Professional Documents
Culture Documents
Dissertation PDF
Dissertation PDF
Michael McClimon
Doctoral Committee
December 9, 2015
ii
Copyright 2016
Michael McClimon
iii
Acknowledgements
This project would not have been possible without the help of many others, each of whom deserves
my thanks here. Pride of place goes to my advisor, Jay Hook, whose feedback has been invaluable
throughout the writing process, and whose writing stands as a model of clarity that I can only hope
to emulate.
Thanks are owed to the other members of my committee as well, who have each played
important roles throughout my education at Indiana: Kyle Adams, Blair Johston, and Brent
Wallarab. Thanks also to Marianne Kielian-Gilbert, who would have served on the committee were
it not for the timing of the defense during her sabbatical. I would like to extend my appreciation to
Frank Samarotto and Phil Ford, both of whom have deeply shaped the way I think about music,
but have no ocial role in the dissertation itself.
I am grateful to the music faculty of Furman University, who inspired my love of music theory
as an undergraduate and have more recently served as friends and colleagues during the writing
process. Special thanks are owed to my theory colleagues, Mark Kilstofte and Dan Koppelman, and
to Matt Olson, without whom I would not have discovered my passion for jazz.
I have the privilege of having been at Indiana University at the same time as many talented
individuals, including (and by no means limited to) Gabe Lubell, Nathan Blustein, Je Vollmer,
Diego Cubero, Mark Chilla, and Garrett Michaelsen. The ideas that coalesced into this
dissertation were formed in part over many years of friendly conversation with these people; these
conversations are among my fondest memories of my time in Bloomington.
I have been lucky to have the unagging support of my parents while I earned three degrees
in music theory, and they are to thank for untold amounts of money and time spent encouraging
my love of music from a very young age.
Finally, words cannot express my love and gratitude to my wife Carolyn for her undying love
and support, in the writing process and in life. Without her, I would never have made it to this
point (nor would I have the privilege of knowing our two cats, who lent their own brand of moral
support to this project).
iv
Michael McClimon
A TRANSFORMATIONAL APPROACH TO JAZZ HARMONY
Harmony is one of the most fundamental elements of jazz, and one that is often taken for
granted in the scholarly literature. Because jazz is an improvised music, its harmony is more uid
and potentially more complex than that of other, notated traditions. Harmony in common-practice
jazz (c. 19401965) is typically represented by chord symbols, which can be actualized by performers
in any number of ways, and which might change over the course of a single performance.
This dissertation presents a transformational model of jazz harmony that helps to explain this
inherent complexity. While other theories of jazz harmony require transcriptions into notation, the
transformational approach enables analysis of chord symbols themselves. This approach, in which
chord symbols are treated as rst-class objects, is consistent with the way jazz harmony is usually
taught, and with the way jazz musicians usually discuss harmony. Though transformational theory
has been applied to later jazz, the aim of this study is rather dierent: the music under
consideration here might be called tonal jazz, in which functional harmonic progressions are still
the rule.
After a general introduction, the rst chapter introduces the transformational approach by
developing a diatonic seventh-chord space. Chapter 2 expands this diatonic space to a fully
chromatic space that focuses on the iiVI progression, laying the foundation for much of the work
that follows. Chapter 3 extends the model to examine music in which root motion by thirds plays
an important role, paying special attention to the way in which harmonic substitution interacts
with more normative jazz harmony. Since the pioneering work of George Russell in the 1950s,
many jazz musicians have drawn an equivalence between chords and scales; Chapter 4 develops a
transformational approach to these chord-scales, enabling analyses of improvisations on tunes rst
analyzed in the preceding chapters. The nal chapter centers on a single harmonic archetype,
Rhythm changes, and brings together the theoretical framework in a series of analyses featuring
solos by Johnny Grin, Thelonious Monk, George Coleman, Sonny Rollins, and Sonny Stitt.
v
Julian Hook, Ph.D.
vi
Contents
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Notes to the Reader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv
1 Introduction 1
1.1 Problems of Jazz Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Theoretical Approaches to Jazz Harmony . . . . . . . . . . . . . . . . . . . . . . . 7
1.3 Transformational Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.4 Aside: Lead Sheet Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.5 Diatonic Chord Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.5.1 Intervals and Transformations . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.5.2 Analytical Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2 iiV Space 38
2.1 A Descending Fifths Arrangement . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.1.1 Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.1.2 Analytical Interlude: Lee Morgan, Ceora . . . . . . . . . . . . . . . . . . 45
2.2 Tritone Substitutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.2.1 Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.2.2 Analytical Interlude: Charlie Parker, Blues for Alice . . . . . . . . . . . . . 53
2.3 A Few Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.3.1 Minor Tonic Chords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.3.2 Other Kinds of Tonic Chords . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.3.3 Interaction with Diatonic Spaces . . . . . . . . . . . . . . . . . . . . . . . . 65
2.3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3 Thirds Spaces 71
3.1 Minor-Third Substitutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.1.1 Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.1.2 Analytical Interlude: Joe Henderson, Isotope . . . . . . . . . . . . . . . . 75
3.2 Major-Third Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.2.1 Introduction: Coltrane Changes . . . . . . . . . . . . . . . . . . . . . . . . 80
3.2.2 Developing a Transformational System . . . . . . . . . . . . . . . . . . . . 83
3.2.3 Analytical Interlude: Richard Rodgers/Lorenz Hart, Have You Met Miss
Jones? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3.3 Parsimonious Voice-Leading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
vii
4 Chord-Scale Transformations 99
4.1 George Russells Lydian Chromatic Concept . . . . . . . . . . . . . . . . . . . . . . 99
4.1.1 Lydian Tonal Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
4.1.2 Chord/Scale Equivalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.1.3 Chord-Scale Theory after Russell . . . . . . . . . . . . . . . . . . . . . . . 113
4.2 A Chord-Scale Transformational System . . . . . . . . . . . . . . . . . . . . . . . 115
4.2.1 Introduction: Scale Theory . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.2.2 A GIS Proper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
4.2.3 Relaxing the GIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.3 Chord-Scale Transformations in Analysis . . . . . . . . . . . . . . . . . . . . . . . 129
4.3.1 Rahsaan Roland Kirk, Blues for Alice . . . . . . . . . . . . . . . . . . . . 130
4.3.2 Gene Ammons and Sonny Stitt, Autumn Leaves . . . . . . . . . . . . . . 138
4.3.3 Joe Henderson, Isotope . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Bibliography 254
Discography 262
Curriculum Vitae
viii
List of Figures
1 Introduction
1.1 A sample lead sheet of Just Friends (John Klenner/Sam Lewis). . . . . . . . . . . 18
1.2 Two piano realizations of Just Friends, mm. 18. . . . . . . . . . . . . . . . . . . 20
1.3 The changes to Autumn Leaves (Joseph Kosma), A section. . . . . . . . . . . . . 21
1.4 The changes to Autumn Leaves, arranged around the diatonic circle of fths. . . . 22
1.5 The Autumn Leaves gis, generated by diatonic step and descending fth. . . . . . 25
1.6 The complete changes to Autumn Leaves. . . . . . . . . . . . . . . . . . . . . . 29
1.7 A transformation network for Autumn Leaves, bridge. . . . . . . . . . . . . . . . 29
1.8 The diatonic cycle of Autumn Leaves, with a hypermetrically displaced copy
spanning the formal boundary at the end of the bridge. . . . . . . . . . . . . . . . 30
1.9 Changes to Alice in Wonderland (Sammy Fain/Bob Hilliard), mm. 116. . . . . . 31
1.10 The changes to Alice in Wonderland, with transformational labels between
harmonies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.11 A diatonic cycle in C major. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.12 Changes to How My Heart Sings (Earl Zindars), mm. 112. . . . . . . . . . . . . 33
1.13 Changes to All the Things You Are (Jerome Kern). . . . . . . . . . . . . . . . . . 34
1.14 Two versions of All the Things You Are, nal nine bars. . . . . . . . . . . . . . . 35
2 iiV Space
2.1 The bridge of All the Things You Are (Jerome Kern). . . . . . . . . . . . . . . . 39
2.2 A transformation network for a single iiVI progression. . . . . . . . . . . . . . . 39
2.3 The underlying transformation graph for a single iiVI progression. . . . . . . . . 40
2.4 Voice leading in the iiVI progression. . . . . . . . . . . . . . . . . . . . . . . . . 42
2.5 A transformation graph and network for a small portion of iiV space . . . . . . . . 43
2.6 The complete iiV space, arranged around the circle of fths. . . . . . . . . . . . . 45
2.7 Changes for the A section of Ceora (Lee Morgan). . . . . . . . . . . . . . . . . . 47
2.8 The A section of Ceora in iiV space. . . . . . . . . . . . . . . . . . . . . . . . . 47
2.9 Changes for the B section of Ceora (Lee Morgan). . . . . . . . . . . . . . . . . . 48
2.10 The complete iiV space, showing tritone substitutions. . . . . . . . . . . . . . . . 50
2.11 The Mbius strip at the center of iiV space. . . . . . . . . . . . . . . . . . . . . . 50
2.12 Voice leading in the TF and TFT transformations, compared. . . . . . . . . . . . . 51
2.13 A transformation network for a small portion of iiV space, with tritone substitutions. 52
2.14 The slide7 transformation from D fmaj7 to Dm7 in mm. 56 of Ceora. . . . . . . 53
2.15 Changes to Blues for Alice (Charlie Parker). . . . . . . . . . . . . . . . . . . . . 54
2.16 An analysis of Blues for Alice in iiV space. . . . . . . . . . . . . . . . . . . . . . 54
2.17 A small portion of iiV space, including minor tonic chords. . . . . . . . . . . . . . 58
2.18 Changes to Solar (Miles Davis). . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.19 An analysis of Solar in iiV space, with C minor tonic. . . . . . . . . . . . . . . . 60
2.20 A generic version of iiV space, with unspecied tonic chords. . . . . . . . . . . . . 61
2.21 Voice leading in the TFblues transformation. . . . . . . . . . . . . . . . . . . . . . . 62
ix
2.22 A small portion of blues iiV space. . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.23 A portion of iiV space, conformed to the white-key diatonic circle of fths. . . . . 66
2.24 An A f -major diatonic iiV space, arranged in descending steps. . . . . . . . . . . . 67
2.25 Detail of diatonic iiV space, showing the slide 7 transformation between key centers
related by half-step. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.26 An analysis of Ceora in diatonic iiV space. . . . . . . . . . . . . . . . . . . . . . 69
3 Thirds Spaces
3.1 Changes to Lady Bird (Tadd Dameron), mm. 15. . . . . . . . . . . . . . . . . . 71
3.2 The complete minor-third representation of iiV space (m3 space). . . . . . . . . . 73
3.3 The toroidal center of minor-third space (the m3-torus). . . . . . . . . . . . . . . . 73
3.4 The BD transformation in m3 space. . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.5 Solo changes for Isotope (Joe Henderson). . . . . . . . . . . . . . . . . . . . . . 76
3.6 An analysis of Isotope(solo changes) in the m3-torus. . . . . . . . . . . . . . . . 76
3.7 Isotope, head (Joe Henderson). . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.8 Changes to Giant Steps (John Coltrane). . . . . . . . . . . . . . . . . . . . . . . 81
3.9 Changes to Infant Eyes (Wayne Shorter). . . . . . . . . . . . . . . . . . . . . . . 82
3.10 Coltranes major-third cycle as a substitution for a iiVI progression. . . . . . . . . 82
3.11 The changes to Countdown (Coltrane), compared with Tune Up (Miles Davis). . 82
3.12 Giant Steps analyzed in Douthetts Cube Dance. . . . . . . . . . . . . . . . . . . 85
3.13 Cohns Four-Cube Trio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.14 Matthew Santas nonatonic cycles. . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.15 A major-third organization of iiV space. . . . . . . . . . . . . . . . . . . . . . . . 88
3.16 Coltrane changes as a iiVI substitution, shown in M3-space. . . . . . . . . . . . . 89
3.17 Changes to Have You Met Miss Jones? (Richard Rodgers/Lorenz Hart). . . . . . 91
3.18a Miss Jones, A section (mm. 18), analyzed in M3-space. . . . . . . . . . . . . . . 92
3.18bMiss Jones, second A section and bridge (mm. 923). . . . . . . . . . . . . . . . 92
3.18c Miss Jones, last two bars of bridge and nal A section (mm. 2332). . . . . . . . . 93
3.19 A transformation network for the bridge of Miss Jones. . . . . . . . . . . . . . . 93
3.20 A portion of Four-Cube Trio, redrawn to include major seventh chords. . . . . . . . 95
3.21 The transformations 3rd, 7th, slide 7 , TF, and TFT in the Four-Cube Trio. . . . . 96
3.22 Parsimonious voice leading of the minor seventh chord. . . . . . . . . . . . . . . . 97
4 Chord-Scale Transformations
4.1 The major and Lydian scales as stacks of thirds and fths. . . . . . . . . . . . . . . 101
4.2 Russells Interval Tonic Justication for the Lydian scale. . . . . . . . . . . . . . . 103
4.3 The seven principal scales of the F Lydian Chromatic scale. . . . . . . . . . . . . . 105
4.4 Russells Lydian Chromatic Order of Tonal Gravity. . . . . . . . . . . . . . . . . . 106
4.5 The four horizontal scales of the F Lydian Chromatic scale. . . . . . . . . . . . . . 107
4.6 The second mode of the C auxiliary diminished scale, in scalar and tertian formations. 111
4.7 A typical VI jazz lick, along with its chord-scale GIS analysis. . . . . . . . . . . . 121
4.8 A resolution from the F blues scale to the F Lydian scale. . . . . . . . . . . . . . . 121
4.9 An ambiguous acoustic scale and two concrete presentations of it. . . . . . . . . . . 124
4.10 A planetary model of the chord-scale GIS. . . . . . . . . . . . . . . . . . . . . . 127
x
4.11 Two presentations of the F acoustic scale in the planetary model. . . . . . . . . . . 127
4.12 Two non-harmonic tones in Kirks solo. . . . . . . . . . . . . . . . . . . . . . . . . 131
4.13 Two examples of enclosures in a iiVI progression. . . . . . . . . . . . . . . . . . 131
4.14 Harmonic generalization in Kirks solo. . . . . . . . . . . . . . . . . . . . . . . . . 132
4.15 The turnaround at the end of Kirks rst chorus. . . . . . . . . . . . . . . . . . . . 133
4.16 Two analytical possibilities for Kirks turnaround. . . . . . . . . . . . . . . . . . . . 134
4.17 Four dierent transformation networks of the turnaround in mm. 1113. . . . . . . . 135
4.18 Three iiVI progressions from Kirks solo on Blues for Alice. . . . . . . . . . . . 137
4.19 Two examples of anticipations, from Gene Ammonss solo on Autumn Leaves. . . 138
4.20 A barline shift from Ammonss solo. . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.21 A barline shift from Sonny Stitts solo on Autumn Leaves. . . . . . . . . . . . . . 139
4.22 An ambiguous barline shift in Ammonss solo, with two possible transformation
networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
4.23 Autumn Leaves, mm. 15, showing substitutions in mm. 34. . . . . . . . . . . . 141
4.24 The substitutions in mm. 34 of Autumn Leaves in iiV space. . . . . . . . . . . 141
4.25 Stitts 2f harmonic generalization, highlighting the implicit diatonic cycle. . . . . . . 143
4.26 A scalar shift highlights a change in function in Joe Hendersons solo on Isotope. . 143
4.27 Two interpolated ii7 chords from Hendersons rst chorus. . . . . . . . . . . . . . . 144
4.28 Signature transformations in the opening of Hendersons fth chorus. . . . . . . . . 145
4.29 A similar passage in chorus 14, with signature transformations. . . . . . . . . . . . 146
4.30 Hendersons solos on G f7E f7C7 in several choruses of Isotope. . . . . . . . . . 146
4.31 Hendersons sixth chorus, featuring contrapuntal motion. . . . . . . . . . . . . . . 148
5 Rhythm Changes
5.1 The basic changes to George Gershwins I Got Rhythm. . . . . . . . . . . . . . . 151
5.2 Several harmonizations of Rhythm changes, mm. 14. . . . . . . . . . . . . . . . . 152
5.3 The rst four bars of Rhythm changes in iiV space. . . . . . . . . . . . . . . . . . 153
5.4 The Serpents Tooth (Miles Davis), mm. 14. . . . . . . . . . . . . . . . . . . . . 153
5.5 Several harmonizations of Rhythm changes, mm. 58. . . . . . . . . . . . . . . . . 154
5.6 Several harmonizations of the Rhythm bridge, mm. 1724. . . . . . . . . . . . . . . 155
5.7 The four Rhythm bridge harmonizations of Figure 5.6 in iiV space. . . . . . . . . 155
5.8 Three possible hearings of the opening of The Serpents Tooth. . . . . . . . . . . 158
5.9 Thelonious Monk, Rhythm-a-ning, head. . . . . . . . . . . . . . . . . . . . . . 159
5.10 Diatonic harmonic generalization in Johnny Grins solo on Rhythm-a-ning. . . . 161
5.11 Altered diatonic generalization Grins solo. . . . . . . . . . . . . . . . . . . . . . 161
5.12 Grins blues generalization in a nal A section. . . . . . . . . . . . . . . . . . . . 162
5.13 A D fm triad as a B f blues subset in Grins solo. . . . . . . . . . . . . . . . . . . 162
5.14 Four possible harmonic contexts for a D f m triad. . . . . . . . . . . . . . . . . . . . 163
5.15 Half-note harmonic rhythm in an A section of Grins solo. . . . . . . . . . . . . . 163
5.16 Grins improvised solo line, along with Ahmed Abdul-Maliks bass line. . . . . . . 164
5.17 A 4-cycle bridge from Grins last chorus. . . . . . . . . . . . . . . . . . . . . . . 166
5.18 A 4-cycle bridge, with chord-scale elaborations that lead more strongly toward the
following harmony. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
5.19 The bridge from Grins eighth chorus, with tritone substitutions. . . . . . . . . . 166
xi
5.20 Monks dominant-cycle A section and blues generalization. . . . . . . . . . . . . . 168
5.21 Monks bridge from chorus 14, using the whole-tone scale. . . . . . . . . . . . . . . 168
5.22 George Coleman, Lo-Joe, melody, mm. 14. . . . . . . . . . . . . . . . . . . . . 171
5.23 The rst four bars of the A section of Lo-Joe, with ensemble. . . . . . . . . . . . 171
5.24 The third and fourth bars of all three A sections of Lo-Joe. . . . . . . . . . . . . 173
5.25 Possible interpretations of the third and fourth bars of the A sections. . . . . . . . . 174
5.26 A section endings in Lo-Joe. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
5.27 Lo-Joe, bridge (mm. 1724). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
5.28 Several possible transformation networks of the bridge. . . . . . . . . . . . . . . . 176
5.29 Sonny Stitt, The Eternal Triangle, head. . . . . . . . . . . . . . . . . . . . . . . 179
5.30 Derivation of the bridge of Eternal Triangle from a standard Rhythm bridge. . . . 180
5.31 Side-slipping in Sonny Rollinss solo. . . . . . . . . . . . . . . . . . . . . . . . . . 180
5.32 A non-sequential bridge from Sonny Rollinss solo. . . . . . . . . . . . . . . . . . . 181
5.33 Two sequential bridges from Rollinss solo. . . . . . . . . . . . . . . . . . . . . . . 182
5.34 A sequential bridge with metric shift from Sonny Stitts solo. . . . . . . . . . . . . 183
5.35 A loosely sequential bridge from Stitts solo. . . . . . . . . . . . . . . . . . . . . . 184
5.36 The end of chorus 11, including rhythm section. . . . . . . . . . . . . . . . . . . . 187
5.37 A harmonic interaction between Rollins and Stitt from chorus 15. . . . . . . . . . . 189
5.38 A convergent interaction between Rollins and Stitt from later in chorus 15. . . . . . 189
5.39 Rollinss outside playing in the second A section of chorus 17. . . . . . . . . . . . . 190
5.40 Rollinss increased chromaticism in the nal A section of chorus 17. . . . . . . . . . 191
5.41 Transcription of chorus 18, including rhythm section. . . . . . . . . . . . . . . . . 193
5.42 One possible interpretation of Rollinss descending line. . . . . . . . . . . . . . . . 195
5.43 Stitts side-slipping at the beginning of chorus 19. . . . . . . . . . . . . . . . . . . 196
5.44 Ensemble arrival on Bmaj7 at chorus 19A2 . . . . . . . . . . . . . . . . . . . . . . . 197
xii
List of Tables
3 Thirds Spaces
3.1 All possible consonant triads and incomplete dominant sevenths in the nonatonic
collection {D, E f , E, Fs, G, A f , B f , B, C}. . . . . . . . . . . . . . . . . . . . . . . 87
3.2 Parsimonious voice-leading among members of Smin , Sdom , and Smaj . . . . . . . . . . 97
4 Chord-Scale Transformations
4.1 Russells principal scale names and their other common names. . . . . . . . . . . . . 105
4.2 Modes of the C Lydian scale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
4.3 The eight principal modal tonics and their associated modal genres. . . . . . . . . . 111
4.4 Levines chord-scale description of C major scale harmony. . . . . . . . . . . . . . . 114
4.5 Levines chord-scale description of C melodic minor scale harmony. . . . . . . . . . 114
4.6 A scale index inspired by Russell, listed from most consonant to most dissonant. . . 118
4.7 Common chords in the modes of the F Lydian Chromatic scale. . . . . . . . . . . . 122
5 Rhythm Changes
5.1 A selection of Rhythm tunes and their composers. . . . . . . . . . . . . . . . . . . 150
xiii
Notes to the Reader
References to recordings in this dissertation are generally made only by giving the performers
name and album title; complete identifying information can be found in the discography.
When discussing a particular piece of jazz, the words piece, composition, and work all
seem out of place. In general, I have adopted the word tune to mean roughly the basic structure
of a work, including (primarily) its melody and chord changes. This is in keeping with the way
jazz musicians use the word: they may refer to a 16-bar tune, a Rhythm tune, or one of my
favorite tunes (all referring to the abstract structure of the tune and not simply the melody).
When I am referencing a particular instantiation of a work (e.g. Bill Evanss recording of Autumn
Leaves from Portrait in Jazz), I will make that clear.
In running text, dominant seventh chords are indicated with just a 7, major sevenths with
maj7, minor sevenths with m7, and half-diminished sevenths with m7 f5. In a minor key the
tonic chord is often played with a major seventh, which is indicated mM7. The progression
Dm7G7Cmaj7 thus indicates a D minor seventh moving to a G dominant seventh moving to a C
major seventh chord. In examples, chord symbols typically follow conventions used by The Real
Book.
There are two ways mathematicians notate function composition: left-to-right and
right-to-left. When combining two functions, f followed g, right-to-left orthography writes
g(f (x)), while left-to-right orthography writes (fg)(x). Right-to-left orthography is familiar to
most readers, and we will use it when talking about transformations as functions: statements of the
form f (x) should be read from right to left. When discussing strings of transformations, however,
right-to-left orthography can be confusing: L(P (x)) indicates the P operation applied to some
object x, followed by the L operation (it is easy to see how the problem can proliferate when there
are multiple operations involved). To avoid this problem, we will use left-to-right orthography
denoted by the symbol ; the operation P followed by the operation L is notated P L. This
combination of notation seems intuitive, but we will be explicit in situations where the
orthography may be confusing.
1. This use follows that of Julian Hook; see Exploring Musical Spaces (New York: Oxford University Press,
forthcoming), Section 2.3.
xiv
Chapter 1
Introduction
When compared to a score by Beethoven (for example), the jazz lead sheet appears strikingly bare.
The Beethoven score species nearly everything one might need to know in order to perform it.
Though the minor detailsdynamics, articulations, phrasing marks, and the likewill dier from
piece to piece, we can usually depend on the presence of some basic information. It is rare for
traditional scores not to include the instrumentation, for example, and a score that did not include
the number of measures or which notes to play in combination with which which other notes
would be very unusual indeed.
And yet, this is the usual state aairs for the jazz lead sheet, which is probably the most
common form of a jazz score. Most lead sheets only include the basic outline of a melody, along
with a set of changes that prescribe the harmonic structure of a piece. Beyond this most basic
instruction, every other aspect is left up to the performers. Of these two elements (melody and
harmony), harmony has a much larger role in determining the course of a particular jazz
performance, so it seems appropriate to focus our analytical attention on it.
Jazz is essentially a harmonic music. In a typical jazz performance, the melody of the piece is
heard only twice (at the beginning and the end), while the harmonic structure is heard
throughout, determining the structure of the performance. Each soloist typically plays one or
more choruses, where each chorus is understood as a single iteration of the pieces harmonic
structure. In marked contrast to a Beethoven score, jazz compositions usually remain unspecied
when it comes to their contrapuntal structure: performers will typically improvise counterpoint
1. This is not to say, of course, that there are not jazz compositions that do specify these minor details. These
compositions are the exception, rather than the rule, in the music in which this study is interested.
1
that ts with the underlying harmonic framework. Harmony is the main restraining factor of a
piece, and its primary method of coherence.
The word jazzwhich has been used at various times to describe McKinneys Cotton
Pickers, Benny Goodman, Sun Ra, John Zorn, Tito Puente, and Brad Mehldau, among many
othersis inescapably vague, so it will be useful at this point to delimit the terms of this study
somewhat. Here I am interested in in what might be called tonal jazz, which begins in the swing
area and continues through hard bop, covering roughly the years 19401965. In this music,
functional harmonic progressions are the norm; tonal jazz is meant in opposition to modal jazz,
where the rate of harmonic change is slower and the harmony is mostly non-functional. This
includes much of the music that most people think of when they hear the word jazz, including
big-band swing (Count Basie, much of Duke Ellingtons music), bebop (Charlie Parker, Dizzy
Gillespie, Thelonious Monk), and the mainstream jazz that followed bebop, known variously as
hard bop or post-bop (John Coltrane, Sonny Rollins, Bill Evans, and many others). I intend
the dates to be exible, especially on the later end; given the strong inuence of the bebop
tradition on jazz and jazz pedagogy, the hard bop style continued to exist well beyond 1965, and
many players today still play in the style.
Now that we have delineated jazz, we should explain exactly what we mean by harmony.
Harmony is of course one of the oldest topics in music theory, and as such has been hotly
contested throughout its history. It is often found in opposition to counterpoint; in this view,
counterpoint is concerned with individual melodic voices, while harmony is concerned with
individual verticalities. In other traditions (most notably the Schenkerian tradition), harmony is
understood to be an outgrowth of counterpoint: verticalities arise primarily through contrapuntal
2. Miles Daviss So What is probably the most well-known modal jazz piece; it is a 32-bar tune in which the
rst chord, Dm, lasts 16 bars, moves to E fm for 8 bars, and back to Dm for the nal 8. The term also describes other
similar pieces, including the rest of Daviss album Kind of Blue, John Coltranes recording of My Favorite Things, and
Herbie Hancocks Maiden Voyage.
3. As Scott DeVeaux puts it, bebop is both the source of the present . . . and the prism through which we absorb
the past. To understand jazz, one must understand bebop. The Birth of Bebop: A Social and Musical History (Berkeley:
University of California Press, 1997), 3.
2
procedures. Furthermore, study of harmony is often broken down by genre: tonal harmony plays
a dierent role than does chromatic harmony in both theoretical research and pedagogy.
When jazz musicians refer to harmony, they are typically referring to the changes
themselves; that is, the chord symbols given on a lead sheet or arrangement. Even when they are
not playing from sheet music, the chord symbol is the basic unit of harmonic understanding for
jazz musicians. The reason for this is largely practical: a chord symbol is a concise way of referring
to a particular sound, and improvising musicians must be able to understand this information
quickly (when reading) and to recall it easily (when improvising).
Since the pioneering work of George Russell in the late 1950s, many jazz musicians conceive of
an equivalence between a harmony (a chord symbol) and a scale. The chord symbol Dm7 might
imply a D dorian scale, for example, rather than simply the notes DFAC. Because any of the
notes of this scale will sound relatively consonant over a Dm7 chord, the chord symbol acts as a
convenient shorthand for a particular way of playing for a jazz improvisor. This equivalence
between chords and scales will be the focus of Chapter 4; for now it enough to note that
understanding jazz harmony often involves more than understanding relationships between
four-voice seventh chords.
When analyzing jazz harmony, it is often dicult to determine exactly what one should be
analyzing. Lead sheets as circulated in fake books can be highly inaccurate, and often cannot be
relied upon as a single source for any particular jazz performance, since it is rare that performers
play directly from a lead sheet with no modications. In the case of jazz standards which may have
4. This is not meant to imply that chromatic harmony is not tonal; rather, studies that focus specically on
chromatic harmony often dierentiate themselves from other tonal theoretical traditions.
5. George Russell, The Lydian Chromatic Concept of Tonal Organization, 4th ed., vol. 1, The Art and Science of
Tonal Gravity (Brookline, MA: Concept, 2001).
6. Fake books are collections of lead sheets that traditionally were compiled anonymously and sold illegally, in
order to avoid paying the copyright owners of the compositions they contained. The name fake book comes from the
fact that with the melody and chord changes, jazz musicians can easily fake a tune they do not know. The most
famous jazz fake book is ironically titled The Real Book, and was compiled in Boston in the early 1970s. In recent years,
fake books have become mainstream, and most of them have now obtained proper copyright permissions. Hal Leonard
now publishes the 6th edition of The Real Book (a nod to the ve illegal editions); many of the notorious errors in the
earlier editions have been corrected and it is now available for purchase legally. Further references to The Real Book in
this document refer to this version unless otherwise noted. For a history of fake books, see Barry Kernfeld, The Story
of Fake Books: Bootlegging Songs to Musicians (Lanham, MD: Scarecrow Press, 2006).
3
originated elsewhere, we might wonder whether should we analyze the original sources. In many
cases, however, the jazz standard version may be signicantly dierent from the original version,
reecting a history of adaptation by generations of jazz musicians. To make matters worse for the
hopeful academic, this knowledge is often secret knowledge, not written down and learned only
from more experienced musicians.
Many published jazz analyses rely on transcriptions of particular performances as a way to
avoid some of these issues. In general, this solution works well, and I will certainly make use of
transcriptions from time to time. This study, however, is interested in harmony more generally,
and transcriptions can confuse matters somewhat. The kinds of questions I am interested in
answering are of the type What can we say about harmony in the piece Autumn Leaves? and less
often of the type What can we say about Bill Evanss use of harmony in the recording of Autumn
Leaves from Portrait in Jazz? Furthermore, even transcriptions are not denitive when it comes
to harmony: the pianist and guitarist might not be playing the same chord; the soloist might have
a dierent harmony in mind than the rhythm section; or the bass player might play a bass line in
such a way that aects our perception of the chordal root. Even in the course of a single
performance, a group might alter a tunes harmonic progression, perhaps preferring some
substitutions during solos and others during the head.
This is a problem without one clear solution, and it may make more sense to use one method
or another depending on the situation. Some compositions have canonical recordingsColeman
Hawkinss recording of Body and Soul, for exampleand in those circumstances determining the
changes is usually unproblematic. Other compositions are more uid, and dierent choruses might
alter the basic structure within the course of a single performance (substituting a V7 f9 for a
V13 s11 chord, for example). In these cases, I am interested in what Henry Martin has called the
7. By the time they become standards, many non-jazz compositions (Tin Pan Alley songs or traditional songs like
Back Home Again in Indiana) have been adapted, typically to additional harmonic motion to provide interest during
solos. Some examples are more radical: the tune Alice in Wonderland (analyzed below), for example, is known to jazz
musicians as a jazz waltz with one chord per bar. The main title music in the 1951 lm from which it was taken,
however, is in 4/4 with a relatively slower harmonic rhythm.
8. The head is what jazz musicians call the statements of the melody in the course of a jazz performance. This
melody is typically played once at the beginning of a performance and again at the end (often referred to as the out
head).
