the end of the section, there is a gradual decline from the high point of

bar 26 (if marc. tutti) to p (flute) in bar 29.

In section D, the harpsichord is marked mf throughout, though small variations

are produced in bars 4 1-53 (pià 1 created through addition of the 4' stop of the lower

manual to the 1/2 8' stop and Coupler) and bars 52-68 (rneno f in left hand, created

through switching to the upper manual). The remaining trio are predominantly p and

though occasional outbursts of f and if still intrude (bb.34, cello, oboe; 41, cello;

266
44, flute, cello; 59, cello; 60, flute). From bar 40, the cello plays pizzicato (apart

from two isolated notes, the quaver d2 in b.53 and the b 5 harmonic in bb.62-4), thus

diminishing its dynamic power.

As can be seen from this brief description, the overall dynamic progress of the

harpsichord part is not so much a "long diminuendo" as an initial alternation between

opposite extremes, balanced by a long passage at an "average" dynamic, as though a

pendulum were left to swing until it came to rest.

(b) Pulse

A second, clearly audible, process working in parallel with the dynamic decay,

is the gradual deceleration of the basic rhythmic pulse in the harpsichord part from

semiquavers in section A to quavers in B and crotchets in D. As with the dynamic

decay, this process takes place in distinct stages rather than continuously; there is no

suggestion of metric modulation (this is reserved for the last movement) or of

continuous deceleration as in the Variations for Orchestra. However, the process is

complicated by other features; (1) while the harpsichord tends to behave

"mechanically", moving in equal note values, the rhythmic behaviour of the other

instruments, like their dynamic behaviour, is much more varied and does not always

match the harpsichord's; (2) the process is temporarily reversed in section C, so that,

just as there is a return to the loud dynamic level of the opening section, there is also a

return to its busy, semiquaver-dominated rhythmic character.

The rhythmic profile of the four brief passages for the harpsichord in A is

essentially the same: each begins with two to four semiquaver attacks, leading to a

267
 sustained chord, and ends with a "tail" of between three and seven semiquavers. The

rhythmic characteristics of the other three instruments are more varied, but in general,

the principal material moves mainly in semiquavers, and other movement is all in note-

values which are multiples of the semiquaver.

In section B the harpsichord plays continuous quavers on one manual,

shadowed by movement in crotchets or dotted crotchets on the other. The flute

doubles the harpsichord in bars 9-10, otherwise, flute, oboe and cello have isolated

phrases, mainly using the dotted quaver-semiquaver figure and long sustained notes.

Section C is characterized by a return to the faster moving pulse of A, with an

added feature of complexity. The phrases of the harpsichord move mainly in

semiquavers, while in the remaining trio, the principal material moves mainly either in

semiquavers or triplet quavers. In a movement which, by Carter's standards, is

relatively restricted in terms of rhythmic resources, the introduction of triplets in the

oboe part in bar 20 is a significant event. In the succeeding bars, triplets appear with

increasing frequency and are played off against semiquavers to produce a more

complex rhythmic texture. Bars 27-8, with five consecutive beats of triplet motion

moving throughout the texture, is the climactic point of this process. In the later

stages of the movement, triplet motion is reduced to a few fragmentary echoes (b.33

oboe, b.48, 50 harpsichord, b.61 flute).

Section D is dominated by the crotchet movement in the harpsichord. Four

sub-sections may be discerned by examining the harpsichord part - (i) 30-36, (ii) 37

40, (iii) 41-53, (iv) 54-68 - each beginning with continuous movement in crotchets,

split between right and left hands and ending with sustained chords. The tendency is

towards longer stretches of movement in equal note values, thus sub-sections (i) and

268
 (ii) begin with respectively ten and thirteen consecutive crotchet attacks. Sub-section

(iii) begins in a similar way, but after twelve crotchets, instigates a gradual increase of

pulse (to triplet quavers in b.48) and decrease (through triplet crotchets in b.50).

Sub-section (iv) consists of no fewer than fifty-six consecutive crotchet attacks and is

thus the culmination of this process. Against this quietly inexorable march, flute,

oboe and cello have material of vely variable phrase-lengths and rhythmic values,

tending increasingly to brief, isolated phrases or single notes (e.g. last 5 bars), but

which nevertheless includes occasional bursts of continuous activity (e.g. cello bb.40-

41, 23 semiquaver run). Between bars 44 and 50 all three move in patterns of

continuous staccato quavers, mimicking the harpsichord's manner (and in particular its

part in section B).

The pattern produced here is essentially one of contrast between sections in

which the harpsichord and the other instruments behave in an approximately similar

fashion (A and C) and those in which their behaviour polarizes (B and D). In

sections A and C, all instruments participate in a texture in which an impression of

continuous but irregular movement is created, while in B and D the harpsichord tends

towards absolute regularity, while the others tend towards greater irregularity. This

interpretation is supported by Carter's use of rubato' and giusto markings. At the

opening of the movement there is a general direction sempre un poco rubato. During

The use of the instruction rubato is unusual in Carter's music at this point. It
certainly has no place in the precisely measured rhythmic relationships of the First Quartet
and even the improvisatory cello line at the opening of the Cello Sonata is marked quasi
rubato rather than rubato. The rubato quality of the latter is conveyed through a variety
of techniques, ranging from the relatively simple, such as the avoidance of movement in
equal note values or on the crotchet beats, to the more complex, such as the use of
varying subdivisions of the beat, polyrhythm and metric modulation. In the Quartet
Sonata, however, only the simplest of these techniques are used.

269
section C, the instruction (sempre rubato) appears on each individual part, perhaps

implying that the continuous quavers in the harpsichord in bars 9-15 should be played

in tempo giusto. Tempo giusto does appear as a general direction at the opening of

section D (b.30), and again as a reminder in the harpsichord part at the beginning of

the second sub-section (b.37). Meanwhile, the other three instruments are again

instructed to play rubato in bars 32-34.

(c) Textural continuity, density and register

The processes discussed thus far are merely the simplest among the many facets

of texture explored by the composer in this movement. Carter frequently uses

register and density to emphasize basic characteristics of textural continuity. In the

flute, oboe and cello parts, there is a textural "decay" from a sustained tutti at the

opening to a fragmented texture at the close, in which the individual lines rarely

overlap and linear continuity evaporates into isolated notes. This process is reversed

in the harpsichord part, which begins with isolated gestures and gradually builds longer

paragraphs. As has been observed with regard to dynamics and pulse, this is not a

smoothly continuous process, since in section B, the harpsichord builds a relatively

sustained passage, while the other instruments have a highly fragmented texture. The

complementary relationship between the continuity processes of the harpsichord and of

the other instruments introduces another important formal factor - the relative

interdependence of the harpsichord and the other instruments.

In section A, the harpsichord initiates and punctuates the material of the other

instruments; its phrases are generally short (between 2.5 and 5.25 crotchets duration),

270
and are separated by rests of similar duration. Each phrase follows a similar pattern

in terms of density and contour: a succession of dyads leads to a sustained chord in the

middle to low register, followed by a rapid descent in a single line of semiquavers (the

very brief second phrase is an exception here). While the harpsichord exploits the

middle to low register (b 1-c5), the flute, oboe and cello play mostly in the middle to

high register (f#3-a6), apart from opening and closing phrases in cello. In contrast to

the harpsichord's well-separated phrases, they play an almost continuous tutti from the

third crotchet of bar 1. There are always at least two parts playing simultaneously,

usually three (the only rests being in the flute: 4.5 crotchets in bb.2-3, and oboe: 3

crotchets in bb.4-5).

In the B section, there is an abrupt change of texture and the harpsichord is

clearly dominant. The flute doubles the harpsichord's melodic line two octaves higher

for two bars, while from 10 to 15, flute, oboe and cello have isolated, fragmentary

phrases, clearly differentiated from the harpsichord's material. This passage thus

prefigures the textural character of later parts of the movement. The harpsichord

plays continuous quaver movement in one hand, sustained 4-part chords in the other,

all within a narrow range in a middle register (b 3-g5). Flute, oboe and cello play

three or four-note phrases, ranging from 1 to 5.5 crotchets in length. There are never

more than two parts together and usually only one plays at a time. All three are

silent for 4.5 crotchets in bars 11-12 and for 2.5 crotchets at the end of bar 14. Their

material is also strongly differentiated from that of the previous section in terms of

register, being in a middle to low range (g# 2-f5). The single pizzicato cello chord in

bar 14 also anticipates the use of this timbre in the movement's final stages.

Section C is perhaps the most complex part of the movement. The texture

271
 tends towards a pattern of alternation between the harpsichord and the other

instruments, with the latter gradually asserting their independence from the former.

As in section A, the harpsichord has several discrete phrases of variable duration,

separated by rests of similar length. The music for the other instruments, as in A, is

mostly presented three parts at once, but its continuity is interrupted by general rests,

in which the harpsichord is allowed to stand out. This tendency culminates in two

brief, climactic "solos", one for the harpsichord (23-4) and one for the trio (26-9).

The latter corresponds with the harpsichord's longest silent stretch in the movement.

Bars 26-29 also effect a climax through the conjunction of the goals of processes

begun earlier; (1) increasing rhythmic variety, through the introduction of triplets (as

discussed above, p.23!) and (2) the expansion of the total pitch-range of the

movement. The pitch extremes of the movement - c2 in the cello, b( 6 in the flute -

are heard in close proximity in bar 27, 4th beat. This is the culmination of a process

of gradual expansion of the pitch rage covered by the flute, oboe and cello. In

section A, the tessitura is predominantly high (above the harpsichord) until the cello

plunges in bar 9. The cello's exploration of its lower register continues as part of a

process of building melodic lines that sweep up or down through registers (see bb. 18-

19, 20-21, 24-25). The bk 6 of the flute has been sounded previously (bb.9, 15, 16).

However, in bar 27, it appears as the goal of a semitonal progression stretching over

six bars. The flute's a6 is sounded in bars 22 and 25, but each time, the ascending

line turns back on itself. In bar 27, however, the progression g#6--bL" 6 in the upper

register of the flute (clearly indicated by the Hauptstimmc markings) resolves the

expectation of upward motion.

After this climactic "sounding-together", in section D, the members of the

272
 instrumental trio become isolated from each other, and the harpsichord quietly asserts

its dominance. Apart from the passage between bars 44 and 49, flute, oboe and cello

do not all play together for the rest of the movement. There are several examples of

overlapping of phrases between two instruments (for example in bb.33-5), but for the

most part, the instruments appear singly, separated by long rests.

The dominance of the harpsichord takes the form of sustained rhythmic motion

(as discussed above) and harmonic "comprehensivenes?. The latter has two stages;

first, the gradual chromatic saturation of a small pitch-space (bb.4 1-53) and second,

the quasi-symmetrical expansion across the entire keyboard range (bb.54-68). 2 These

processes will be examined in the following two paragraphs.

Between bars 30 and 44, the harpsichord resumes a similar texture to that

found in section B, with a "melodic outline" of continuous crotchets forming a

sustained chord of gradually shifting pitch and density (up to seven notes) all within a

restricted middle register (b3 4). However, the third statement of the "melodic

outline", beginning in bar 41, leads to a gradual cluster-like chromatic saturation of a

hand's span on the keyboard. The spacing of the twelve-note sonority which is

formed in bars 46-7 and 48-9 is interesting, however, in that it gives priority to one

pitch among the twelve - bb. The major seventh from at' 4 to g 5 is saturated, but the

bt' is displaced from its "natural" position near the bottom of the cluster and appears

above the g. The focal priority of bt' is also emphasized by the other instruments;

the fourth f-bt appears in octaves between cello and flute in bar 47 and is repeated in

bar 48 in three octaves on all three melody instruments. The bL 6 of the flute in bar

2
Both of these processes may be compared with the "cyclonic" orchestral clusters of
the second movement of the Piano Concerto (1965).