4
ideal changes; a hypothetical set of chords that we can use as a basis for understanding the many
variations that might occur in actual performance. These changes represent a sort of Platonic
model of a composition; individual performances of Autumn Leaves can be seen as instances of
some ideal Autumn Leaves. Determining these ideal changes is often a process of mediating
published lead sheets, recorded versions, and other sources; throughout this study I have tried to
clarify exactly what harmony I am analyzing in any given example.
In an attempt to answer some of these questions, this dissertation presents a transformational
model of jazz harmony. While on the surface a transformational model may seem abstract and
far-removed from the concerns of performing jazz musicians, harmony in jazz ts together nicely
with David Lewins famous transformational attitude. A jazz musician does not typically think
of harmonies as a series of points in space, but rather as a series of characteristic gestures between
them. Rather than focusing on an underlying tonality, a jazz musician often tries to make the
changesto fully engage with the sound of each individual harmony.
There is often quite a large gap between the way jazz is most commonly taught (in jazz studios
and pedagogical books) and the way it has traditionally been understood by music theorists.
Another goal of the present study is to use transformational methods in an attempt to narrow this
gap, by bringing theoretical and mathematical rigor to materials that are often ignored by the
theory community, and by applying established theoretical principles in a way that corresponds
closely with the understanding of jazz musicians.
While we can never claim to know what jazz musicians think, we might get somewhat closer
to an answer by examining jazz pedagogical materials. In the late 1960s, jazz began to be accepted
into the academy, and many young jazz musicians began learning to play the music in schools,
9. Henry Martin, Charlie Parker and Thematic Improvisation (Lanham, MD: Scarecrow Press, 1996), 56. See
also David J. Heyer, Vocabulary, Voice Leading, and Motivic Coherence in Chet Bakers Improvisations (PhD diss.,
University of Oregon, 2011), 3639.
10. In places where I refer to a tune generically, I have provided at least two references to relatively
straight-ahead recorded examples in the Appendix A.
11. David Lewin, Generalized Musical Intervals and Transformations (New York: Oxford University Press, [1987]
2007), 159 (hereafter GMIT).
5
rather than exclusively from older musicians. To supplement this teaching, a great deal of
pedagogical material has appeared that aims to teach young musicians how to play jazz.
Unfortunately, there is little interaction between these pedagogical materials and music
theoretical materials. Pedagogical materials, such as the recent Berklee Book of Jazz Harmony, often
do not have bibliographies or mention recent work in the theoretical literature. Likewise, most
theoretical work does not refer to these harmonic handbooks which are staples of jazz education.
Mark Levines Jazz Theory Book, widely regarded in the jazz education world as the book on jazz
theory, does not appear in any of the bibliographies in a special jazz issue of Music Theory Online
(18.3), for example.
In recent years the theory community has embraced pedagogical materials as a means of
uncovering how historical musicians might have thought about their own music. Though jazz
pedagogues do not typically publish articles in music theory journals or otherwise consider
themselves to be music theorists, per se, the goal of their harmonic textbooks is quite similar to
the goals of pedagogical books in music theory: to teach students how to think (or hear, or
perform) in a particular style of music.
David Lewin points out in his introduction to transformational theory that when considering
a particular musical passage, we often conceptualize along with it, as one of its characteristic
textural features, a family of directed measurements, distances, or motions of some sort. I
certainly hear these characteristic motions when I listen to jazz, and I think it is these motions that
jazz pedagogues are emphasizing when they teach students to make the changes. Despite its
somewhat hostile mathematical appearance, transformational theory is an eective means of
exploring these families of intuitions. Modeling sets of chord changes as transformations between
12. The Thelonious Monk Institute of Jazz maintains a web page on the history of jazz education in America at
http://www.jazzinamerica.org/JazzResources/JazzEducation/Page.
13. A noted exception here is Andy Jaes Jazz Harmony (Tbingen: Advance Music, 1996), which features an
extended bibliography that includes many theoretical works.
14. This issue of Music Theory Online is a Festschrift in memory of Steve Larson (September 2012), guest edited
by Stephen Rodgers, Henry Martin, and Keith Waters.
15. See, for example, Robert Gjerdingen, Music in the Galant Style (New York: Oxford University Press, 2007); and
Giorgio Sanguinetti, The Art of Partimento: History, Theory, and Practice (New York: Oxford University Press, 2012).
16. Lewin, GMIT , 16.
6
harmonic objects, for example, allows the theoretical discourse to draw on the ways in which jazz
musicians teach harmony, and can bring these two disparate areas somewhat closer together.
Studies of jazz harmony in recent years have primarily taken the form of Schenkerian analyses that
seek to uncover large-scale voice-leading structures in order to dene tonality. Schenkerian analysis
has proven to be an extremely useful tool for analyzing tonal music, and Steve Larsons pioneering
work in applying its methods to jazz has undoubtedly expanded the eld of jazz studies and brought
jazz analysis into the theoretical mainstream. While theorists may disagree on exactly how we
should apply these Schenkerian techniques, hardly anyone seems to doubt that they are the best
way to examine tonal structures in jazz.
The touchstone of the Schenkerian jazz literature is Steve Larsons Analyzing Jazz: A
Schenkerian Approach, which is the culmination of his work of the previous decades. In this and
all of his work, Larson advocates what might be called an orthodox Schenkerian approach. He
treats the extended tones common in jazz (sevenths, ninths, elevenths, etc.) as standing in for tonic
members at some deeper structural level. Steven Strunks important article on linear intervallic
patterns in jazz also uses an orthodox approach, as does the work of Daniel Arthurs, David Heyer,
and Mark McFarland.
17. Steve Larson, Analyzing Jazz: A Schenkerian Approach (Hillsdale, NY: Pendragon Press, 2009). This book
makes use of much of his earlier work, including Strict Use of Analytic Notation, Journal of Music Theory Pedagogy
10 (1996): 3171; Schenkerian Analysis of Modern Jazz: Questions about Method, Music Theory Spectrum 20, no. 2
(October 1998): 20941; and Composition versus Improvisation?, Journal of Music Theory 49, no. 2 (October 2005):
24175.
18. Larson, Analyzing Jazz, 6.
19. Steven Strunk, Linear Intervallic Patterns in Jazz Repertory, Annual Review of Jazz Studies 8 (1996): 63115.
See also, for example, Daniel Arthurs, Reconstructing Tonal Principles in the Music of Brad Mehldau (PhD diss.,
Indiana University, 2011); David J. Heyer, Applying Schenkerian Theory to Mainstream Jazz: A Justication for an
Orthodox Approach, Music Theory Online 18, no. 3 (September 2012); and Mark McFarland, Schenker and the Tonal
Jazz Repertory: A Response to Martin, Music Theory Online 18, no. 3 (September 2012). Arthurss work on Mehldau
is something of an outlier, for reasons explained further in note 38.
7
At levels close to the background, these orthodox analyses of jazz do not appear signicantly
dierent than analyses of classical music, and indeed that is part of their appeal. Because most
jazz is basically tonal music, these theorists are interested in showing its connection to the
European classical tradition by using the same techniques to analyze both of them. This is
especially important to note, since Schenkerian analysis has a well-known ethical component. For
Schenker, the compositions that were well-described by his theory were judged to be masterworks.
By showing that jazz compositions can also be understood with these Schenkerian techniques, the
implicit conclusion is that they too should be judged to be masterworks. Underlying these
orthodox approaches, I think, is a desire to legitimize a place for jazz in academic music theory (I
return to this point below).
An opposing group of theorists also supports the use of Schenkerian analysis, but argues that
it should be adapted to account for tonal features specic to jazz. Principal among this group is
Henry Martin, who advocates for the use of alternative Urstze in jazz, including ascending or
gapped Urlinien, non-triadic descents, and chromatic or neighbor-note Urlinien. He argues that
jazz pieces are often inuenced by a more African-American aesthetic that favors repetition and
rhythmic interplay over voice-leading motion through descending linear progressions, and thus we
should not feel obligated to adhere to traditional Schenkerian techniques.
Martin draws in part on James McGowans dialects of consonance in jazz, which describes
dierent contextually stable notes of the tonic triad. McGowan argues that consonance in jazz is
stylistically dened, and that in certain styles we might hear extended tones as consonances, even
20. Throughout this document, the word classical used in this sense is used in its generic sense to stand for
tonal music that participates in the Western art music tradition. This denition is imperfect, but is useful for
distinguishing the music of Bach, Mozart, Beethoven, Schumann, and Wagner from that of Ellington, Basie, Parker,
Monk, and Coltrane. I do not mean to imply that jazz cannot participate in the Western art music tradition; certainly
at least some of it does.
21. Nicholas Cook, Schenkers Theory of Music as Ethics, Journal of Musicology 7, no. 4 (October 1989): 41539.
22. Henry Martin, Schenker and the Tonal Jazz Repertory, Tdschrift voor Muziektheorie 16, no. 1 (2011): 1617.
This work draws on earlier work of his own as well as that of others, including Allen Forte, The American Popular
Ballad of the Golden Era, 19241950 (Princeton University Press, 1995); and David Neumeyer, Thematic Reading,
Proto-backgrounds, and Registral Transformations, Music Theory Spectrum 31, no. 2 (October 2009): 284324.
23. Martin, Schenker and the Tonal Jazz Repertory, 7.
24. See James McGowan, Dynamic Consonance in Selected Piano Performances of Tonal Jazz (PhD diss.,
Eastman School of Music, 2005); and Psychoacoustic Foundations of Contextual Harmonic Stability in Jazz Piano
Voicings, Journal of Jazz Studies 7, no. 2 (October 2011): 15691.
8
though they would be dissonant in classical music. He describes three principal dialects: the
added sixth, common in early jazz and Tin Pan Alley standards; the minor seventh, particular to the
blues; and the major seventh, common in later jazz performance. For Schenker, the background
is derived from the consonant tonic triad; Martin is able use these stylistically determined
denitions of consonance as support for his stylistically informed background structures.
McGowans work is among the minority in recent years that does not feature a Schenkerian
bent; in addition to his dialects of consonance, he is interested in applying (paleo-) Riemannian
functional analysis to jazz. Despite this Riemannian focus, however, he is not interested in
transformational analysis: David Lewins name appears nowhere in his bibliography. Closer to my
own interests is John Bishops 2012 dissertation, which also turns to transformations and
mathematical group theory to close the gap between jazz pedagogy and jazz theory. Bishop
focuses primarily on the interplay of pure triads in jazz, and connects harmony to chord-scales via
these triads.
Other important non-Schenkerian models of jazz harmony are found in earlier works of
Martin and Strunk. In his dissertation, Martin advocates a syntactic approach based on the circle
of fths, in which chains of descending fths point towards a tonic pitch. Steven Strunks early
theory of jazz harmony is a layered approached that draws on the Schenkerian concept of analytical
levels to uncover a single tonal center for a jazz tune. These older models tend to fall more in line
with how jazz musicians themselves discuss harmony, and we will return to them below.
25. McGowan, Dynamic Consonance, 7679 and throughout. The dialects are particularly clear in nal tonic
chords, especially in the case of the characteristic tonic major-minor seventh of the blues.
26. Ibid. On the distinction between paleo- and neo- Riemannian analysis, see Steven Rings, Riemannian
Analytical Values, Paleo- and Neo-, in The Oxford Handbook of Neo-Riemannian Music Theories, ed. Edward Gollin
and Alexander Rehding (New York: Oxford University Press, 2011), 487511.
27. John Bishop, A Permutational Triadic Approach to Jazz Harmony and the Chord/Scale Relationship
(PhD diss., Louisiana State University, 2012).
28. Henry Martin, Jazz Harmony (PhD diss., Princeton University, 1980); see also his Jazz Harmony: A
Syntactic Background, Annual Review of Jazz Studies 4 (1988): 930.
29. Steven Strunk, The Harmony of Early Bop: A Layered Approach, Journal of Jazz Studies 6 (1979): 453.
30. I have a suspicion that these earlier models of harmony tend reect jazz musicians intuitive understanding
because they are not interested in legitimizing jazz analysis for the academy, as I suggest above. Both Strunk and
Martins articles appear in the Journal of Jazz Studies/Annual Review of Jazz Studies (the journal was renamed in 1981),
while more recent articles on jazz harmony have appeared in music theory journals: Music Theory Spectrum, Journal of
Music Theory, and the Dutch Journal of Music Theory, for example. As jazz research has moved from the fringes into
the theoretical mainstream, I think it has grown more removed from jazz practice itself. As I mention above, one of the
9
Given the prevalence of Schenkerian techniques in jazz analysis and its proven explanatory
power in other tonal repertories, it is reasonable to ask why a dierent approach like
transformational theory is useful or necessary. In order to answer this question, I would rst like
to problematize the Schenkerian focus of recent jazz analysis. Steve Larsons Schenkerian Analysis
of Modern Jazz: Questions about Method will serve as a useful foil; it is one of the fundamental
articles in the eld, and its titular questions will help to guide the discussion here.
In the article, Larson asks three main questions we must answer if we are to take seriously the
suggestion that such Schenkerian analysis is appropriate for jazz:
1. Is it appropriate to apply to improvised music a method of analysis developed for the study of
composed music?
2. Can features of jazz harmony (ninths, elevenths, and thirteenths) not appearing in the music
Schenker analyzed be accounted for by Schenkerian analysis?
Larsons answer to the rst question is yes. Many of Schenkers own methods were of course
developed for improvisatory music, and even if they had not been, they have proven to be
explanatory for such music. I agree with Larson on this point, and dont have anything in
particular to add. Certainly we should not expect that our theories can prove useful only for the
music for which they are designed; in fact as theorists we generally hope that the opposite is true,
and that our theories have broader applications than originally intended!
The second question has generated much more discussion in the literature regarding exactly
how we might apply Schenkerian methods to jazz, already discussed above. While this discussion is
goals of the present study is to show how we might narrow this gap while still approaching the music with the
necessary theoretical rigor.
31. This article appears with only slight changes as the second chapter of Larsons book, Analyzing Jazz. Garrett
Michaelsen critiques Schenkerian analysis in similar ways in Analyzing Musical Interaction in Jazz Improvisations of
the 1960s (PhD diss., Indiana University, 2013), 711.
32. Larson, Schenkerian Analysis of Modern Jazz, 210.
33. In the opening of Free Composition, Schenker refers to improvisation as the ability in which all creativity
begins; Free Composition [Der freie Satz], ed. and trans. Ernst Oster (New York: Longman, [1935] 1979). Schenkers
rst Erlauterungsausgabe (explanatory edition) was indeed of an improvisatory work, Bachs Chromatic Fantasy and
Fugue.
10
mostly one of the mechanics of analysis, it raises another point in which I am interested: is a jazz
musicians conception of musical space (or musical structure) the same as a classical musicians? The
orthodox Schenkerians argue that it is not, and that jazz is a fundamentally triadic music (since at
some deep structural level all of the extended tones are reduced away). Those that favor a modied
approach disagree with this characterization, but still agree that Schenkerian analysis is the best
way to approach jazz harmony. My own reactions to this question overlap with my answer to
Larsons third question, so I will return to it shortly.
Larsons third question regards compositional or improvisational intent. The argument is
perhaps obvious: because Schenkerian analysis depends on uncovering long-range voice-leading
plans, how could improvising musicians possibly hold such plans in their memory while playing?
At some level, we might not even be interested in the answer to this question. Schenkerian analysis
has proven explanatory, after all, for music that was doubtlessly composed without Schenkerian
methods in mind (namely, music written before Schenkers birth). Nevertheless, Larson spends a
great deal of time addressing this particular point, so we should see to it here as well.
After dismissing the possible intentional fallacy of this question, Larson turns to pianist Bill
Evans as an example of how a jazz musician could produce complicated long-range voice-leading
patterns while improvising. To do so, he relies heavily on an interview that Bill Evans gave on
Marian McPartlands radio program, Piano Jazz. Evans discusses how he always has a basic
structure in mind while playing:
McPartland: Well, when you say structure, you mean like, one chorus in a certain style,
another . . .
Evans: No, Im talking about the abstract, architectural thing, like the theoretical
thing.
Evans goes on to demonstrate how he has certain structural features in mind (harmonic and
melodic arrivals). Larson then shows how Evanss accompanying commentary can be understood as
34. The interview was recorded on November 6, 1978. The program is available online at
http://www.npr.org/2010/10/08/92185496/bill-evans-on-piano-jazz, and was released on CD under Evanss name as
well: Marian McPartlands Piano Jazz Radio Broadcast (The Jazz Alliance TJA-12038-2, 1993).
35. Larson, Schenkerian Analysis of Modern Jazz, 219.
11
explaining voice-leading events like prolongation and linear progressions, and provides
voice-leading analyses of his playing.
I think that Larsons use of Schenkers methods to analyze the music of Bill Evans is well
justied, but I am less sure of the extent to which they are applicable to jazz more generally.
Larson anticipates this objection, noting that it might be oered on two grounds: rst, that
Evans was unusually talented as an improviser; and second, that his way of thinking was radically
dierent from that of other jazz musicians. Certainly Evans was not a typical jazz musician: he
was white, and studied at the Mannes College of Music, at that time a cradle of Schenkerian
activity in this country. Larson suggests that Evans was such an inuential pianist that his
Schenkerian improvisational tendencies might have inuenced other musicians with whom he
played. While this may be true, Evans was probably not seen as inuential until after 1960, and this
study is concerned primarily with music before that time (or at least, with musicians whose style
was well established by that time).
Larson allows that the rst objection is justied: That Evans was an unusually talented
improviserand that Schenkerian analysis can show thisis a principal argument of this article.
This statement is representative of the legitimizing enterprise of the application of Schenkers
methods to jazz mentioned above. It also contains a dangerous implication, made explicit in
Larsons closing paragraphs:
Much jazz improvisation lacks the relationships that reward long-range hearing, and
consists, as [Andr] Hodeir observes, of disconnected bits of nonsense. . . . But the
fact that jazz musicians often say that a jazz improvisation should tell a story suggests
that many jazz musicians are concerned with creating and experiencing global
relationships. That they do not always achieve this goal in performance is not
surprisingthe task is dicult. But there are exceptions.
36. See especially the table in Larson, Schenkerian Analysis of Modern Jazz, 229.
37. Ibid., 239.
38. Brad Mehldau, who is the focus of Daniel Arthurss work, is quite similar to Evans, in that he is a
classically-trained white pianist with a strong acknowledged inuence of the European classical tradition.
39. Bill Evans made his rst recordings as a leader in 1958, but he was not widely known until his appearance on
Miles Daviss Kind of Blue in 1959. Shortly after he left Davis, he won wide acclaim with his trio with Scott LaFaro and
Paul Motian, whose groundbreaking rst album was Portrait in Jazz, released in 1960.
40. Larson, Schenkerian Analysis of Modern Jazz, 239.
12
Is Schenkerian analysis applicable only to jazz performances that are exceptions?
No, Schenkerian analysis may be applied to any jazz performanceand it may show the
shortcomings of that performance.
Far from retaining Schenkers methods but not his epistemology, his specic insights into
music but not the system of beliefs that supported them (as Nicholas Cook suggests that modern
Schenkerians often do), Larson seems to be using Schenkerian analysis to judge certain
performances as masterworks and others as inferior, much in the way done by Schenker
himself. Performances by musicians who do not share Evanss interest in the abstract,
architectural thing may well be excellent performances when judged by other value systems.
Furthermore, Schenkerian analyses of jazz often focus on what turns out to be the least interesting
part of a jazz piece. Jazz is essentially tonal music, so it is not at all surprising (to me, at least) that
it is often possible to reveal an Ursatz from a particular performance.
Whether or not jazz musicians are thinking in a Schenkerian manner is not really the point,
however. As Steven Rings puts it in the introduction to his book, any analytical act will . . . leave
a surplusa vast, unruly realm of musical experience that eludes the grasp of [a] single analytical
model. Corners of that vast realm may nevertheless be illuminated via other analytical approaches,
but those approaches will leave their own surpluses. And so on. Schenkerian analysis typically
focuses on long-range voice leading in order to reveal an underlying diatonic framework, while
deemphasizing (some would say reducing) surface details, including much harmonic chromaticism.
This focus on harmony as a rst-class object is something that is at the heart of much jazz
pedagogical material, and might help constitute a dierent set of analytical values by which we can
understand jazz. Rather than having to consign musicians who we cannot understand with
Schenkerian analysis to a second tier of appreciation, we can instead try to understand them on
13
something of their own terms. A Schenkerian analysis of Coltranes Giant Steps solo might reveal
that it is lacking an artistically convincing relationship among structural levels, but it would likely
be dicult to nd a jazz musician who did not hold the composition up as an example of Coltranes
supreme mastery of the music.
Given the current dominance of Schenkerian theory in the study of jazz harmony, we might ask
what transformational thinking brings to the table. Transformational theory in recent years has
focused on neo-Riemannian analysis, with particular emphasis on ecient voice leading and
non-functional, chromatic progressions. Steven Rings has written that this focus on
neo-Riemannian theory has led to a view that some works are divvied up into some music that is
tonal . . . and some that is transformational. Continuing, he argues that to do so is to
misconstrue the word transformational, treating it as a predicate for a certain kind of music, rather
than as a predicate for a certain style of analytical and theoretical thought. As he is right to point
out, there is nothing about transformational theory that necessitates its restriction to this locally
chromatic music; his book uses the theory to explain, as he says, specically tonal aspects of tonal
music. It is this use of transformational that I wish to bring to bear on jazz, which is essentially
tonal music.
Fundamental to the Schenkerian approach is the relatively equal balance of harmony and voice
leading; for jazz musicians, though, this balance is heavily weighted toward the harmonic.
Schenkerian analysis tends to deemphasize certain harmonies with the aim of exposing an
underlying diatonic framework. This goal is in contrast with the typical goal of a performing jazz
musician, for whom individual chords have rst-class status.
Transformational theory too, often treats harmonies as rst-class objects, and thus makes it
especially appropriate for analyzing jazz. Often, only because transformational theory can be used
45. Larson, Schenkerian Analysis of Modern Jazz, 241. Though there are many published analyses of Giant
Steps, I am unaware of any that use Schenkerian methods.
46. Rings, Tonality and Transformation, 9.
14
for more than simply examining harmony: Lewins Generalized Interval Systems (explained in
detail below) only require a set of elements, a group of intervals, and a function mapping pairs of
elements of the set into the group of intervals. Most commonly the elements of the set are
harmonies, but they do not have to be.
Mathematical music theories have become especially widespread in recent years. Many of
these models focus only on triads; while these models are valuable, nearly all chords in jazz are (at
least) seventh chords. Because the neo-Riemannian literature is relatively well-known, it will be
useful here to limit our focus to those theories that deal in some way with non-triadic music. This
work falls basically into two categories: work that deals exclusively with a single type of chord, and
work that deals with musical objects of dierent types.
Most of the studies dealing with a single type of chord are concerned with members of set
class (0258), the half-diminished and dominant seventh chords. In a 1998 article, Adrian Childs
develops a model for these chord types that is closely related to standard neo-Riemannian
transformations on triads. Edward Gollins article in the same issue of the Journal of Music
Theory explores three-dimensional Tonnetze in general, with special focus on the dominant and
half-diminished seventh chords.
In general, neo-Riemannian-type operations on the (0258) tetrachords turn out to be
somewhat less useful than their triadic counterpoints, owing to the symmetry of set class (0258).
Any one tetrachordal Tonnetz can only show a subset of all of the (0258) tetrachords, while the
47. One of the main subjects of Rings book is a gis that describes heard scale degrees, and does not include
harmony at all; see Tonality and Transformation, 4199, and throughout. Lewin provides several examples of
non-harmonic gises in GMIT , 1624.
48. As evidenced in part by the number of books published in the last several years; in addition to Rings, there is
Richard Cohn, Audacious Euphony: Chromatic Harmony and the Triads Second Nature (New York: Oxford University
Press, 2012); and Dmitri Tymoczko, A Geometry of Music: Harmony and Counterpoint in the Extended Common
Practice (New York: Oxford University Press, 2011).
49. Adrian P. Childs, Moving Beyond Neo-Riemannian Triads: Exploring a Transformational Model for Seventh
Chords, Journal of Music Theory 42, no. 2 (October 1998): 18193.
50. Edward Gollin, Some Aspects of Three-Dimensional Tonnetze, Journal of Music Theory 42, no. 2 (1998):
195206. Because the terms are in common use in the transformational literature, Tonnetz and the plural Tonnetze
are rendered without italics throughout this study.
51. Gollin refers to the dierences between the two-dimensional triadic Tonnetz and his three-dimensional
tetrachordal version as degeneracies (Ibid., 200). Childs cubic representation only shows 8 of the possible 24 (0258)
tetrachords: those related by parsimonious voice leading to a single diminished seventh chord (Beyond
Neo-Riemannian Triads, 188).
15
familiar triadic Tonnetz of course shows all 24 major and minor triads. Recognizing this limitation,
Jack Douthett and Peter Steinbach present a model that also includes minor sevenths and fully
diminished seventh chords, using a digram they refer to as the Power Towers. While Douthett
and Steinbachs description accounts for two of the three main types of seventh chords commonly
used in jazz (it is missing the crucial major seventh), all of these neo-Riemannian models focus on
parsimonious voice leading. While this focus is valuable, it will not prove to be terribly useful for
the functional harmony in which this study is interested.
The other group of transformational models consists of what Julian Hook has termed
cross-type transformations: he extends Lewins denition of a transformation network to allow
for transformations between objects of dierent types. This category of transformations contains
the inclusion transformations (discussed by both Hook and Guy Capuzzo) which map a triad into
the unique dominant or half-diminished seventh chord that contains it and vice versa. Also
included in this category are more general approaches for relating set classes of dierent
cardinalities, including Joseph Strauss formulation of atonal voice leading and Clifton Callenders
split and fuse operations. Finally, Dmitri Tymoczkos continuous tetrachordal space can
accommodate all four-note chords, but, as Hook notes, Tymoczko downplays (and sometimes
ignores) the transformational aspects of his geometric models.
Though I have mentioned these cross-type transformational works only in passing here, we
will return to some of them in some detail below, where they will be more relevant. Because Hook
52. Jack Douthett and Peter Steinbach, Parsimonious Graphs: A Study in Parsimony, Contextual
Transformations, and Modes of Limited Transposition, Journal of Music Theory 42, no. 2 (1998): 25556.
53. Cross-Type Transformations and the Path Consistency Condition, Music Theory Spectrum 29, no. 1 (April
2007): 140. Lewins denition of a transformational network is in GMIT , 19397.
54. Julian Hook, Uniform Triadic Transformations, Journal of Music Theory 46, nos. 1/2 (SpringAutumn
2002): 57126; Guy Capuzzo, Neo-Riemannian Theory and the Analysis of Pop-Rock Music, Music Theory Spectrum
26, no. 2 (October 2004): 17799.
55. Joseph Straus, Uniformity, Balance, and Smoothness in Atonal Voice Leading, Music Theory Spectrum 25,
no. 2 (October 2003): 30552; Clifton Callender, Voice-Leading Parsimony in the Music of Alexander Scriabin,
Journal of Music Theory 42, no. 2 (1998): 21933.
56. An explanation of this four-dimensional space is found in Tymoczko, A Geometry of Music, 93112; much of
this work is based on Dmitri Tymoczko, The Geometry of Musical Chords, Science 313, no. 5783 (July 2006): 7274;
and Clifton Callender, Ian Quinn, and Dmitri Tymoczko, Generalized Voice-Leading Spaces, Science 320, no. 5874
(April 2008): 34648. Hook points out Tymoczkos relationship to transformational theory in his review of Tymoczkos
book, 1314.
16
does not strictly dene what constitutes a type in a cross-type transformation, his formulation
will allow us to apply transformational procedures rigorously in situations where we might wish to
consider objects to be members of dierent types, even though they may be identical in some other
typological system.
As mentioned in Section 1.1, jazz musicians often begin learning a particular tune with a lead sheet,
and the changes found there form the harmonic foundation of a particular performance. Because
this dissertation is interested in jazz harmony generally, lead sheets serve as a useful abstraction of
the countless possible instantiations of any one tune, and considering them them briey here will
prove fruitful for the rest of this study.
A lead sheet typically gives only a melody and a set of chord changes: Figure 1.1 gives the Real
Book lead sheet for John Klenner and Sam Lewiss Just Friends. It is a very typical example, and
nearly everything about the page is designed to make it easy for a jazz musician to fake a
performance of the tune on the bandstand: the anonymous compilers of the book provide the
composer and lyricists names and a sample recording; the music is split into four-measure chunks
to make the phrases and the form of the tune clear; and there is almost no extraneous
informationeven the key signature is omitted on all lines but the rst. The melody is given for
the head, and the changes are provided for the rhythm section (usually piano, bass, and drums, but
sometimes other instruments like guitar) and for solos. All other aspects of the tune need to be
negotiated prior to performance (how the tune will begin and end, for example).
57. In particular, it will be useful to consider the IIm7 f5 chord (a half-diminished seventh) as a dierent type
than the V7 chord (a dominant seventh) given their dierent functional roles in jazz harmony, despite the fact that
they are of the same set class.
58. It is worth mentioning that not all jazz musicians read music; jazz is largely an aural tradition, and many early
jazz musicians did not read, instead learning the music by ear. For students learning jazz today, however, learning to
read chord changes is an essential part of their training. For an ethnomusicological survey of jazz musicians
relationships with lead sheets, see Paul Berliner, Thinking in Jazz: The Innite Art of Improvisation (Chicago:
University of Chicago Press, 1994), 7176.
59. This lead sheet is taken from the older, illegal Real Book (249). While the newer Hal Leonard edition
maintains most of the original selections, in some cases they do not: Just Friends does not appear until vol. 4 of the
Hal Leonard Real Book.
17
Figure 1.1. A sample lead sheet of Just Friends (John Klenner/Sam Lewis).
18
The Real Book uses standard conventions for labeling chords. A chord symbol consists of the
chord root (referred to by a letter name), and a symbol indicating the quality: the most common of
these are are the dominant seventh (simply 7), minor seventh (7) and major seventh
(maj7). Thus, the opening of Just Friends (G7Cmaj7) consists of a G dominant seventh
chord moving to a C major seventh chord. This abstraction is extremely useful for a performing
musician, but leaves something to be desired if pressed into use as an analytical system. The chord
symbols do not explicitly tell us, for example, that G7Cmaj7 is a typical VI progression in C
major; that kind of knowledge is implicit for experienced musicians and analysts.
Complicating the problem somewhat is that chord symbols are imprecise by design. In most
situations, jazz musicians do not want to be told exactly what notes they should play (if they did,
they probably would not have become jazz musicians); instead, they treat chord symbols only as
guidelines. A G7 chord would certainly include the root, third, and seventh (G, B, and F), but
might also include the s11 (Cs), f9 (A f ), or s5 (Ds), depending on the situation: the melody
might suggest certain alterations, a performer might prefer some alterations over others, or an
improvisor may work themselves into a dissonant portion of a solo where a bare dominant seventh
in the piano would sound especially out of place.
To illustrate the exibility of chord symbols, Figure 1.2 gives two realizations of the rst eight
measures of Just Friends. The rst is Mark Levines, from early in his book on jazz piano; it uses
what he calls three-note voicings (the root, third, and seventh). The second realization is my
own, and features many alterations to the basic outlines given by the chord symbols. Both of these
realizations are valid interpretations of the given chord symbols, and are meant to reinforce the
point that chord symbols, while only a guideline, indeed represent something important about a
given harmonic progression. For all of their imprecision, chord symbols represent a reality for
performing jazz musicians, and as such will be foundational for our work on harmony here.
60. Jamey Aebersold publishes a free book on his website that serves as an introduction to the very basics of jazz;
the section titled Nomenclature is especially useful for deciphering chord symbols. Jazz Handbook (New Albany, IN:
Jamey Aebersold Jazz, 2010), 15, http://www.jazzbooks.com/mm5/download/FQBK-handbook.pdf.
61. Mark Levine, The Jazz Piano Book (Petaluma, CA: Sher Music, 1989), 21. Realizing chord symbols is perhaps
most important for pianists and guitarists (who are most often charged with realizing them in performance), and any
introductory text on these instruments will be lled with voicings to be used in dierent situations.
19
a)
G7 Cmaj7 C 7 F7
#
& 44 n ww w
3
.
b b ww w
?# 4
4 nw
w w w
#
Gmaj7 Bb7 Eb7
ww w
3
&
w w b b ww w.
?# bw
w w bw
b)
# 4
G7b13 Cmaj13 C13 F13
& 4 b ww w
3 b9
.
w w b b wwww b www
? # 44 n ww ww
w
w w w n ww
#
Gmaj #11
13 Bb13 Eb9#5
w w
3 #11
&
# ww w b b wwww n n www.
? # ww ww nw
w w bw
bw
20
1.5 Diatonic Chord Spaces
It will be easiest to introduce the transformational approach to jazz harmony developed in this
study by way of an example. Much of this dissertation will be interested in the development of
various musical spaces and motions that are possible within them. This kind of approach was rst
developed by David Lewin in Generalized Musical Intervals and Transformations, and a review of
his approach will be useful before moving on to more involved examples.
Figure 1.3 shows the chord changes to the A section on the jazz standard Autumn Leaves. Jazz
musicians sometimes refer to this progression as a diatonic cycle: it uses only seventh-chords
found in the G-minor diatonic collection. As in classical music, the leading tone is raised in the
dominant chord so that the resulting chord is D7, not Dm7. We can easily arrange this
progression around the familiar circle of fths, placing the tonic G minor at the top of the circle
(see Figure 1.4).
While this arrangement around the diatonic circle of fths makes intuitive sense, it can also
represent what Lewin has called a Generalized Interval System (gis). Generalized Interval Systems
are Lewins way of formalizing the directed measurements, distances, or motions that we often
understand as characteristic textural features of a given musical space. Though we will unpack
A-7 b5 D7 G-
62. The Real Book gives these changes in E minor, but most recorded performances are in G minor. I have
transposed the given changes to reect the most common performance key. Steven Strunk analyzes both Autumn
Leaves and How My Heart Sings (analyzed below) as examples of 107 linear intervallic patterns in Linear
Intervallic Patterns in Jazz Repertory, 9697.
63. Emile De Cosmos etude book titled simply The Diatonic Cycle (North Bergen, N.J.: EDC Publications,
1970) gives many dierent solo patterns possible over the diatonic cycle in all twelve major and minor keys.
64. Lewin, GMIT , 16.
21
G-
D7 C-7
A-7 b5 F7
Ebmaj7 Bbmaj7
Figure 1.4. The changes to Autumn Leaves, arranged around the diatonic circle of fifths.
this denition using the harmonies from the A section of Autumn Leaves as an example, Lewins
formal denition is as follows:
A Generalized Interval System (GIS) is an ordered triple (S, IVLS, int), where S, the
space of the GIS, is a family of elements, IVLS, the group of intervals for the GIS, is a
mathematical group, and int is a function mapping S S into IVLS, all subject to the
two conditions (A) and (B) following.
(A): For all r, s, and t in S, int(r,s)int(s,t) = int(r,t)
(B): For every s in S and every i in IVLS, there is a unique t in S which lies the interval
i from s, that is a unique t which satises the equation int(s,t) = i.
The rst element in a gis is a family of elements, S, which Lewin also calls a musical space.
Preceding this formal denition, he gives a number of examples of musical spaces, including the
familiar diatonic and pitch and pitch-class spaces, along with less familiar musical spaces like
frequency space, time point space, and various durational spaces. In our Autumn Leaves
example, we are interested in the (unordered) set of harmonies in the G-minor diatonic collection:
S = {Gm, Am7 f5, B fmaj7, Cm7, D7, E fmaj7, F7}.
With the rst element of a gis satised, we must then dene a group of intervals (ivls).
Though we could perhaps imagine a number of dierent ways to dene intervals among elements
of the set S (a point to which we will return later), the most obvious is to measure distances in
65. Lewin, GMIT , Denition 2.3.1 (26). GMIT contains many terms that Lewin renders in all capitals; I have
rendered them here in small capitals (except when quoting directly) in order to reduce their typographical impact.
66. Ibid., 1625.
67. The tonic harmony is given as it is in The Real Book, simply as Gm. In performance, a musician might
choose to play this chord with a major seventh (GmM7), a sixth (Gm6), or even a minor seventh (Gm7; this is less
likely since minor seventh chords are most commonly ii chords, not tonics).
22
diatonic steps between chord roots. Because we are interested in abstract chord roots and not the
actual pitches played by some bass player or pianists left hand, we will use diatonic pitch classes.
This has the eect of modularizing the set of possible intervals, changing ivls from
{. . . , 2, 1, 0, 1, 2, . . .} (as it would be in diatonic pitch space) to {0, 1, . . . , 6} (diatonic pitch
class space). Arithmetic in this group is mod-7, exactly in the way that arithmetic using the more
familiar chromatic pitch class space, {0, 1, 2, . . . , 11}, is mod-12.
Lewin species that ivls must be a mathematical group, and we will take care here to show
that ivls = {0, 1, . . . , 6} is indeed such a group. A group is a set of elements, G, and a binary
operation, , that satises the four group axioms:
68. In chromatic pitch space, we might say that the interval between C4 and A3 is 3, while the interval between
C4 and A5 would be +21. In chromatic pitch class space, however, the interval is calculated mod-12 (because octaves
are equivalent), and both of these intervals are equal to 9.
69. Lewins denition of a mathematical group is spread over several pages in the rst chapter of GMIT (46).
For other denitions in works of music theory, see Rings, Tonality and Transformation, 1213, and Julian Hook,
Exploring Musical Spaces (New York: Oxford University Press, forthcoming), Ch. 5. Any introductory mathematical
text on group theory will contain the group axioms; see, for example, Israel Grossman and Wilhelm Magnus, Groups
and Their Graphs (Washington, D.C.: Mathematical Association of America, 1964), 1014 and Nathan Carter, Visual
Group Theory (Washington, D.C.: Mathematical Association of America, 2009), 51.
23
number gives 0, mod-7): 21 = 5; 11 = 6; and so on. The integers modulo n are labeled Zn , so
we may also refer to ivls in our Autumn Leaves example as the group Z7 .
The last element of a gis is an interval function that maps S S into ivls. In other words,
the interval from one element of S to another must be a member of the group Z7 . In our Autumn
Leaves example, the interval from one element of S to another is simply the number of steps in
the G-minor diatonic collection (always counting upward) between the two elements. Thus,
int(D7, Gm) = 3, since G, the root of the second chord, lies 3 diatonic steps above D, the root of
the rst. Likewise, int(Am7 f5, B fmaj7) = 1; int(F7, E fmaj7) = 6; and so on.
We now have all of the elements of a gis, but we must still prove that Lewins conditions A
and B are satised, which we will do by example. Condition A states that for all r, s, and t in S,
int(r,s)int(s,t) = int(r,t). In our Autumn Leaves example, int(Cm7, D7) and int(D7, E fmaj7)
(both interval 1) must combine to equal int(Cm7, E fmaj7) (interval 2). Second, for every chord s in
S and every interval i in Z7 , there must be a unique chord t which satises the equation
int(s, t) = i. For example, there must be exactly one chord that lies 2 units above Cm7: namely,
E fmaj7. It is easy to conrm that the two conditions are also true for any choice of r, s, t S.
Before shifting our focus from generalized intervals to transformations, I want to return to the
structure of ivls, the group Z7 . Above, we measured intervals by the number of G-minor diatonic
steps between chord roots: a single step corresponded to the interval 1. We can also say that the
group Z7 is a cyclic group, generated by the interval 1. (Cyclic groups are notated Cn , where n is the
size of the group; we could therefore label the group in question C7 .) Counting diatonic steps is
not the only way we might consider measuring intervals in the space, however. Given the
ubiquitous descending fths of Autumn Leaves, we might instead like to measure distance by the
number of descending fths between chord roots. Though this generation has no noticeable eect
on the abstract group structure of ivlsthe group is C7 in either caseit does aect the last
element of a gis, the interval function. As shown in Figure 1.5, int(Am7 f5, D7) = 3 when
70. The group C7 can be generated by any of its members, since 7 is prime. In general, a cyclic group can be
generated by a member only if the two are relatively prime. The group C12 , for example, can be generated only by 1, 5,
7, or 11; put another way, only the chromatic scale and circle of fths (or fourths) cycle through all 12 pitches in the
chromatic octave before returning to the starting point.
24
G- G-
F7 A-7b5 D7 C-7
Figure 1.5. The Autumn Leaves gis, generated by diatonic step (left) and descending fifth (right).
measured by diatonic steps (the left gure), but int(Am7 f5, D7) = 1 when measured by descending
fths (the right gure).
Generalized Interval Systems are but one part of Lewins project; we will now turn our
attention to the transformations of the books title. Much has been made of the
transformational attitude that accompanies the shift from generalized intervals to
transformations that occurs in the later chapters of GMIT. As Lewin has it, gis thinking
represents a Cartesian, observer-oriented position, examining musical objects as points in abstract
space. This is in contrast to the transformational attitude, which Lewin describes as much less
Cartesian in what is perhaps the most-cited portion of the book:
Given locations s and t in our space, this attitude does not ask for some observed
measure of extension between reied points; rather it asks: If I am at s and wish to
get to t, what characteristic gesture (e.g. member of strans) should I perform in order
to arrive there?
The gis perspective outlined above is an intervallic perspective: we developed a system that
allowed us to say that the distance from Am7 f5 to D7 is 3. By replacing, as Lewin does, the
71. To avoid confusion, we will use only the gis that measures distance in diatonic steps from this point on.
72. Lewin, GMIT , 159. Henry Klumpenhouwer argues that the dichotomy between Cartesian gis thinking and
what he calls anti-Cartesian transformational thinking is the central theme of GMIT ; see In Order to Stay Asleep
as Observers: The Nature and Origins of Anti-Cartesianism in Lewins Generalized Musical Intervals and
Transformations, Music Theory Spectrum 28, no. 2 (Fall 2006): 27789. For more on the transformational attitude,
see Rings, Tonality and Transformation, 2429; Ramon Satyendra, An Informal Introduction to Some Formal
Concepts from Lewins Transformational Theory, Journal of Music Theory 48, no. 1 (Spring 2004): 1023 and 11617;
and Julian Hook, David Lewin and the Complexity of the Beautiful, Intgral 21 (2007): 17277.
25
concept of interval-in-a-gis with the concept of transposition-operation-on-a-space, we can
convert this gis statement into a transformational one: transposing the root of Am7 f5 by three
diatonic steps gives D7. The transformational statement is more active, replacing distance metrics
with verbs like transpose.
Though there is certainly a dierence in the language used in gis statements and that used in
transformational statements, Lewin takes care to note that the two attitudes are not diametrically
opposed:
26
use the lowercase, tk . This diatonic transposition operator will be used in almost the same way,
however: the operation tk is that transposition which transposes the root of a chord k steps inside
the G-minor diatonic collection. With this understanding, the conversion works as expected: the
interval between Am7 f5 and D7 is 3, so the operator t3 maps Am7 f5 onto D7 in the space
S = {Gm, Am7 f5, B fmaj7, Cm7, D7, E fmaj7, F7}. The same statement may also be written as
t
Am7 f5
3
D7, if we want to emphasize the idea of a transformation as a mapping operation.
Now that we have explored the underlying mathematics, we are in a position to reexamine the
A section of Autumn Leaves, (as given in Figure 1.3), as well as the circle-of-fths arrangement
in Figure 1.4. The progression is, quite simply, a series of t3 operations within the G-minor
diatonic space; each chord root descends by diatonic fth. In a mod-12 chromatic space (like the
ones we will begin developing in the next chapter) it can be dicult to make sense of this
harmonic progression. The chord qualities can be confusingtwo major sevenths in a row
followed immediately by a half-diminished seventhand it is sometimes hard to remember where
the tritone in the bass falls in the progression, given the prevalence of perfect-fth bass motion in
chromatic space. Understood in the context of this diatonic space, however, the harmonic motion
becomes much clearer.
It is reasonable to pause at this point and ask what advantages this transformational approach
brings. After all, we began our discussion of Autumn Leaves by noting that jazz musicians
sometimes refer to the progression of the A section as a diatonic cycle, and it may seem as
though we have gone through a great deal of mathematical rigmarole simply to arrive back at our
starting point. There would seem to be very little dierence between describing a progression as a
diatonic cycle and describing it as a series of t3 transformations in the G-minor diatonic set of
seventh chords. And yet, this is almost exactly the point. Transformational theory allows us a
78. The diatonic transposition operator tk is described in Julian Hook, Signature Transformations, in Music
Theory and Mathematics: Chords, Collections, and Transformations, ed. Jack Douthett, Martha M. Hyde, and
Charles J. Smith (Rochester: University of Rochester Press, 2008), 13940.
79. The intervals chosen for measurement in the gis do have an impact on the associated transformational
system: if we had used the descending fths generation as described above, the the operator t1 (not t3 ) would map
Am7 f5 onto D7, since the interval in the underlying gis would be dierent.
27
means to formalize what is often intuitive knowledge for jazz musicians, thereby narrowing the gap
between the way jazz musicians discuss harmony and the way music theorists often do.
Before concluding this chapter, I want to examine Autumn Leaves in a bit more detail, then
move on to a few other analytical examples. The full chord changes to Autumn Leaves are given
in Figure 1.6. The opening A section, discussed at length above, can be understood as a series of t3
operations in the G-minor diatonic set. The bridge modulates to the relative major, but we can
still understand this passage using the same transformational system. After repeating the G-minor
iiVI progression that concludes the A section in the rst four measures of the bridge, the entire
progression is transposed up a third (a larger-scale t2 ) to repeat the iiVI in the key of B f .
Despite this modulation, the connections from chord to chord are all t3 operations, continuing the
chain that has been present since the beginning (see Figure 1.7).
There are three passages in this piece that are not simply t3 operations: the connection from
Gm to Am7 f5 that begins the bridge; the connection from B fmaj7 at the end of the bridge to
Am7 f5 that follows; and the third and fourth bars of the nal section, Gm7C7Fm7B f7. The
rst of these, GmAm7 f5, is a t1 transformation. The bridge begins by retracing the same
harmonic ground as the last four bars of the A section (a iiVI progression in G minor); the
connection to the bridge, then, can be understood as reversing the two descending fths that
1 1
ended the A section. This observation can be represented algebraically (t
3 t3 = t4 t4 = t1 ) or
graphically (by taking two steps counterclockwise in the circle of fths in Figure 1.4).
The next transformation that breaks the series of t3 operations is the t6 from B fmaj7 to
Am7 f5 at the end of the bridge. This t6 is easily understood as a combination of two t3 operations,
by imagining that there is a missing E fmaj7 chord in the last bar of the bridge. Interpolating
80. Lewins denition of a transformation network appears is at GMIT , 196. We will delay an in-depth
explanation until the next chapter, at which point it will be more relevant.
81. Recall the discussion in the preface (page xiv): the symbol here represents left-to-right function composition.
82. This observation has not gone unrecognized by jazz musicians: at least one recording (Vince Guaraldi, on the
album A Flower is a Lovesome Thing ) includes the VImaj7 chord in this measure, though most do not.
28
A C-7 F7 Bbmaj7 Ebmaj7
A-7 b5 D7 G- :
B A-7 b5 D7 G-
C-7 F7 Bbmaj7
A-7 b5 D7 G-
t33 t3 t3 t3 t3
A-7 b5 D7 G- C-7 F7 Bbmaj7
t2
this chord allows us to hear the progression (beginning on Cm7) as identical to that of the rst
eight bars, now displaced to span a formal boundary, as shown in Figure 1.8.
The only four chords in Autumn Leaves that cannot be understood in the gis developed
above are those in the progression Gm7C7Fm7B f7. While the chord roots belong to the
G-minor collection, the qualities are incorrect. This progression consists of two iiV
progressions, the rst in F and the second in E f (a tonic that, like the one at the end of the bridge,
does not actually appear in the music). These iiV progressions are best situated in chromatic,
rather than diatonic, space; for now, we will pass over this progression until we develop such a
space in the next chapter.
It is relatively rare for tunes to be as systematically diatonic as Autumn Leaves (though Bart
Howards Fly Me to the Moon comes close); instead, pieces often make use of diatonic cycles in
83. Hearing this parallelism also requires hearing the Gm7 in the third bar of the nal A section as equivalent to
the cadential Gm in the rst. The chord symbol Gm is ambiguous by nature; if a performer wanted to bring out this
parallelism, she could perhaps play the tonic chords in the rst A section as minor seventh chords.
29
C-7 F7 Bbmaj7 Ebmaj7
t3
A-7 b5 D7 G- :
A-7 b5 D7 G-7
Figure 1.8. The diatonic cycle of Autumn Leaves, with a hypermetrically displaced copy spanning
the formal boundary at the end of the bridge.
84. The most well-known recording of Alice in Wonderland is probably on Bill Evanss Portrait in Jazz. These
changes are as given in The Real Book, and as played by Evans. This gure gives the changes to the second ending of
the rst sixteen measures (the rst ending contains a iiV in D minor to return to the opening).
85. The interval functions are identical, with the caveat that diatonic steps are to be counted in the A-minor
collection, not the G-minor.
86. The E f7 chord is a tritone substitute for A7, which is of course the dominant of the following Dm chord.
This harmonic motion is not easily understood in either the A-minor or C-major diatonic spaces; we will return to
tritone substitutes in the next chapter.
30
D-7 G7 Cmaj7 Fmaj7
B-7 b5 E7 A- Eb7
D-7 G7 Cmaj7
Figure 1.9. Changes to Alice in Wonderland (Sammy Fain/Bob Hilliard), mm. 116.
D-7 G7 Cmaj7
Figure 1.10. The changes to Alice in Wonderland, with transformational labels between harmonies.
1.10). After a C-major iiV in mm. 910, the harmony moves up a step to Em7Am7. The
progression G7Em7 seems to move backwards in the cycle, almost as if realizing the phrase is in
danger of arriving at C major too soon. We can capture this intuition by choosing to understand
the progression not as a forward-directed t5 , but as an algebraically-equivalent combination of
three ascending fths: t3 1 t3 1 t3 1 (which we could also write as t3 1 raised to the third
3
power: (t3 1 ) ).
Throughout GMIT, Lewin is clear that transformational theory is a means of expressing our
intuitions about a musical passage in a mathematically rigorous way. As he puts it: If I want to
change Gestalt 1 into Gestalt 2 . . . , what sorts of admissible transformations in my space
(members of strans or otherwise) will do the best job? Our explication of diatonic
87. It is this Am7 that conrms we are in a C-major diatonic space; it would be more typical to transpose the
iiV progression exactly to Em7A7.
88. Steven Rings refers to Lewins intuitions as apperceptions; see Tonality and Transformation, 1721.
89. Lewin, GMIT , 159.
31
C maj7
G7 Fmaj7
D-7 B-7b5
A-7 E-7
seventh-chord spaces may appear to stem from the desire to label everything a t3 and move on,
condent that we could justify our labels mathematically if called upon to do so. This, though,
could not be further from the truth; developing the space allows us a powerful means to capture
intuitions (or apperceptions) like the one above. Though the operation t5 does map G7 to Em7 in
1 1 1
the C-major diatonic space, hearing this connection as t
3 t3 t3 instead represents the idea of
stepping backwards through the circle of descending fths (an observation that may not be easy to
show using other methods of analysis).
Both examples of outright diatonic cycles we have seen thus far have been in minor keys. In
general, minor-key cycles are easier to use than those in major keys (a C-major diatonic cycle is
given for reference in Figure 1.11). In minor, the raised leading tone in the V chord means that the
cycle consists of a iiVI in the relative major and a iiVI in the tonic joined by the VImaj7 chord
(as we saw in the second cycle in Autumn Leaves, Figure 1.8). In a major-key cycle, the lone
half-diminished seventh chord is followed immediately by three minor seventh chords (in C major,
Bm7 f5Em7Am7Dm7), which by comparison is a relatively unusable tonal progression.
Nevertheless, Earl Zindarss standard How My Heart Sings (the changes to which are given
in Figure 1.12) does indeed contain a major-key cycle. Unlike the previous two examples, this
tune begins on the iii chord, at which point it begins the t3 cycle. Once again, it is the Am7 chord
90. The canonical recording of this tune is again by Bill Evans, on the album How My Heart Sings! Evans seems
to have a propensity towards tunes with diatonic cycles: Portrait in Jazz also contains a very well-known recording of
Autumn Leaves.
32
E-7 A-7 D-7 G7
Figure 1.12. Changes to How My Heart Sings (Earl Zindars), mm. 112.
that alerts us that this progression takes place in diatonic, rather than chromatic, space (A7 would
seem to make more harmonic sense, as the dominant of the following D minor). After reaching
the tonic in m. 5, the t3 s continue making their way to A minor, four measures later.
Notably, the chord in m. 8 is an E7 f9this chord belongs not to C major, but rather to A
minor. At some point, then, the progression shifts from taking place in a C-major cycle (the E
chord is a minor seventh in the opening bar) to an A-minor cycle. Unlike the abrupt shift to the
relative major in Alice in Wonderland (signaled by E f7), this modulation is gradual, using a
traditional pivot chord. In a transformational reading, the progression maintains the t3 sequence
throughout the rst nine measures, but the underlying diatonic set changes almost imperceptibly
from C major to A minor. Zindars uses this modulation in order to negotiate the unusual
succession of chord qualities in the major diatonic cycle: by beginning the progression on iii in a
major key and modulating to the relative minor before returning to it, he is able to have his cake
and eat it toothe chord root E appears rst as a minor seventh and again later as a dominant of
the relative minor.
One more example will suce to conclude our discussion of diatonic cycles: Jerome Kerns
All the Things You Are (the changes are given in Figure 1.13). Though this tune does not
contain a cycle as explicit as the examples we have seen so far, viewing this piece through the lens
of diatonic seventh chord space reveals relationships that may otherwise go unnoticed.
91. Exactly which chord functions as the pivot is of little importance, so long as it happens before the E dominant
seventh. I am inclined to hear the Fmaj7 as a pivot (functioning simultaneously as IV and VI), in order to keep both
iiVI progressions (in C major and A minor) intact.
92. These changes are again taken from The Real Book. The Am7 f5 chord in the sixth bar of the A section is
not included in some charts of this tune, so I have put it in parentheses here. (The older, illegal Real Book as well as
another illegal fake book called simply The Book both omit this chord.)
33
A F-7 Bb-7 Eb7 Abmaj7
D bmaj7 G7 Cmaj7
B A-7 D7 Gmaj7
Figure 1.13. Changes to All the Things You Are (Jerome Kern).
All the Things You Are begins in the key of F minor, and progresses through a diatonic
cycle (a chain of t3 operations in an F-minor diatonic seventh-chord gis) until arriving on D fmaj7
in m. 5. At this point, the chord roots continue to descend by diatonic fth in the key of F
minor, but the qualities of the chords rooted on G and C have been altered, from Gm7 f5 and C7
(as they would be in the F-minor gis) to G7 and Cmaj7. This arrival on a C major chord, rather
than a C dominant seventh, has the eect of a half cadence in the prevailing key of F minor. The
half-cadential C-major chord also serves as a linking chord to the next phrase, which contains a
diatonic cycle in the key of C minor. Like the rst A section, this phrase also veers away from the
cycle to end in a half cadence on G major.
The bridge of this tune is usually described as being made up of two iiVI progressions, the
rst in G major and the second in E major. While this is true, we might also understand this
93. I tend to hear the opening of All the Things You Are in F minor, though it is certainly possible to hear it in
A f major instead. In either case, the diatonic gis is nearly identical; the only dierence is the quality of the C chord
(Cm7 in A f , and C7 in F minor). Readers who prefer to hear the opening in A f can easily make the necessary
alterations to the commentary here.
34
progression as an alteration of a diatonic cycle in E minor. The bridge begins on a iv chord, which
initiates a iiVI in the relative major. After two bars of Gmaj7, a t6 takes us to F sm7 f5. Just as in
the bridge to Autumn Leaves, we can understand this t6 as a combination of two t3 operations,
interpolating a missing Cmaj7 chord. Though the cycle of the bridge (and its associated gis) is in
the key of E minor, the nal chord is an E major seventh, a kind of Picardy third; in this way the
bridge, like the A sections, can end on a major chord.
The nal A section of All the Things You Are is an expanded version of the rst, now
ending in the overall tonic of A f . It begins, after a linking C7 s5 chord, by outlining a series of t3
operations that lead to D fmaj7. At this point we might expect the t3 s to continue, leading to
Gm7 f5C7Fm (the tonic of the cycle), but instead we see G f7Cm7B 7. This progression is
clearly not diatonic, so we cannot say much about it at this point. It is interesting to note, however,
that the harmony four bars from the end is a B fm7, which is exactly where the chain of t3 s would
have arrived, had it continued (see Figure 1.14). After this phrase expansion, the piece closes with a
iiVI in the key of A f .
a)
b)
Figure 1.14. The final nine bars of All the Things You Are.
a) The changes as written.
b) A hypothetical version that continues the t3 cycle in F minor/A f major.
94. Again, this implicit diatonicism has not gone unnoticed; Ahmad Jamal arrives emphatically on Cmaj7 in the
fourth bar of the bridge on Jamal at the Pershing, Vol. 2.
95. If the linking dominant did not appear at the end of the bridge, we would see the succession Emaj7Fm7.
This succession is remarkably similar to Lewins slide operation: retaining the third of the chord while moving the
root and fth by half-step; see GMIT, 178. Here, the seventh is also retained as a common tone; we will return to this
operation (which we will call slide7 ) in the next chapter.
35
Again, we might pause to ask what is gained by hearing All the Things You Are in diatonic,
rather than chromatic, space. After all, there is never a clear statement of a diatonic cycle in the
manner of Autumn Leaves or Alice in Wonderland, and music in chromatic space (as we will
begin to see in the next chapter) still tends to descend by fth. Without the guiding inuence of F
minor, though, the succession of chord qualities at the beginning is dicult to make sense of: two
minor chords, followed by a dominant seventh, then two major chords, all seemingly unrelated to
the phrase-concluding tonic C major. Understanding the rst 8 bars as diatonic, ending with a
tonicized half cadence, makes sense of the chord qualities, and eliminates the dicult-to-explain
third relation A fMCM that results from a desire to hear both VI progressions in the A section
as tonic-dening.
Hearing All the Things You Are diatonically allows us to listen to sections of music at once:
the rst eight bars are in F minor, the next eight in C minor, the bridge in E minor, and the last
twelve return to F minor before shifting to the relative major, A f , for the nal cadence. When we
hear the tune as a chain of t3 operations in shifting diatonic spaces, our attention is drawn to the
connections between the spaceskey areasthemselves, rather than the (comparatively boring)
series of descending diatonic fths that occur within them. While transformational analyses are
often accused of privileging chord-to-chord connections to the detriment of long-range hearing, in
this case the gis framework developed above allows us to hear over longer distances where
chord-to-chord connections may fail to do so (while, crucially, still recognizing the importance of
chord-to-chord connections for a jazz musician aiming to make the changes).
Though they do appear occasionally, most jazz tunes do not contain diatonic cycles, and thus
we will need to expand the transformational approach introduced here to account for a larger
portion of jazz practice. Chapter 2 will outline an approach to chromatic space that will help to
understand the iiVI progressions that we passed over in our discussion of diatonic space.
96. Henry Martin analyzes the A section simply as a series of descending fths (taking note of the aberrant root
tritone D f G) ending on C major. His analysis assumes a chromatic space, in which the harmonic progressions are
more dicult to make sense of: he notes that a certain tonal ambiguity pervades this piece (Jazz Harmony, 1519.).
97. Hearing the third-relations does have the nice side eect, absent from our analysis here, that all of the
phrase-ending major-sevenths in the rst sixteen bars (A f , C, E f , G) spell out the tonic major-seventh chord; ibid., 19
makes this observation.
36
Transformational approaches, and neo-Riemannian theories in particular, have ourished partly
because of their ability to explain non-functional progressions that contain primarily root motion
by thirds. Chapter 3 will draw upon this literature to approach jazz, especially common after bebop,
that is more dependent on thirds than fths for structure. Though harmony is crucially important
to performing jazz musicians, much of the jazz pedagogical literature equates chords with scales: a
Dm7 chord is functionally equivalent to a D dorian scale, for example. Chapter 4 will develop a
transformational approach for these chord-scales, treating scales as rst-class harmonic objects.
Finally, Chapter 5 will bring the theoretical work of the early chapters to a close, by taking a close
analytic look at tunes based on George Gershwins I Got Rhythm.
37
Chapter 2
iiV Space
Because most jazz is not purely diatonic, we need to expand our transformational system to account
for more chromatic examples. This chapter will begin that work, taking the very common iiVI
progression as its basis. Most of the spaces we will develop in this chapter will fall under the broad
category of fths spaces, but at the end of the chapter we will have occasion to return to the
diatonic space of Chapter 1 to see how they might be enhanced with the chromatic spaces
introduced here.
2.1.1 Formalism
The most common harmonic progression in jazz is undoubtedly the ii7 V7 I7 progression
(hereafter, simply iiVI, or often just iiV). It is the rst progression taught in most jazz method
books, and the only small-scale harmonic progression to have an entire Aebersold play-along
volume dedicated to it. The progression is so prevalent that many jazz musicians describe tunes in
terms of their constituent iiVs; a musician might describe the bridge of All the Things You Are
(shown in Figure 2.1) as being iiV to G, iiV to E, then VI in F. Given the importance of this
progression for improvising jazz musicians, it seems natural to use it as the basis for developing a
more general transformational model of jazz harmony.
Earlier
. versions of this chapter were presented at the annual meetings of the Music Theory Society of the Mid-Atlantic
(Philadelphia, PA, March 2013) and the Society for Music Theory (Milwaukee, WI, November 2014). I am grateful for
the many helpful comments and questions I received at these conferences (especially those from Stefan Love, Brian
Moseley, and Keith Waters).
1. Jamey Aebersold, The IIVI Progression, Jamey Aebersold Play-A-Long Series, vol. 3 (New Albany, IN: Jamey
Aebersold Jazz, 1974). The Aebersold play-along series is a staple of jazz pedagogues; most contain a selection of tunes,
along with a CD of a rhythm section so that students can practice with a recording. The iiV volume is number three
of well over 100, and includes the phrase the most important musical sequence in jazz! on the cover.
2. Our diatonic analysis of All the Things You Are in the previous chapter notwithstanding, the ubiquity of
iiVI progressions means that many jazz musicians are apt to hear the progression as successions of iiVs, even in
cases where a diatonic pattern may be present.
38
A-7 D7 Gmaj7
F-7
Figure 2.1. The bridge of All the Things You Are (Jerome Kern).
ii7 TF
V7 TF
C
Figure 2.2 shows a transformation network for a single iiVI progression; we will begin by
developing the formal apparatus for this progression, after which we can begin to combine iiVI
progressions to form a larger musical space. This gure, with its combination of general Roman
numerals and specic key centers, is designed to reect how jazz musicians tend to talk about
harmony; we might read this network as a iiVI in C. The combination of Roman numerals and
key areas bears some similarity to Fred Lerdahls chordal-regional space, but Figure 2.2 is a
transformation network, while chordal-regional space is strictly a spatial metaphor.
We have encountered one transformation network already (Figure 1.7 in the previous chapter),
but we have yet to dene the concept formally. Transformation networks are a major part of David
Lewins project in GMIT, and have been thoroughly covered in the literature, so we will need to
consider the formalism only briey here. A transformation network consists of objects of some
kind (here, they are chords) represented as vertices in a graph, along with some relations
(transformations) between them, represented as arrows. In Lewins denition, all of the objects in
a transformation network must be members of a single set S, and the transformations must be
3. The triangle on the C chord in this gure indicates a major seventh. The triangle (instead of maj7) is
intended to save space and reduce clutter in the graphical representations.