273
 48 marks the high point of the passage, while the crescendo to nip and the sustaining

of the bt 6 for four beats in the oboe also contribute to the prominence of this pitch.

The texture adopted in bars 44-49 is essentially an expansion of the flute's doubling of

the harpsichord in bars 9-10. In the later passage, the melody instruments double one

another freely (exploiting every possible combination), always adhering to the notes of

the cluster sustained by the harpsichord and occasionally doubling its "melodic

outline".

In the final sub-section of the movement (54-68), the fourth statement of the

harpsichord's "melodic outline" begins as before within the narrow span of a perfect

fifth (a3-e4), but the hands gradually diverge from one another, eventually coming to a

halt when a range of four and a half octaves has been opened up (bL 1 -f6), these pitches

representing the harpsichord's registral extremes for this movement. This process of

gradual expansion of pitch range is a massive expansion of the semiquaver gesture

with which the harpsichord begins the movement, a relationship which is confirmed by

examining the harmonic content of the beginning and end of the movement (as

discussed in Chapter 5).

(d) Summary

Carter's metaphor for the form of the movement is, like those used in

describing several of his later works, drawn from the natural world. 3 The image of a

See, for example, his comments on the Variations for Orchestra: "1 have tried to
give musical expression to experiences anyone living today must have when confronted by
so many remarkable examples of unexpected types of changes and relationships of
character, uncovered in the human sphere by psychologists and novelists, in the life cycle
of insects and certain marine animals by biologists, indeed in every domain of science and

274
 "splashing gesture" and its "subsiding ripples" suggests an organic wave-pattern of the

type discussed in Schiffs study [Schiff 1983: 47-50]. However, the essentially

mechanical nature of the harpsichord (especially its inability to graduate changes of

dynamic and tone colour smoothly), coupled with Carter's avoidance of the possibility

of fluid transitions using metric modulation, lead to a rather less continuous, more

clearly divided form. The structure of the movement may be best represented in the

diagram below:

Al I C

B D

This schematic representation emphasizes the relationship between the dynamically and

rhythmically more animated A and C sections (dominated by the rubato style of the

melody instruments) on one hand, and the more restrained B and D sections

(dominated by the giusto style of the harpsichord) on the other. It also suggests a

kinship with the kind of collage or cinematic "intercutting" technique which Carter was

evolving during this period and which he practised on a large scale in the First Quartet

and on a smaller scale in Variation 7 of the Variations for Orchestra and in the second

alt" [WUC 308-91, the Double Concerto: "The work is built on a large plan, somewhat
like that of Lucretius's De Rerum Natura, in which its cosmos is brought into existence by
collisions of falling atoms [...]The coda is the dissolution of this musical cosmos." [CEL
243, 246], and the Concerto for Orchestra "[the music I was writingj seemed to revolve
around four main ideas[.. .1(1 ) the drying up of autumn, [...J; (2) the swiftness and
freshness of the winds that blow away the old and bring in the new; (3) the exhortation of
a shaman-poet calling for rebirth and a destruction of worn-out things; and, finally, (4) the
return of spring and life." ICEL 277]

275
 and third movements of the Quartet Sonata. Interestingly, in its internal proportions,

it represents the exact opposite of the first movement of the Piano Sonata. Whereas

the earlier work proceeded through a succession of sections of increasing concentration

and intensity, in the later work, the sections are increasingly expansive.

3. Motivic and harmonic material

The preceding discussion has strongly suggested that it is essentially rhythmic

and dynamic, rather than motivic and harmonic, factors which generate the

movement's form. Carter's approach to thematic material in this movement is in

some respects closer to his later "athematic" works of the 1960s and '70s, than are

even such sophisticated pieces as the First Quartet and the first movement of the Cello

Sonata. In those works, despite the fact that thematic ideas are in a state of restless

transformation and recombination, the articulation of structure relies to a large extent

upon the recurrence and sometimes extensive recapitulation of recognizable harmonic

or melodic motives (that is, "ordered" statements, not just recurrent set classes). In

the Quartet Sonata's first movement, however, although pitch-structuring processes are

discernable, motives rarely achieve a "definitive" form which can be referred back to,

and, compared with the movements analyzed in Chapters 8 and 9, there is almost no

recapitulation of the type described above. Perhaps the closest relatives of this

movement among Carter's oeuvre are those of the Eght Etudes and a Fantasy that are

principally concerned with texture rather than thematic processes.

As in many Carter works, the opening contains the seeds of both the harmonic

276
language and the gestural vocabulary of the movement. However, it should be borne

in mind throughout the ensuing discussion that, while different gestural or

"behavioural" types are evolved for the harpsichord on one hand and the remaining

trio on the other, in the domain of harmony there is no such opposition. Instead, as

in most of the works before the Second String Quartet, there is a homogeneity of

harmonic fabric, in which all parts share equally in the exploration of the intervallic

potential of a "key-chord".

(a) Rhythmic and pitch-contour cells

The thematic material of the movement is perhaps best understood as being

generated through the continual transformation and combination of a handful of

extremely small and simple rhythmic and pitch-contour cells, which are contained

within the first bar of the harpsichord part. Most of the significant motivic ideas of

the movement are evolved during the opening A section (bb. 1-8), which will be the

focus of the following discussion.

The initial rhythmic "head-motive" - four semiquavers leading to a minim -

proves to be highly plastic, partly because of its ambiguous metrical stress; the

semiquavers appear during the rest of the passage both as an anacrusis and as a

downbeat. The first four phrases of the harpsichord each offer a different variant on

this basic idea, for example, by adding or subtracting semiquavers from the initial

group, and bringing forward the attack on the long note (see Ex. 10.1). The

distinctive pitch-contour of this head-motive, a widening zig-zag, is also subjected to

transformations. The original motive is symmetrical about the pitches d 4 and eLi 4 , but

277
 subsequent appearances in the harpsichord distort this symmetry4.

The harpsichord's "tail-motive" - a sequence of descending semiquavers

escaping from a tied-over long note and avoiding metrical stress either at the beginning

or end - is similarly elastic, ranging from three to seven notes and being omitted

entirely in bar 3 (also, see Ex.lO.l).

These variations in length and shape are exaggerated in the parts for the flute,

oboe and cello in bars 1-8 (see Ex.lO.2). The head-motive never appears exactly as

on its first occurrence; its least radical variant is that heard in the flute and oboe in the

first bar, extended by a single extra semiquaver anacrusis. Much more elaborate

extensions of the semiquavers as an anacrusis can be found in the oboe in bars 2-3',
34.41 and 5161. The other most significant variants on the rhythm of the motive are

the contraction of the semiquaver group to a single semiquaver, producing a "scotch-

snap" figure, and its inverse, a dotted-quaver-semiquaver anacrusis. These two

rhythms, together with the augmentation of the semiquavers to quavers, come to

dominate the texture during the last three bars of the section, thus gradually preparing

for the quaver pulse of section B. The zig-zag contour of the head-motive is also

influential in shaping these lines, whether in passages which refer unambiguously to

the rhythm of the harpsichord's first bar (such as the cello part in b.2, or the oboe in

bar 4), or in slower-moving lines (such as the flute part in bb.5-6).

The harpsichord's "tail-motive" plays a lesser part in the formation of the flute,

oboe and cello lines, although its descending contour, greatly elongated, can be

detected in passages such as the flute part of bars 3342 and the cello part of bars 5471

The implications of this symmetry, and the exact intervallic content of this motive
will be explored later.

278
 Analysis of the harpsichord's opening gesture reveals the origins of other

important cells generated in section A. Dividing the harpsichord's first two dyads

into a central falling semitone (e -d) encased within a falling interval class 7 (gl -b)

produces two intervallic cells which have further consequences. Examples 10.3 and

10.4 illustrate Carter's use respectively of the falling semitone and the falling "perfect

fifth" (and their inversions), within section A. Still other cells derive from the flute

part of bars 1-2 (itself a variation on the harpsichord's first phrase); the upward leap

of a major tenth from b4 to d#6 in bar 1 reappears prominently in flute or oboe in bars

3, 4, 6 and 7-8 (see Ex. 10.5), and the conjunct motion d#-c#-b#-c# of bar l-2' is

developed extensively by the oboe (see Ex. 10.6).

These first eight bars are typical of the movement as a whole in their protean

re-shaping of simple intervallic cells, a process in which all instruments participate,

creating a polyphonic texture. The net result of Carter's technique here is that

practically all the melodic motion, whether by leap or step, and all the rhythmic ideas,

which variously divide the crotchet beat into groups of semiquavers, are perceived as

belonging to the thematic fabric. As has been discussed above, the only significant

rhythmic cell not introduced here is the triplet quaver idea, which is incorporated into

the texture from bar 20 onwards, and the only exceptions to the general procedure are

those passages in which a mechanical regularity of pulse obtains.

(b) The "key-chord" and its relatives.

There is very little discussion in the literature on Carter of the harmonic

language of the Quartet Sonata. Schiff states that the Sonata's "harmonies derive

279
 from the opening bars of the harpsichord" (Schiff 1983: 1651. This statement is

accompanied by a chart labelled "Harmonic Scheme"(see Ex.10.7), which illustrates

some of the symmetrical structures contained within the harpsichord's first four

semiquavers. The pitch-configurations labelled in Schift's chart as "chords" may be

interpreted as sets 4-17, 4-20 and 4-7, and their similar intervallic structure of ic4+ic4

is explicitly shown.

As has previously been discussed (see Chap.5, pp.124-5), these three

tetrachords are subsets of the hexachord 6-20 and it is the latter which seems the most

likely candidate as "key-chord" or source-set for the movement, by virtue of its

properties of symmetry, the tightly unified nature of its Kh subcomplex and its

prominent appearances within the movement as a referential sonority. These

appearances have already been noted (see p.129 and Exx. 5.15, 5.16 and 5.17). As

in the Cello Sonata, Carter does not restrict himself to a vocabulary consisting entirely

of members of the Kh(6-20) subcomplex. The tetrachord 4-18, which has a close

relationship with the "quasi-seventh" chords 4-17, 4-19 and 4-20 (see Chap.5, p.116)

is a significant part of the movement's make-up. Another important group is formed

by those tetrachords which share with 4-7, 4-17 and 4-20 a symmetrical pairing of

identical interval classes, particularly ics 1, 3 and 5, which are constituents of the

interval vector of the "key-chord" 6-20 g. The important trichordal additions to Kh(6-

20) are 3-2 and 3-5, which share with 3-3 and 3-4 an intervallic structure containing

one instance of ic 1. Table 10.1 summarizes the significant members of the

movement's harmonic vocabulary.

The intervallic structure ic4 + ic4 has been omitted here since it merely duplicates the
transitive group 4-7, 4-17 and 4-20 and adds the "whole-tone" tetrachords 4-21 [0,2,4,6]
and 4-28 [0,2,6,8], neither of which are used in this movement.

280


Table 10.1. Set-class vocabulary of Quartet Sonata, first movement.

Kh(6-20)

6-20 10,1,4,5,8,91 3-3 [0,1,41

5-21 [0,1,4,5,81 3-4 [0,1,5]
4-7 [0,1,4,51 3-11 [0,3,71
4-17 [0,3,4,71 3-12 10,4,81
4-19 10,1,4,81
4-20 [0,1,5,81

"Triad + seventh chords" Symmetrical tetrachords

4-17 [0,3,4,71 4-3 [0,1,3,41 Ic! +jcl/ic3+ic3

4-18 [0,1,4,7] 4-7 [0,1,4,51 id +icl/ic4+ic4
4-19 [0,1,4,81 4-8 [0,1,5,6] Ic! +icl/ic5+1c5
4-20 10,1,5,81 4-9 [0,1,6,71 id +icl/ic6+ic6
4-10 [0,23,5] ic2+ ic2fic3 + 1c3
4-23 [0,2,5,7] ic2+ ic2Jic5+ 1c5
4-26 [0,3,5,8] ic3 + ic3lic5 +ic5
4-28 [0,3,6,9] ic3+ic3/ic6+ic6

Tnchords including o Ic!

3-2 [0,1,3]
3-3 [0,1,4]
3-4 [0,1,5]
3-5 [0,1,61

The following examples will demonstrate Carter's deployment of these sets

within passages of contrasted character, namely sections A and B of the movement.

Example 10.8 shows the harpsichord part of section A, which may be segmented

straightforwardly. The example shows clearly Carter's manipulation of sets formed

from the (often symmetrical) combination of interval classes 1, 3, 4 and 5. Each of

the four phrases begins with two dyads of identical interval class, the first three using

ic3, the last ic4. The tetrachords thus formed are 4-17, 4-3, 4-28 and 4-17 again (the

281
last phrase therefore presents the complementary pairing of intervals to the first). As

the head-motive of the first phrase widens out, two forms of set 4-20 appear, still

distributed symmetrically about the d 4-eL' axis. Since these tetrachords share the

pitch classes f and c, when they are sustained in combination, they form set 6-20.

The cello triad g 2-d3-bti completes several symmetries; (1) a form of 4-20 combined

with the e? on the second beat in the harpsichord; (ii) this form of 4-20, combined

with the initial 4-17 of the harpsichord, completes another form of 6-20; (iii) this

hexachord is in turn the complement of that formed by the symmetrically paired

tetrachords on the second quaver. Thus, the first two beats of the movement present

all twelve notes in a highly structured fashion. However, the symmetry about the d-eI?

axis is slightly disturbed by the displacement of the cello's d and g into lower octaves.

This slight "skewing" of strictly symmetrical presentation in the first phrase is

exaggerated in the subsequent phrases by similar registral displacements and distortions

of interval.

The third phrase arrives at a sustained harmony which is asymmetrical, but

which nevertheless shows an intervallic structure built from ics 3 and 5. Taking the

whole pitch-class content of the harpsichord part of bar 4 produces the hexachord 6-27

[0,1,3,4,6,9], a set whose interval vector [225222] is unique among hexachords in its

maximization of interval class 3. The fourth phrase also has an asymmetrical central

chord, although again its construction from ics 3 and 5 is obvious and the interplay of

overlapping versions of the tetrachords 4-8 and 4-19 show its integration with the rest

of the harmonic fabric.

In each case, the "tail-motive" of the harpsichord's phrases are formed from

subsets of 6-20. The first phrase uses two overlapping forms of 5-21 (or non-

282
overlapping forms of 4-19), the third and fourth phrases single statements of 4-20 and

5-21 respectively.

Example 10.9 examines the flute, oboe and cello parts in the same section.

Though these are clearly not as tightly organized as the harpsichord part, the

coordinating role of certain sets, especially 4-18 and 4-19, can be seen. A particular

feature of these primarily melodic rather than harmonic parts is the repeated use of

sets 3-2 and 3-3 as quasi-scalic connective material. Example 10.10 shows the

development of a particularly important idea, deriving from the use of set 3-2 in this

section, which establishes a strong connection between the melodic and harmonic

dimensions of the movement. Initially, the oboe figure in bar 2 appears as a

development of the motive in the flute in bars l-2' which outlines trichord 3-2: d#-c#-

b#-c#. The pervasiveness of 3-2 in the oboe line in bars 1-5 gives rise to a quasi-

symmetrical falling and rising linear pattern, which is taken up by the harpsichord in

bar 16. This achieves its definitive form in bars 30-31, where the motive receives its

clearest expression as the "melodic outline" of the harpsichord part and also gives rise

to the symmetrical pairing of ic3s to form 4-17, the movement's opening sonority.

Example 10.11 demonstrates the integration of the musical language in section

B of the movement. This example is a more detailed analysis of the passage

examined in Ex.5.9, and is derived from the harpsichord part of bars 9-15 and the

flute part of bars 9-10. Since these parts consist of an unbroken string of equal note

values, some of which are sustained in a gradually changing four-note chord, the

problem of segmentation has been approached in the following way; the melodic line

has been subjected to a process of imbrication in order to extract each group of three

or four pitch-classes and each successive harmony has been recorded as an individual

283


unit. Table 10.2 shows a statistical summary of the data in Ex.10.11, which, though

crude, clearly illustrates the central importance of the sets listed in Table 10.1.

Furthermore, the culmination of the passage in a melodic statement of 6-20 and a

harmonic one of 4-19, combining with 4-17 to produce its complement 8-19, is a

striking example of Carter's control over the possibilities of his harmonic material.

Table 10.2 Statistical summary of trichords and tetrachords shown in Ex. 10.11.
Sustained tetrachords (duration in quavers/number of separate occurrences):
4_7* 3/1 4_17* 5/2 4-22 3/1
4-8 10/3 4-18 2)1 4-27 211
4-13 3/1 4-19' 11/4
4-15 6/2 420* 3/1

Melodic tetrachords (number of separate occurrences):

4-4 1 4-11 1 417* 2 4-22 1

4-5 1 4-13 2 4-18 2 4-24 1
4.7* 1 4-15 2 419* 6 4-27 1
4-8 2 4-16 1 420* 4

Melodic trichords (number of separate occurrences):

3-1 1 3-5_ 5 3-9 1
3-2 1 3-6 1 3-10 1
3-3* 3 3-7 3 311* 9
3-4* 8 3-8 0 312* 3

* = Kh(6-20) members
-= otherrelatedsets(seeTable 10.1)

Example 10.12 shows the music of the other instruments during bars 10-16.

Though clearly contrasted in character with the harpsichord, in terms of set-class

content, their material is clearly related, thus demonstrating that at this stage in his

development, Carter was not yet thinking in terms of contrasted harmonic

vocabularies.

284
4. Large-scale pitch-organization

It is doubtful whether harmony, voice-leading or thematic process can offer

much illumination of the large-scale progress of the movement. However, these areas

will be briefly investigated in order to contrast with Carter's approach here with that in

his previous works.

(a) Thematic recall

Between the four main sections of the movement, there is very little

recapitulation of material. The flute, oboe and cello parts at the opening of section C

(bb.154-17') recall bars 1-2 (see Ex.10.13) while the harpsichord refers to the oboe

part of bar 5. However, after this brief reminiscence, the material develops in new

directions. Occasional symmetrical gestures help to shape the movement. Example

10.14 shows a particularly clear and striking one which combines symmetry of pitch

with that of form. The descending form of 5-5 in the harpsichord in bar 19 is

literally inverted in bar 25, and this pair of brief gestures surrounds the larger

statement of bars 21-24. Towards the end of the movement there are a few ghostly

shadows of previously heard melodic shapes, for example, the oboe part in bar 43

recalls the flute in bars 1-2, the flute in 53-57 suggests its own line from 4-5, and the

cello in 59 relives the surging semiquavers of the oboe part in bar 2 (see Ex.10.15).

285
 (b) Pitch fields

In this work, Carter has clearly advanced beyond even the most highly

chromatic passages of the movements analyzed in Chapters 8 and 9. Analysis in

terms of diatonic fields would therefore be entirely inappropriate. Nor, on the other

hand, apart from isolated passages, such as bar 1, bars 23-4 and bars 41-50, does

Carter set about the systematic aggregation of all twelve pitch-classes. However,

referring back to Ex. 10.11, we can see a different example of chromatic

complementation. A tabulation of the occurrences of each pitch-class in the

harpsichord part from the beginning of bar 9 to the end of the second beat of bar 15

produces the results shown in Table 10.3.

Table 10.3 Pitches occurring in Ex. 10.11.

(number of occurrences)
c 8 e 2 g# 0
2 f 0 a 1
d 7 f# 6 bb 5
eb 5 g 8 b 6

This clearly shows the focal importance of the pitch g, which is sustained in the

harpsichord for two long stretches, first in the middle of the harmony, then as its bass.

On the third beat of bar 15, it forms the accented high point of the staccato quaver

line. The p-c content and distribution in this passage also has a faintly "G

minor/major" colour, particularly strongly felt in bars 9-10 and 13-14. The absence

of f and g# from the harpsichord part during this passage contributes to the feeling of

286
 "g-centredness". However, these two pitches feature prominently in the other

instrumental parts (see Ex. 10.12, cello b. 10, oboe b. 11 and oboe & cello b. 14), thus

making up the full chromatic complement.

(c) Linear and registral connections

The problems of discerning genuine middleground and background motives in

post-tonal music such as this have been discussed in Chapter 6. The observations

which follow will attempt to show voice-leading connections which are close to the

surface of the music and reasonably palpable.

The most obvious means of unifying the musical texture is the passing of

pitches and/or intervals between instruments and registers. For example, in bars 19-

20 the e 6-d6 at the end of the flute phrase is taken up at the beginning of the oboe

phrase two octaves lower (see Ex. 10.16). A similar use of this pervasive semitone

motive occurs in bars 32-35, as the motive g-f# is passed from flute to cello and c-b

from oboe to flute (see Ex.l0.17).

Over a longer period, the repetition of a particular p-c in a particular register

or the linking of pitches in stepwise progressions are the clearest means of establishing

structural connections. Example 10.18 illustrates the use of these procedures in

identifying a "middleground motive". There are two basic criteria for the abstraction

of the p-cs f#-g#-a-g#-f# from the flute line of bars 3-8. Firstly, they stand out from

the surrounding texture by virtue of their registral extremity (emphasized by wide "up-

beat" leaps). Secondly, they appear to parallel a more obvious "foreground" pattern

which appears in this passage - a stepwise ascent and/or descent through a minor third,

287
 alternating half- and whole-steps (i.e. two overlapping statements of 3-2 producing 4-

1). The flute and oboe line of bars 1-2 (d#-c#-b#-c#) and the cello line in bars 4-5 (f-

g -g-{c}-a) present this idea at "foreground" level, while the registral (and metrical)

connection of the oboe's b-bk in bar 3 to the flute's b#-c# in the previous bar presents

the motive at an intermediate level. The importance of this idea is confirmed by its

reappearance in condensed form in bar 16 (flute and oboe) in a passage which recalls

the opening in rhythm, contour and texture.

It is difficult, however, to get beyond these tentative observations. The motive

identified here does not recur, apart from the foreground "ghost" in the oboe in bar

43, nor is it possible to single out one of its pitches as structurally more significant in

the sense that it might form an element of a motive on an even higher plane of

abstraction. It is perhaps best to regard it as an example of unconscious clinging on

to a mode of musical coherence that the composer was soon to outgrow.

5. Conclusion

The formal and harmonic characteristics of this movement clearly indicate a

further stylistic development from the Cello Sonata. Processes of change - many of

these involving factors such as rhythm, register and texture, rather than motives and

harmonies - have replaced patterns of repetition as the fundamental basis of form.