4. Fred Lerdahl, Tonal Pitch Space (New York: Oxford University Press, 2004), 9697.
5. Lewins denition of a transformation network is in Generalized Musical Intervals and Transformations (New
York: Oxford University Press, [1987] 2007), 196. For a relatively concise summary, see Steven Rings, Tonality and
Transformation (New York: Oxford University Press, 2011), 11016.
39
Smin Sdom Smaj
TF TF
Figure 2.3. The underlying transformation graph for a single iiVI progression.
functions from S into S itself. The transformations in Figure 1.7 are indeed Lewinian
transformations (mappings in the G-minor diatonic set), but the iiVI transformation network is
more complex.
The transformation TF in Figure 2.2 is in fact a cross-type transformation, as dened by Julian
Hook. Hook expands Lewins denition of a transformation network to include objects of
dierent types, necessary to dene transformations in the iiVI progression. The progression
contains musical objects of three types of diatonic seventh chords: minor, dominant, and major
sevenths (in the key of C major, the progression is Dm7G7Cmaj7). Using Hooks relaxed
denition, we are free to dene transformations from any set of objects to any other; to understand
the gure above, we need to dene the transformation TF such that it maps ii7 chords to V7
chords, and V7 chords to I7 chords.
Before dening the transformations, however, we rst need to dene the sets themselves. To
help with this, Figure 2.3 shows the underlying transformation graph of the transformation
network in Figure 2.2. Throughout GMIT, Lewin is careful to distinguish transformation graphs
from transformation networks: a graph is an abstract structure, showing only relations
(transformations) between unspecied set members, while a network realizes a graph, specifying
the actual musical objects under consideration. Because cross-type transformation graphs contain
objects of dierent types, a node in a cross-type transformation graph must be labeled with the set
from which the node contents may be drawn (even in the abstract transformation graph). In
40
Figure 2.3, the nodes are labeled simply Smin , Sdom , and Smaj , which we can understand as the sets
of minor, dominant, and major seventh chords, respectively.
While at its core the iiVI progression contains three types of seventh chords, in reality a
jazz musician might add any number of extensions or alterations to this basic structure. Given this
practice, dening the archetypal progression as being composed of four-note set classes (seventh
chords) seems unnecessarily restrictive. In order to allow for some freedom in the chord qualities,
we will consider only chordal roots, thirds, and sevenths; these pitches are sucient to distinguish
the three chord qualities in a iiVI.
In this chapter, we will represent a chord with an ordered triple X = (xr , xt , xs ), where xr is the
root of the chord, xt the third, and xs the seventh. The denitions of the three sets are as follows:
The denitions are intuitive, and have clear musical relevance: ii7 chords have a minor third
(interval 3) and minor seventh (interval 10), V7 chords have a major third and minor seventh
(intervals 4 and 10), and I7 chords have a major third and major seventh (intervals 4 and 11).
Dening the chords this way rather than as four-note set classes oers the great advantage of
exibility. Using the ordered-triple representation, the progressions Dm7G7Cmaj7 and
Dm9( f5)G7 f13 s9 f9Cmaj7 s11 are understood as equivalent, since the roots, thirds, and sevenths
are the same: both progressions are represented (2, 5, 0)(7, 11, 5)(0, 4, 11). Because the sets are
dened in pitch-class space, the three sets all have cardinality 12: each pitch class is the root of
exactly one ii7 , V7 , and I7 chord.
10. In fact, many jazz piano texts begin with three-note or shell voicings, consisting only of chordal roots,
thirds, and sevenths; see, for example Mark Levine, The Jazz Piano Book (Petaluma, CA: Sher Music, 1989), 1722;
and Joe Mulholland and Tom Hojnacki, The Berklee Book of Jazz Harmony (Boston: Berklee Press, 2013), 21112.
11. Here and throughout this chapter, pitch classes are represented as mod-12 integers, with C = 0; all
calculations are performed mod-12.
41
D-7 G7 Cmaj7
&
seventh 0 -1 5 -1 11
third 5 0 11 0 4
root 2 5 7 5 0
TF TF
With the space of the nodes dened, we can now formulate the transformation representing a
iiVI, which we will call simply TF:
Again, these denitions are designed to be musically relevant; the voice-leading diagram in
Figure 2.4 illustrates this more clearly. The root of the second chord is a fth below the root of
the rst (yr = xr + 5), the third of the second chord is a semitone below the seventh of the rst
(yt = xs 1), and the seventh of the second chord is a common tone with the third of the rst
(ys = xt ). In Lewins transformational language, if a jazz musician is at a ii7 chord and wishes to
get to a V7 chord, the transformation that will do the best job is TF: move the root down a fth
and the seventh down a semitone to become the new third. (Recall that we may also write
TF
ii7 V7 , rather than TF(ii7 ) = V7 .) Note that the transformation TF is also valid between V7 and
I7 (the second equation above, involving sets Sdom and Smaj ). TF is both one-to-one and onto for
sets of ordered triples; it maps each ii7 to a unique V7 , and each V7 to a unique I7 . As such, its
inverse (TF1 ) is well dened, and allows motion backwards along the arrows shown in the
transformation graph in Figure 2.3.
12. This gure represents what Joseph Straus calls transformational voice leadings in his study of atonal voice
leading; Uniformity, Balance, and Smoothness in Atonal Voice Leading, Music Theory Spectrum 25, no. 2 (October
2003): 30552.
42
TF TF TF TF
Smin Sdom Smaj ii7 V7 G
TF TF TF TF
Smin Sdom Smaj ii7 V7 C
TF TF TF TF
Smin Sdom Smaj ii7 V7 F
Figure 2.5. A transformation graph (left) and transformation network (right) for a small portion of
iiV space.
It is worth mentioning here that TF and TFT (which we will dene in the next section) are
well-dened operations for any ordered triple of members of the integers mod-12 (i.e., a member
of the set Z12 Z12 Z12 ). There is nothing mathematically incorrect about the statement
TF TF
(0, 1, 2) (5, 1, 1) (10, 0, 1), for example, but this succession has little musical relevance for
the applications under consideration here. Because Hook does not formally dene what he means
by a type, the formulation allows for situations like this one, in which the three types are all
members of a single larger set. The advantage for dening TF as a cross-type transformation is
that the content of a single node in the transformation graph is restricted to members of a
12-element set of specic ordered-triple congurations.
With this understanding of the transformations involved in a single iiVI progression, we
can continue to see how we might connect multiple progressions in order to form a larger iiV
space. Because root motion by descending fth is extremely common in jazz, we might consider
connecting iiVI progressions by descending fth; Figure 2.5 illustrates this arrangement both as
a transformation graph and a transformation network. This descending fths arrangement means
13. Hook himself makes this clear, noting that for any two sets S and T it is possible to dene a single-type
transformation in the union set S T, though it is not always clear how a function dened on one set should be
extended to cover both. He also notes that even when a single-type transformation is possible, the cross-type approach
is often simpler and more natural, which certainly seems to be the case here. Cross-Type Transformations, 5n8.
43
that all of the chords sharing a root are aligned vertically (directly below Gmaj7 is G7, which is
itself above Gm7). This arrangement allows us to dene two more transformations, which we will
call simply 7th and 3rd:
Like the TF transformation, the 7th and 3rd transformations have clear musical relevance: each
lowers the given note by a semitone. Although adjacent progressions are connected by descending
fth, the T5 labels connecting adjacent ii7 chords and I7 chords are shown in gray in the graph (and
omitted in the network, and in later examples), since these chords are not often directly connected
in jazz.
By extending the network of Figure 2.5, we arrive at the entirety of iiV space, as shown in
Figure 2.6. Because iiV space includes cross-type transformations, it does not easily form a
Lewinian gis. Considered more generally, though, it is easy to see that by considering a single
iiVI progression as a unit, iiV space maps cleanly onto ordinary pitch-class space. As Figure 2.6
makes clear, we can consider the iiVI in C as being three perfect fths above the iiVI in E f (or
put transformationally, the T3 operation transforms a iiVI in C to one in E f ). This formulation
does not allow us a means to say, for example, that the ii chord in C is x units away from the V
chord in E f , but because iiVIs are rarely split up, falling back on normal pitch-class distance is
sucient in most situations.
14. It would be possible to form a gis by considering all ordered triples as the group, as suggested above. While
this is possible, dening an interval function in this group is much more dicult: such a function would need to
account for the 36 ordered triples in iiV space (ii7 , V7 , and I7 chords) as well as the many more (1692) that are not
included in the space. Such a function is conceivable, but would not in any case reect the musical realities iiV space
is interested in portraying.
15. iiV space is a directed graph, so in circumstances where the pitch-class distance metric is somehow not
sucient, we can instead rely on the standard way of measuring distance in a directed graph: by counting the number
of edges in the shortest path between two chords. The distance from ii7 of C to V7 of E f is then 4: ii to V in C (1
edge), then 3 T5 s to V of E f .
44
C
G F
V7
V7
7
V
ii7
ii 7
7
ii
D Bb
7
V
V7
7
ii
ii 7
TF 7TH
A V7 ii7
T5 ii7 V7 Eb
3RD
ii
7 ii 7
V
7 V7
E Ab
ii 7
ii
7
ii7
V7
V
7
V7
B Db
Gb/F#
Figure 2.6. The complete iiV space, arranged around the circle of fifths.
Though we will return to the formalism a bit later, we have dened enough of iiV space at this
point to see how it might be useful in analysis. To do so, we will examine Lee Morgans
composition Ceora, rst recorded on the 1965 album Cornbread. The changes for the A section
are given in Figure 2.7, and the accompanying moves in iiV space are shown in Figure 2.8.
Ceora is in the key of A f major, and begins with the progression IiiVI in the rst three bars,
16. These changes are taken from The Real Book, and reect what is played on the Cornbread recording. In this
gure, the circle indicates the tonic, while the numbers on the labels indicate the order of transformations. Because
Ceora uses only a part of the space, the circle of Figure 2.6 has been squared o here so that the labels are easier to
read.
45
staying within a single horizontal slice of iiV space. This is followed immediately by a iiVI
progression in D f , a fth lower (mm. 45).
At this point, we might expect the iiVs to continue in descending fths, but the potential ii7
chord in G f is substituted with Dm7, its tritone substitute, which then resolves as a iiV in C.
Instead of resolving to C major in m. 7, this iiV resolves instead to C minor : both the seventh and
third of the expected Cmaj7 are lowered to become Cm7. (This progression is extremely common,
and is one of the principal means of maintaining harmonic motion in the course of a jazz tune.)
A similar progression in B f follows, leading to a B fm7 chord in m. 9. We then hear a iiV
progression in the tonic in mm. 910, but the expected A fmaj7 does not materialize; the E f7
chord moves instead to Cm7 as ii of B f (a northwesterly move in the space). This progression
repeats in mm. 1113, leading once again to the Dm7 chord rst heard in m. 5. The repeated
upward motions in the space have the eect of ramping up the tonal tension in the passage; not
only do the dominants fail to resolve as expected, but their stepwise rising motion takes the music
far away from the tonic A f . To release this harmonic tension, the iiV in C resolves at m. 15 to C
minor, at which point the harmonic rhythm doubles and the progression follows the normal
descending fths pattern to reach the tonic that begins the B section in m. 17.
The B section of Ceora (shown in Figure 2.9) follows much of the same trajectory as the A
section until the last four bars; the only dierences are the addition of the f5 in the Cm7 and the
s9 in the F7 in mm. 1112 of the section. Because we have dened chords and transformations only
in terms of chordal roots, thirds, and sevenths, neither of these changes aect our transformational
reading of the passage. Instead of ramping up to ii7 of C as in the A section, the iiV in B f
resolves to B fm7 in m. 12. This B fm7 becomes the ii chord of a iiVI in tonic, resolving in m. 15
of the section. A nal iiV in the last measure provides additional harmonic interest, and functions
as a turnaround to lead smoothly back to A fmaj7 to begin the next chorus.
At this point, we have successfully mapped all of the chords in Ceora to their associated
locations in iiV space; it is reasonable to ask, though, whether this mapping of chords to space
17. We will return to the concept of tritone substitutes in the next section.
46
Abmaj7 Bb7 Eb7 Abmaj7 Eb7 Ab7
Dbmaj7 D7 G7 C7 F7s9
Bb7 Eb7 C7 F7
D7 G7 C7 F7 Bb7 Eb7
ii7 8, 16 V7 9, 17 C
ii7 V7 F
15
ii7 V7 E b
13
ii7 2, 12, 20 V7 3, 21 A b
1 4
ii7 5 V7 6 D b
7
ii7 V7 Gb
47
Abmaj7 Bb7 Eb7 Abmaj7 Eb7 Ab7
Dbmaj7 D7 G7 C7 F7s9
locations should even count as analysis. After all, iiV space contains each minor, dominant, and
major seventh chord exactly once, so we did not even need to make any decisions as to where in the
space a particular chord should go. Have we, in fact, learned anything about Ceora from our
exploration of iiV space?
The answer, I think, is yes. Although we could criticize iiV space for being simply a
particular arrangement of common harmonic progressions in jazz, similar arrangements have
proven themselves useful in many areas of music theory: the circle of fths, the neo-Riemannian
Tonnetz, the pitch-class clock face, and countless others. One of the benets of iiV space is
that it allows us to easily visualize common harmonic motions in jazz. The succession of chord
symbols that make up the changes to Ceora may make immediate sense to an experienced jazz
musician, but iiV space allows others to to make sense of these relationships more clearly. The
fact that our analysis in iiV space may seem obvious is in fact a feature, not a bug; such a criticism
reveals that iiV space, with all its mathematical formalism, can clarify information that may
otherwise remain hidden in the raw data of the chord symbols.
18. For a study of many dierent musical spaces (and a defense of their use), see Julian Hook, Exploring Musical
Spaces (New York: Oxford University Press, forthcoming), Chapter 1 and throughout.
19. Music analysis is in many ways similar to the eld of data visualization: both often involve revealing the
underlying structure of what might otherwise seem like an undierentiated stream of data. As Edward Tufte tells us,
visualizations are often more precise and revealing than other, mathematical means of analyzing data. The Visual
Display of Quantitative Information, 2nd ed. (Cheshire, CT: Graphics Press, 2001), 1314.
48
2.2 Tritone Substitutions
2.2.1 Formalism
There is an important aspect of jazz harmony that has not yet been considered in our discussion of
iiV space. Crucial to harmony beginning in the bebop era is the tritone substitution: substituting
a dominant seventh chord for the dominant seventh whose root is a tritone away. Because
tritone-substituted dominants are functionally equivalent, both the progressions Dm7G7Cmaj7
and Dm7D f7Cmaj7 may be analyzed as iiVI progressions in the key of C.
This functional equivalence means that a tritone-substituted dominant can act as a shortcut to
an otherwise distant portion of iiV space. In the circle-of-fths arrangement of Figure 2.6, keys
related by tritone are maximally far apart (diametrically opposed on the circle), but in jazz practice,
G7 and D f7 are functionally identical (both dominant-function chords in C major). To account for
this progression in our space, we need to somehow bring these chords closer together; one solution
is to connect two segments of the space by T6 in a sort of third dimension, as shown in Figure
2.10. The topology of this space is more complicated than the ordinary circle of fths, however.
Once a progression reaches the bottom of the front side of the gure, it reappears at the top of
the back side (G f at the bottom is listed again as F s at the top); likewise, progressions
disappearing o the bottom of the back side reappear at the top of the front side (C major is given
in both locations).
This arrangement of key centers is topologically equivalent to a Mbius strip, which is
somewhat easier to see by focusing only on the dominant seventh chords, as shown in Figure
2.11. By wrapping this gure into a circle and gluing the left and right edges together with a
half-twist (so that the two G7 chords and the D f7/C s7 match up), we arrive at the desired Mbius
20. The tritone substitution has been discussed extensively in the literature, so we will not discuss it at any length
here. See, for example, Nicole Biamonte, Augmented-Sixth Chords vs. Tritone Substitutes, Music Theory Online 14,
no. 2 (June 2008); Henry Martin, Jazz Harmony: A Syntactic Background, Annual Review of Jazz Studies 4 (1988): 11;
and Dmitri Tymoczko, A Geometry of Music: Harmony and Counterpoint in the Extended Common Practice (New York:
Oxford University Press, 2011), 36065.
21. A similar diagram can be found in Werner Phlert, Basic Harmony, trans. Jrgen Krohn and Norman Bowie
(Werner Phlert Publications, 1989), 5, and can be seen implicitly in Figure 1-1 of Martin, Jazz Harmony.
49
ii7 V7 F #
ii7 V7 C
ii7 V7 B
ii7 V7 F
ii7 V7 E
ii7 V7 Bb
ii7 V7 A
ii7 V7 E b
ii7 V7 D
ii7 V7 A b
ii7 V7 G
ii7 V7 D b
ii7 V7 C
TF T5 T6 3RD 7TH
ii7 V7 Gb
T5
G7 C7 F7 Bb7 Eb7 Ab7 Db7
T11
T6
Cs7 Fs7 B7 E7 A7 D7 G7
T5
50
D b7
& b
D-7 G7 Cmaj7 D-7 Cmaj7
b
seventh 0 -1 5 -1 11 0 5 11 5 11
third 5 0 11 0 4 5 6 5 6 4
root 2 5 7 5 0 2 -1 1 -1 0
TF TF TFT TFT
strip. Though the underlying topology is easier to visualize this way, it is dicult to include all of
the other progressions (the iiVs themselves) in this diagram, so we will continue to use the
three-dimensional version of Figure 2.10, with the understanding that this topology remains in
eect. In any case, the arrangement of keys into the front and back sides is arbitrary, and may be
repositioned as necessary; it is often convenient to have the tonic key (when there is one) centrally
located at the front of the space.
While we could navigate this space using only the transformations TF and T6 , it is convenient
to dene another transformation to help with a common progression like Dm7D f7Cmaj7. We
will call this transformation TFT , to highlight its relationship to the more normative TF:
The TFT transformation represents a tritone substitution, but it transforms bass motion by fth
into bass motion by semitone (not bass motion by tritone); the voice-leading diagram in Figure
2.12 claries the relationship with the ordinary TF. Because TF and T6 commute, TFT can be
considered as either TF followed by T6 , or vice versa. With this new transformation, we can
understand the progression A fm7D f7Cmaj7 as a substituted iiVI in C:
TF TF
A fm7 D f7
T
Cmaj7.
51
TFT TFT
ii7 V7 G
T5
E
7
ID
SL
3RD
TF TFT
ii7 V7 C
T11 7TH
TF TF
ii7 V7 F
Figure 2.13. A transformation network for a small portion of iiV space, with tritone substitutions.
The introduction of tritone substitutes complicates the space somewhat; Figure 2.13 shows a
transformation network of the same portion of the space as in Figure 2.5, but with some chords
replaced with their tritone substitutes (shown in green). The relationship between a substituted
dominant in G major (A f7) and the unsubstituted ii7 chord in C (A fm7) is still, of course, a 3rd
transformation. The substituted V7 in C moving to the diatonic V7 in F changes the transposition
from a descending fth to a descending half-step, as indicated by the T11 arrow.
Perhaps most interesting in this tritone-substituted space is the new relationship between a
major seventh chord and the substituted ii7 in the progression a fth below (in this gure, between
Gmaj7 and A fm7). Normally there is no voice-leading connection between these two chords, but
with the substituted ii7 , the third and seventh are both held as common tones, and the root and
fth of the chord both ascend by semitone (from GBDF s to A fC fE fG f). Because of its
similarity to the standard slide transformation, with the addition of the common tone seventh, we
will call this transformation slide7 :
22. Determining the structure of the underlying transformation graph of this network is straightforward, so I
have not included a gure of it here.
23. Though we are dening chords as ordered triples in this chapter, I have included the fth in this description
to highlight the relationship to the triadic slide, which maintains the root and fth of a triad while changing the
quality of the third.
24. The slide transformation was introduced by David Lewin (GMIT , 178), but has since become a part of of
the standard set of Neo-Riemannian transformations. slide7 is dened here only as a transformation from I7 chords
to ii7 chords, but of course the triadic slide is an involution (two successive applications of slide to any triad will
result in the same triad).
52
ii7 V7 D b
T6
E7
ii7 V7 G ID
SL
ii7 V7 G b
T6
ii7 V7 C
Figure 2.14. The SLIDE7 transformation from D fmaj7 to Dm7 in mm. 56 of Ceora.
In jazz this progression occurs frequently when moving between key centers related by half step,
though it is uncommon in classical music. We have encountered this transformation once already:
the motion from D fmaj7 to Dm7 in mm. 56 of Ceora is indeed a typical slide7 transformation
(see Figure 2.14).
Equipped with these new tritone-substitution transformations, we can now analyze somewhat
more complicated music; Charlie Parkers Blues for Alice will serve as a useful rst example (the
changes are given in Figure 2.15). The essential structure of the blues is present: the tune arrives
on a subdominant in m. 5, and on a home-key iiV in m. 9 of the twelve-bar form.
Parker elaborates this basic structure with a series of stepwise descending iiV progressions
(see Figure 2.16). The rst of these is a diatonic descent: m. 2 jumps from the tonic F major to a
25. The slide7 transformation can be found in a chromatic sequence in the second movement of the Faur string
quartet, mm. 3639. Julian Hook analyzes this passage from a number of dierent mathematical perspectives in
Contemporary Methods in Mathematical Music Theory: A Comparative Case Study, Journal of Mathematics and
Music 7, no. 2 (2013): 89102.
26. This progression is often known as the Bird Blues, though Mark Levine calls it the descending blues in
The Jazz Theory Book (Petaluma, CA: Sher Music, 1995), 228. Like many sets of Parker changes, several dierent
versions exist; the changes here represent a mediation of these sources. The Real Book gives Am7 (a vi chord) instead
of Fmaj7 in m. 11; Levines Jazz Theory Book gives D f7 (a tritone substitute) instead of G7 in the second half of m. 3;
and the Charlie Parker Omnibook omits both D7 chords (the rst in the second half of m. 7, the second in m. 11) and
the tonic in m. 11 is an F7. Most of these dierences are minor, and over the course of a recorded performance the
changes might vary among all of these versions. Other compositions that contain this progression include Parkers own
Conrmation, Sonny Stitts Jack Spratt, and Toots Thielemanss Bluesette.
53
F maj7 E7b5 A7 D7 G7 C7 F7
G7 C7 Fmaj7 D7 G7 C7
ii7 V7 Ab
ii7 V7 D
(8, from m. 5)
2 3
ii7 V7 D b ii7 V7 Ab
9
ii7 V7 G ii7 V7 10D
ii7 V7 G b ii7 V7 D b
ii7 V7 C ii7 V7 G
4 5 11 12
ii7 13 V7 G b
1
14
ii7 V7 F ii7 V7 C
ii7 V7 Bb ii7 V7 F m. 11
6 7 15 16
ii7 V7 E b
(8, to m. 6)
ii7 V7 A b
Figure 2.16. An analysis of Blues for Alice in iiV space: mm. 15 (left), mm. 611 (right).
54
iiV in D, which resolves (via the 7th and 3rd transformations) to a Dm7 chord as the ii7 of C
major. We rst saw this progression in Ceora (mm. 67), where we noted that it was a very
common way of maintaining harmonic motion; instead of a iiV resolving to its tonic, it resolves to
the minor seventh chord with the same root. Because this progression is so common, it is useful to
dene it as its own transformation, which we will call EC (for evaded cadence). Unlike TF, EC is
useful only as a transformation from V7 chords to ii7 chords:
EC is of course equivalent to TF 7th 3rd, but only when the starting chord is a V7 chord (a
member of Sdom ). In iiV space, EC can be represented by starting on a dominant, then following
one arrow to the right and two arrows downward. The structure of the space immediately shows
that EC is impossible beginning on a ii7 chord; we can follow a single arrow to the right, but there
is only one downward arrow from a V7 chord.
This pattern of stepwise descending iiVs continues until arriving at the subdominant B f in
m. 5, which includes the standard blues alteration of the lowered seventh. The intuition that this
B f7 is in fact a stable harmony, rather than a descending-fth transposition of F7, can be captured
T
somewhat in our transformational labels. Instead of labeling this progression F7
5
B f7, we might
TF 7th
instead label it as F7 B f7; this designation expresses the notion that the B f7 chord is
heard as a resolution to a stable chord (the TF transformation) that has merely been inected with
the lowered seventh (the 7th transformation). Combining this with the rest of mm. 25, it is easy
to construct a transformation network:
TF EC TF EC TF TF 7th
Em7 f5 A7 Dm7 G7 Cm7 F7 B f7
27. At least, impossible if we want to stay within the three sets we are studying in this chapter. Like the other
transformations we have dened, EC is an admissible transformation on the set of all mod-12 ordered triples: if we
begin with a Dm7 chord, (2, 5, 0), EC gives us the triple (7, 10, 4), which of course is not a major, minor, or dominant
seventh chord.
28. The major-minor seventh chord as a stable chord is characteristic of the blues; see, for example,
James McGowan, Psychoacoustic Foundations of Contextual Harmonic Stability in Jazz Piano Voicings, Journal of
Jazz Studies 7, no. 2 (October 2011): 15859 and throughout. This fact is somewhat obscured in iiV space, since
major-minor sevenths appear in the space only as V7 chords; we will return to this limitation later in Section 2.3.2.
55
After this B f7 chord, Parker uses a chromatic stepwise pattern of iiVs (mm. 510), which we
can understand as a tritone-substituted version of the earlier descending fths pattern:
TF EC TF EC TF EC
B fm7 E f7
C
Am7 D7
C
A fm7 D f7
C
Gm7
(Here, ECC is the chromatic variant of EC, equivalent to TF slide7 , applied to a dominant
seventh chord.) Once this sequence arrives on Gm7 as the ii chord of the tonic F, there is a iiVI
progression in the home key. After the resolution in m. 11, the progression moves backwards
through fths space to begin a VIiiV turnaround to F major to begin the next chorus.
So far, we have not said very much about the rst two chords in Blues for Alice:
7th T
Fmaj7Em7. In iiV space, these chords are relatively far apart (4 edges): Fmaj7 F7
6
T 3rd
B7
5
E7 Em7. Because iiV space prioritizes functional relationships, chord progressions
that are close in terms of voice leading often appear quite distant in the space. In reality, a musician
would probably not think of this move as being distant, since the two chords are so close to one
another in pitch space: to rephrase again in Lewinian terms, the characteristic motion that does
the best job in taking us from Fmaj7 to Em7 is something like move the root down a half step and
both the third and seventh down a whole step. This transformation is easy enough to dene, but
would not be part of iiV space proper; inevitably, the space cannot tell us everything we want to
know about jazz harmony. The space is designed to show typical harmonic motions, so
progressions that do not seem to lie well in iiV space demand other explanations (and indeed,
often they are voice-leading explanations).
As it has been developed thus far, iiV space has a glaring omission: it requires that all tonic
chords be major sevenths. Certainly there are jazz tunes in minor keys, and thus there is a need to
29. A turnaround is what jazz musicians call a short progression that leads from a chord (often a tonic chord)
back to itself. They appear most commonly at the ends of forms, and provide harmonic interest during solos, when a
player might play several choruses in a row. The iiV appears frequently in this formal location, as do many of its
variants: viiiV, iiiVIiiV, iii fIII7ii fII7, etc.
56
account for the tonic minor. We have already seen iiV progressions that resolve to minor
chordswe called that transformation EC in the previous sectionbut the only minor chords in
the space are ii7 chords, not tonics. One of the advantages of iiV space is that it is easily extended
to account for harmonic features specic to particular situations; in this section, we will do just
that to allow for stable minor tonic chords.
The minor iiVi progression is usually played as ii7 f5V7 s9i mM7. Because we are
working with ordered triples of only roots, thirds, and sevenths in this chapter, the alteration of
the fth and ninth have no eect on our denitions of Smin and Sdom (dened in Section 2.1 above).
We do, however, need to formally dene the set of minor-major seventh chords (chords with a
minor third and major seventh):
With this denition in place, we can explore how this set interacts with the three sets we have
seen already in this chapter. The 3rd transformation works intuitively, and transforms a major
seventh chord to a minor-major seventh with the same root:
Likewise, the 7th transformation transforms a minor-major seventh into a minor-minor seventh
with the same root:
It will also be useful to dene versions of the TF and TFT transformations that transform a
dominant seventh into a minor tonic, equivalent to TF 3rd or TFT 3rd. We will call them
30. The extensions used for the dominant chord in a minor iiV is quite exible: Aebersolds IIVI Progression
gives the quality as 7 s9, but Mark Levine usually gives the chord symbol simply as alt. Levine includes the minor
iiV in the category of melodic minor scale harmony, and alt. is short for the altered scale (the seventh mode of
melodic minor). The G altered scale is GA f B f BnCsE f FnG, and is sometimes called the diminished
whole-tone scale, since it begins as an octatonic scale and ends as a whole-tone scale, or the super-locrian scale, the
locrian mode with a atted fourth. This sound could be expressed with a number of dierent chord
symbolsG7( f9 s9 s11 f13) or G7( f5 s5 f9 s9), for exampleso jazz musicians typically write alt. See The Jazz Theory
Book, 7077. We will return to this equivalence between chords and scales in Chapter 4.
57
TF TF
ii7 V7 G7
tf
g7
3RD
T5
3RD 7TH
TF TF
ii7 V7 C7
tf
c7
3RD
3RD T5 7TH
TF TF
ii7 V7 F7
tf
f 7
3RD
Figure 2.17. A small portion of iiV space, including minor tonic chords.
simply tf and tfT (the lowercase here is meant to parallel the use of lowercase letters to indicate
minor triads):
Note that unlike the standard TF and TFT transformations, tf and tfT only transform V7 chords to
I7 chords; the same transformations do not hold for ii7 to V7 .
Figure 2.17 shows a small portion of iiV space that includes minor tonic chords. Because
most jazz tunes do not contain exclusively minor chords, this gure gives both major and minor
tonic chords in every key. The transformations dened in the previous paragraph are readily
apparent in the space, with the exception of the 7th transformation from a minor-major seventh
to a minor seventhGmM7 moving to Gm7 as ii7 of F major, for example. Though we will not do
so here, determining how to ll in the gure with tritone substitutions, or to conform it around
the circle of fths in the manner of Figure 2.6, is easy enough to imagine (if not to draw, given the
added complexity of the minor-major sevenths).
58
C G7 C7
Fmaj7 F7 Bb7
C[maj7]
By way of a brief example, Figure 2.18 gives the changes for Miles Daviss Solar. This tune is
in C minor, though that is not immediately apparent from the changes themselves; in the
canonical recording of this piece (from Daviss own Walkin ), the C minor chords are played as
minor-major sevenths, and the piece ends on a CmM7 chord. The fact that the only tonic chord
appears in the opening bar of the form gives performances of this tune even more of cyclical
quality than is usual in jazz. By not arriving on a tonic at the end of the short 12-bar form, Davis
achieves a formal overlap at every chorus: the opening tonic serves simultaneously as the harmonic
resolution of the previous chorus and the formal beginning of the next.
The analysis in iiV space is mostly unremarkable, but it is given in Figure 2.19. Note that
this gure has replaced Cmaj7 at the top of the space with a minor tonic, CmM7, and as such
there is no arrow given between the C-minor tonic and V7 /F. This analysis, though, is not possible
in the iiV space of the previous section, since the C-minor harmony of the rst bar is decidedly
not a ii7 chord (it would be ii of B f , and there is no B f major harmony in the piece at all).
59
tf
ii7 V7 c
11 12
1
ii7 V7 F
2 3 4
ii7 V7 Bb
ii7 V7 E b
5 6 7
ii7 V7 A b
TF 3RD 7TH T5
ii7 V7 D b
8 9
10
(SLIDE 7 to ii of cm)
We have now solved the problem of tonic chords that happen to be minor-major seventh chords,
but in fact the problem is more general: it would be nice to have some way of allowing for any kind
of tonic chord we might nd in real music. As mentioned in Section 1.2, James McGowan has
argued for what he calls three dialects of consonance in jazz (extended tones we might consider
consonant): the added sixth, the minor seventh, and the major seventh. Both of the approaches
in this chapter so far have focused only on the major-seventh dialect, when it appears atop both
major and minor triads. Many Tin Pan Alley tunes end with tonic add-6 chords (which appear
31. James McGowan, Dynamic Consonance in Selected Piano Performances of Tonal Jazz (PhD diss., Eastman
School of Music, 2005), 7679.
60
TF RESI
ii7 V7
3RD T5
TF RESI
ii7 V7
3RD T5
ii7 TF V7 RESI
Figure 2.20. A generic version of iiV space, with unspecified tonic chords.
nowhere in iiV space), and as we noted in our discussion of Blues for Alice, it is very common
for a blues tonic to be a major-minor seventh chord (which appear only as V7 chords in the space).
The solution to this shortcoming of the space is to introduce some general transformation
(which we might call ResI) that could be redened as needed for each style. The generic space
would then appear as it does in Figure 2.20. This space is still arranged in perfect fths, and the
basic shape of the iiVI progressions is still present, but the quality of the tonic chords is
unspecied. Before using this space in analysis, of course, we must actually dene what we mean
by a tonic chord in a given situation. Because iiV space contains cross-type transformations, this
means we need to dene both the set of tonic chords and the transformation ResI, from Sdom to
the set of tonics. (By dening ResI to be equivalent to TF and dening tonic chords to be
members of Smaj , for example, the generic space here becomes the specic layout of iiV space rst
presented in Figure 2.10.)
Again, it will be easiest to demonstrate exactly how this generic space can be actualized by
means of an example. In our analysis of Blues for Alice in Section 2.2.2 above, we noted that the
32. The name of the transformation ResI is inspired by Steven Ringss use of the transformation ResC in the
rst chapter of Tonality and Transformation, 25-27.
33. In practice, ResI will almost always be a transformation that moves the root of a V7 down a perfect fth. In
theory, however, there is no limitation on the denition of ResI. It is possible, for example, to construct a space
where tritone substitutes are normative by dening ResI to be equal to TFT ; in this case, the gray arrows in Figure
2.20 would represent the transformation 7th T6 .
61
G7 C7
& b
seventh 5 -1 10
third 11 -1 4
root 7 5 0
TFblues
B f7 chord in m. 5 served as the resolution of the iiV in the preceding bar, but contained the
lowered seventh, which is typical for the blues. There, we tried to capture the intuition that the
B f7 was stable by labeling the transformation as TF 7th: a resolution merely inected with the
lowered seventh. This transformation, though, still results in the B f7 as a dominant seventh chord
(it appears in the space only as V7 of F).
The generic ResI transformation oers a better solution, in that we can dene a blues TF,
which resolves a V7 to a tonic major-minor seventh:
Note that TFblues is equivalent to T5 , but is dened in a way to demonstrate its similarity to TF (see
the voice leading in Figure 2.21; TFblues is undened on ii7 chords). We have also dened the set
SIMm7 , the set of tonic major-minor seventh chords; this seems intuitive, but is somewhat
complicated. SIMm7 is exactly equivalent to Sdom in the language of set theory, they are the same
set. The dierence between them is not structural, but interpretive: SIMm7 is the set of tonic
major-minor seventh chords, while Sdom is the set of dominant major-minor sevenths. This
distinction allows us to capture the dierence between B f7 as a stable resolution (as it is in m. 5 of
Blues for Alice) and B f7 as V7 of E f (as in m. 8 of Solar, for example).
62
This sort of interpretive analysis lies at the heart of Steven Ringss work in Tonality and
Transformation; the gises he develops there are designed to capture the intuitions that collections
of pitches can be heard (or experienced) dierently in dierent contexts. We can adapt this work
slightly to capture the intuition that tonic major-minor sevenths are experienced dierently than
dominant major-minor sevenths; Rings would say that the two sets have dierent qualia.
The tonal gis Rings develops in his second chapter consists of ordered pairs of the form (scale
degree, acoustic signal); as he has it, the notation (7, x) . . . represents the apperception: scale
degree seven inheres in acoustic signal x. Rings goes on to describe sets of these ordered pairs,
which we will use to capture our intuitions about the varying roles of the B f7 chord, as shown
below:
(4, A f )
(f7, A f )
(7, D)
(3, D)
(5, B f )
(1, B f )
B f7 as dominant B f7 as tonic
Here, the left gure (read bottom to top as root, third, seventh) represents B f7 as a dominant
seventh of E f (i.e., with 5, 7, and 4) while the right gure represents the same three pitch classes
as a tonic major-minor seventh (with 1, 3, and f7). Ringss system of heard scale degrees allows
us to distinguish between the sets SIMm7 and Sdom : the B f7 in m. 5 of Blues for Alice is a
member of SIMm7 , while the B f7 in m. 8 of Solar is a member of Sdom .
With the distinction between tonic and dominant minor-major sevenths worked out, we can
now specify the generic space of Figure 2.20 to create what we might call a blues iiV space; a
small portion of this space is shown in Figure 2.22. This space, though, presents another
complication: the top arrow marked with a question mark represents a transformation from B f7 as
63
TF TFblues
ii7 V7 Bb7
3RD T5 ?
TF TFblues
ii7 V7 Eb7
3RD T5 ?
TF TFblues
ii7 V7 Ab7
tonic to B f7 as dominant. The pitch classes remain the same, but the quale of the chord changes
from tonic to dominant, so this transformation is not simply the identity (T0 ).
Because this transformation is one of quale, we can turn to Ringss tonal gis for an
explanation. Intervals in this gis are measured with ordered pairs, like the elements themselves:
the rst element is a scale-degree interval (measured upward), and the second is a pitch-class
interval. Pivot intervals are those intervals where the second element of the pair is 0. In the
situation here, we have what Rings would call a pivot fth between the two B f7 chords:
(f7, A f )
(4, A f )
(5th,0)
(3, D)
(7, D)
pivot fth
(1, B f )
(5, B f )
The pitch-class interval here is 0, since both chords contain B f , D, and A f , and the scale-degree
interval is a 5th (1 to 5, 3 to 7, and f7 to 4). With this transformation dened, we can now more
fully realize our intuitions about the short passage in Blues for Alice (mm. 46):
64
2.3.3 Interaction with Diatonic Spaces
The preceding consideration of other kinds of tonic chords has taken us relatively far aeld from
the starting point of this chapter, and indeed these extensions are not necessary to understand most
tonal jazz. For many purposes, the conventional space developed in Sections 2.1 and 2.2 will be
sucient. What is missing in our treatment so far, though, is the concept of a global tonic. This
dissertation, after all, is interested in tonal jazz, and most of this music (and certainly all of the
examples in this chapter) is in a key. To this point, we have acknowledged this fact only by
mentioning the key of a particular tune in our analytical commentary, or circling the tonic chord in
a representation of the space. Dening iiV space as a fully chromatic space has many advantages:
it is rare that every chord in a tune can be understood in a single key, and it is convenient not to
have to switch continually between diatonic collections. Moreover, chromatic spaces are much
more regular than their diatonic counterparts: chromatic step size is consistent, while diatonic step
size varies between one and two half steps.
Still, given the exploration of diatonic transformational systems in Section 1.5, it seems wise
to consider what a diatonic iiV space might look like. We rst made the space chromatic by
arranging individual iiVI progressions in descending fths (recall Figure 2.6). We could instead
arrange the space according to the diatonic circle of fths, as shown in Figure 2.23. This space
looks much like the chromatic space, with the exception of the diminished fth between 4 and 7,
where the regular transformational structure of the chromatic space breaks down. The change of
the descending perfect fth (T5 ) to a diminished fth (T6 ) means that all of the transformations
3rd T T
linking these two key areas must all be combined with T1 : C7
1
C sm7, C7
6
F s7, and
7th T
Fmaj7
1
F s7.
As noted in Chapter 1, diatonic space (the cyclic group C7 ) can be generated by any of its
members, while chromatic space (C12 ) can only be generated by the members 1, 5, 7, or 11
40. For more on diatonic systems generally, see John Clough and Gerald Myerson, Variety and Multiplicity in
Diatonic Systems, Journal of Music Theory 29, no. 2 (Autumn 1985): 24970.
41. These transformations are all relatively parsimonious, and seem in some way related to the slide7
transformation introduced above. We will delay a discussion of these parsimonious aspects of these transformations
more generally until the next chapter.
65
ii7 V7 C
ii7 V7 F
3RD T1 T6 7TH T1 NB: not T5 !
ii7 V7 B
ii7 V7 E
TF 3RD 7TH T5
Figure 2.23. A portion of iiV space, conformed to the white-key diatonic circle of fifths.
(half-steps and perfect fourths/fths). Diatonic iiV space, then, oers the interesting possibility
of departing from the fths-based space used so far in this chapter, in favor of some other
organization of the space (since any interval we might choose will generate the entire space). To see
how such an organization might allow us to capture dierent kinds of analytical insights, I want to
return to briey to Lee Morgans Ceora.
In the analysis of Ceora in Section 2.1.2, we saw that the whole tune takes place in four key
areas: D f , C, B f , and the tonic A f . Given this organization, we might consider arranging iiV
space in descending diatonic steps, as shown in Figure 2.24. (The entire gure could be wrapped
around a circle so that the identical iiVI progressions in A f at the top and bottom of the gure
line up.) This arrangement into steps means that the key areas used in the tune are adjacent in the
space; in the chromatic space of Figure 2.8, they were separated by an intervening fth.
This gure is structurally a bit dierent than the other spaces explored in this chapter, so it
will be helpful to examine it in some detail before returning to Ceora. The arrangement into
descending steps means that we can no longer align all of the seventh chords sharing a root; only
66
ii7 V7 Ab
(t6 is half-step)
SLIDE7
ii7 V7 G
ii7 V7 F
ii7 V7 Eb
ii7 V7 Db
(half-step)
SLIDE7
ii7 V7 C
ii7 V7 Bb
7TH 3RD
TF t6
ii7 V7 Ab
Eb
t6 ( T10 = T5 T5)
(root T0 ) 7TH 3RD
Eb7 Ab7 Db
t6 ( T11 = T6 T5)
(root T1 ) 7TH 3RD T1
( SLIDE7 )
D7 G7 C
Figure 2.25. Detail of diatonic iiV space, showing the SLIDE7 transformation between key centers
related by half-step.
67
the major and minor seventh chords sharing a root are adjacent in the space. The Gmaj7 and
Gm7 (as ii7 /F) chords are close to one another, for example, but G7 (V7 /C) is farther removed.
The key areas in this gure are not connected by T5 , but instead by t6 : all of the roots of the major
seventh chords (reading down the right side of the gure) are members of the 4-at diatonic
collection. This t6 operation aects only chord roots, unlike the tk operator in Section 1.5.1; all of
the resulting chords are major sevenths. Though the diatonic distance between key areas is
consistent, the chromatic distance varies: there are two points in the space connected by half steps
t t
rather than whole steps (compare, for example, Fmaj7
6
E fmaj7 and D fmaj7
6
Cmaj7).
As we saw above, the tritone appearing in diatonic space alters the transformational structure
somewhat: transformations spanning this tritone must must be combined with T1 . Here, the
relationship between most I7 and ii7 chords is the transformation 7th 3rd, but between the keys
of D f and C (as well as A f and G), it is 7th 3rd T1 . This transformation is in fact equivalent
to the transformation slide7 (see the detail in Figure 2.25); this diatonic origin is one of the
reasons the slide7 transformation is so common in tonal jazz.
All of Ceora takes place in a relatively small portion of diatonic iiV space; Figure 2.26 gives
an analysis of the A section in this space (the changes can be found in Figure 2.7). The analysis
of course looks very similar to our analysis in Section 2.1.2, but the stepwise arrangement of the
space helps us to show dierent analytical insights. Moves in the space that are relatively close in
the chromatic fths arrangement in Figure 2.8 appear much larger in this arrangement (the move
from A fmaj7 to E fm7, marked 4 in this gure), and vice versa (the slide7 from D fmaj7 to
Dm7, marked 7).
It is worth noting at this point that although we have adapted iiV space to show aspects of
diatonicism, iiV space is still chromatic. The transformations are still dened on ordered triples
42. This gure has been skewed somewhat to conserve space on the page. If it were drawn in a manner parallel
with standard iiV space, a ii7 chord would be directly below the I7 chord with the same root. As is the case
throughout this study, the particular visual representation chosen for a given space does not aect the abstract
structure of the space itself.
43. Ceora is perhaps even more diatonic than this space implies, since Cmaj7 and B fmaj7 never appear in the
music, while D fmaj7 does. Thus the chord qualities strongly suggest A f major: I and IV (A f and D f ) both appear as
major sevenths, while ii and iii (B f and C) appear only as minor seventh chords (unstable ii7 chords).
68
ii7 5 V7 6 Db
7
SLIDE7
8, 16 9, 17
ii7 V7 C
15
ii
7 10, 14, 18 V7 11, 19 Bb
13
ii
7 2, 12, 20 V7 3, 21 Ab
1
4
of mod-12 (not mod-7) integers, and nothing in the iiVI progressions themselves has changed.
We used the guiding inuence of a diatonic collection in this section only to choose the key centers
we showed in particular representation of the space. This use reects the construction of jazz itself;
tunes are often globally diatonic (in a key), while locally chromatic, using iiVI progressions to
tonicize other key areas to a much greater extent than is usually seen in classical music. The reason
for this is largely practical. The head-solos-head form of most jazz means that we hear the same
progression repeated many times (Morgans recording of Ceora runs about 6 minutes, for
example), and using only pitches from the A f -major diatonic collection would quickly become
boring.
The arrangement of iiV space in Figure 2.24, combining chromatic and diatonic operations,
is mathematically complicated. As Steven Rings notes, transformation networks involving both
chromatic and diatonic operations violate Lewins formal denition of a transformation network,
since they act on dierent sets. The underlying transformation graph is not path consistent, since
the transformation slide7 TF TF is in general not equal to the transformation t6 . Put
another way, putting Gmaj7 in the top row of Figure 2.24 while leaving the transformational labels
69
unchanged does not work: obeying the t6 arrow requires Fmaj7 to occupy the row below, but
following the other path would give F smaj7. The graph is, however, realizable: it is possible, as the
gure itself attests, to ll in the nodes such that the arrows do make sense. Both Rings and Hook
have shown that transformation networks that are not path consistentlike the diatonic iiV space
developed in this sectioncan nevertheless be analytically productive.
2.3.4 Summary
By using the iiVI as the basis of the transformational spaces developed in this chapter, we can
now understand a large swath of tonal jazz harmony. Because this progression is so ubiquitous,
many jazz tunes can be well understood using only the spaces here (perhaps with some adaptations,
as suggested in this nal section). Treating chords as ordered triples of root, third, and seventh
allowed us to dene transformations in a way that is still valid when the actual form of a chord
might dier greatly from instance to instance. This is a useful abstraction, and we will continue to
use it in the next chapter, where we will also consider relationships among our sets of ordered
triples more generally.
70
Chapter 3
Thirds Spaces
The space developed in the last chapter was organized primarily in descending fths, and works
well for most tonal jazz. Motion by thirds, both major and minor, is also a common (though less
frequent) occurrence, and will be the focus of this chapter. Harmonic motion by thirds is one of
the main emphases of non-jazz transformational theorists; this chapter will allow the opportunity
to explore connections between our approach to jazz harmony and the existing neo-Riemannian
and transformational literature.
3.1.1 Formalism
The most common substitution for the dominant in jazz is undoubtedly the tritone substitution
(discussed in Section 2.2), but the minor-third substitution is also relatively common, especially in
the bebop era. By way of example, Figure 3.1 gives the changes to the opening ve bars of Tadd
Damerons Lady Bird. Normally the progression in mm. 34 would function as a iiV in the key
of E f , but in m. 5 it resolves instead to C. What might have been a iiVI in C (Dm7G7) does
not appear; the iiV is transposed up a minor third to become Fm7B f7.
The identical tendency tones shared by tritone-substituted dominants makes them relatively
easy to explain, but minor-third substitution is more dicult. Jazz harmony textbooks often do
not provide an explanation for the phenomenon: Jerry Coker, for example, simply states that the I
71
chord . . . is often preceded by IV-7 to fVII7, instead of the usual V7 chord. The Berklee Book of
Jazz Harmony places the fVII7 chord in its chapter on modal interchange (what might also be
called modal mixture), and notes that its function is ambiguous: the chord has dominant quality
but not dominant function, since it lacks the leading tone (of the following tonic). Unlike the
tritone substitution, there is no strong voice-leading rationale for the minor-third substitution; it
is simply a progression that bebop players often used, and that we as analysts must now contend
with.
There is, though, a certain similarity between the tritone and minor-third substitutions: just
as the tritone evenly divides the octave, so too the minor third evenly divides the tritone. In the
previous chapter, introducing tritone substitutes to iiV space eectively divided the space in half,
splitting the complete space into front and back sides (recall Figure 2.10). Repeating this
process again results in a space that looks something like Figure 3.2 (which we will call m3
space). The introduction of minor thirds once again changes the topology of the space. This is
somewhat easier to see by focusing again only on the centrally-located dominant seventh chords;
while the tritone version of iiV space is topologically equivalent to a Mbius strip, the
minor-third version of the space is equivalent to a torus (see Figure 3.3).
Figure 3.3 looks remarkably similar to a more familiar toroidal gure common in music
theory: the neo-Riemannian Tonnetz. Despite the surface similarity though, the two are quite
dierent. Like the Tonnetz, the dominant sevenths at the center of m3 space are arranged into axes
of perfect fths (verticals), minor thirds (horizontals), and major thirds (northwestsoutheast
diagonals, not shown in Figure 3.3). Crucially though, the vertices in the m3-torus are dominant
1. Jerry Coker, Elements of the Jazz Language for the Developing Improvisor (Miami: Belwin, 1991), 82.
2. Joe Mulholland and Tom Hojnacki, The Berklee Book of Jazz Harmony (Boston: Berklee Press, 2013), 12324.
3. That is, there is no strong voice-leading from fVII7 to I7 in the same way as there is in the tritone substitution,
7 7
fII I . We will return to the voice leading of minor third substitutions below.
4. Turning Figure 3.3 into a torus involves gluing the top and bottom edges together and the left and right edges
together; the dotted line representing perfect fths then wraps around the surface of the torus in a continuous line (as
though you had wrapped a barbers pole around a doughnut).
5. In fact, the graphs of the note-based Tonnetz and the m3-torus here are isomorphic. This fact introduces some
tantalizing possibilities, but none turn out to be terribly interesting since, as Richard Cohn has shown, the consonant
triad is unique among trichords in its capability for parsimonious voice leading; Neo-Riemannian Operations,
Parsimonious Trichords, and Their Tonnetz Representations, Journal of Music Theory 41, no. 1 (Spring 1997): 17. As
such, the m3-torusmade of (026) trichordsdoes not show common-tone relations.
72
ii7 V7 A
ii7 V7 F#
ii7 V7 Eb V7 D
ii7 V7 C V7 B
V7 Ab V7 G
ii7 V7 F V7 E
V7 Db V7 C
ii7 V7 Bb V7 A
V7 Gb
TF T5 T3 ii7 V7 E b
Figure 3.2. The complete minor-third representation of iiV space (ii7 chords omitted on rear levels
for clarity), or m3 space.
C7 Eb7 Gb7 A7
F7 Ab7 B7 D7
Bb7 Db7 E7 G7
73
ii7 V7 Eb
BD
ii7 V7 C
V7 Ab
ii7 V7 F
seventh chords (ordered triples), not individual notes; the m3-torus does not represent a
parsimonious voice-leading space. The neo-Riemannian Tonnetz can also be drawn to represent
triads instead of individual notes, but the resulting graph (Douthett and Steinbachs chicken-wire
torus) no longer resembles the m3-torus here.
The minor-third arrangement of iiV space makes it easy to dene a transformation to
represent the minor-third substitution. Because jazz musicians often refer to the minor-third
substitution as the backdoor substitution, we will call this transformation BD:
Musically, there is no compelling reason to dene BD such that the third is calculated from the
previous chords seventh and vice versa; there is no voice-leading connection between them as there
is in the TF transformation. Dening them this way, however, allows the same function to model
both the transformation from minor to dominant seventh and from dominant to major seventh.
In the space, BD is represented as a diagonal line moving frontward between two layers; see
Figure 3.4. With this denition, we can easily understand the progression in mm. 35 of Lady
TF BD
Bird: Fm7 B f7 Cmaj7.
6. Jack Douthett and Peter Steinbach, Parsimonious Graphs: A Study in Parsimony, Contextual
Transformations, and Modes of Limited Transposition, Journal of Music Theory 42, no. 2 (1998): 24648. The
chicken-wire torus is the dual graph of the more common note-based Tonnetz; both are shown in Dmitri Tymoczko,
The Generalized Tonnetz, Journal of Music Theory 56, no. 1 (2012): Figure 1.
BD
7. The denition of Smin Sdom is not given, since it is relatively rare in jazz. It models a progression like
iv7 V7 , which is much more common in classical music.
74
As noted above, the minor-third substitution is not explained in terms of its voice leading in
the jazz pedagogical literature, since it does not contain the tonic-dening 47 tritone. There is
precedent, however, for the backdoor substitution in classical music. Dmitri Tymoczko has written
about third substitution more generally, noting that it is often derived by replacing a root-position
chord with one in rst inversion (e.g., replacing IV with ii6 ). Though he does not mention minor
third substitution in his chapter on jazz, he identies the voice-leading proximity of third-related
chords as fundamental, since they can substitute for one another without much disrupting the
musics contrapuntal or harmonic fabric. The minor-third substitution in jazz, then, does have
some voice-leading rationale: the V7 and fVII7 chords are connected to each other by a minimal
amount of voice-leading work (a point to which we will return in the nal section of this chapter),
but this voice-leading proximity does not manifest in the surface motion from, for example,
B f7Cmaj7. This logic, though, would seem to suggest that a chord a minor third below the
dominant can be substituted just as well (e.g., E7Cmaj7 standing in for G7Cmaj7), but this
progression is extremely rare in jazz.
Minor-third space, as it has been developed so far, may seem like merely another arrangement of
iiV space. One might reasonably ask why it merits a section in this chapter, instead of being
merely an extension of the space like those explored in Section 2.3. In a situation parallel to that of
19th-century chromatic tonality, jazz after 1960 began to use more chromatic progressions
especially those built on thirdsin what still might be called tonal jazz. In these compositions,
there is still a prevailing sense of key, but local harmonic progressions depart from the
descending-fths norm of earlier jazz. Exploring minor-third space, and understanding why it
8. Dmitri Tymoczko, A Geometry of Music: Harmony and Counterpoint in the Extended Common Practice (New
York: Oxford University Press, 2011), 232.
9. Ibid., 283.
10. Keith Waters has called this period jazzs second practice (Chick Corea, Postbop Harmony, and Jazzs
Second Practice [paper presented at the annual meeting of the Society for Music Theory, Charlotte, NC, November
2013]). Waters, along with J. Kent Williams, has explored post-tonal jazz harmony using the Tonnetz and
hyper-hexatonic systems familiar from classical theory; see Modeling Diatonic, Acoustic, Hexatonic, and Octatonic
75
C7
F7 Bb7 C7 A7
10 9 4,8 11
Eb7 Gb7 A7 C7
5 1
Ab7 B7 D7 F7
6 2
7 3
Db 7
E
7
G 7
Bb7
Harmonies and Progressions in Two- and Three-Dimensional Pitch Spaces; or Jazz Harmony after 1960, Music
Theory Online 16, no. 3 (August 2010).
11. Isotope was rst recorded on Hendersons album Inner Urge, released in 1965. It uses a slightly dierent set
of changes for the solos than it does for the head (often referred to simply as the solo changes and head changes).
This is often the case when the head changes are complex, fast-moving, or contain unusual extensions to account for
specic melody notes. These changes are taken from The Real Book; the C chord in m. 7 is played as either a major or
a dominant seventh on the Inner Urge recording, so analyzing it as C7 here seems reasonable.
76
moving directly from IV to I in m. 7 (which would be typical for a blues), Henderson moves rst to
B f , resulting in a variant of the backdoor progression B f7C7 in mm. 67. The tonic function of
C7 is extended by moving to A7 in m. 8, a minor third away. What would be a string of T5
operations, A7D7G7C7, is disrupted slightly by the tritone substitution of A f7 for D7
(represented in the m3-torus by replacing a chord with the chord two spaces to its left or right).
Once the progression returns to the tonic C7 in m. 11, Henderson uses a complete minor-third
cycle as a turnaround, maintaining tonic function for two measures before beginning the next
chorus.
The logic of the minor-third substitution means that the B f7 chord in m. 6 is understood as
a substitute for true dominant G7. Likewise, the A7 chord in m. 8 seems to substitute for the
tonic in the same way; in an ordinary 12-bar blues, mm. 78 would both contain tonic. This
correspondence suggests an interesting possibility: the top row of harmonies in Figure 3.6 all seem
to have tonic function, while the two chords used in the bottom row both act as dominants. The
harmonies in the middle row, then, seem to serve as predominants (or subdominants), appearing
in Isotope just before the G7 and B f7 chords.
This functional analysis helps to make sense of the unusual turnaround in the last two bars of
the tune. Turnarounds are inherently prolongational structures, a way of providing harmonic
interest between a chorus-concluding tonic and the next chorus-beginning tonic. While most
turnarounds use functional harmony (a iiVI progression or some variant), Henderson uses a
non-functional minor-third cycle. Coming as it does at the end of the chorus, which makes liberal
12. We could easily dene a BDblues transformation, similar to TFblues from the last chapter; for now, we can
simply understand this variant as BD 7th.
13. The question of the meaning of function is is a dicult one. Here I mean the term as Brian Hyer does: it
is not what a chord does that matters, but what it is: a functional designation names a chords being (What Is a
Function?, in The Oxford Handbook of Neo-Riemannian Music Theories, ed. Edward Gollin and Alexander Rehding
[New York: Oxford University Press, 2011], 109, emphasis original).
14. The word prolongational in this sentence is admittedly problematic, given its Schenkerian implications. By
using it here I mean only that at some deeper level the turnaround is harmonically superuous, as it occurs after the
main tonal conclusion of the chorus (an observation conrmed by the fact that the turnaround is usually omitted in
the last head). I do not mean to imply that the turnaround in the last two bars of Isotope is harmonically
uninteresting; indeed, it is the most distinctive feature of the piece.
77
C7 Eb7 D7 G7 C7
4 j
& 4 j .. # J j b # j . b b
3
F7 Bb7 C7 E7
b b b
& # b b # n
A
5
n .
E b7 D7 G7 C7 A7 Gb7 E b7
& J b b b n n . # .
Ab
b b
9
n b b .
3
use of minor-third substitutions, we are primed to hear this cycle as a unique way of maintaining
tonic function while avoiding the use of a functional harmonic progression.
This minor-third motion is seen rst in the head changes, given in Figure 3.7. (Because the
changes are altered in the head in order to t the melody, this gure gives the melody as well.)
Most of the alterations between the head and solo changes occur in the rst four bars of the tune;
the remaining dierences are relatively insignicant. While the solo changes give only C7 in the
rst four bars, the head changes elaborate this harmony with an alteration of a IiiVI
progression: the II7 chord in m. 2 is preceded by E f7. While it would be easy to write o this
chord as an upper-neighbor harmony to the more structural II chord, doing so would minimize
the important role of the minor-third substitution in the rest of the tune. What appears at rst to
be an inconsequential embellishment (substituting E f7 for C7) gains in signicance throughout
the tune, rst becoming realized in the backdoor progression in mm. 57 and reaching its fullest
expression in the turnaround that ends the chorus.
15. The most obvious dierences in mm. 512 of the tune are the slash chords in mm. 89. The chord symbol
Em7/A indicates an E minor seventh chord played with an A in the bass; the resulting sound is an A7 chord with a
suspended fourth (D replaces Cs). The older Real Book gives the same change as A7sus4. The only other slight
alteration is the addition of the ii chord, Dm7, in m. 10.
78
Hendersons tour of the m3-torus in Isotope is interesting for a number of reasons. First, he
is able to create a sense of tonal function and harmonic progression while using almost entirely
major-minor (dominant) seventh chords. This is a feature common to many blues tunes, but is
stronger in Isotope given the pervasive use of minor-third substitutions. Second, this is a tune
that does not seem to make much sense in the descending-fths iiV space of the previous chapter.
While certainly some of the tune makes use of harmonic motion in fths, the backdoor
progression in mm. 67 and the nal turnaround would appear as seemingly random, nonsensical
jumps in iiV space.
Finally, the progression of Isotope is not one that is easily explained using neo-Riemannian
theory as it is usually applied to classical music. Constructing a Tonnetz usually relies on having
two varieties of musical objects under consideration (commonly major and minor triads, or
half-diminished and dominant seventh chords), while our m3-torus only uses major-minor
sevenths. Neo-Riemannian theories often focus on smooth voice leading; measured in these terms,
the B f7 and C7 chords of mm. 56 are quite distant from one another, despite their functional
equivalence to a VI motion. The turnaround, most unusual from a tonal perspective, is actually
quite typical of patterns usually analyzed in neo-Riemannian terms: the major-minor sevenths
found there are all minimal perturbations of a single diminished seventh chord, and each can be
connected to the next with a minimal amount of voice-leading work (two semitones moving in
opposite directions).
16. Exactly how far apart B f7 and C7 depends on how one chooses to measure voice-leading distance, and
whether we consider major-minor sevenths in the usual way, as four-note chords, or in the way we have been doing so
here, as ordered triples of root, third, and seventh. In Jack Douthetts Four-Cube Trio, for example, B f7 and C7 are
maximally far apart4 semitones; see Richard Cohn, Audacious Euphony: Chromatic Harmony and the Triads Second
Nature (New York: Oxford University Press, 2012), 15758.
17. The notes in these four dominant sevenths form an octatonic collection, and have been studied fairly
extensively in the literature. See, for example, ibid., 15258; Douthett and Steinbach, Parsimonious Graphs, 24546;
and Tymoczko, A Geometry of Music, 371.
79
3.2 Major-Third Spaces
Root motion by major third is one of the most dicult harmonic motions to explain using
traditional tonal methods; it is with this kind of music that transformational methods have proven
to be most useful. The increasing use of these progressions in nineteenth-century harmony has a
parallel in jazz, as Keith Waters has shown; in both, the harmonic vocabulary is familiar, but
harmonic progressions are often unfamiliar. Nonfunctional harmony is a dening feature of some
post-bop jazz, including much of the music of Chick Corea, Herbie Hancock, Wayne Shorter, and
others. This dissertation, though, is interested specically in tonal jazz, and we will remain
careful during this discussion to avoid straying too far aeld from this topic.
The locus classicus for root motion by major third in jazz is of course John Coltranes Giant
Steps, rst recorded in 1959 on the album of the same name. Much of the use of nonfunctional
harmony in jazz that develops after 1960 can be traced back to Giant Steps; Keith Waters has
outlined this lineage in selected compositions of Wayne Shorter, Bill Evans, and Herbie Hancock.
Given this inuence, Giant Steps will serve here as a useful foil for major-third cycles in jazz
more generally. Though we will delay a proper analysis of the tune until we have developed some
formalism, a short overview will be useful at this point.
18. The examples are too numerous to list here, but for an overview, see Matthew Bribitzer-Stull, The AfCE
Complex: The Origin and Function of Chromatic Major Third Collections in Nineteenth-Century Music, Music
Theory Spectrum 28, no. 2 (Fall 2006): 16790.