The harmonic language itself is fundamentally atonal, with only the slightest of

references to conventional tonal chords and functions. The composer has created a

harmonic vocabulary with a strong sense of focus and unity stemming from the

symmetrical properties of the key-chord, but which contains a high degree of

288
intervallic variety. Perhaps most importantly, Carter has taken the idea of contrasting

instrumental characters a stage further by pitting the quasi-mechanical aspect of the

harpsichord against the more expressive, "human" qualities of the other instruments.

289
CHAPTER 11: CONCLUSION

The foregoing chapters have illustrated something of the complexity of Carter's

musical language by isolating its elements and then showing how they are combined in

individual works. A summary of these findings is required in order to clarify the

argument of the thesis, but it should be borne in mind that each of the works studied

here is unique in its particular combination of compositional materials and techniques

and that generalizations applied to an oeuvre which covers such varied terrain

inevitably risk distorting this uniqueness and minimizing the differences between

works. All of this is to say that in analyzing works of a "transitional" nature, there

must be a balance between evaluating the works themselves and tracing the continuity

of the process of transition. The teleology inherent in the latter must be tempered by

an acknowledgement that there may be "processes" at work, rather than a single

process, and that these may progress at different rates and are not necessarily

continuous at all.

Two ideas shared by Carter with other composers of transitional music are the

reinterpretation of certain sets and chords and the gradual weakening of tonal function.

In the area of pitch vocabulary, it is clear that Carter turned away from a language in

which diatonic formations were the norm towards one in which all twelve pitch-classes

circulate freely, and hence from a world in which the distinction between consonance

and dissonance operated to one in which contrasts of intervallic "colour" or

"character" were more important. Bridging across this division is Carter's use of

families of p-c sets, related through their intervallic content, in order to achieve the

twin goals of unity and variety of vocabulary. The most prominent of these families

290
of sets are clearly those combining interval classes 3, 4 and 5, which can be used to

create a diatonic context or an "atonal" one. These families of sets may be notionally

coordinated through relation to a central "key-chord" or source-set, though in practice,

the prominence and role of such a set varies considerably from work to work.

In the area of voice-leading, a gradual process of loosening may be observed.

Techniques which bear an obvious resemblance to those of tonal prolongation (octave

coupling, neighbour- and passing-notes, arpeggiation) give way to the linear

association of pitches in patterns determined by the context of the work rather than by

pre-existing concepts. In the later works, pitch centricity ceases to operate except at

the very local level, thus weakening the sense of a "middleground" defined in terms of

voice-leading. In common with the music of many other post-tonal composers,

factors such as symmetry and complementation occasionally play a role in the small-

scale organization of pitch materials.

Carter showed an early preference for forms which rely on continuous

evolution rather than sectional juxtaposition. This is not to say that his music does

not sometimes exhibit a kaleidoscopic variety of ideas and textures (the second

movement of the Quartet Sonata is a good example), but that this is usually

incorporated into an overarching dynamic process. Thus the conventional structures

of "sonata form" and "scherzo and trio", which lie behind the first movement of the

Piano Sonata and the second movement of the Cello Sonata respectively, are already

subjected to the pressures of evolutionary change which mean that their final sections

are very far from being literal repeats of earlier material. The first movement of the

Quartet Sonata offers a particularly transparent example of a work whose form is

created through the setting-in-motion of certain processes of change, with little or no

291
reference to existing formal types.

Carter's approach to thematicism is inextricably bound up with formal issues.

An evolutionary formal process requires adaptable thematic material; brief, sharply

characterized motives, rather than fully-fledged themes, and rows of pitches with

variable rhythmic values are Carter's favoured resources during this period. In the

later works of the period, there is often a certain "neutrality" to the compositional

materials used; the polyrhythmic counterpoint of passages in the Cello Sonata, the

Quartet Sonata and the First Quartet is created through superimposing streams of

pitches moving in equal note values, many of which fall into quasi-scalic or quasi-

arpeggio patterns. This has the effect of diverting attention from the actual pitch-

content of the passage to the rhythmic processes.

The end result of this study is to show that Carter's works of 1945-55 throw up

a particularly rich nexus of compositional and analytical issues, demanding another

new combination of the elements of p-c set theory and neo-Schenkerian analysis,

together with considerations of other factors peculiar to the composer.

292
APPENDiX 1: Cross-Reference of Carter's Listing of Three- to Six-Note Chords
with Forle's List of Prime Forms of Pitch-Class Sets

Carter Forte

numbering p-c ordering numbering

3-1 [0,4,8] 3-12

3-2 [0,3,6] 3-10
3-3 [0,2,41 3-6
3-4 [0,1,2] 3-1
3-5 [0,2,7] 3-9
3-6 [0,3,7] 3-11
3-7 [0,1,6] 3-5
3-8 [0,2,6] 3-8
3-9 [0,1,5] 3-4
3-10 [0,2,5] 3-7
3-11 [0,1,4] 3-3
3-12 [0,1,3] 3-2
4-1 [0, 1,2,31 4-1
4-2 [0,1,6,71 4-9
4-3 [0,2,3,51 4-10
4-4 [0,2,5,7] 4-23
4-5 [0,3,6,9] 4-28
4-6 [0,1,2,7] 4-6
4-7 [0,1,3,6] 4-13
4-8 [0,1,4,5] 4-7
4-9 [0,1,3,4] 4-3
4-10 [0,1,5,61 4-8
4-11 [0,2,4,6] 4-21
4-12 [0,2,6,8] 4-25
4-13 [0,3,4,71 4-17
4-14 [0,3,5,8] 4-26
4-15 [0,1,5,8] 4-20
4-16 [0,2,4,8] 4-24
4-17 [0,1,2,4] 4-2
4-18 [0,1,4,6] 4-Z15
4-19 [0,1,5,7] 4-16
4-20 [0,1,2,5] 4-4
4-21 [0,1,4,7] 4-18
4-22 [0,1,2,6] 4-5
4-23 [0,1,3,71 4-Z29
4-24 [0,3,4,8] 4-19
4-25 [0,2,3,7] 4-14
4-26 [0,1,3,5] 4-11
4-27 [0,2,4,7] 4-22
4-28 [0,2,3,6] 4-12
4-29 [0,2,5,8] 4-27

293
Carter Forte
numbering p-c ordering numbering

5-1 10,1,2,3,4] 5-i

5-2 [0,2,3,4,6] 5-8
5-3 [0,3,4,5,8] 5-Z37
5-4 [0, 1,2,6,8] 5-15
5-5 [0,1,3,5,6] 5-Z12
5-6 10,2,4,6,8] 5-33
5-7 [0,2,4,7,9] 5-35
5-8 [0,1,4,7,81 5-22
5-9 [0,2,4,6,91 5-34
5-10 [0,1,3,4,8] 5-Z17
5-11 [0,1,2,3,5] 5-2
5-12 [0,1,2,3,61 5-4
5-13 [0,1,2,3,71 5-5
5-14 [0,1,2,4,5] 5-3
5-15 [0,1,2,4,6] 5-9
5-16 10,1,2,4,71 5-Z36
5-17 [0,1,2,4,8] 5-13
5-18 [0,2,3,4,7] 5-11
5-19 [0,1,3,4,6] 5-10
5-20 [0,1,3,4,7] 5-16
5-21 [0,1,4,5,8] 5-21
5-22 [0,1,3,5,7] 5-24
5-23 [0,1,3,5,8] 5-27
5-24 [0,2,3,5,81 5-25
5-25 [0,2,3,5,7] 5-23
5-26 [0,2,4,5,8] 5-26
5-27 [0,1,2,5,6] 5-6
5-28 [0, 1,2,5,7] 5-14
5-29 [0,1,2,5,8] 5-Z38
5-30 [0,1,2,6,7] 5-7
5-31 [0,1,3,6,7] 5-19
5-32 [0,1,3,6,81 5-29
5-33 [0,1,3,6,9] 5-31
5-34 [0,1,3,7,8] 5-20
5-35 [0,1,4,5,7] 5-Z18
5-36 [0,2,3,6,8] 5-28
5-37 [0,1,4,6,8] 5-30
5-38 [0,1,4,6,91* 5-32

*Jfl [Schiff 1983] this set appears as [0,1,4,7,91. However, the true prime form of
the latter is [0,2,3,4,7], making it identical with Carter's 5-18 or Forte's 5-11.

294

Carter Forte
numbering p-c ordering numbering prime form
(if different from Carter's ordering)
6-1 [0,2,4,6,8,10] 6-35
6-2 [0,1,4,5,8,91 6-20
6-3 [0,1,3,4,5,81 6-14
6-4 [0,1,2,3,4,5] 6-1
6-5 [0,2,3,4,5,71 6-8
6-6 [0,2,4,5,7,9] 6-32
6-7 [0,1,2,6,7,8] 6-7
6-8 [0,1,4,5,7,9] 6-31
6-9 [0,1,3,5,7,9] 6-34
6-10 [0,1,2,4,6,8] 6-22
6-11 [0,1,4,5,6,8] 6-16
6-12 [0,2,3,4,6,8] 6-21
6-13 [0,1,2,4,5,8] 6-15
6-14 [0,2,3,5,6,9] 6-27 [0,1,3,4,6,9]
6-15 [0,1,3,6,7,91 6-30
6-16 [0,1,2,3,6,71 6-5
6-17 [0,1,2,5,7,81 6-18
6-18 [0,2,3,5,7,9] 6-33
6-19 [0,1,2,3,4,61 6-2
6-20 [0,1,2,3,5,7] 6-9
6-21 [0,1,3,5,6,9] 6-Z28
6-22 [0,1,3,4,7,9] 6-Z49
6-23 [0,1,2,3,4,8] 6-Z37
6-24 [0,1,2,4,5,6] 6-Z4
6-25 [0,1,2,5,7,9] 6-Z48
6-26 [0,1,3,5,7,8] 6-Z26
6-27 [0,2,3,5,6,8] 6-Z23
6-28 [0,2,3,4,6,9] 6-Z45
6-29 [0,1,3,4,6,7] 6-Z13
6-30 [0,1,2,3,6,9] 6-Z42
6-31 [0,1,4,6,7,9] 6-Z50
6-32 [0,1,3,6,8,9] 6-Z29
6-33 [0,1,2,5,6,7] 6-Z6
6-34 [0,1,2,3,7,8] 6-Z38
6-35 [0,1,2,4,7,8] 6-Z17
6-36 [0,1,2,5,6,81 6-Z43
6-37 [0,1,3,4,7,81 6-Z19
6-38 [0,1,2,5,6,9] 6-Z44
6-39 [0,1,3,4,6,8] 6-Z24
6-40 [0,1,2,4,6,9] 6-Z46
6-41 [0,2,3,4,5,8] 6-Z39
6-42 [0,1,3,4,5,7] 6-Z10
6-43 [0,1,3,5,6,8] 6-Z25

295
Carter Forte
numbering p-c ordering numbering

6-44 [0,1,2,4,7,9] 6-Z47

6-45 [0,1,2,3,6,8] 6-Z41
6-46 [0,1,2,4,6,7] 6-Z 12
6-47 [0,1,2,4,5,71 6-Zil
6-48 [0,1,2,3,5,8] 6-Z40
6-49 [0,1,2,3,5,6] 6-Z3
6-50 [0,1,2,3,4,7] 6-Z36

296


APPENDIX 2: Classification of Diatonic Sets

TYPE A TYPE B TYPE C

5-35 [0,2,4,7,9] 6-32 [0,2,4,5,7,9] 7-35 [0,1,3,5,6,8,10]

Subsets: Subsets: Subsets:

Tnchords: Trichords: Trichords:

3-6 [0,2,4] as for 5-35 + as for 6-32 +

3-7 [0,2,5] 3-2 [0,1,3] 3-5 [0,1,6]
3-9 [0,2,7] 3-4 [0,1,5] 3-8 [0,2,6]
3-11 [0,3,7] 3-10 [0,3,6]

Tetrachords: Tetrachords: Tetrachords:

4-22 [0,2,4,7) as for 5-35 + as for 6-32 +

4-23 [0,2,5,7] 4-10 [0,2,3,5] 4-8 [0,1,5,61
4-26 [0,3,5,8] 4-11 [0,1,3,5] 4-13 [0,1,3,6]
4-14 [0,2,3,7] 4-16 [0,1,5,7]
4-20 [0,1,5,8] 4-21 [0,2,4,6]
4-27 [0,2,5,81
4-Z29 [0,1,3,7]

Pentachords: Pentachords:

5-35 + asfor6-32 +
5-23 [0,2,3,5,7] 5-Z12 [0,1,3,5,6]
5-27 [0,1,3,5,8] 5-20 [0,1,3,7,8]
5-24 [0,1,3,5,7)
5-25 [0,2,3,5,8]
5-29 [0,1,3,6,8]
5-34 [0,2,4,6,9]

Hexachords:

6-32+
6-Z25 [0,1,3,5,6,8]
6-Z26 [0,1,3,5,7,8]
6-33 [0,2,3,5,7,9]

297
APPENDIX 3: Formation of Tetrachords from Pairs of Non-overlapping Dyads

id + id ic3 + ic3
ic and pc set interval ic and pc set interval
structure vector structure vector

13 34
I'I{ fllJ 4-1 [321000] 3({ H 3 1 4-3 [2121001
02 01

14 35
i'l{ }l'I 4-3 12121001 (3({ }J3 4-10 [122010]
03 02

15 37
l'l{ }l'l 4-7 [201210] 3J{ }Il 4-17 [102210]
04 04

16 38
l'I{ H') 4-8 1200121] )3({ }(3( 4-26 [012120]
05 05

17 39
l'l{ H') 4-9 [200022] (3{ fl31 4-28 [004002]
06 06

ic2 + ic2 ic4 + ic4

icandpc set interval ic and pc set interval
structure vector structure vector

23 45
12){ })2) 4-1 [3210001 4){ ))4) 4-7 [201210]
01 01

25 47
121{ }121 4-10 [1220101 )4({ ))4) 4-17 [102210]
03 03

26 05
)2){ })21 4-21 [030201] (41{ })4) 4-20 [101220]
04 81

27 46
)2) { } (2) 4-23 [021030] (4({ ))4) 4-21 [030201]
05 02

28 06
121{ })21 4-25 [0202021 (4) { } 4) 4-25 10202021
06 82

298
ic5 + ic5 id + ic2

icandpc set interval ic and pc set interval

structure vector structure vector

56 14
1 5 1{ }1 5 1 4-8 [200121] l'l{ } 1 2 l 4-2 [221100]
01 02

06 15
1 5 1{ }1 5 1 4-9 [200022] l'l{ })23 4-11 [121110]
71 03

51 16
II{ }1 5 1 4-20 [101220] I 1 I{ }1 2 1 4-Z15 [111111]
08 04

57 17
lI{ )1 5 1 4-23 [021030] I'I{ }1 2 1 4-16 [1101211
02 05

58
1 5 1{ )II 4-26 [012120] id + ic3
03
icandpc set interval
structure vector
1c6 + ic6
23
icandpc set interval l'I{ )1 3 I 4-1 [321000]
structure vector 10

67 15
J6( }1 6 1 4-9 [200022] I'I }1 3 1 4-4 [211110]
01 02

68 16
16){ }161 4-25 [020202] I 1 I }1 3 1 4-13 [112011]
02 03

69 17
J6J{ )1 6 1 4-28 [004002] I'I{ }(3( 4-18 [102111]
03 04

18
I l l { } 1 3 1 4-20 [101220]
05

299
id + ic4 id + ic6

ic and pc set interval ic and PC set interval

structure vector structure vector

24 26
'I{ }141 4-2 12211001 1'I{ }j6 4-5 12101111
10 10

16 36
I l I{ H41 4.5 [2111101 I'I{ }161 4-12 11121011
02 20

18
I'I{ }141 4-19 11013101
04 ic2 + ic3

17 ic and PC set interval

I'I }141 4-Z29 [1111111 structure vector
03
24
2 1{ }131 4-2 12211001
id + ic5 01

ic and pc set interval 26

structure vector 121{ }131 4-12 [112101]
03
25
I'J{ }151 4-4 [211110] 27
10 1211 }131 4-22 [0211201
04
17
I'I{ }I5I 4-6 12100211 28
02 1211 }131 4-27 (012111]
05
35
I 1 L( }151 4-10 1122010]
20

30
I'I{ }I5I 4-14 [1111201
27

40
I'I{ H51 4-17 (1022101
37

300
ic2 + ic4 ic2 + ic6

icandpc set interval Ic and pc set interval

structure vector structure vector

34 27
121{ H"I 4-3 [212100] 1211 )161 4-6 [210021]
10 01

25 36
121{ }141 4-4 [211110] 1211 )j6) 4-13 [112011]
01 10

27 46
12K }141 4-14 [1111201 12K }161 4-21 [030201]
03 20

28
121( )141 4-24 (020301]
04 ic3 + ic4

50 ic and pc set Interval

121{ H"I 4-26 [0121201 structure vector
38
35
1 3 K )1 4 1 4-11 [121110]
01

ic2 + ic5 36
1 3 K }1 4 1 4-12 [112101]
icandpc set interval 02
structure vector
40
26 1 3 K }1 4 1 4-19 [101310]
12K }II 4-5 [210111] 18
01
50
35 1 3 11 )1 4 1 4-27 [0121111
12K }5j 4-11 [121110] 28
10

40
12K }J5J 4-22 [021120]
27

30
J2J{ }1 5 1
4-Z29 [111111]
17

301


ic3 + ic5 ic4 + ic5

ic and pc set interval ic and PC set interval

structure VeCtOr structure vector

45 46
1 3 1{ }II 4-7 [2012101 4 1{ }j5J 4-Z15 11111111
10 01

36 50
j3) }Il 4-13 11120111 4K }l 5 I 4-16 [110121]
01 17

37 41
1 3K }ll 4-14 1111120] 4 k }1 5 1 4-19 [101310)
02 08

40 47
3K )1 5 1 4-18 [1021111 4K }1 5 1 4-22 [021120]
17 02

50
I 3 1 { } 1 5 1 4-23 [021030)
27 1c4 + ic6
Ic and pc set interval
structure vector

ic3 + ic6 56
)4j{ }1 6 1 4-8 [2001211
icandpc set interval 10
structure vector
47
46 1 4 1{ }1 6 1 4-18 [102111]
1 3 k }1 6 1 4-Z15 [1111111 01
10
48
37 II{ }1 6 1 4-24 [020301]
1 3 k }1 6 1 4-Z29 [111111] 02
01

ic5 + ic6
ic and PC set interval
structure vector

57
1 5 1{ }1 6 1 4-16 [1101211
01

58
lK ))61 4-27 [012111)
02

302
Set Possible intervallic combinations
4-1 1+1 1+3 2+2
4-2 1+2 1+4 2+3
4-3 1+1 2+4 3+3
4-4 1+3 1+5 2+4
4-5 1+4 1+6 2+5
4-6 1+5 2+6
4-7 1+1 3+5 4+4
4-8 1+1 4+6 5+5
4-9 1+1 5+5 6+6
4-10 1+5 2+2 3+3
4-11 1+2 2+5 3+4
4-12 1+6 2+3 3+4
4-13 1+3 2+6 3+5
4-14 1+5 2+4 3+5
4-Z15 1+2 3+6 4+5
4-16 1+2 4+5 5+6
4-17 1+5 3+3 4+4
4-18 1+3 3+5 4+6
4-19 1+4 3+4 4+5
4-20 1+3 4+4 5+5
4-21 2+2 2+6 4+4
4-22 2+3 2+5 4+5
4-23 2+2 3+5 5+5
4-24 2+4 4+6
4-25 2+2 4+4 6+6
4-26 2+4 3+3 5+5
4-27 2+3 3+4 5+6
4-28 3+3 6+6
4-Z29 1+4 2+5 3+6

303
BIBLIOGRAPHY

Abbreviations:

JMT Journal of Music Theory

M4 Music Analysis
MQ Musical Quarterly
MiS Music Theory Spectrum
NCM Nineteenth-Century Music
PNM Perspectisvs of!*w Music

AGAWU, Kofi, 1992: "Theory and Practice in the Analysis of the Nineteenth-Century
Liec!', M411:1, pp 3-36

AYREY, Craig, 1982: "Berg's "Scheideweg": Analytical Issues", M4 1:2, pp 181-201

BABBITT, Milton, 1955: "Some Aspects of Twelve-Tone Composition", The Score

12, pp 53-61
1960: "Twelve-tone invariants as Compositional Determinants", MQ xivi, pp
246-259
1961 : " Set Structure as a Compositional Determinant", JMT5/i, pp 72-94

BAKER, James M., 1982: "Coherence in Webern's Six Pieces for Orchestra Op.6",
M754, pp 1-27
1983: "Schenkerian Analysis and Post-Tonal Music", in BEACH, David (Ed.):
Aspects of Schenkerian Theory (New Haven: Yale University Press)
1986: The Music of Alexander Scria bin (New Haven: Yale University Press)
1990: "Voice-leading in post-tonal music: suggestions for extending Schenker's
theory", M49:2, pp 177-200
1993: "Post-Tonal Voice-Leading", in DUNSBY, Jonathan (Ed.): Models of
Musical Analysis: Early Th.entieth-Century Music (Cambridge, Mass. and
Oxford: Basil Blackwell)

BENJAMIN, William, 1974: Review of FORTE, Allen The Structure of Atonal Music,
PNM 13/i, pp 170-190
1979: "Ideas of Order in Motivic Music", MiS 1, pp 23-34

BERGER, Arthur, 1953: Aaron Copland (Oxford: Oxford University Press)

BERNARD, Jonathan W., 1981:" Pitch/Register in the music of Edgard Varese", MiS
3, pp 1-25
1983: "Spatial Sets in Recent Music of Elliott Carter", M4 2:1, pp 5-34
1987: The Music of Edgard Varese (New Haven: Yale University Press)
1988: "The Evolution of Elliott Carter's Rhythmic Practice", PlM26/ii, pp
164-203
1990(a): "An Interview with Elliott Carter", PNM28/ii, pp 180-2 14
1990(b): Review of HARVEY, M4 9:3, pp 344-355

304
 1993: "Problems of Pitch Structure in Elliott Carter's First and Second String
Quartets", JMT
1995: "Elliott Carter and the Modern Meaning of Time", MQ 79/4, pp.644-82
1996: "Poem as non-verbal text: Elliott Carter's Concerto for Orchestra and
Saint-John Perse's V.fnds", in Analytical strategies and musical interpretation:
Essays on nineteenth- and tientieth-century music, ed. Craig Ayrey and Mark
Everist (Cambridge: Cambridge University Press)

BORETZ, Benjamin, 1970: "Conversation with Elliott Carter", PNM8/ii, pp 1-22

CARTER, Elliott, 1973: Elliott Carter: Sketches and Scores in Manuscript (New
York: The New York Public Library and Readex Books)
1977: The Witings of Elliott Carter: An American Composer Looks at Modern
Music, compiled, edited and annotated by Else and Kurt Stone (Bloomington
and London: Indiana University Press)
1997: Elliott Carter: Collected Essays and Lectures, 1937-1995, edited by
JonathanW.Bernard (Rochester, N.Y.: University of Rochester Press)

CONE, Edward T., 1962: "Stravinsky: The Progress of a Method", PJVM1/i, pp 18-
26

COWELL, Henry, 1930: 1*w Musical Resources (New York: A.A.Knopf)

DAHLHAUS, Carl, 1987: Schoenberg and the New Music, trans. Derrick Puffett and
Alfred Clayton (Cambridge: Cambridge University Press)

DEMBSIU, Stephen and STRAUS, Joseph N. (ed.), 1987: Milton Babbitt: Words
About Music (Madison, Wisconsin and London: University of Wisconsin Press)

DUNSBY, Jonathan and WHITTALL, Arnold, 1988: Music Analysis in Theory and
Practice, (London: Faber)

DUNSBY, Jonathan (ed.)1993: Models of Musical Analysis: Early Tiientieth-Century

Music (Cambridge, Mass. and Oxford: Blackwell)

EDWARDS, Allen, 1971: Flawed Words and Stubborn Sounds: A Conivrsation th

Elliott Carter (New York: W.W. Norton)

EPSTEIN, David, 1979: Beyond Orpheus: Studies in Musical Structure (Cambridge,

Mass.: M.I.T.Press)

FORTE, Allen, 1973: The Structure of Atonal Music (New Haven and London: Yale
University Press)
1978: "Schoenberg's Creative Evolution: The Path to Atonality", MQ LXW/2,
pp 133-176
1987: "Liszt's Experimental Idiom and Music of the Early Twentieth Century",
M'MX/3, pp 209-228, also in KERMAN (ed.), 1990

305
 1988: "Pitch-Class Set Theory and the Origin of Modem Harmonic Species",
JMT 32/il, pp 187-270
1990: "Mussorgsky as Modernist: The Phantasmic Episode in Boris Godunov",
MA 9:1, pp 3-45
1991: "Debussy and the Octatonic", MA 10:1-2, pp 125-169

GAMER, Carlton, 1973: Review of EDWARDS in PNM 1 1/u, pp 146-155

GRIFFITHS, Paul, 1995: Modern Music and After: Directions Since 1945 (Oxford:
Oxford University Press)

HARVEY, David LH.., 1989: The Later Music of Elliott Carter: A Study in Music
Theory and Analysis (New York and London: Garland Press)

HASTY, Christopher, 1981: "Segmentation and Process in Post-Tonal Music" MTh 3,

pp 54-73

HAYES, Malcolm, 1990: Review of HARVEY and recordings of Carter's Music,

Tempo 173, pp 57-59

HYDE, Martha, 1996: "Neoclassic and Anachronistic Impulses in Twentieth-Century

Music", MTh 18/2

JOHNSON, P., 1978: "Symmetrical Sets in Webern's Op.IO, no.4", PNM 17/i, pp
219-229

KATZ, Adele T., 1947: A Challenge to Musical Tradition: A I*w Conception of

Tonality (London: Putnam)

KELLER, Hans and COSMAN, Milein, 1982: Stravinsky Seen and Heard (London:
Toccata Press)

KERMAN, Joseph (ed.), 1990: Music at the Thin of Century: A 19th-Century Music
Reader (Berkely, Los Angeles, Oxford: University of California Press)

KIES, Christopher R., 1984: A Discussion of the Harmonic Organization in the First
Movment of Elliott Carter's "Sonata for ñoloncel10 and Piano" in the Light of
Certain Deielopments in 19th and Early 20th Century Music (dissertation,
Brandeis University)

LANGER, S.K., 1953: Feeling and Form (London: Routledge and Kegan Paul)

LERDAHL, Fred and JACKENDOFF, Ray, 1983: A Generati Theory of Tonal

Music (Cambridge, Mass.: M.I.T. Press)

306
LOCHHEAD, Judy, 1994: "On the 'framing' music of Carter's First String Quartet",
in Musical Transformation and Musical Intuition, ed. R.Atlas and M.Cherlin
(Roxbury, Mass.: Ovenbird Press)

MELLERS, Wilfrid, 1987: Music in a New Found Land; Themes and Developments in
the History of American Music (London: Faber)

MESSING, Scott, 1988: Neoclassicism in Music: From the Genesis of the Concept
through the Schoenberg/Stravinsky Polemic (Ann Arbor: U. M. I. Research
Press)

MORGAN, Robert, 1991: Twntieth Century Music (New York and London:
W.W.Norton)

NEUMEYER, David, 1986: The Music of Paul Hindemith (New Haven: Yale
University Press)

NICHOLLS, David, 1990: American Experimental Music, 1890-1940 (Cambridge:

Cambridge University Press)

NORTHCOTF, Bayan, 1980: "Elliott Carter" entry in The New Grove Dictionary of
Music and Musicians ed. Stanley Sadie (London: Macmillan)

OLMSTEAD, Andrea, 1985: Roger Sessions and His Music (Ann Arbor: UMI
Research Press)

PARKS, Richard, 1989: The Music of Claude Debussy (New Haven: Yale University
Press)

PEARSALL, Edward R., 1991: "Harmonic Progressions and Prolongation in Post-

Tonal Music", MA 10:3, pp 345-355

PERLE, George, 1981: Serial Composition and Atonality (Los Angeles: University of
California Press)
1990: The Listening Composer (Berkeley, Los Angeles and Oxford: University
of California Press)

POPLE, Anthony, 1989: Skryabin and Stravinsky 1908-1914: Studies in Theory and
Analycis (New York & London: Garland Press)
1991: Berg: ''iolin Concerto (Cambridge: Cambridge University Press)
1996: "Misleading voices: contrasts and continuities in Stravinsky studies" in
Analytical strategies and musical interpretation: Essays on nineteenth- and
twentieth-century music, ed. Craig Ayrey and Mark Everist (Cambridge:
Cambridge University Press)

POFFETT, Derrick, 1986: "The Fugue from Tippett's Second String Quartet", M4
5:2-3, pp 233-264

307
ROSEN, Charles, 1975: Arnold Schoenberg (New York: Viking)

RUSS, Michael, 1990: "The Mysterious Thread in Mussorgsky's Nurserf , M4 9:1,

pp 47-65
1996: "Reading a Mussorgsky Song", NCM2O/l, pp 27-45

SALZER, Felix, 1962: Structural Hearing: Tonal Coherence in Music (New York:
Dover)

SAMSON, Jim, 1977: Music in Transition: A study of tonal expansion and atonality,
1900-1920 (London: Dent) (2nd edn. 1993)

SCHENKER, Heinrich, 1979: Free Composition, trans. E.Oster (New York:

Longman)

SCHIFF, David, 1979: "Carter in the Seventies", Tempo 130, pp 2-10

1982: "In Sleep, In Thunder: Elliott Carter's Portrait of Robert Lowell",
Tempo 142, pp 2-9
1983: The Music of Elliott Carter (London: Eulenberg)
1989: "Carter's New Classicism", College Music Symposium 29, pp 115-122
1990: Review of first performance of Carter's Violin Concerto, Tempo 174, pp
22-24

SCHOENBERG, Arnold, 1970: Fundamentals of Musical Composition, ed.Gerald

Strang and Leonard Stein (London: Faber)
1975: Style and Idea, ed. Leonard Stein (London: Faber)
1983: Structural Funtions of Harmony, ed.Leonard Stein (London: Faber)
1995: The Musical Idea and the Logic, Technique and Art of its Presentation,
ed. P.Carpenter and S.Neff (Columbia University Press)

SESSIONS, Roger, 1979: Roger Sessions on Music: Collected Essays ed. E.T.Cone
(Princeton: Princeton University Press)

STRAUS, Joseph N., 1982(a): "A Principle of Voice-Leading in the Music of

Stravinsky", MiS 4, pp 106-124
1982(b): "Stravinsky's Tonal Axis", JMT2612, pp 261-290
1987: "The Problem of Prolongation in Post-Tonal Music", JMT31/1, pp 1-21
1990: Remaking the Past: Musical Modernism and the Influence of the Tonal
Tradition (Cambridge, Mass. and London: Harvard University Press)

VAN DEN TOORN, Pieter C., 1983: The Music of Igor Stravinsky (New Haven: Yale
University Press)
1987: Stravinsky and "The Rite of Spring": The Beginnings of a Musical
Language (Oxford: Oxford University Press)
1997: "Neoclassicism and its definitions" in Music Theory in Concept and
Practice, ed. James Baker, David Beach, Jonathan W. Bernard (Rochester,
N.Y.: Rochester University Press)

308
WHI1TALL, Arnold, 1982: "Music Analysis: Descriptions and Distinctions": An
Inaugural Lecture in the Faculty of Music (University of London, King's
College)
1983: "Webern and Atonality: The Path from the Old Aesthetic", Musical
Times cxxiv, pp 733-737
1987: "The Theorist's Sense of History: Concepts of Contemporaneity in
Composition and Analysis", Journal of the Royal Musical Association 112, pp
1-20
1988: Music Since the First World War (London: J.M.Dent & sons)
1989: "Review Survey: Some Recent Writings on Stravinsky", MA 8:1-2, pp
169-176
1990: The Music of Britten and lippett: Studies in Themes and Techniques
(Cambridge: Cambridge University Press)

WINTLE, Christopher, 1985: "Kontra-Schenker: Laigo e mesto from Beethoven's

op.10 no.3", M4 4: 1-2, pp 145-82

309
 r e

•7 II •- -

—i
f.

2:1. • -

FeU

I j 1-

reQ • r.t %U* k#.* tLte t

7 _

II'

rn
&p . & i the pt.r Z
• ti.:t j4,(

H m
I TTJ

lW-44 12STAVE

:1
310
 (0jJ

1*

,-' I—'

ILL

(r fAf)

U' - tc.vo wi.ct U .

C,- D

( f__ (J
TAVE
 . .'i

- a A.*tn

Y,.., .-f
11rjt,

I-'.

- r - ---- - 3 4•

__IT_--i-"

Q
(7J&) - ____( fj)
( ____________ _____ ______ ________________________

_ r1
vk wLre

__ _b ' ____________________
K11____ ,__ _____________
(/k7 ot4
_______

-
.- •

- J__-- t_-4--4 •:: .______ j-i £:

1— - - _____- - ____- ___-,_-.

L211
F" .,-

PtN A4 12 SlAVE
3l.
 _E.iJ_. P O S%d.*p,ii, I—

Andante (Ji,9)
Con SOnorit

Ic

-.

l I' ii-
motto IegCto
j IL•-

I.--

a ) meno ereic.
2.0 __ 1' / / 2 2 Jf

P
IeflO TflOSSO (JS3)

cL4 • VI.&o -OA4t IJI.

(c&) -

________ p t&04, O.AJ* /i/i-is

(I

(1:,)

P4IJ.A4 12ST4/E
Et.i_ 9t&.o S0* ,J1 23-Si

Mèno mosso (.1 = 63)

H
I- 4-

'1 ;j

:f!t P
I,
ft
J-]

nt.

-r'ib : S
33 303/. off

I (===—)
Con pedale
 (b&k.

8 Z'1 ; -

wr p Ji __
40
cresc.

>. > 1r'_ mob inten.samnte

______________________________________

____-

S00.fIJ8—S

L
0 '0 1 S 1 6 I

dOLIL(.

h..- .

. p.
o o g
S

CI.

Ex.t .. CtU. Sc*G /If'-(o

q I
^ to

,-.--., ..

2.