19. Waters, Chick Corea, Postbop Harmony, and Jazzs Second Practice.
20. Waterss work in particular has focused extensively on this music, though he is hardly alone. See, for example,
Patricia Julien, The Structural Function of Harmonic Relations in Wayne Shorters Early Compositions: 19591963
(PhD diss., University of Maryland, 2003); Steven Strunk, Notes on Harmony in Wayne Shorters Compositions,
196467, Journal of Music Theory 49, no. 2 (Autumn 2005): 30132; Keith Waters, Modes, Scales, Functional
Harmony, and Nonfunctional Harmony in the Compositions of Herbie Hancock, Journal of Music Theory 49, no. 2
(Fall 2005): 33357; and Waters and Williams, Jazz Harmony after 1960.
21. Many people have discussed Giant Steps in the literature; the most substantial work in this area is Matthew
Santas Nonatonic Progressions in the Music of John Coltrane, Annual Review of Jazz Studies 13 (2003): 1325. Guy
Capuzzo compares Santas analysis to one done by Pat Martino in Pat Martinos The Nature of the Guitar : An
Intersection of Jazz Theory and Neo-Riemannian Theory, Music Theory Online 12, no. 1 (February 2006). See also
David Demsey, Chromatic Third Relations in the Music of John Coltrane, Annual Review of Jazz Studies 5 (1991):
14580; and Matthew Goodheart, The Giant Steps Fragment, Perspectives of New Music 39, no. 2 (July 2001): 6395.
22. Keith Waters, Giant Steps and the ic4 Legacy, Intgral 24 (2010): 13562.
80
Bmaj7 D7 G maj7 Bb7 Ebmaj7 A-7 D7
The changes to Giant Steps are given in Figure 3.8. The major-third construction of the
tune is readily apparent: the three tonal centers are B, G, and E f , as evidenced by the major
seventh chords. These local tonics are all preceded by their dominants (mm. 13, 57) or by
complete iiV progressions (all other locations). Though the distance between the key centers is
unusual, the individual progressions are not. Giant Steps is not exactly tonal, but neither is it
really atonal. When I listen to the piece, at least, the impression is not one of nonfunctional
harmony, but rather of tonal harmony used in an unconventional way. This distinction is easiest to
understand with a counterexample: Figure 3.9 gives the changes to Wayne Shorters ballad Infant
Eyes. Here there are no iiVs, and the only VI progressions occur across formal boundaries.
Shorters use of harmony does seem nonfunctional, and gives the piece a oating quality that Giant
Steps does not have. Rather, Giant Steps is strongly forward-directed: all of the dominant
chords push toward their respective tonics, and although the global tonic may be in question, local
tonic chords are crystal clear.
23. Frank Samarotto has suggested to me in connection with an unpublished paper of his that Giant Steps is
chromatically coherent, while locally diatonic. As such, it represents an example of his hypothetical Type 4 coherence,
in which areas of diatony occur only in local isolation and in which some other (presumably post-tonal) coherence
might be in eect (Treading the Limits of Tonal Coherence: Transformation vs. Prolongation in Selected Works by
Brahms [paper presented at the annual meeting of the Society for Music Theory, Madison, WI, November 2003]).
24. These changes are from Mark Levine, The Jazz Piano Book (Petaluma, CA: Sher Music, 1989), 30. Infant
Eyes appears on Shorters album Speak No Evil (1964).
25. Logical arguments could be made for both B and E f as the prevailing key of Giant Steps. Given the
organizing inuence of the major-third cycle, I am not sure the question is so important; the tune uses tonal
progressions, but may not be in a key.
81
G-7 F-7 Ebmaj7 A7b9
E-7 A7 D maj7
Bb: VI Gb: VI D: VI
Figure 3.10. Coltranes major-third cycle as a substitution for a iiVI progression. (Adapted from
Levine, The Jazz Theory Book, 359.)
D-7 G7 C maj7
D-7 Eb7 Abmaj7 B 7 E maj7 G7 C maj7
C-7 F7 Bbmaj7
C-7 Db7 Gbmaj7 A7 D maj7 F7 Bbmaj7
Figure 3.11. The changes to Countdown (Coltrane), compared with Tune Up (Miles Davis).
82
The major-third cycle of Giant Steps is the most well-known example, but Coltrane rst
developed the progression as an elaborate substitution over a standard iiVI progression; this
particular set of substitutions is often referred to as Coltrane changes. Figure 3.10 shows how
this process works: the goal of the progression (in this case, Dmaj7) is shifted to the fourth bar;
then, major seventh chords related by major third are placed on the downbeats (B fmaj7 and
G fmaj7); nally, all of the major sevenths are preceded by their own dominants. This process can
clearly be seen in Coltranes composition Countdown, which is based on the changes to Miles
Daviss Tune Up (see Figure 3.11). Coltrane changes can be superimposed over any four-measure
iiVI progression, so they can be found not only in Coltranes own compositions, but also in his
improvisations on other tunes and his reharmonizations of standards like Body and Soul.
Because Coltrane changes can be considered a iiV variant, it is logical to include them in this
study, even though we may be slightly pushing the limits of tonal jazz. We now have a
preliminary understanding of how the substitution works, but it still remains to incorporate it into
the transformational system under development here. First, though, there is a bit of unnished
business to take care of: in Section 1.3, we touched on several transformational approaches only
briey, promising to return to them at a point when they would be more relevant. Given the
central role of harmonic motion in thirds in many neo-Riemannian theories, it seems appropriate
to fulll that promise at this point.
26. Exactly how Coltrane devised this substitution set is dicult to say: authors have at various times pointed to
classical sourcesespecially Nicolas Slonimskys Thesaurus of Scales and Melodic Patternsas well as the music of
Thelonious Monk, Dizzy Gillespie, and Tadd Dameron, among others. One source that is nearly always cited is the
tune Have You Met Miss Jones? (Richard Rodgers/Lorenz Hart), which is analyzed in Section 3.2.3 below. For a
review of these possible origins, see Demsey, Chromatic Third Relations, 14857; and Lewis Porter, John Coltrane:
His Life and Music (Ann Arbor: University of Michigan Press, 1998), 14547.
27. Nearly every discussion of Coltrane changes includes the Countdown/ Tune Up pairing. See, for example,
Demsey, Chromatic Third Relations, 15962; Mark Levine, The Jazz Theory Book (Petaluma, CA: Sher Music,
1995), 35960; or many others. Countdown was also recorded on Giant Steps, and Tune Up can be heard on
Cookin with the Miles Davis Quintet (1957).
28. Demsey provides a list of third-relations in jazz tunes in an appendix to Chromatic Third Relations, 17980.
Coltranes famous take on Body and Soul is found on the album Coltranes Sound (1960), which also features two
original tunes that make prominent use of the major-third cycle: Central Park West and Satellite.
83
While not quite neo-Riemannian, Tymoczkos general notion of third substitution (discussed
above in connection with the backdoor substitution) also includes major-third substitution, and
thus we might consider all of the local tonics in a Coltrane-changes progression as third
substitutions for the true tonic. These substitutions, according to Tymoczko, are explained by the
voice-leading proximity of third-related chordsa factor that plays a critical part in the work of
Richard Cohn. In both of Cohns models of triadic space in Audacious Euphony, hexatonic cycles
and Weitzmann regions, three M3-related major triads combine with three minor triads to form a
six-chord system. Thus, the three tonal centers of Giant Steps can be contained in a single
hexatonic cycle or a single Weitzmann region.
Understanding the rest of Giant Steps in terms of one of these systems, though, leaves much
to be desired. Cohns systems, and others like them, are fundamentally triadic, which is clearly a
problem for understanding jazz, with its saturation of seventh chords (and beyond). To account for
this incongruity, we must either adapt the music to t our analytical system, or adapt our analytical
system to t the music. The rst option is clearly a nonstarter: Figure 3.12 gives a non-example of
Giant Steps analyzed in Jack Douthetts Cube Dance. While this analysis makes the
major-third cycle of the tonic chords clear, reducing the chords to triads loses the detail of the
chord qualities (both major-seventh and dominant-seventh chords become major triads), as well as
their functional relationships. Despite the prominent emphasis on M3-cycles in these theories,
then, using these triadic systems as an analytical basis for our work here would not seem to be the
answer.
There are of course neo-Riemannian theories involving seventh chords, but these turn out to
be not so helpful for our purposes here either. Theories that include only the (0258)
tetrachordshalf-diminished and dominant seventhsare obviously not suitable for analyzing the
29. Cohn, Audacious Euphony. Hexatonic cycles are discussed primarily in Cohns chapter 2, and Weitzmann
regions in chapter 4; chapter 5 combines these models into a single system which is then used throughout the rest of
the book. Lewis Porter mentions Weitzmanns treatise on the augmented triad as a possible inuence on Coltrane
(John Coltrane: His Life and Music, 146).
30. In this gure + indicates major triads and indicates minor triads. Cube Dance appears in Douthett
and Steinbach, Parsimonious Graphs, 254, and is one of Cohns primary models of triadic space; see Audacious
Euphony, 86109 and following.
84
C E G# d s)
or
ch
(ii
a-
C+
e- E+ f- F+
c- Ab+ cs- A+
g#- Cs+
Eb G B END C# F A
START G+ f #-
B+ b- D+ d-
g- F#+ bb-
E b+
eb-
Bb+
)
ds
(V
r
ho
ch
c
or
(I
ds
D F# Bb )
Figure 3.12. Giant Steps, mm. 15, analyzed in Douthetts Cube Dance. Begin by following the solid
arrows, then continue with the dashed arrows.
Figure 3.13. Cohns Four-Cube Trio (Audacious Euphony, Fig. 7.16, 158.) The black triangles
indicate minor seventh chords, while the hollow stars indicate French sixth chords.
85
multiple seventh-chord types in jazz. The seventh-chord analogy of Cube Dance is what Cohn
calls the Four-Cube Trio, which is shown in Figure 3.13. As is readily apparent, Four-Cube Trio
does not contain any major seventh chords, so it would also create problems if pressed into use for
analyzing Giant Steps.
There is, though, a more fundamental problem with these neo-Riemannian theories of
seventh chords, at least when approaching M3-cycles in jazz. As we have mentioned, many
neo-Riemannian theories focus on ecient voice-leading, and parsimonious relationships among
seventh chords can be understood as minimal perturbations of fully-diminished seventh chords.
This does not generate major-third cycles (as in the triadic case), but rather partitions the octave
into minor thirds. The three dominant sevenths appearing in Giant Steps, for example, are in
three dierent towers in the Four-Cube Trio, which does not reect the organizing inuence of
major thirds in the way that the triadic Cube Dance does.
Some theorists have turned to neo-Riemannian theory to explain jazz progressions, though;
closest to our intent here is Matthew Santas nonatonic system for analyzing Coltrane. In a
parallel with Cohns hexatonic systems, Santa draws from the nonatonic (or enneatonic) collection
in order to explain Giant Steps in terms of parsimonious voice leading. Figure 3.14 shows one of
Santas cycles, along with three-voice parsimonious realization. (In this gure, note that the
triangle indicates a major triad, not a major seventh chord.) This nonatonic system seems to be a
convincing analysis of the opening of Giant Steps, but it comes up a bit short as a general
theoretical system. First, Santa considers only major triads and incomplete dominant seventh
chords; all major seventh chords are reduced to triads, and minor seventhslike the ii7 chords of
Giant Stepsare simply ignored. The cycle in Figure 3.14 is generated by the collection {D,
31. This gure is taken from Cohns book, and has several errors, the most important of which is that the C at
the 10 oclock position should be a Cs, forming a fully-diminished seventh chord. 4-Cube Trio was originally devised
by Jack Douthett, and is very similar to the Power Towers graphic in Douthett and Steinbach, Parsimonious
Graphs, 256 (which omits the French sixth chords). For more on its history, see Cohn, Audacious Euphony, 157n15.
32. This is a central thesis of Cohns chapter 7 (see especially Audacious Euphony, 14858), and gures prominently
in Tymoczkos geometric theory (A Geometry of Music, 97112).
33. Santa, Nonatonic Progressions in Coltrane.
34. Santa omits the fth of dominant seventh chords, as we have been doing here. The reasons, though, are
dierent: the stated reason is to keep the cardinalities of the chords the same, but he does not mention why he chooses
not to use major seventh chords (rather than triads) and complete dominant sevenths, for example. Santa notes later that
86
Eb
& b b # n # # n b b
n b
G B
D7
Figure 3.14. Matthew Santas nonatonic cycles: the Western nonatonic cycle (left), and a
three-voice parsimonious realization (right). (Adapted from Nonatonic Progressions in Coltrane,
14.)
E f , E, Fs, G, A f , B f , B, C}, but it does not contain all of the triads or incomplete dominant
sevenths in that collection (see Table 3.1). Santas 4-cycle system, then, is somewhat misleading,
since any triad or dominant seventh can be located in two dierent nonatonic collections.
Having brought up all of these approaches only to show their shortcomings, though, the
question remains: what should a transformational system that includes major-third relations look
like? Although the relationship of M3-cycles and smooth voice leading is valuable, so far in this
study we have focused primarily on functional relationships, and it would seem foolish to abandon
that approach here. As we noted above, Giant Steps does contain a major-third cycle, but within
that cycle the progressions are functional: it is locally diatonic, but globally chromatic.
including the fth involves one of the three missing notes from the nonatonic collection, which is acceptable because
the fourth voice is not essential to the voice leading of the cycle (Santa, Nonatonic Progressions in Coltrane, 15).
87
ii7 V7 E
ii7 V7 Ab
ii7 V7 C V7 A
V7 Db
ii7 V7 F V7 D
V7 Gb
ii7 V7 Bb V7 G
V7 B
ii7 V7 E b V7 C
V7 E
TF T5 T8
ii7 V7 A b
Figure 3.15. A major-third organization of iiV space (ii7 chords omitted on rear levels for clarity).
As it turns out, we can once again adapt iiV space in order to show organization by major
third, as shown in Figure 3.15. This gure looks very similar to the minor-third organization in
Figure 3.2, but the relationships between layers have changed. Here, the layers of the space are
arranged in descending major thirds (T8 ), while the descending fths arrangement is otherwise
unchanged. This arrangement means that all of Giant Steps happens in a single horizontal slice
of the space. This organization of iiV space reects our intuitions about the organization of this
tune and others like it: by maintaining the integrity of the iiVI progressions and instead altering
the relationships between them, we can keep both the local functional progressions important to
improvising musicians and reect the unusual chromatic organization of the tune itself.
35. The major-third gure represents a kind of cross-section of the minor-third torus: to see this clearly, locate
the key areas C, A f , and E on both Figure 3.2 and Figure 3.15.
36. In fact, the rest of the gure is unnecessary for Giant Steps; the piece is easier to understand using a
subgraph of the complete M3-space that contains only the iiVI progressions in B, G, and E f .
88
ii7 V7 E
ii7 V7 Ab
ii7 V7 C
The arrangement into T8 -related layers also helps to clarify the function of Coltrane changes
as a substitution for a iiVI progression (see Figure 3.16). As usual, we could dene a
transformation to help explain this substitution. Though the T8 between major seventh chords is
certainly important, the most unusual surface feature in the substitution is the jump from a
major-seventh chord to the dominant seventh whose root is a minor third higher; it is this
harmonic move that gives the progression its forward momentum. We might call this
transformation CS (for Coltrane Substitution):
With this transformation, we can understand the entire iiVI substitution as follows:
TF TF
Dm7 G7 Cmaj7
becomes
TF T TF CS TF CS TF
Dm7
8
E f7 A fmaj7 B7 Emaj7 G7 Cmaj7
This M3-space is well equipped to show the logic of major-third cycles, though other kinds of
tonal relationships are more dicult to see: both tritone substitutes and minor-third substitutions
are maximally far away in M3-space, for example. While this is indeed a limitation, the fact that
iiV space and its variants share the same essential features means that each can be substituted for
another as needed for a given analytical situation. The spaces developed so fariiV space, its
tritone-substituted variant, minor-third space, and now major-third spaceall reect the basic
89
descending-fths orientation of tonal jazz, but each prioritizes a particular secondary relationship.
Maintaining the same basic structure in our analytical apparatus allows us to understand a wide
range of music as variations on a basic, functionally harmonic theme. There is no need for a great
switching of context from the logic of tritone substitutions to the logic of Coltrane changes, nor is
there a need to invoke set classes of dierent cardinalities to justify dividing the octave into three or
four equal parts. This presentation is developed in a rough parallel with the music itself: jazz
musicians did not (and do not) discard everything they learned from bebop when approaching a
progression like that of Giant Steps; rather, each is part of a single, coherent through line of
tonal jazz.
3.2.3 Analytical Interlude: Richard Rodgers/Lorenz Hart, Have You Met Miss
Jones?
37. These spaces represent all but one of the equal partitions of the octave: while we could easily construct a
whole-tone space, it would not have very many applications to tonal jazz. In cases where whole-tone relationships
seem important, it is usually not problematic to consider a whole tone as a combination of two perfect fths, which are
readily shown in all of the other spaces.
38. These changes are again taken from The Real Book, and are the standard changes for the tune; nearly all
recordings agree with this set of changes.
90
A Fmaj7 Fs 7 G-7 C7
A Fmaj7 Fs 7 G-7 C7
Figure 3.17. Changes to Have You Met Miss Jones? (Richard Rodgers/Lorenz Hart).
Gm7. This Gm7 initiates a home-key iiV in mm. 34 that resolves deceptively to Am7 in m. 5,
at which point the piece begins a iiiviiiV turnaround to return to Fmaj7 for the repeat of the A
section.
The second A section begins like the rst, but the end is altered so that the bridge can begin
on the subdominant, B f (Figure 3.18b). The bridge of this tune is its most well-known aspect, and
contains the major-third cycle. This organization by major third is readily apparent in the space, as
the music seems to break free of the descending fths to elaborate the subdominant with a
sequence that moves into the rear layers of the space and then returns. After arriving on B f in the
rst bar of the bridge, the tune moves to a iiVI in G f (a major third lower), followed by a iiVI
in D (yet another major third lower). After making its way to the rear of the space, it begins to
work its way back up the chain of thirds, nishing the bridge with a return to a iiVI in G f . This
G fmaj7 chord moves via a slide7 transformation to a home-key iiVI, which returns to Fmaj7 to
begin the nal A section. This nal section (Figure 3.18c) is nearly the same as the rst, but altered
slightly in the last four bars to arrive more strongly on tonic in the penultimate bar of the form.
39. Levine, The Jazz Theory Book, 85.
91
ii7 V7 G
5 2
ii7 V7 C
6
1
ii 3,7
7
V 8 7
F ii7 V7 D
4
ii7 V7 Gb
ii7 V7 Bb
ii7 V7 Eb
ii7 V7 G
ii7 V7 C
(15 as
in A) 6
15
ii
7
V 7
F ii7 13 V7 14 D
12
ii7 10,16 V7 Gb
11,17
9
ii
7
V 7
Bb
7 8
ii7 V7 Eb
Figure 3.18b. Miss Jones, second A section and bridge (mm. 923).
92
ii7 V7 G
8 5,9
ii7 V7 C
4
ii 2,6,10
7
V 3,11
7
F ii7 V7 D
7
ii7 V7 Gb
1
ii7 V7 Bb
ii7 V7 Eb
Figure 3.18c. Miss Jones, last two bars of bridge and final A section (mm. 2332).
Bridge Final A
ii7 V7 Bb T7
T8
T8 ii7 V7 Gb SLIDE7 ii7 V7 F
T8
T4
T11
ii7 V7 D
93
A more detailed transformation network for the bridge of Miss Jones is given in Figure
3.19. In this network, transformations actualized in the music are shown as solid arrows, while
others are shown with dotted arrows. (As usual, the unlabeled arrows are TF transformations.) A
complete T8 -cycle is thwarted when the Dmaj7 chord moves instead back to G f , but the dotted
arrow shows that another T8 move would have completed the cycle. It also claries the return to F
major in the last A section: a larger-scale T11 from G f to F is accomplished via the slide7
transformation from G fmaj7 to Gm7.
The analysis of Have You Met Miss Jones in M3-space demonstrates how the logic of
major-third cycles in jazz is not independent from that of standard fths-based harmony, but
instead an extension of it. Any of our spaces would illustrate the A sections of Miss Jones equally
well, but the construction of M3-space allows us to better understand the bridge. In ordinary iiV
space (Figure 2.10 on p. 50), the iiVI progressions related by major third are maximally far apart;
analyzing the bridge in that space would make the T8 -related iiVIs seem like nonsensical
harmonic motions. Rearranging the basic space as we have done in this section shows that the
M3-cycle participates in a coherent way within the logic of the otherwise typical harmony of the
tune.
40. The bubble notation used in Figure 3.19 is rst used in GMIT, 2056. It is, as Lewin describes it, a
network-of-networks: each bubble here represents a single iiVI network (the unlabeled arrows are again TF
transformations), and these networks are connected by larger-scale transpositions. In this gure, the slide7
transformation breaks through the bubble itself, and describes a transformation directly from G fmaj7 to Gm7.
94
Bb7b5/E7b5 Cs7b5/G7b5
cs-
c# E7
bb bb- C#7
g Bb7
g-
e G7
C E
A e-
(m7b5) (m7) (dom. 7) C#
Fs Bb
Cs E G Bb Eb D F Gs B
G
(M7) (M7)
(dom. 7) (m7b5)
E b7 ab
Fs7 b
A7 d
C7 f
Figure 3.20. A portion of Four-Cube Trio, redrawn to include major seventh chords. All lines,
regardless of style, represent a single voice moving by half-step.
our transformations are dened on ordered triples of the form (root, third, seventh), not on triads
or seventh chords. As such, we will need to take a few steps in order to connect our work here with
the literature that deals with these basic chord types directly.
For the moment, let us set aside the ordered-triple representation we have been using here
and return to four-note seventh chords proper. Parsimonious relationships among seventh chords
are shown in Douthetts Four-Cube Trio (recall Figure 3.13); Figure 3.20 redraws a portion this
gure to include major seventh chords. Because we are interested in this gures application to
jazz, the minor seventh chords have been labeled with root names, and the French sixth chords are
labeled as dominant seventh chords with atted fths (a favorite chord of Thelonious Monk).
41. In addition to including the major seventh chords, this gure corrects some errors in Cohns version
(Audacious Euphony, Fig. 7.16, 158): in his version, the voice leading to and from the French sixth chords and minor
seventh chords is incorrect. My thanks to Thomas Cooke-Dickens for helping to nd many of these errors.
42. The French sixth chord, set class (0268), is of course invariant when transposed by tritone, so both roots are
given in this gure.
95
Bb7b5/E7b5 Cs7b5/G7b5
cs-
c# E7
3RD
TF bb bb- C#7
TFT
g Bb7
g-
e G7
7TH
C SLIDE7 E
A e-
(m7b5) (m7) (dom. 7) TF C#
Fs TFT Bb
Cs E G Bb D F Gs B
Eb G
(M7) (M7)
(dom. 7) (m7b5)
E b7 ab
Fs7 b
A7 d
C7 f
Figure 3.21. The transformations 3RD, 7TH, SLIDE7 , TF, and TFT in the Four-Cube Trio.
By way of illustration, Figure 3.21 shows some of the transformations we have dened in the
last two chapters in the Four-Cube Trio. The most important thing to note about this gure is
that it is not a voice-leading graph of the transformations as they have been dened. Because all of
our transformations are dened on ordered triples (ignoring the fth of a seventh chord), mapping
them into four-note space can be misleading. In our ordered triple representations, for example,
there is no distinction between half-diminished and minor seventh chords, nor between the French
sixths and the dominant sevenths (i.e., B f7 is indistinguishable from B f7 f5). This gure is
included only to demonstrate that some of the transformations do represent single voice-leadings
(the 3rd, 7th, and slide7 transformations), while others do not. It is also worth noting that the
slide7 transformation is the only one we have dened in which the voice-leading ascends
(indicated by a clockwise motion in the gure).
96
G-7 G7 Bb7 G-7 G7
&
b n b b &
b n
Figure 3.22. Parsimonious voice leading of the minor seventh chord, with its fifth (left) and without
(right). Common tones are shown with hollow noteheads.
Table 3.2. Parsimonious voice-leading among members of Smin , Sdom , and Smaj . In this table, r, t, and s
indicate the root, third, and seventh of a chord, respectively.
The parsimonious picture that emerges after we collapse the chords which have the same
ordered-triple representation is much less interesting. Most damaging to the structure of the
Four-Cube Trio is that the minor seventh chords become more discriminating: considered as an
ordered triple, a minor seventh chord is only connected to one dominant seventh chord, not two
(see Figure 3.22). This, plus the collapse of the French sixths into dominant sevenths, means that
every chord is connected to exactly one other by single-half-step voice-leading. (For reference, a
complete voice-leading roster for the ordered-triple representation is given in Table 3.2.)
43. This is not terribly surprising; our ordered triples are trichords of set classes (015), (016), and (026). Richard
Cohn has shown (Neo-Riemannian Operations, Parsimonious Trichords, and Their Tonnetz Representations) that
the consonant triad, (037), is unique among trichords in its ability to form parsimonious relationships. Although his
work there does not examine parsimonious relationships among members of dierent set classes, the fact that our
three types are not nearly even means that we should not expect to nd very many of these relationships.
97
We could generate the entire set of 36 ordered triples using the single-voice half-step
1
transformations 7th, 3rd, and slide
7 :
The resulting graph, however, is simply a circle, and does not mirror the rich voice-leading
network of the Four-Cube Trio. We could redene all of the other transformations in terms of
6
TF slide 1 (7th 3rd slide 1 ) 7th
these single voice-leadings (Dm7 G7 becomes Dm7
7
7 G7), but
these decompositions would not seem to give much insight into tonal jazzwhich is, after all, the
aim of this study.
Throughout the course of the last two chapters, we have developed a fairly complete
transformational system for jazz harmony. Taking iiV space as our starting point, we have seen
how it can be altered in various ways to show dierent aspects of standard tonal jazz harmony. To
this point, though, we have focused almost exclusively on chord symbols (via their abstraction into
ordered triples). Of course, jazz harmony involves quite a bit more than relationships among
three-note chords: these basic structures are altered in various ways by rhythm section members,
and we have yet to say anything at all about the role harmony plays for an improvising musician.
To do so, we will need to expand our harmonic universe somewhat; the next chapter begins to take
steps in that direction.
44. The inverse is needed for slide7 because it was dened as a transformation from a major seventh chord to the
minor seventh whose root is a half-step higher. A cycle that uses the non-inverse slide7 transformation would require
inverses on both the 3rd and 7th transformations.
98
Chapter 4
Chord-Scale Transformations
For jazz musicians, a chord symbol implies more than just the root, third, and seventh; its fth,
ninth, and potentially other chord members are implied as well. Instead of considering the
extended harmonies common in jazz as stacks of thirds, musicians often describe harmony in a
more linear fashion, using a scale to stand in for a chord symbol. What is often called chord-scale
theory is a major part of jazz pedagogy, and cannot be ignored as we try to approach a general
theory of jazz harmony. This chapter begins with an introduction to chord-scale theory, both in its
original form and its later pedagogical adaptations, and then continues to incorporate it into the
transformational system developed thus far; nally, we will see how the theory allows us to make
analytical insights that can go beyond the chord-symbol-based analysis of previous chapters.
The ultimate origin of chord-scale theory is George Russells Lydian Chromatic Concept of Tonal
Organization, rst published in 1953 but revised several times throughout Russells life. Russell
was a jazz pianist, drummer, and a well-known composer and arranger; he devised the majority of
the Lydian Chromatic Concept while hospitalized for tuberculosis in 194546. The inuence of
the Concept (as it is often called) is dicult to overstate. Joachim-Ernst Berendt and Gnther
Heusmann describe it as the rst work deriving a theory of jazz harmony from the immanent laws
1. George Russell, The Lydian Chromatic Concept of Tonal Organization, 4th ed., vol. 1, The Art and Science of
Tonal Gravity (Brookline, MA: Concept, 2001); hereafter, simply LCC . References to the book in this dissertation
will be to this nal edition unless otherwise noted. Though a full reception history of the Lydian Chromatic Concept is
beyond the scope of this project, it is worth noting that the later editions focus more heavily on the theory than the
earlier editions, which were more practical in nature. In the original edition, the section on theoretical foundations
appears after eight initial lessons; in the 2001 edition, this material has been moved front and center to Chapter 1.
Furthermore, the last edition was reframed as the rst volume in what was to be a multi-volume set; at the time of
Russells death in 2009, only the rst volume had been published.
2. Encyclopedia of Popular Music, s.v. George Russell, last modied July 4, 2006,
http://www.oxfordmusiconline.com/subscriber/article/epm/48476.
99
of jazz, not from the laws of European music, and the blurbs on the back cover contain praises
from musicians including Gil Evans, Ornette Coleman, Eric Dolphy, and Toru Takemitsu.
Despite its importance to jazz theory, though, Russells work has not received much attention
in music-theoretical scholarship on jazz. Dmitri Tymoczko, for example, does not mention Russell
at all in his survey on the pedagogical use of chord-scales in jazz, and the only mention of Russell in
his book is in a footnote unrelated to Russells contributions to chord-scale theory. There may be
many reasons for thisRussells serpentine and hard-to-follow prose are probably not least among
thembut regardless, an introduction to Russells theories as he conceived them will be in order
here. The Lydian Chromatic Concept can be divided into two main components, which we will
treat separately in the following sections: Lydian tonal organization and chord/scale equivalence.
Russells central insightindeed, the Concept itselfis that the Lydian scale, rather than the
major scale, serves a fundamental role in equal-tempered music. He oers many explanations, but
this central idea is easiest to demonstrate, as he does, with an example. Figure 4.1a reproduces
Russells rst example; he provides the following instructions and explanation:
Sound both of the following chords separately. Try to detect the one which sounds a
greater degree of unity and nality with its tonical [sic] C major triad. . . . In tests
performed over the years in various parts of the world, the majority of people have
repeatedly chosen the second chordthe C Lydian Scale in its tertian order. (LCC 1)
The lowest note of a stack of six perfect fths is what Russell calls the Lydian tonic; the BF
tritone in the C major scale disrupts the perfect symmetry of the fths (LCC 4; see Figure 4.1b).
3. Joachim-Ernst Berendt and Gnther Huesmann, The Jazz Book: From Ragtime to the 21st Century, 7th ed.,
trans. H. and B. Bredigkeit et al. (Chicago: Lawrence Hill, 2009), 602. Takemitsus fascination with the Concept is
discussed at length in Peter Burt, Takemitsu and the Lydian Chromatic Concept of George Russell, Contemporary
Music Review 21, no. 4 (2002): 79109.
4. Dmitri Tymoczko, The Consecutive-Semitone Constraint on Scalar Structure: A Link Between Impressionism
and Jazz, Intgral 11 (1997): 13579; and A Geometry of Music: Harmony and Counterpoint in the Extended Common
Practice (New York: Oxford University Press, 2011), 366n13. My intent here is not to single out Tymoczko, but only to
note that modern theorists who engage directly with Russells ideas do not always mention his work. The most complete
treatments of the Concept in modern scholarship are Burts previously-cited article on Takemitsu and Brett Clements
work on Frank Zappa (who, while inuenced by jazz, is not part of the jazz mainstream that is the focus of this study);
see A New Lydian Theory for Frank Zappas Modal Music, Music Theory Spectrum 36, no. 1 (Spring 2014).
100
www # wwww
a) b)
www ww # ww
& w
ww
w w w
? ww ww
w w ww ww
w w
Figure 4.1. The major and Lydian scales, as stacks of thirds and fifths (Russells Examples I:1, I:2a,
and I:5a).
He does concede that C is also understood as tonic in the major scale arrangement, but that it does
not sound resolved, since the presence of the Lydian do on the major scales fourth degree
permanently denies [it] that possibility (LCC 4). For Russell, the major scale is always in a state
of tension, wanting to resolve to the Lydian.
Closely allied with the Lydian tonic is the concept of tonal gravity, which Russell describes as
the fundamental principle of the Concept. In a stack of fths, tonal gravity ows downward: the
tone Fs yields to B as its tonicFs and B surrender tonical authority to E, and so on down the
ladder of fthsthe entire stack conferring ultimate tonical authority on its lowermost tone, C
(LCC 3). The concept of tonal gravity provides the justication for the primacy of the Lydian
scale, since the major scale cannot be constructed by generating perfect fths from its tonic.