N-M 12 STAVE
(3
Th76
 x. S . P... o&t () ir ,

T- -.-----

-__--- -r

('b) 2.o - 2. [Lpko citio1 _X -- - --________________________

'1

0 ___

I-

N-A4 STAVE 32j

 Lx. 3 . _ r;04, (G.5) I L'.3

• . 1

-5.-
TITT'
1
I

- -----5-- - -S ___
____________ - ___

----5- ___

__ , ___

______ I

-2J

_____ -
, ii I

pI-o

(
\j

1PJM 12STAVE
In
is

S •?_
] , Io ct - 1 (LLu')
A - - - -- itS

-- .- I

QDSt I
(HpL&&.')

I i ._i3 4...iS
4•4 ML.'
(1.-ti tP -t . 4 -1.1

-(1
_r

t4..g L1-t 4r4 4'-t'! ''

Lj._I_ L4-I '41 LL.(

PANA4 12STAVE
 -p..,

L-( -L'j

_____ _____ 1i ____ _______

,L',_l'
.

A/I • . , s _______ ._+._i

IA _____________________ ____________________________ . ,
______-_-... ..
r', j ________________ ______________________
I I • I I

I 3 it
_____________________ _________________ 0 _____________ V .4-

S •
______ ____ ________

I -T 7-
=

___

:
L -i - 9 q -2. k-1(. - ' , 4 .1- '4 4-"
2' - & .L ,, ,, 4, , 4

M : - ...-. ________

3- it-U 3,-3 _i' 3-M 1-5 3' 3-j 1-9- i-'6 4- -2.

FW+M STAW


.1_I,.
-' 3-u 3 3 .. ¼-t
k-"t 1-; -u 1 u .4l -24 '1. L
34

4 Lq k -%
________ 1 4?
_________
.i' 4.

?V '
S•
$ j ••'•$
S
. • 4: : : . . I I • - - *
* . I
'- p.-.- •_ --

"p ti

I S
I
-
S •
-'
S
1 - —
±' L
• • • it S . - S 4t =
-. I —, c__ __________________ —
c}
it:
• • I I-I • _________________
•
I I

.x- —
I'
XQ5 _
----
(-________ 4. __
__ , ____________
____________________ _L /E
k-14
__________
, 4
+-Z L.-to
•
__.
I

3-
'1'
(4.-il
4,
4_I
,,
-IL'
-t3
'' 44
3-•;
_____
_______
11 __________
•

_________________
0
I__I 1I''3
• I
• _______- S

L0 I',
'Li
tkt L
I •_ I - S • 4-: • _________ _____________
+ • ____________ _____________
____________________________________________________________________________ ______________________________ I

LI -'1 '-S
1. 4 4 ______
'1 1
$ — t I 'I 1
/
t S I • b. _________
(ç) I S •
I I 2l •S
'S.- - 'S._ —
'—'-I. 2.f as
7-$-
p __ ______ ________ ________

I_______ _________ _________

S____________________
IT LI, b... - (7-4-
11

It',

___ rs ____

IWM4S1M


Ek4Etii4-' vcis , i (g.e42o)

- ______
- . jJ 1 * i---- dl'

11

2.

-I. tr T I I'
4—It
I

3 -i 3-c, 4-il l,.-l1 4'

-"
I
-I • —

4-
•1-

-V
It V

11

,,1,
__y
4, 4r I 4, .L £
4- - i -
= — --. —

- .-.. -_- .-
. '- - -_- - -
2
-
Ii -
10
- I., -. —f.'. __ .

-r -r I I

4, i t- 1-33-8
4, 4' r 4 1. 4 4, 4'
± . __' ,_•'.
p 4. .4 - .,4_ _4 _. _4 dZ L

I— I :;L;
I ' I
I____ r
3-3 '3-1 3-L

N-A4SIA)E


I- ii 1
3 1-3 3 it 1- 1-i
3 .1 3
j, I
-

- r1

.-ti 4-, ! 3-S

I :' -t+

149
I4-* tjr4
±
- ± -i--'

M4 S11JE 3


X . . 'Ctt frD u Ethm& C,r

oiii-v I'DC&t

Cit " tI)

Ci44i

1' - J .1_I. - --______________________________

E6-43 L34 t:+si C34s1

$ 1k (LaO C.EkpL,I ,a-,ui

t._IQ __
4 -'

IW'J-A4 sr 3o
 p

x.s.Is q.u..*-&,&*v,i,'

_____ ,E, 13-2.'4

to,I ,I0,
E ,, ' ,;

N-MSTAV
E.3t G *&*1-,t, .o-6t
LI.
I

- 44I
t .S ( t St R'*rzt - c.
ti

(L) 4k14s5 k-t lit

t.-Is I.—s
4,
ki
ir j•. - -;;.

Cc) T-?

kA4 sTdE 2.
.) 11-12-

,1
1-3° ,l'J
I-
h-il I hII lI.U4 ________
1 .4. ' .. i - I T I I
—1 •• i-

1—k

'I
r1 rZ 1 U±

-
tt
.
_____________
?° t '' I fl' • "-
3-S 3-Il 6 3-11 111
I

$48-. [ o 1 , 1i.,, s,I 1 ?t,23 1ix l,.g -12. • ig

I. -tP3

JM STJE
 E,c .S .20 r .StA QM3t6C, lEt.

)1112/P
(OL) 2- 88

lit18 6- ix 1-31

(•) (ciictt .-4z)

LcLcik

(J)12.8 -3°

•! -S 3-Il
—o 3-Il

'-3-3
T-+-' T

-3 It-IS

PNI-A4STJE
 L.. U S

z.-iV S-3° 6-U s-io Ii L-1.

542:j,(I._(ci
(it-'1)
.j.. i,r (i-rgf) (i-i's- 4,
s! .4 T ,L L 'r

1 ..tt. (pc..iis,

9-31 (x-1tS,i,j -t
(I-6,'4-IE'
(,4clt-IT 5.ti

- 4, 4-

-•l

é-U (' -1' -%ZC1.x'q-2 5,bc't4 f U 1S (ze41is, L%k42) ?-I (l?c P3 PS, tX
f x'-t 4, k..-Ig
( 41

-0-

-r -I-

',. -2.O -2J(*c4.4IS1 -iz)

. p '¼ 61 I-l(..;tc
'--4.

U '-' -

I ' ___
II
¼
,_,-$-

FW'J-A4 ST
 I

(&p
,L- A
PS
2

:z' t :T "::
(V ?. t* zk ,c- &I -
3

A ' •1.
-
(-'-I - -

(C_______ - Ps 3
- 2.

-I
z? (.z

STAVE
 c& ?ot.rrJ 2S-

'I

,,_. _,___ I-'

1--
* --4' 1. - - -

J ,-._J .-_.t ,-f

y —f JL
7Jj

1
f -
t

A4 ST/E I
 (cs) (a .OV.DJ* ,.

V' • ' i- g t,i —I

cg!_

Ech.3 (& P.i.o S0.i,II 'S

S
A

I
I I I

o': X' ! () G: 6 6

IW-M STAVE
I c (LU.o&&i, , 3(.Lj.

.----- r i_---- a

fl
I--

I (b) - - -
1v_
FNA4 STM -..--' --------'
(€


L8-V-

• () IA O SI.$t, I, _•:1.
I
I --L

U (4 .5')

f I \ -I

' 1 ('q)•
i L

(c)(t(lO S,.i.. , ]il1t)-iJ

• &- 32 (±?
I

1-- I - - --rç
/ l's
I I • II; -

I I --
- !.
-çF i'z-S'.. - -
-. -
jli—'3
t •1 —t '4.-I'

5-24 6-2.0

IN-M STAVE
 5-12
''I'

-32 (fL!) c-U.

3-'

IW1-M PSTAJE
 )
,2-°3

() ait 0 SCML.th I, 8-O

-
- I'.-

-- 1
L_.__J

14
I I
(c)&1lo ur.sb. ,I 33-36

( ..I - - I) -

A4STAVE
( fl Wi-c tL- Il-i

A- p%

- --- ci - - - - p
4.
:1•' 1;
c'-.
r --
ct,r _v\ -p. -
I vrIa\ -

(i-) (I)
$
(w -I

VSTAVE
 jt;w Z2 - - -
E,c.C' (&)

() 4.r4lL . LLt_twt ,1 O -tOG

IJ - I I
-
I ' I Is I"

-6• (.1 - -
•1 - -

FW •A4 QSAVE
(C) V0t
 CI) P.v-o So.4u,t, 'c—Lb

Mt

- J? LI'

L ______J I

' u i t to
•
• ii
3-Ii '-III
it
- - 1•

P4'J-A4 'QSFAVE
 I-- . -

-
r

(.

-- II 4 +11=

N-A4 ST
 - •-..- --- ,.- -'

(5 )CL(L. O4.* ,i

-1 23

. r v_i.

—• I
b?i

cv QujJtt*, Ui , 3- -".4°
'tI E—
] -.I -1 [

J - I ^

'f
(;) (11O

_______________ a a

(Ic')

fl
Q '.utc51r..,.kL,

c'*. nr g _________

I:
1

_____
3o%At fc't ,

N-M STi
 (c) (LU-. 4-S
- (fl (.tU,!,U-'

(t'ott.flt. '- k

(-f ') Q tQkSb.t.&tD, I!J, 'd2.-S

_______________
.g Wt U- k 3-4. _ 1
1

-I 4.IL -- IL

1
-4--1_ f

(li) rr,2-

N-M 12STAV
 B:. . ci (c) Pj0...a Stxb, L • -

r 1 1
I LJ1-1TJ iL—c s

-
-
i

L--

c' (tUaSoo.kX,][, &-

'tic

FN-M STAVE

2
 (S C.) L1.'- ,LLI ,

ti - I

I
-
I
F'.--

p*vo (g(o
1.19

- 5 1___.;
(_J- 5
S
Lti.

N-A4 STAVE
 -.
cklo (& CLLL. , U-I4-o

(L1 t Qu.Ut ,-

v1.I

- i:;.
( Q,j,ttk to, •

-' _± — L.

tX.b."

N-A4 STA
 I-

- -

I I

I I -

CIa II-- Ii

STAVE I
r
(•f t Q.&rt**,E

_
g -

fl .-----'-.-- ,'--.

flj-1f

1X
N
r
L* -J) .t1
r

--
Ft.

.6.11 1!, %O ill

-,. 7. -s . .-.-..- ____

I - - - - I I
. .+_p_.... _^

I - -

V

I - 1-Bi - -I--I--1

N-M 12 STA


tc TL&&e. c..4L) ,3-I

( 1 r t iii (

_____________________________ $1
.' I

'4) i-1 I' I I ç i'°i

•r
__________
I"
(

IA
4 -'

e(.. 4L&.

N-M STVE


I
OL

_____ &

T.
j
L 'I L g1- b
L L

1
_________ C
- C
____-___----, -5 .' t I..,. - 9 9 , 10

I _________________
It
ck
& tI!'tl.

*1 1
9 I.

v. C
4

C g I

-
I t LAW.

c- L-

,- ;.i.

c_ '- (
-f 4:
IM0. I.d.

I -'
IC -
Afl1. ________________________
I /
-
, - .—, '" I 1

L_ 1
•1 CL
C
N-A4 VS1
 ,j ,
.

L
_________ 1
-' -&- a •

1"

o2
t.
1
-.

(i

r q I.

C. •'
& ft C :)

(s

a 12.

C
-j

-J
S I _________
________ _______
& o L___
___ . L I
j ___
I 6.
L

Co.

6.1 r

NA4 2SI4VE

3S
 L.

I . L
D.1

N i i3.