Its theoretical justications aside, the Lydian scale has an almost mystical quality to Russell,
which can sometimes be o-putting. A somewhat longer passage from the Concept will help to
illustrate Russells fascination with the scale, as well as his usual circuitous mode of presentation
(the emphasis and non-bracketed ellipses are original):
The Lydian Tonic, as the musical Star-Sun, is the seminal source of tonal gravity
and organization of a Lydian Chromatic scale. [. . .] Unity is the state in which the
Lydian Scale exists in relation to its I major and VI minor tonic station chords, as well as
5. I am using the term scale here as Russell does: he always refers to the Lydian as a scale rather than a mode.
Dmitri Tymoczko has argued against this usage, saying that there is a widespread tendency to elide the dierence
between scale and mode (A Geometry of Music, 366n14). To call the Lydian a mode would imply that it is simply a
reordering of a major scale, though, and for Russell the two are fundamentally dierent objects.
101
those on other scale degrees. Unity is . . . instantaneous completeness and oneness in the
Absolute Here and Now . . . above linear time.
The Lydian Scale is the musical passive force. Its unied tonal gravity eld,
ordained by the ladder of fths, serves as a theoretical basis for tonal organization within
the Lydian Chromatic Scale and, ultimately, for the entire Lydian Chromatic Concept.
There is no goal pressure within the tonal gravity eld of a Lydian Scale. The Lydian
Scale exists as a self-organized Unity in relations to its tonic tone and tonic major chord.
The Lydian Scale implies an evolution to higher levels of tonal organization. The Lydian
Scale is the true scale of tonal unity and the scale which clearly represents the
phenomenon of tonal gravity itself. (LCC 89)
Russells logic is, of course, circular: the Lydian tonic is by denition the note that is the
bottom of a stack of six perfect fths, and the principle of tonal gravity confers a special status on
the bottom of a stack of six fths (conveniently, the Lydian tonic). Partly for this reason, this part
of Russells theory has not really been taken seriously by modern scholars. He never gives a reason,
for example, that the stack should not be extended further: would the lowest note of a stack of
seven fths not be imbued with even more tonal gravity? This complication reappears when Russell
later presents the complete Lydian Chromatic Order of Tonal Gravity, given here starting on F
(LCC 12):
F C G D A E B Cs Af Ef Bf Gf
What should be a perfect fth from B to Fs is replaced by a whole step (BCs), so that the pitch
forming a minor ninth with the Lydian tonic (G f ) does not appear until the last note. This sleight
of hand also prevents there from being more than one succession of six perfect fths in the series.
If it had continued in perfect fths, there would be by denition seven Lydian tonics!
Considering these inconsistencies, we might ask if there is anything worth saving in Russells
ideas. He is probably right that most people prefer the sound of the stack of thirds on the right of
Figure 4.1a, with the Fs. And the Lydian scale does have some practical advantages over the major
scale. As Russell points out, s4 appears before n4 in the harmonic series (he does not mention that
f7 appears before n7 ). The Lydian scale is also unique in that it is possible to form all twelve
6. Though Russell attributes this preference to the Lydian scale, we could probably point to another reason: the
second chord does not contain the dissonant minor ninth between Fn and E, the third of the chord.
7. Russell does recognize that s4 in the harmonic series is not pure. Describing this potential problem, he notes
that in relation to a C fundamental, the eleventh overtone, represented as Fs has a frequency of 551 cents: 1/100th of a
102
C Lydian Scale
&
#
Octave Maj 7th Min 7th Maj 6th Aug 5th Fifth Aug 4th Fourth Maj 3rd Min 3rd Maj 2nd Min 2nd
&
()
C major scale
Figure 4.2. Russells Interval Tonic Justification for the Lydian scale (his Example I:9).
interval types with the tonic; Russells explanation of this fact is reproduced in Figure 4.2 (the F in
the major scale is shown with a hollow note because it is the true Lydian tonic for the scale).
Though we may not share Russells preoccupation with the Lydian scale, we should not let his
idiosyncratic views distract us from the more important points of the theory. Theorists are, of
course, used to adopting worthwhile theoretical ideas from authors without assuming their entire
worldview. We regularly practice Schenkerian analysis without adopting Schenkers views on the
superiority of German music, and even 18th-century authors like Johann Phillip Kirnberger
accepted Jean-Phillipe Rameaus fundamental bass without necessarily espousing his more
contentious thoughts on harmonic generation or subposition. And yet, with the Concept, this does
not seem to have taken place. If we grant Russell these inconsistencies, though, his ideas prove to
be remarkably useful, as we shall see.
semitone closer to Fs than Fn. In fact, the overtone series fails to list the Fn tone even when carried out to twenty
harmonics (LCC 3).
8. Joel Lester, Compositional Theory in the Eighteenth Century (Cambridge, MA: Harvard University Press, 1992),
24041.
9. My own suspicion is that Russells work is too new to be considered historically, but not new enough to be
taken seriously as modern scholarship. I discuss this idea at length in Reconceptualizing the Lydian Chromatic
Concept : George Russell as Historical Theorist (paper presented at the annual meeting of the Society for Music
Theory, St. Louis, MO, October 2015); see also Kyle Adams, When Does the Present Become the Past? A
Re-examination of Presentism and Historicism (paper presented at the annual meeting of the Society for Music
Theory, Charlotte, NC, November 2013).
103
Before moving on to chord-scales proper, we should rst examine the scales themselves, as
their initial presentation in the Concept is entangled with the discussion of the nature of the
Lydian scale. The scales are generated (more or less) from the chromatic order of tonal gravity,
which is given again here in its generic form:
I V II VI III VII +IV +V fIII fVII IV fII
When taken together, the entire series represents the Lydian Chromatic Scale, the foundation of
the titular Concept.
The Lydian Chromatic (or LC) scale contains eleven member scales, each of which is
chosen, Russell says, for three reasons:
10. Russell denes ingoing only in passing, saying that all music conceived within the equal tempered system
maintains a closer (more ingoing) relationship to one tone than to all others, regardless of the musics style or genre
(LCC 9).
11. Russells idiosyncratic names have mostly fallen out of use, since they are long and dicult to remember. It is
easier for an improvising musician, for example, to recall the halfwhole diminished scale than it is to remember the
auxiliary diminished blues scale (named, incidentally, for the fact that it shares f3, n3, and f7 with the blues scale).
The Lydian dominant scale is an exception: though Russell does not use the term himself, it is clearly derived from
his work.
12. The remainder of this paragraph is a gloss on the material in LCC 1217.
104
Lydian &
#
&
Lydian
augmented
b
&
Lydian
diminished
b
Lydian
at seventh &
# b
&
Auxiliary
augmented
b b n #
&
Auxiliary
diminished
b
Auxiliary
diminished blues & b b n
Figure 4.3. The seven principal scales of the F Lydian Chromatic scale.
105
12 TONE ORDER
OUTGOING TONAL GRAVITY LEVEL Aux.
Dim.
11 TONE ORDER Blues
SEMI-OUTGOING TONAL GRAVITY LEVEL Aux.
Dim.
10 TONE ORDER
SEMI-OUTGOING TONAL GRAVITY LEVEL Lydian
b7
9 TONE ORDER CONSONANT NUCLEUS
SEMI-INGOING TONAL GRAVITY LEVEL Lydian
Dim.
7 TONE ORDER Lydian
INGOING TONAL GRAVITY LEVEL Aug.
Lydian Aux.
Aug.
LYDIAN
TONIC
I V II VI III VII +IV +V bIII bVII IV bII
F C G D A E B Cs Gs Ds As Ess
1 2 3 4 5 6 7 8 9 10 11 12
(Af) (Ef) (Bf) (Gf)
Figure 4.4. Russells Lydian Chromatic Order of Tonal Gravity (his example II:3).
it. The consonant nucleus also provides a (tautological) explanation for the missing fth in the
order of tonal gravity: the skipping of the interval of a fth between the seventh and eighth tones
of the Lydian Chromatic Scale allows the ve basic chord categories of Western Harmony to be
assimilated by its Nine-Tone Order, Semi-Ingoing Level, in the logical order of their development
in Western Harmony and the Lydian Chromatic Scale (LCC 16).
The other four of the eleven member scales are known as the horizontal scales, and are
shown in Figure 4.5. For Russell, horizontal is used in opposition to the vertical generation of
the Lydian scale. Because the major scale is not a stack of perfect fths, he considers it to be
generated in a dierent direction. All of the horizontal scales have n4; Russell only includes them
because of their historical and/or sociological signicance. The horizontal scales do not, as we
shall see, generate chords in the same way as the vertical scales, and for Russell they exist in a
constant state of tension between the false tonic and the true Lydian tonic.
13. Exactly what Russell means by seventh is unclear; he likely means the dominant seventh, though all four
standard seventh chord types (major, minor, diminished, and half-diminished) appear in the consonant nucleus.
106
b
Major &
b b
Major
at seventh &
b #
Major
augmented fth &
b ( n )
African-American
blues & ( ) b n b n
Figure 4.5. The four horizontal scales of the F Lydian Chromatic scale.
Given that most of Russells ideas on Lydian tonal organization have disappeared from modern
chord-scale theory, it is reasonable to ask why so much space has been devoted to them here.
There are two reasons. First, much modern scholarship does not seriously engage with Russells
ideas, and as a result most theorists are not familiar with its rst incarnation. Because the original
presentation of chord-scale theory is tied up with that of Lydian tonal organization, understanding
the former is important in order to make sense of the latter. Second, and more importantly, one of
the goals of this dissertation is to take jazz musicians conceptions of harmony seriously:
chord-scale theory is an integral part of the way jazz is taught, and therefore many practicing
musicians understand harmony in terms of this theory. Later in this chapter, we will seek to revive
some of Russells initial formulation of the Concept as we develop a transformational system of
chord-scales.
14. The counterargument is perhaps obvious: if the ideas about Lydian tonal organization have fallen by the
wayside, why bother trying to resuscitate them? As I hope to show in the following sections, Russells systematic
approach is useful as a means of formalizing what has since become implicit knowledge.
107
4.1.2 Chord/Scale Equivalence
While Russell may have understood Lydian tonal organization to be the most important part of his
new theory, the part that has survivedourished, evenis his novel conception of chord/scale
equivalence. Russells rst mention of the concept explains its inception:
In a conversation I had with Miles Davis in 1945, I asked, Miles, whats your musical
aim? His answer, to learn all the changes (chords), was somewhat puzzling to me
since I feltand I was hardly alone in the feelingthat Miles played like he already
knew all the chords. After dwelling on his statement for some months, I became
mindful that Miless answer may have implied the need to relate to chords in a new way.
This motivated my quest to expand the tonal environment of the chord beyond the
immediate tones of its basic structure, leading to the irrevocable conclusion that every
traditionally denable chord of Western music theory has its origin in a parent scale.
In this vertical sense, the term refers to that scale which is ordainedby the nature of
tonal gravityto be a chords source of arising, and ultimate vertical completeness; the
chord and its parent scale existing in a state of complete and indestructible chord/scale
unitya chordmode. (LCC 10)
What Davis was looking for is essentially a way of determining what notes he could play over a
given chord. Simply knowing the chord tones no longer seemed to be enough, since various
extensions and alterations can change the sound of the chord. Chord-scale theory is ultimately,
then, an improvisational expedient: a single scale stands in for a chord symbol. Chord symbols
with alterations are represented by dierent scales, and thus musicians do not necessarily have to
keep track of all of the individual chord tones.
Russells later explanation of the concept is uncharacteristically clear:
The chord and its parent scale are an inseparable entitythe reciprocal sound of one
another. . . . In other words, the complete sound of a chord is its corresponding mode
within its parent scale. Therefore, the broader term chordmode is substituted for what
is generally referred to as the chord. (LCC 2021)
15. Dmitri Tymoczko makes the argument that something like chord-scale theory existed in the music of the
Impressionists, and suggests that jazz musicians may have discovered it by listening to Debussy and Ravel (The
Consective-Semitone Constraint, 152 and 173). Given the initial reception of the Concept as the rst real theory of
jazz, I am skeptical of this claim, and treat Russells ideas as their rst appearance. Certainly, though, the tenets of
chord-scale theory can be applied to other musics: Russell himself examines passages of Bach, Beethoven, and Ravel,
among others.
108
It is important to understand that for Russell, the two termschord and scaleare truly
equivalent: one does not substitute for the other, rather one is the other. Here, we arrive at the
reason for the inclusion of this material in a dissertation about jazz harmony. If we take Russell
seriously (and I am arguing that we should), a harmony is a scale, and vice versa. The two ideas are
inseparable, and a study of harmony in jazz would be incomplete without a commensurate
discussion about scales.
At this point a brief overview of Russells brand of chord-scale theory is appropriate. This
material accounts for the majority of the length of the Concept, so we will be careful here to avoid
going into all of its painstaking detail. Determining the scale that belongs with a particular chord
is a multi-step process: rst, identify the parent Lydian scale; then, determine the harmonic genre
based on the characteristic modes of the Lydian scale.
Russell goes through all seven modes of the Lydian scale, identifying the principal chords of
each mode. These chords represent the purest form of the mode, and the basis for the chord/scale
matching process. An overview of these is given in Table 4.2, which merits a bit of commentary.
First, the order of modes in the table follows Russells order of presentation, which (though he does
not explain it) roughly coincides with the frequency of each principal chordmode in jazz practice.
Second, the sub-principal chords are those which are also representative of a given mode; they
do not contain all the tones of [the] relative Principal Chordmode, but they still exist in a state
of unity with [the] parent Principal scale (LCC 23). Last, those modes with b in their names
refer to bass notes: the III Major (IIIb) group refers to major chords with the third in the bass.
The treatment of Mode V here bears special mention. The fth mode of the Lydian scale is of
course the ordinary major scale, which Russell took great pains to show earlier was not a
16. In order to avoid confusion, we will generally respect the distinction between chord and scale from this point
on. Where we have occasion to refer to the explicit chord/scale equivalence, we will use the term chord-scale rather
than Russells chordmode. Russells interchangeable use of chord, chordmode, and mode tends to confuse more
than it helps, and chord-scale is the generally accepted term in modern scholarship and pedagogy.
17. This process is simplied somewhat by the inclusion of a foldout chart in the book (in the rst published
edition, it was referred to as the Lydian slide rule). The chart is somewhat dicult to understand, though, and
including it here would likely confuse matters.
18. This table summarizes the material in LCC 2329.
109
Principal
Mode Spelling chordmode Sub-principal chords
CM (triad), Cmaj6,
I Major C D E Fs G A B Cmaj13s11
Cmaj7, Cmaj7f5
Am (triad), Am6,
VI Minor A B C D E Fs G Am13
Am7, Am9, Am11
III Major
E Fs G A B C D Cmaj13s11/E C/E, Cmaj7/E, etc.
(IIIB/Minor +5)
+IV Minor f5 Fsm7f5, Fsm7f5f9,
Fs G A B C D E Fsm11 f9
Seventh b5 f13 Fsm11f5f9
chord-generating scale. Its role as the fth mode of the Lydian scale is only to act as support for
the consonant Lydian harmony. The principal chordmode for this scale is the same as that of the
Lydian proper, with the fth in the bass. This chord and its relatives are by nature unstable (cf. the
cadential 64 chord), and this instability allows Russell to avoid a potential complication of his theory.
After the explanation of the modes of the Lydian scale, Russell goes on to show how his seven
vertical scales give rise to other kinds of chords. To do so, he introduces another bit of
terminology, the Primary Modal Genre (pmg):
All of the principal chordmodes in Table 4.2 are pmgs, and the six other vertical scalesLyd.
augmented, Lyd. diminished, Lyd. at seventh, Aux. augmented, Aux. diminished, and Aux. dim.
bluesgenerate similar assemblages of chordal types.
19. His treatment of Mode II is similarly problematic, as Lydian Mode II (a vertical scale) is distinct from the
Major at seventh scale (a horizontal one).
110
s9
D13 f9
b www
f5
b n w
& b # b # wwww
? w
Figure 4.6. The second mode of the C auxiliary diminished scale, in scalar and tertian formations.
Figure 4.6 gives an example of how this works in practice. The left-hand side of the gure
shows the second mode of the C auxiliary diminished scale, and the right-side gives its vertical
expression as an altered dominant chord: D13 s9 f9 f5. Russell works through all of the modes of
the six other vertical scales, and the chart included with the book lists almost all of these. The
eight pmgs fall into general categories, which are shown in Table 4.3 and will be sucient for our
purposes here.
It is by now, I hope, apparent how the Concept can simplify matters somewhat for an
improvising jazz musician. Once the process of matching chord symbols with scales is learned, it is
20. There are eight pmgs rather than seven because Russell needs to account for the sharp fth in the Lydian
augmented scale. Other altered scale degrees (fII, fIII) are seen simply as alterations and are not counted among the
principal genres. Russell does not explain why, but the reason is probably related to the fact that the +V genre gives
rise to an important class of chords (7+5), while the others do not.
111
a relatively simple matter to determine what notes will sound good over, for example, a D13 s9 f9 f5
chord. Russell seems to have taken Daviss wish to learn all the changes to heart; he takes care to
note that indeed all of the harmonies of Western music can be found somewhere in the chart, and
notes that many non-traditional harmonic colors can be found as well (LCC 29).
At the same time, though, it is probably also apparent that Russells system is somewhat more
complicated than it needs to be. His ideas about Lydian tonal organization are in fact the source of
much of this complication. To see the extent to which the two are entangled, take Russells
explanation of how to nd the parent scale for an unadorned E f7 chord (a relatively
straightforward example):
Over the roman numerals of the scales of Chart A are listed dierent chord families. For
example, over roman numeral II of the Lydian Scale are listed 7th, 9th, 11th, and 13th
chords. They belong to the same family: the (II) seventh chord family of a Lydian
Chromatic Scale.
The E f7 chord is found in this family above roman numeral II of the Lydian Scale
in the right column of Chart A. The Lydian Scale is therefore the parent scale of the
E f7 chord.
Place the root of the E f7 chord on roman numeral II, and E f becomes the second
degree of that chords parent scale.
Think down a major 2nd interval; if E f is the second degree of the parent scale, D f
is the rst degree. Therefore D f is the tonic (root) of the E f7 chords parent scale. This
tonic is called the Lydian tonic. For the E f7 chord, D f is the Lydian Tonic and the
parent scale is D f Lydian. (LCC 59)
This is quite a long process to determine that the most ingoing (consonant) scale for an E f7 chord
is the second mode of the D f Lydian scale. The equivalence of chords and scales is genuinely useful
for improvising musicians, but the Lydian organization is more abstract. Faced with this situation,
jazz musicians made the obvious simplication: over an E f7 chord, play the E f Mixolydian scale.
Russells theory becomes interesting, though, when we realize that any of the member scales
of the Lydian tonic can stand in for the ordinary Lydian. Once you have determined that the
parent scale of an E f7 chord is D f Lydian, then it becomes easy to substitute more complicated
scales built on the same Lydian tonic. If you wanted to create a more dissonant (outgoing) sound,
you might instead play the second mode of the D f Lydian at seventh scale; the second mode of
the D f auxiliary augmented blues scale would be more dissonant still. Because the seven principal
112
scales form a spectrum of consonance to dissonanceas Russell frames it, there is a progression of
unity from ingoing to outgoingthe Concept provides a means of measuring how closely a
particular progression or improvisation stays to a particular Lydian tonic. This idea is the core of
what we might recover from Russell, and we will return to it when we begin to develop a
transformational system in the next section.
Russells fundamental insight about the nature of chords and scales was revolutionary, and now
forms the basis for much of modern jazz pedagogy. Most of the later sources for chord-scale
theory, as it has come to be called, do not contain any mention of the Lydian generation of the
tonal system, or go through the fuss of nding a parent Lydian scale and its associated pmg. Some
of these texts do not mention Russell at all, which we might take as evidence that (not unlike
Rameaus fundamental bass) chord/scale equivalence is such a natural way of thinking about music
that it was taken for granted and no longer associated with its original author. Given the inuence
of this theory in jazz pedagogy, it will be worthwhile to sketch a brief outline of the literature here,
if only to show how it diers from Russells conception.
Each text uses slightly dierent variations on the theory, but Mark Levines Jazz Theory Book
will serve here as a surrogate for the theory in general. Levine divides his chapter on chord-scale
theory into four parts: major scale harmony, melodic minor scale harmony, diminished scale
harmony, and whole-tone scale harmony. For each of these families, he describes the modes of the
given scale and the harmonies (chord symbols) associated with them.
21. Mark Levine, The Jazz Theory Book (Petaluma, CA: Sher Music, 1995), 3194. Dmitri Tymoczko summarizes
several textbooks views on chord-scale theory in The Consective-Semitone Constraint, 17479. Other texts that
discuss chord-scale theory at length include Andy Jae, Jazz Harmony (Tbingen: Advance Music, 1996);
Joe Mulholland and Tom Hojnacki, The Berklee Book of Jazz Harmony (Boston: Berklee Press, 2013);
Jamey Aebersold, Jazz Handbook (New Albany, IN: Jamey Aebersold Jazz, 2010),
http://www.jazzbooks.com/mm5/download/FQBK-handbook.pdf; and Richard Graf and Barrie Nettles, The Chord
Scale Theory and Jazz Harmony (Advance Music, 1997). I am eliding the minor dierences in these discussions, since
they will not aect the transformational system in the next section. John Bishop provides a good overview of the
distinctions in A Permutational Triadic Approach to Jazz Harmony and the Chord/Scale Relationship (PhD diss.,
Louisiana State University, 2012), 7781.
113
Mode Chord symbol
Ionian Cmaj7 (avoid 4)
Dorian Dm7
Phrygian Esus f9
Lydian Fmaj7 s4
Mixolydian G7 (avoid 4); Gsus
Aeolian Am f6
Locrian Bm7 f5
Table 4.4. Levines chord-scale description of C major scale harmony (Jazz Theory Book, 34).
Many later chord-scale theorists describe avoid notes in scales; these are notes that are
dissonant with the underlying harmony and should generally be avoided in improvisations except as
non-harmonic tones. This is an idea that is not explicit in the Concept, but many of the notes that
are described as avoid notes can be traced back to the fact that later theorists do not take the
Lydian scale as their starting point (4 is usually described as an avoid note on a major seventh
chord). Levines rst-choice chord-scales for major scale harmony and melodic minor scale
harmony are shown in Tables 4.4 and 4.5, respectively, with avoid notes listed as needed. Levine
does not give the modes of the diminished and whole-tone scales (for obvious reasons), and notes
that the diminished scale represents 7 f9 and fully-diminished harmonies, while the whole-tone
scale can be played over 7 s5 and 7alt. harmonies.
22. Levine and others often describe only the rst choice (Russell would say most ingoing) for matching a scale
with a chord; Russells system of substituting other member scales sharing the same Lydian tonic is generally not
present in the later texts. For jazz musicians the melodic minor scale refers only to its ascending form, which is played
both ascending and descending in jazz.
114
From this brief description, we can see how Russells theory is more-or-less stripped of its
philosophical underpinnings and used simply as a pedagogical and performance tool. Though there
are vestiges of the Lydian conception of tonal space (Levine names the Lydian augmented and
dominant scales), what remains is only the idea of chord/scale equivalence. On the one hand, this
is certainly simpler: gone is the complicated derivation of Lydian parent scales, in its place a simple
one-to-one matching of scales with chord symbols.
At the same time, though, something seems lost. For Russell, Lydian organization of tonal
space was not incidental, but in fact the most important idea in the book (its title, after all, is The
Lydian Chromatic Concept of Tonal Organization, and not Chord/Scale Equivalence and Jazz
Improvisation or the like). Rather than simply writing o Russells more unusual ideas as eccentric
ramblings, we will aim in the next section to reincorporate some of them, in an eort to assimilate
some of the rst jazz theory back into modern scholarship.
Now that we have explored Russells theory in some detail, we can take it as a basis on which to
construct a transformational system. Focusing on chord-scales as rst-class objects will allow us to
take seriously the idea that scales are harmony, and will enable analytical observations about the
way improvising musicians might understand harmonic structure.
First, though, it will be useful to take a brief tour through other analytical approaches that
incorporate scales. Dmitri Tymoczko dedicates much of A Geometry of Music to the study of scales,
and applies them analytically to both twentieth-century music and jazz. He is interested primarily
in voice-leading among scales, and constructs voice-leading spaces among the diatonic, acoustic,
23. Tymoczko, A Geometry of Music, Chapter 4 and throughout. This work is a culmination of much of his earlier
work that incorporates scales; see, for example, The Consective-Semitone Constraint and Scale Networks and
Debussy, Journal of Music Theory 48, no. 2 (October 2004): 21994.
115
harmonic major and minor, hexatonic, octatonic, and whole-tone scales. His conception of scales
is somewhat dierent than Russells, though; for Tymoczko, a scale is a ruler, and provides a way
of measuring musical distance. The Concept s view of scales, on the other hand, overlaps
somewhat with Tymoczkos notion of macroharmony: for Russell, a scale acts more like a set of
notes that are all available for improvisation. In general, Russell is not interested in common-tone
connections between chords or their abstract structure, and accordingly we will not have much
occasion to draw on Tymoczkos work here.
Other authors have applied chord-scale theory to jazz, though in somewhat dierent ways
than we will do here. Garrett Michaelsen, for example, draws on Tymoczkos work on scalar voice
leading to construct networks for the music of Wayne Shorter. Michaelsen does take seriously the
notion that chords and scales are equivalent, but his work is more interested in determining how
scalar structure can bring structure to harmony that is not obviously functional. Stefan Loves work
on parsimonious connections is valuable for teaching students about chord-scales, but falls
somewhat short for our purposes here, since it does not include all of the scales Russell identies.
The work that intersects most closely with our work here is John Bishops dissertation, which
incorporates chord-scales into a triadic transformational system. Bishop is inuenced by
chord-scale theory as it is taught at the Berklee College of Music, which is dierent in some ways
from Russells theory outlined above. He is also interested in triadic approaches to improvisation;
116
in his theory, chord-scales exist as a means of generating these triads. In this section we will not
restrict our focus to triads, but will instead consider chord-scales as objects unto themselves.
One of the problems facing any scale theory is the need to account for scales of dierent
cardinalities. Tymoczkos common-tone theory provides a means of connecting a means of relating
the whole-tone (6 notes), diatonic, acoustic, harmonic major (all 7), and octatonic (8) scales via
split and merge operations, and indeed these scales account for six of Russells seven vertical
scales. Tymoczko does mention the ascending melodic minor scale, but it does not merit a place
in his diagram since it is simply a mode of the acoustic scale. His system does not, though, account
for Russells African-American blues scale (hereafter, simply the blues scale), which has either 8
or 10 notes, depending on whether 2 and n7 are included. While we could incorporate this scale
into the common-tone systemit is two splits and a semitone displacement from an octatonic
collectionchord-scale theory as it is usually taught does not focus on common tones between
chord-scales, but rather on determining what scale captures the sound of a particular chord.
Russells Lydian tonic system, despite all of its seemingly unnecessary complexity, provides a
simple solution to the cardinality problem. Because all non-diatonic scales have a Lydian scale as
their ultimate source (their parent scale), this means we can understand these other scales as
alterations of some diatonic collection. All of the scales in Figure 4.3, for example, are derived
from the F Lydian diatonic collection: the D melodic minor collection (Lyd. augmented), F
harmonic major (Lyd. diminished), F acoustic (Lyd. at seventh), whole-tone (wt1 , aux.
augmented), and two octatonic scales (oct02 and oct12 , the auxiliary diminished scales). Russells
31. These triadic approaches also tend towards music that is less clearly tonal, and were devised partly as a means
of moving beyond the standard chord-scale approach we are examining here.
32. Tymoczko, A Geometry of Music, 13435.
33. Russells blues scale contains more pitches than the blues scale as described by some others; Mark Levine gives
the C blues scale as CE f FFsGB f C (The Jazz Theory Book, 219). Joe Mulholland and Tom Hojnacki give
Russells version (with 2 and n7) with the caption a more complete blues scale, but note that there is no single blues
scale. Rather, there is a large variety of scales that share common blues characteristics (Berklee Book of Jazz Harmony,
135). In practice, melodies derived from the blues scale focus heavily on f3 and f7, and are usually unambiguous, as we
will see in the analyses below.
34. Indeed, this lack of focus on common-tone connections in the pedagogical literature is the main impetus for
Loves Model of Common-Tone Connections.
117
0. Lydian (diatonic)
1. Lydian augmented
2. Lydian diminished
3. Lydian f7
4. Whole-tone
5. Whole-half diminished
6. Half-whole diminished
7. Blues scale
Table 4.6. A scale index inspired by Russell, listed from most consonant to most dissonant.
four horizontal scales, of which the blues scale is the most important, also have a Lydian tonic, and
can be understood as still further variations on the Lydian collection.
This reduction to a single diatonic collection will be the rst step in devising a transformational
system for chord-scales. Instead of referring to a scales parent Lydian tonic as a Lydian scale, we
will instead refer to it by a key signature: the D Lydian collection is 3s, the E f Lydian collection
2f, the F Lydian collection simply n, and so on. This notation is in common use and, helpfully,
eliminates some of the awkwardness of having to refer constantly to the Lydian mode. The second
mode of the F Lydian scale is of course the same as the G Mixolydian scale, and both refer to the
collection n.
It is not yet clear, though, how Russells other member scales might be incorporated into this
system. To do so, we will rst introduce the concept of a scale index, shown in Table 4.6 (they are
numbered from 0 to 7 for reasons that will become clear shortly). Several things are worth noting
about this table that dier from Russells presentation. First, some of the scale names have been
changed to reect their common usage; we no longer need to remember, for example, which of the
diminished (octatonic) scales is the blues variant. I have maintained Russells names when they
35. The shift in terminology from the Lydian scale to the Lydian collection here is deliberate, but not
theoretically signicant. In practice, Russell treats a scale as a collection: a group of notes from which to generate
chord tones or improvisations. In this section we will be more interested in scales as collections, rather than (say) their
function as musical rulers.
36. These names have been chosen to reect common jazz usage, so scale 3 is the Lydian f7 scale rather than the
acoustic scale.
118
clarify the relationship to the parent scale: scale 1 remains Lydian augmented rather than
melodic minor, since the D melodic minor scale has F Lydian, not D Lydian as its parent scale.
There are eight scales in the scale index, but Russell gives eleven member scales. The reason
for this is a practical one: two of the horizontal scales are simply diatonic modes (the major scale
and major at seventh), and the third is a major scale with an additional s5 (major augmented
fth), which we can usually understand as a chromatic passing tone. The blues scale, though,
does appear frequently in jazz, and merits its own place here. This scale is given last in the order
because it is one of Russells horizontal scales, which are inherently more outgoing than their
vertical companions.
With some mathematical sleight of hand, we can dene a chord-scale gis using the scale
index along with a diatonic collection as above. Elements of this gis have the form
diatonic collection, scale name; the F Lydian collection is described by the pair n, Lydian, while
2f, Lyd. f7 describes the E f acoustic collection (E f Lydian is in the 2f diatonic collection, and
the Lydian f7 scale on E f adds the pitch D f ). Creating an ordered-pair gis of course requires us
to show that both elements are part of a mathematical group. Though this will not be its nal
form, we will dene this chord-scale gis formally here, with the knowledge that it will be relaxed
in the next section. It is important to realize that this gis is designed to reect Russells own
conception of chord/scale equivalence. There are only three distinct octatonic collections, for
example, but the gis contains 24 distinct diminished scales: whole-half and half-whole diminished
scales on all twelve Lydian tonics.
37. Admittedly, Russell himself would likely object to this characterization since, as noted in the rst section of
this chapter, the Lydian and major scales are fundamentally dierent objects. In practice, many jazz musicians do avoid
4 in major scales, and so whether a given passage is Lydian or major is often ambiguous.