___________
I
- ! t

C-

'4-L -L

_7

-J
-$

- ) -it 3z n ,- . •

STM
 - 0(1
1. .. - Ill!

rf' .!

'--I

•-

FN-A4 STAVE

co
 1

¼
4- i- '-

bc. . S

C') St*Oh*. :i• i - H'3

r
^ r
3 S

Lf

jjTr (__

t' Ir,
1Ikr £
j 1"

SvuIz3. 1 r, S6-bD
.
L

() $o*.&kti. : '
4.
S ,—.\


14 () ii 8. -

Pir
(lb(

Ti.'.. -
U br -

1___ 4r ,i., ----

!.
4-

(W £.. ..*. i i-I.

- 2.

1•
':1.

N-A4 STAVE
31
 40
40 5

.i 04j, ':' 4
-3 .
13

- 0

1
'4-
-

-' ._1-1

L#IIJ - .,- -0-

- -

x3? (L &o.ot-/JJ/I-T
I-

-J '_jj-

.1._ 4 L i .L i •
1

Cdi1, &#.Lk, fjy/..i

_-I., 1 ---'

FWA4 STAVE
 &..i f Qu.sti* /i31/Si3-

UI-S

°•lrt r lr
VU. __

c4
I 'ItWhI IM.;
.1
'. (

,vt.I

-? '44f

______ ct Qrt / f
&L Vc.. 442

1.1 _.. - - . - -.
__ -9-
'-' L
I

+
Lt4 - qq.
VI. 1,4?

t. 'I.

. . p. '-
.___________) - .-, L... "'
..l -

3. -

P4M4 STit
34

L'* .1 •( (.l,.stL* .c'tD.f f l- -

.
I-

, - p.'

.12J.) ;&
nA..

________ ___J I _j
I it

JLt10 'IL I L( —SO, V1&.

\2' '

- I

S:

I J I

0 3:
AA.j
M Q ST

'5
 • 2. 0 1jJ .( c.f hk Iir.

%ZI JL.I
-
T
'A

r r• r" t'
_____________________ ______________________
Li t41 "1
___ ___
'4

jt1 —I
I Th - - p
________________________________ _________________________________

_i_ss

A________________ _________________
'201
_________

(c 3 ft u (i It')
a.. - - - , VLI

Ot&L.

:P3L v
( h 1I't.'fl_ç _________________________

-_.
'IC.--------

________________
:3

STAVE
106
 IIjPMLJ.. LO1 1" lii
•-G ,( I t
k LJ
J
f'T 7

L S tJ rNLI tj1 c
_________

_______
,o. ..

•
.11t(° Ft Qirtv I •iii fi - ' 1.
Vc. 1':?
fl - I -I
,

- Ii. i7.( - lt171.( •

(i') 8-1
Ytv.•I 2 -

4 (
4VLIL
q-s—
1 V

MM 'LIt
I
1T

r1 I _H1i.I '7

If
(J40
,
ii f_r_c II
l% OO

I
MM iQO
MMSo

• I
14 ((rç
______________________
rjIi'Cc1r
VL. t

T2STM
 t,L%ç ç flu /('1) -u

V. P3t. k4 J. :1

I I t H t in I i n I I ll
C,,)

( tR3- . .kM O8

v(.1. *_4I_. ii I -.

3o(f _:. I
(c_\
!c. - \la - - I,. r

3. -

p r M. 'i-
(&
t? Li.

FN-A4 STM

9
 •• ' .f.r O.tt (Su. [Sc.L .ff , iiS:
tz0 ..jfl (&) '
'
-
,)
iT
b.1 j:'.
C).

- ((
j O = VtO

(
J %'i'

(vi')

.613 cL J Lit

I 1 L.e i L

FI444 12STM
 i (,_) C'
(1') J_ 'S
•/__%
t'•h'h.Itt
± ' 1- -e a

I f• 1
(iü)' 'tV_•.

4
'Si
Ftpt ir'
(ji) 110

(v :V'cO

(e) r: j5
't° 1ttz

J:
,,._

NA4 STIE
 tX•-t .It(J't'.v. .jy (JjfIO - fll

I 1 4. 81

1°
I
__t z3s4! IS

3' 33 1$ q

/
11' 31 41/% 1 3 3 3 .31 q.

- Ltt
q 31 is 3' 3 1,i, 40
0 3 £. ___..__- _________________________

PsW-M ST4

3
 * % - - -- - --

p.3. O#4* ,a, ii-e

rilatco
F..-

aff mp detache I

I-

0g.!

13 2.
 - I I v JVP.#t' .
•

EEl ticr

l@J @E @@@1

o =
ci

12 STAVE
 E,'P(CL.
c- - -' I-
I t

- 11 - - •-
- -. -

-
3 8 1 10 II '2.

() iI.'i.-iz.
1.
6
SIc.. 2. 3
- . 1 - . - I -

8 O
1

(C') /ta-8l
it Li. 8

10 I' 12.

(L)I (los -I'S

I t4.
S £ Ct')
I 1 - g . i1.'

s—. 3 -, r 3 - 1
-' i - - - -.' - - - a —':-• _., - •1 r— 3

3 1 1 tO

N-M STAVE
 '4•• ' U(U

bet

! ______

- 2O

? :

Fit.

-P

2.i Ut

( k
-,L

PAN-M 12STAV

-S
x. 2. ' (&' C.L.k s... , It, t—'

Vivace (L 208)

Clhk*i ]I ,
Misterioso (1L 132)
 2.SS a
(i;) (osk:,i PD—o , ,

- If

Tenioo orimo es:r

Ex8. ( C0LJ. PI*LO , — ISO

1-1 . _,_. I_ - -

(b) C. rtjr Pi.*o I, tot- o4-

nt,_

coJi
ff

r
- - - - - -
 , VL -
Eic.8 .'+ ?uo

Tempo I
(Maestoso)
m
S

',ress down .,llen

sost. ped.
Qf/-01? I: ,leno /

I -

if
FPfl\


(& cc) _________

4-u
4-t6 i- -' (I"t
- 1

,- ,-.

(- 3j-3 q-
Iiil
_______

IL-J- 4-23

(e) 3S

1-

PN-M Sit
 4 ..t

21

• •
•
?.

o io S S it
2.

t -t'
(1')
'°-
-4-

'i- • . . ,

'4 -1

L -

?oc/f(—ti.3
.- ---
11'
I-
•

IWA4 QSTAV
 • 1 (L1.o /31
(;' ..
r

(ii) Ia-uI 7
7
I -

(ii' U- '1.
n-fr. nIJ

FWi-A4 ST1E

311
!.L.L1. ItUJb-'"'

U— I

- •spres. -.------------------- •c,,sc.

_____________ _______ - r r
p I I I - - I 'r
Ec.1 .2jo#
.1

II

------ 8--,

2g
'c.1.-1 Lto.oj,ii, cl-so

'I
 6cA.S, CtU0 *,J1

cij S?-8
'IC.. S
()
L -1 -i- P

S S
--

c S - S

t.p i.

_____ CLAL 5,L#.*P

1114. •

0
I. •,

. -Lc
Wd-A4 STAVE
 (f_li

iIf44 '2 SIM

 t% L . J 2o

V. to't 41' 2.oS

k

-7-

\t

C.'. -'
7

1_ti l-'t 31'

3 3-tt i'4 3_ti

I I I '1 1 I
3-2.. .3-3
I-

I 1 ( I C Vc. ç
I A II- I .. I ..

L.,
I I ..

r i i-.l -

N-A4 STAVE

\h.

t k . t3 t Jf'f
I 1' L I I
• ' k
u-ID
I
-

t- I
Cc') r___________ I

• U _____
L_

(U)'Ic. 4-
11,_b L-IOF
L.-3 r

-_I
UrIL_ ••1, I

Ctl(*

- to

L I
 '1. 10 ,IE
) -1
- Vt. 1

II I

-1 ' I
I I

(1) 3_t --
r-J.....i..____L_,
Vt.
' r -i
g . ( L I 1 t-i---, I c-i-i
________ -, -
i-i I (

;41.q- .

t+A4 STAVE

310 I
 ______ Ski& //o-iii.
16 ',-tS (o,,',')

AMP-6629-47
 17

AMP-662947
 _.A,.-4•tJ'
- j4LJ

(____. I,4_

- k4 tt L*

1
1 __- Tct
-k4.

k&J. -' _______

1-M ?2SThV

3g'


OIq .
3-7.' °.1 • 1-

I' I

- -

. -..

LJ
- -

&LbA
t. - - L -

'I -.

tO , auJtL* I. I -

- C. .7 (__\- (

N-A4 STM

If
 2 O
4-1-
- (r

1. t( C

L
to I ,

O 1.
¶x. CLet 'L (

W4M 12STA

c
 t,r. Ia QuaJt&* ..ScMJ,*i¼ 8 (U#.ccLc&)

V- - .4.

Ib, \(

k-to

A (4

I I I
- I 4- I
1k-ta
1 C
v*^ "i- _________

JW-44 s1

3(

S5 3-

':1 t

3-3
- —3

jLUL.
1

t7a
-1

-Sc -t

---J -

-4-
L ____________ _______
1

I_____________

-2.

LILS ' __ JTLJTh

4 -'S
________________________________ ____
_______ ______ S
_________ _____________ _____________
______________________
______________________________

-4- _______________

34
___________ ___________
..4- -
I
-_-

-
___________ ______ -
____

(J
-I

-U
I?-.-
-------
__________I
___________
_____________ ______________

N-t4 'QSTAVE

- I,

3.
= -.. ,II_.

I
2-3 L
2 -1

FW-A4 S1
 (o) 2-
0'

______ I I
3-1 3-2 3-2-

- -

-
If r---.L_ r-_-- çj•'tt-22

I- -I 3-2.
c. - to

I'

—Io

FN-M STAVE


E,i.o.i Qv.. rt1-.So&oi,t,_t (*;Ik. 't-tD)

4-
- 4-I -

L- ( .-L] I I - - i_:
+ - 2. t I 4-8

3S
_
St
__
'3-'i.

r 4 . , I •
, •I
3-li _l-q I

4 -t k-I (.tt L.-IS 4-8 '-Lo 9-S

- I _____-_ _ I -
- :.

7 -.-.----- • .-..--.- • _ 4- - - - - I •
___
- -- .- .- 4-

L4-i
- 2.0
-It Li-itS r '410 (4-LO

J IL
I
I
-
I _______________ 4--IS t - I

3-u
It
3-S 3-S 1 3-5 :i -3
1
I 7 )Y-. •)1 c.1 , ç
•
•:g, I. • •
I
S
• . 1 • e
• I I
_ I I _
II I 33 3_to 3_3 i j 3,,
3- 3_It.
3-iL
3-;

I-t t4 -t; L4..!' 'j-S 1-%'( I,-; ,.-I'I lr-I

I. - #-4f

N-A4 STAVE

L.Oo
 flL

Ex. La. t3 ( -'- '

(a) I-2
1rri 4m

IT

LI ')

I,,
IWtM VS1


- (0 .1* Qu4t** &v..QJ', L

Icl_2.o
- I
I1j)c

I t+ - -
-.- -
4
- I S-

0._iS , t&* -Sv&±,

'p..
. ..-.-

IJ
4t-3

H _
() n

' - ffe—===--f

('' -' '
S. . i

I
0 • -

L cr- L.
j

(c) -
. OL.

-1 --

I . , Q tti.* t

4 7
- S__

IWtA4 STAVE
 1:1

32
o(.

SIc: -

Lj

NA4 STtáIE

(LONDON)