38. In fact, the blues scale is probably more common than some of Russells vertical scales.
39. This construction shares some similarity with Julian Hooks spelled heptachords, in which the white-key
diatonic collection is inected by any number of accidentals. Some, but not all, elements of the gis here could be recast
as spelled heptachords, since the diminished scales require a letter doubling. Julian Hook, Spelled Heptachords, in
Mathematics and Computation in Music, Proceedings of the Third International Conference of the Society for
Mathematics and Computation in Music, Paris, June 2011, ed. Carlos Agon et al. (Heidelberg: Springer, 2011), 8497.
119
The rst element of the pair is a key signature, which have been studied in a transformational
context by Julian Hook. Because we are interested in collections in jazz (often a non-notated
music), we will consider enharmonically equivalent key signatures (like 6s and 6 f) to be identical.
There are, then, only twelve key signatures (isomorphic to the group Z12 ), operated on by the
sharpwise and atwise transformations, sn and fn , which add n sharps or ats to a key signature,
s f2 s
respectively; we might write 1s
1
2s, 2f 4f, or 1f
3
2s.
The scales in the scale index do not obviously form a group, but we can (temporarily) dene
the eight scales to be isomorphic to Z8 , the integers mod 8. The scales do form a progression from
consonance to dissonance, and for Russell it is true that the whole-tone scale is in some sense
further away from the Lydian tonic than the Lydian augmented scale. They are not, however,
cyclic in any meaningful way: it is not as though, for example, the blues scale is the most dissonant
scale and if you take one more step you arrive back at the Lydian scale. Nor are the metaphorical
distances of consonance between the scales really consistent: though Russell does not dene them
(and we will not attempt to do so here), the consonant distance between the two diminished scales
seems much less than the distance between the whole-tone and whole-half diminished scales. If
we accept these limitations of the scales group structure, though, we gain all the benets of a gis.
Intervals between scales are calculated as integers mod 8, using the scale labels from Table 4.6: the
interval from a Lydian augmented scale to a whole-tone scale is 3 (= 4 1), from a whole-tone
scale to a half-whole diminished scale is 2, and so on.
We will call a transformation between chord-scales R, after Russell; these transformations
have the form Rsignature transformation, scale index interval. The transformation acts on
40. Julian Hook, Signature Transformations, in Music Theory and Mathematics: Chords, Collections, and
Transformations, ed. Jack Douthett, Martha M. Hyde, and Charles J. Smith (Rochester: University of Rochester Press,
2008), 13760. In this section we will not apply the full power of Hooks transformations, since they operate not only
on key signatures themselves but on pitches (what he calls oating diatonic forms).
41. Ibid., 142. We could also write fn as sn 1 , and the entire set of signatures can be generated by s1 , adjusting for
enharmonic equivalence as necessary.
42. This is apparent in his description of the tone orders of the Lydian Chromatic scale, as reproduced in
Figure 4.4. We could also dene a gis using these tone orders, but such a gis would lose some distinctions between
scales (the Lydian at seventh and auxiliary augmented are both representatives of the 10-tone order).
43. This limitation is not as signicant, and in fact is the normal state of aairs for diatonic intervals, where, for
example, intervals of both 3 and 4 semitones are called thirds.
120
G7 Cmaj7
#
&
Rs1, 0
n, Lyd. s, Lyd.
Figure 4.7. A typical VI jazz lick, along with its chord-scale GIS analysis.
Fmaj7
&b b n b b n
Re, 1
n, Blues n, Lyd.
Figure 4.8. A resolution from the F blues scale to the F Lydian scale.
elements of the gis in a pairwise fashion in the usual way. A passage like the one in Figure 4.7, for
example, expresses the Rs1 , 0 transformation: the collection changes but the scale does not.
Figure 4.8 shows a passage which begins with the F blues scale and resolves to the F Lydian
collection, representing the transformation Re, 1, where e indicates the identity element.
This initial pass at a chord-scale gis is a useful rst approximation, but there are some aspects of it
that are somewhat unsatisfactory. How, for instance, should we determine what scales match with
what diatonic collections? In some cases the answer is clear, but in others it is not. In Figure 4.7
above, for example, we labeled the G7 chord as n, Lyd., rather than, say, 2s, Lyd. f7; the 4 that
would conrm either is absent. The problem seems to become even more intractable when we
encounter the symmetrical scales: a diminished scale has eight possible parent diatonic collections.
Here again, George Russell provides a solution. Table 4.7 presents a portion of the foldout
chart from the Lydian Chromatic Concept in somewhat simplied form (it may be useful to
compare this table with Table 4.3 on p. 111). The top of this table gives the eight scales in the scale
121
Diatonic Lyd. augmented Lyd. diminished Lyd. f7 Whole-tone WH diminished HW diminished Blues
FGABCDE FGABCsDE FGAfBCDE FGABCDEf FGABCsEf FGAfBfBCsDE FGfAfABCDEf FAfABfBCDEf
122
D f7s5, D f7f5 D f7s5, D f7s9
+V
Table 4.7. Common chords in the modes of the F Lydian Chromatic scale.
E7f9 E7f9 E7alt.
VII Phrygian
Alternate scale names are given in italics; less common chords are shown in gray.
index of the previous section, while the left side lists the modal tonics. Only modes that give rise
to common chords are shown in this table; notably absent are modes III and V, which are given by
Russell as tonic chords with altered bass notes. The most common chords and scales (which are
also the most ingoing) appear on the left side of the table, and rarer chords and scales appear
nearer the right side. Alternate names for scales, when they exist, are given in italics in the
appropriate box.
It is important to realize that for Russell the modal degrees (what he calls primary modal
tonics) are roughly equivalent to functional categories. This matters here because it helps us to
determine a scales parent diatonic collection. All of the chords in the top row of the table are
rst-mode scales, and act like tonic chords. The F7 in the top row of the Lydian f7 column thus
represents a major-minor seventh chord acting as tonic (Russell actually gives this chord symbol as
Maj f7 or Maj 9th f7). Likewise, dominant chords appear mostly in mode II, minor seventh
chords appear in mode VI, and half-diminished sevenths in mode +IV.
Of course, the pairing of modes and scales is still not unique, and is ultimately a question of
analysis. Consider the scale in Figure 4.9a. This is an F acoustic scale, which can appear as an F
Lydian f7 scale or as the second mode of an E f Lydian augmented scale. The gis allows us to
show this, since F acoustic as tonic is a dierent gis member than F acoustic as dominant.
Figure 4.9b places the ambiguous scale in the context of a iiVI progression in B f (the last four
bars of an imaginary solo on George Gershwins I Got Rhythm). Here, the acoustic scale clearly
functions as a dominant, and would be labeled 2f, Lyd. aug.: the parent collection is E f Lydian
(2f), and this is a mode of the Lydian augmented scale (number 2 in the scale index of Table 4.6).
In contrast, Figure 4.9c places the fragment in an F blues iiVI progression (the end of Charlie
44. The dominant chords that appear in mode +V are related by tritone to those in mode II. The fact that these
tritone-related dominants appear along with the whole-tone and diminished scales is no accident: because these scales
are symmetrical at the tritone, these scales are particularly eective when soloing over dominant seventh chords.
Dmitri Tymoczko discusses this practice explicitly in A Geometry of Music, 36568.
45. It is reasonable to wonder why we did not collapse the Lydian augmented and Lydian f7 scales into a single
mode, as we did with the horizontal major scale and the vertical acoustic scale. The aim in this section is to show that
the two scales do in fact function dierently, which justies their presence as separate vertical scales.
123
b
a) &
Cm7 F7 Bbmaj7
b b
b) & b b
J . .
b .
Gm7 C7 F7
b
c) & b
Figure 4.9. An ambiguous acoustic scale and two concrete presentations of it.
Parkers Nows the Time, perhaps). Because the collection now functions as a tonic chord, it
represents the gis member n, Lyd. f7.
The intuitions captured by the chord-scale gis here are not quite like those represented by
other theories of chord-scales. Indeed, the fact that both n, Lyd. f7 and 2f, Lyd. aug. refer to
the same 7-element set of pitches is not immediately apparent in the gis itself. Theories that
prioritize voice-leading would likely include the F acoustic collection only once, since the voice
leading from this scale to itself is maximally ecient (no voices move at all). Nor is it enough
simply to label the scale as the F Lydian dominant scale, as this does not capture the dierence in
function between the passages in Figure 4.9bc.
The gis in fact is one of functional or heard chord-scales. In this way, it is more like Steven
Ringss gis for heard scale degrees than theories of scalar voice leading. Rings argues that scale
degrees are a perceived, rather than inherent, quality of music, and thus listeners have dierent
experiences when hearing A4 as 1 and A4 as 7. In many ways, this distinction is like the one at
the heart of what Russell was trying to accomplish in the Lydian Chromatic Concept. For Russell,
the F acoustic collection played over a tonic major-minor seventh chord really is a dierent entity
46. Steven Rings, Tonality and Transformation (New York: Oxford University Press, 2011), 4450 and throughout.
47. Ibid., 42.
124
than the same collection played over a dominant major-minor seventh chord. This is what Russell
means when he writes of a state of complete and indestructible chord/scale unity (LCC 10): if an
F7 chord can function in more than one way, so too can its corresponding scale.
Previous chapters have constructed various musical spaces in which to analyze passages, and
this gis can be turned into a space as well. Before doing that, though, we will relax its denition
such that the scale indexes are no longer isomorphic to Z8 . We noted above that this isomorphism
was somewhat articial, and the space becomes more intuitive if we simply use the non-modular
integers 07 as elements of the space (which we will call S for now). The resulting space, though,
runs afoul of the formal requirements for a gis, which requires that ivls form a mathematical
group. The set S under addition does not form a group, since it is not closed (6 + 5 is not a
member of S ) and elements do not have inverses.
This is of course one of the well-known limitations of gises: the musical spaces must be both
symmetrical and homogeneous. A gis cannot account for musical spaces that are discontinuous or
have boundaries, which is the case here: it is not possible to conceive of a scale in the system
which is more ingoing than the Lydian scale, or more outgoing than the blues scale.
Nevertheless, the transformations still seem to be reasonable reections of intuitions about the
nature of chord-scales. As Hook notes, the narrative portions of Lewins analyses [in GMIT ]
generally far transcend the logical consequences of the group structure, so the fact that a
mathematical group does not underlie this no-longer-gis should not dissuade us from exploring its
analytical potential.
48. Many authors have commented on this limitation, which Lewin rst observes in GMIT, 27. Dmitri
Tymoczko is probably the most vocal in his opposition, and proposes incorporating a distance metric into the denition
of what he calls a Lewinian interval system (Generalizing Musical Intervals, Journal of Music Theory 53, no. 2 [Fall
2009]: 24546); in other places, he has suggested that relaxing some of the restrictions on a gis is anti-Lewinian
(Lewin, Intervals, and Transformations: a Comment on Hook, Music Theory Spectrum 30, no. 1 [Spring 2008]:
16468). For more on this tension, see Rachel Wells Halls review of GMIT (Journal of the American Musicological
Society 62 no. 1 [Spring 2009]: 20522); Julian Hook, David Lewin and the Complexity of the Beautiful, Intgral 21
(2007): especially 18586; and Rings, Tonality and Transformation, 1920.
49. Admittedly, the outgoing boundary is more permeable than the ingoing: we could conceive of a scale that is
more dissonant than the blues scale (the total chromatic, perhaps) and incorporate it into the system, while for Russell
the Lydian boundary is absolute.
50. Hook, David Lewin and the Complexity of the Beautiful, 185.
125
Lewin does allow for semigroups of transformations, but the scale index transformations do
not form a semigroup either, since a semigroup must still be closed under the group action. All
possible intervals for the scale indexes (the integers 07) are contained in the set
{7, 6, . . . , 6, 7}, which forms neither a group nor a semigroup. It does contain the additive
identity (0), and every element has an inverse, so we can use this set under addition in practically
the same way. Every interval is well-dened, but all intervals are not possible from every scale. For
example: int(Lyd., Whole-tone) = 4, and int(Whole-tone, Lyd.) = 4, but there is no scale which
satises x in the statement int(Blues, x) = 4, since there is no scale that is four levels more
outgoing than the blues scale.
With these caveats, this space can be visualized using the diagram in Figure 4.10. In this
gure, the diatonic (Lydian) scale is centrally located, with more outgoing scales located further
toward the outside; these concentric circles combine with the ordinary circle of fths to divide each
scale into diatonic wedges. The gure is inspired by Russells description of the Lydian as a
musical Star-Sun (LCC 8) and the ultimate source of tonal gravity. More outgoing scales have
more gravitational potential energy, as it were, and are more dissonant with the underlying diatonic
collection. We can use this gure to map the two presentations of the F acoustic scale of Figure
4.9; such a mapping is given in Figure 4.11. This visualization makes clear that the second
presentation (a tonic F7 chord) is more outgoing than the rst (a dominant F7 chord), as well as
the shift in underlying diatonic collection (E f Lydian vs. F Lydian).
Figure 4.11 also reveals that perhaps the objects in the system could be more informative. The
right side of this gure represents a iiVI in F with the sequence
Re, 0 Rs1 , 3
f, Dia. f, Dia. n, Lyd. f7.
That is, the G Dorian scale and C Mixolydian scale are both represented by the pair f, Dia.. On
one level, this makes sense: both scales are modes of the B f Lydian (or F major) scale. Still, since
a chord-scale is supposed to represent a complete and indestructible unity, it seems appropriate
51. The diatonic collections are shown in atwise order traveling clockwise. This corresponds to the way the circle
of fths (or fourths) is usually presented in jazz textbooks; see, for example, Jerry Coker, Elements of the Jazz Language
for the Developing Improvisor (Miami: Belwin, 1991), v.
126
n
s f
2s 2f
3s 3f
Di
g.
a.
au
yd .
. d L yd
L
im.
7
L yd . f
4s WT 4f
i m.
HW WH d
di m .
5s Blues 5f
6s/6b
n n
es f es f
Blu Blu
. .
dim dim
W
H dim. W
H dim.
H H
W WT 2f W WT 2f
7 7
d. f d. f
Ly m. Ly m.
di di 2
d. ug. d. ug.
. a a. 2 . a
Ly y
Ly y
a.
L
Di Di
d
1 1
3f 3f
f, Dia. n, Lyd. f7
Figure 4.11. Two presentations of the F acoustic scale, shown in red, in the planetary model
(compare Figure 4.9).
127
to add some information about the chord into the notation itself. As it stands, information about
the chords themselves is separate from the transformations, and there is no way to distinguish the
progression above (a iiVI in F) from, say, the nonsensical progression Em7 f5Fmaj7Bm7 f5.
We could solve this problem in several dierent ways, but the most obvious is to include the
chord symbol itself in the chord-scale representation. This results in what we will call a
chord-scale triple of the form chord symbol, diatonic collection, scale name. This construction will
allow us to draw on the work done in the previous chapters developing a system of transformations
for chord symbols; we will still call the resulting transformations R, but the rst element of the
new triple will be a transformation between chord symbols. The F-major iiVI of Figure 4.9c
thus becomes
RTF, e, 0 RTFblues , s1 , 3
Gm7, f, Dia. C7, f, Dia. F7, n, Lyd. f7.
Formally, the transformations of the last two chapters act on ordered triples of chord root, third,
and seventh; the additional scale information in a chord-scale triple enriches this sparse three-not
representation. Recall that the chord symbol transformations are cross-type transformations, and
thus even if we had not already relaxed the gis of the previous section (removing the cyclic group
Z8 ), the new version with chord symbols cannot form a gis.
While adding chord symbols to the system does clarify matters, it also complicates them. In
particular, the planetary model of Figure 4.10 no longer represents the musical space accurately.
The addition of chord symbols means that a copy of the planetary model exists at every location a
chord symbol appears in the previous chapters. Such a visual space would be forbiddingly
compleximagine the thirds spaces of Figures 3.2 or 3.15 redrawn with the additional chord-scale
models. This is a sacrice made consciously so that the chord-scale triples and R transformations
are clear in the text. In practice, we can use either the planetary model or the chord spaces of the
previous chapters as the situation demands, with the knowledge that both representations exist
52. We will generally use this new form of the R transformation, but in cases where there is a need to distinguish
between this version and the version used just above, 2-element R versus 3-element R works nicely.
53. In fact, the situation is even more complex, since a single chord symbol can support only a subset of the 96
chord-scales. Thus every chord symbol would contain a dierent partial copy of the planetary model. Only the chords
listed in Table 4.7, for example, would be able to show the scales from the n diatonic wedge.
128
simultaneously as part of the single conceptual chord-scale space. With the nal version of the
chord-scale transformational system in place, the stage is now set to turn toward actual jazz
performance.
Analyzing jazz performance is inherently more complicated than the lead-sheet analysis done in
previous chapters. Because jazz is primarily an improvised music, the analyses here will rely on
transcriptions, which carry with them their own set of problems. As Steve Larson notes, any
transcription is also a kind of analysis: in many cases it is not at all clear how a particular recorded
sound should be (or indeed, whether it can be) rendered in Western musical notation. In the
transcriptions in this dissertation, I focus primarily on pitches and rhythms, since they are most
relevant to the discussion of harmony. As such, many of the most important aspects of a
performancedynamics, articulation, timbre, intonation, and so onare absent from the
notation.
The three short analyses that follow will serve as preludes to the longer analyses we will pursue
in the next chapter; each introduces certain issues of analysis to be explored in more detail later.
All are solos on tunes analyzed in the rst three chapters, allowing us the opportunity to discover
how these abstract chord progressions are realized in improvised performance. They are also solos
by tenor (and soprano) saxophonists: Rahsaan Roland Kirk, Gene Ammons, Sonny Stitt, and Joe
Henderson. This selection reects some of my own preference for saxophonists, but also permits a
basis for comparison: dierent instruments have dierent idiomatic patterns. Notably absent from
this list are Charlie Parker and John Coltrane, undoubtedly the two most well-known jazz
saxophonists. As noted in Chapter 1, this dissertation is interested in jazz harmony in the general
54. Complete transcriptions for all solos analyzed in this chapter and the next can be found in Appendix B.
55. Steve Larson, Analyzing Jazz: A Schenkerian Approach (Hillsdale, NY: Pendragon Press, 2009), 12.
56. This is to say nothing of the fact that perhaps the most important aspect of a jazz solo is that it is improvised;
transcriptions must necessarily only focus on a single recorded performance.
129
sense; by focusing on musicians who do not commonly appear in works of music theory, we gain
insight into the lingua franca of jazz, rather than the particulars of Parkers or Coltranes style.
Multi-saxophonist Rahsaan Roland Kirks recording of Blues for Alice from We Free Kings (1961)
will act as an introduction to some of the issues of analyzing improvised performance. The
complete transcription of the performance can be found on p. 210, and the analysis of the chord
progression of this tune is in Section 2.2.2. Kirk often plays multiple saxophones simultaneously;
each instrument is given its own sta in the transcription.
One of the main problems of analyzing chord-scales is determining exactly which notes
should be taken as part of the scale, and which are simply embellishing. If one of the principal
arguments of this chapter is that scales are harmony, then the question becomes one of
determining what is non-harmonic. Sometimes it is obvious that notes are embellishing, but other
cases are not so clear. The Gs in Figure 4.12a, for example, clearly functions as a chromatic passing
tone between G (the fth of the Cm7 chord) and A (the third of F7). Figure 4.12b presents a more
complicated case: is the E f in the scale, with En serving as a chromatic passing tone, or vice versa?
If we choose E f as the main note, the scale implied is B f Mixolydian, while En gives a B f Lydian
dominant scale. The choice has analytical implications, as the two scales represent two dierent
locations in chord-scale space. It is important to note that these kinds of non-harmonic tones are
not quite like ordinary non-harmonic tones. Larson (and many others) would argue that both the
E f and En are non-harmonic, since at some deeper level they would reduce to either D or F
(neither is part of a four-voice B f7). Because we have broadened the denition of harmony to
include chord-scales, our idea of what is non-harmonic must also change accordingly.
57. Thomas Owens (among others) refers to the bebop-inspired style as jazzs lingua franca (Bebop: The Music and
its Players [New York: Oxford University Press, 1995], 4). To avoid focusing on Parker and Coltrane is not to say that
the performers examined here are not also great musicians; all of them (but especially Stitt and Henderson) are highly
regarded among jazz musicians.
58. Larson, Analyzing Jazz, 510.
130
a) b)
b n b
Cm7 F7 Bf7 Bf7
& b # n b &b
Figure 4.12. Two non-harmonic tones, from m. 112 (2:52) and m. 17 (1:00), respectively, of Kirks solo
on Blues for Alice.
n b
Gm7 C7 Fmaj7
& b n b #
Figure 4.13. Two examples of enclosures (marked with brackets) in a iiVI progression, from
mm. 3335 (1:19) of Kirks solo.
One kind of embellishing gure merits special attention, which Jerry Coker calls the
enclosure, where a pitch is approached by semitone on either side. Two examples of enclosures
appear in the iiVI progression in Figure 4.13. The rst appears before the B f on beat 2 of the
rst bar, and the second before the resolution to A at the end of the passage. What is interesting
about enclosures from the chord-scale perspective is that usually only one of the neighbors is
non-harmonic (i.e., not part of the scale). In the rst example, only the Bn is truly non-harmonic,
since the A is a part of the (very clear) G Dorian scale. Likewise, the Bn in the next measure is a
chromatic passing tone in the C Mixolydian scale between C and B f , while Gs is a lower neighbor
to the following A.
Another issue arises when an improviser chooses to play a single scale over several chord
changes; Coker calls this harmonic generalization. This is a technique often used at faster
tempos or with complex chord changes (or both), since it allows the performer to slow down the
eective harmonic rhythm. Often the process results in notes that clash with the underlying
59. Coker, Elements of the Jazz Language, 5054. Cokers manual is intended for student improvisers, but is
analytically useful as something of an encyclopedia of idiomatic melodic devices in improvisations, since he provides
many examples of each technique. He is one of the only authors I am aware of to discuss this aspect of performance;
most others simply do not mention it at all, or hand-wave the problem away, saying that the knowledge will come by
listening to and transcribing many recorded performances.
60. Ibid., 4549.
131
w b b - n .
Fmaj7 Em7 A7 Dm7
Tenor sax &b
3
Soprano sax
& b b. b b - n .
(Manzello)
3
Figure 4.14. Harmonic generalization in Kirks solo: the F blues scale appears over four dierent
harmonies (mm. 2527, 1:10).
harmonic progression, so Coker also notes that it is less than ideal. Kirk uses this technique at
the beginning of his third chorus, where he plays the F blues scale over four chord changes (Figure
4.14). B f , A f , and Fn are all relatively dissonant over Em7, but here the coherence of the F blues
scale allows us to focus on the gure as a single unit rather than hearing chord-to-chord. This
phenomenon is easy to capture in the transformational system; all of the transformations have the
form R__, e, 0, since the chord changes while the scale does not.
The nal issue Kirks solo brings to light is that of chord substitution. Previous chapters have
discussed harmonic substitution at length, but only in the context of tunes. The process of
substitution that happens during solos is more complicated: dierent substitutions can happen in
dierent choruses, a soloist might use a substitution while the rhythm section does not (or vice
versa), and so on. This is an aspect of performance that is not so easy to capture in chord-scale
spaceor at least, in any single transformational label. In general, transformational labels show
only a single interpretation: a passage is either represented by one chord-scale triple or another,
not both simultaneously. We can, however, examine a single passage in multiple, dierent ways,
rst exploring one interpretation then another. Steven Rings calls this process of analysis prismatic,
where phenomenologically rich passages are refracted and explored from multiple perspectives.
132
Fmaj7 Dm7 Gm7 C7 Fmaj7
b b
& b # #
j n
Figure 4.15. The turnaround at the end of Kirks first chorus (mm. 1113, 0:52).
As an example of this prismatic approach, take the turnaround from Kirks rst chorus, shown
in Figure 4.15. Over the nominal Dm7 chord, Kirk plays the notes E f and D f . These two notes
seem incompatible with Dm7; none of Russells scales contain a chromatic trichord (necessary if
the scale is to include the chord root). One possibility is that Kirk is drawing on the total
chromatic gamut, or perhaps some other scale Russell did not recognize (like the enneatonic scale).
This analytical possibility seems unlikely, though; drawing on a high-cardinality scale seems
excessive to explain a mere two pitches. Another, more likely, possibility is that Kirk takes the
tritone substitution here, playing the notes E f and D f over an implied A fm7 chord (or perhaps
A f7, as a tritone-substituted dominant of the following Gm7). Still another possibility is that Kirk
uses a harmonic generalization, playing a diminished scale fragment (E f D f CB f AG) over a
C7 harmony that is implied over the full one-and-a-half bars of the turnaround.
Figure 4.16 shows these last two analytical possibilities in the planetary model of the previous
section. The rst involves hearing the E f D f succession as a diatonic tritone substitution for
Dm7. In this hearing, all of the chord-scales are diatonic, but the diatonic collection shifts
greatly from the rst chord to the second. (There are only three objects in this model because
moving diatonically from Gm7 to C7 does not involve a scalar shift.) Figure 4.16b, on the other
64. The potential hearings given here are not exhaustive, of course. Experienced jazz listeners and analysts must
bring along their knowledge of jazz harmonic progression and elaboration in order for this kind of analysis to make
sense; the transformational system admits of many logically consistent analyses that may not eectively describe the
music under consideration. It seems unlikely to me, for example, that the E f here is a atted 9th over Dm7, with the
D f acting as a chromatic passing tone, simply because f9 is not an extension that is commonly played over minor
seventh chords.
65. The diagram here assumes A f7, since it is more common to substitute dominant sevenths than minor
sevenths. An analysis with A fm7 would be similar, except that the second chord would be in the 6f collection.
133
a) n b) n
es f es f
Blu Blu
. .
dm
i dm
i
W
H dim. W
H dim.
H H 2
W WT W WT
7 2f 7 2f
d. f d. f
Ly im. Ly im.
d d
d. ug. d. ug. 1
.a . 3 .a .
Ly y
Ly y
L
L
ia ia
d
d
D
D
1
3f 3f
2
4f 4f
5f 5f
Figure 4.16. Two analytical possibilities for Kirks turnaround, in mm. 1113:
a) As a tritone substitution for Dm7.
b) As a harmonic generalization for C7.
hand, combines a small diatonic shift with a large leap in the scalar dimension, moving from the
diatonic collection to the wholehalf diminished scale.
We might also construct dierent transformation networks for this passage, as shown in
Figure 4.17. Letter a is a strictly chronological network (time ows from left to right), which
shows the basic transformations involved in the tritone-substitute hearing. Letter b redraws this
network to reect the organization of iiV space: the iiVI is in a single horizontal line, while
Dm7 is a level higher, representing its position as a ii7 chord in C. This network also claries that
the starting and ending Fmaj7 chord is the same point in the space. Letter c, on the other hand,
reimagines the chord transformations as taking place in a F major diatonic space (similar to that of
66. Recall from Table 4.7 that a diminished scale played over a dominant chord is always the wholehalf
diminished scale. Conceptually, the parent scale of C7 is B f Lydian, so the scale in question is the second mode of the
B f wholehalf diminished scale (which is of course the same as the C halfwhole diminished scale).
67. All of the networks here are what Lewin calls gural networks (Musical Form and Transformation: Four
Analytic Essays [New York: Oxford University Press, (1993) 2007], 4553); since the R transformations do not form a
semigroup, a formal network (showing all possible transformations) is not possible. Hall suggests in a footnote
(Review of GMIT and MFT, 213n16) that the logical structure of the group is unnecessary in gural networks; the
networks of this gure support this claim.
134
a)
Fmaj7, n, Dia. Dm7, 5f, Dia. Gm7, f, Dia. C7, f, Dia. Fmaj7, n, Dia.
b)
Dm7, 5f, Dia. RT9 7TH 3RD, f5, 0
RT5, s4, 0
Gm7, f, Dia. C7, f, Dia. Fmaj7, n, Dia.
c)
Fmaj7, n, Dia.
Rt3, s1, 0
C7, f, Dia.
Gm7, f, Dia.
Rt3, s4, 0
Dm7, 5f, Dia.
d)
Dm7, f, WH dim.
Rt5, f1, 5
Rt3, e, 0
Rt3, e, 0
Rt3, s1, -5
C7, f, WH dim.
Figure 4.17. Four dierent transformation networks of the turnaround in mm. 1113.
135
Section 1.5), while still hearing the tritone substitute. Finally, letter d represents the
diminished-scale hearing: both transformations in the middle column contain e, 0 as their
second and third elements, since the scale does not change.
It may seem as though the preceding pages have reached for the pile-driver to kill the gnat, so
to speak; the entire discussion was brought about by the two notes E f D f . These notes are not
of any particular signicance in Kirks solo, and indeed the entire turnaround is relatively ordinary.
As listeners and analysts, though, we can conceive of multiple ways of hearing this particular
passage, and chord-scale transformations oer a way to explore these interpretations. None of the
networks of Figure 4.17 is more correct than the other, and none proposes to be the structure of
this particular passage. As Rings observes, to the extent that [transformational] analyses reveal
structures at all, they are esthesic structures rather than immanent structures. Throughout his
book, he emphasizes that a particular analysis is more a record of an analytical encounter with the
music than the music itself. Though we will not often focus so intently on such a small fragment of
music, adopting Rings attitude means that we can: the lens of transformational theory can zoom
in and out as needed.
Before leaving Kirks solo, it will be instructive to look at his improvisations on a few iiVI
progressions, if only because they are so ubiquitous in jazz. Figure 4.18 gives three of these, each
taken from the last four bars of a chorus of Blues for Alice. Letter a reproduces Figure 4.13,
and uses the most typical iiVI chord-scale pattern:
RTF, e, 0 RTF, s1 , 0
Gm7, f, Dia. C7, f, Dia. Fmaj7, n, Dia..
The G Dorian and C Mixolydian scales are both in the f collection, while the parent scale for
Fmaj7 is the n collection; this results in the typical s1 VI resolution. Figure 4.18b is likewise all
diatonic, and has an identical transformational structure. Its interest comes in the double
68. The pile-driver/gnat metaphor comes from Richard G. Swifts response to Lewins rst published article
(Letter to the editor, Journal of Music Theory 4, no. 1 [April 1960]: 128; quoted in Hook, David Lewin and the
Complexity of the Beautiful, 162).
69. Rings, Tonality and Transformation, 37, emphasis original.
70. Figure 4.18c is missing beats 24 of the third bar because this iiVI leads into the turnaround in Figure 4.15,
which is discussed at length above and would distract from the topic at hand.
71. Absent any additional information we will assume Russells most ingoing scale for chords. Since there is no B
(at or natural) in any of the third bars of Figure 4.18, we assume F Lydian.
136
Gm7 C7 Fmaj7
n b
a)
Chorus 3
& b n b #
(mm. 3335, 1:19)
b)
# n b
Chorus 2
(mm. 2123, 1:05) &b
j j b n b
& b b
c)
Chorus 1
(mm. 911, 0:50)
Figure 4.18. Three iiVI progressions from Kirks solo on Blues for Alice.
enclosure gure that appears over Gm7: the downbeat B f in m. 22 is approached by two
chromatic neighbors on each side. The A and C are both in the G Dorian scale, while Gs and Bn
are simply embellishing, and do not take part in the chord-scale transformations. Figure 4.18c uses
a diminished scale over Gm7 and includes D f over C7, giving a dierent transformation network:
RTF, e, 4 RTF, s1 , 2
Gm7, f, HW dim. C7, f, Lyd. dim. Fmaj7, n, Dia..
Here, the scales get more ingoing over the course of the progression from the halfwhole
diminished scale (implying f13 and f5) to the Lydian diminished scale (f9) before reaching the
diatonic Fmaj7.
It should be apparent that these analytical fragments of excerpts from Kirks solo do not
constitute an analysis of the solo as a whole. Nothing has been said about the overall shape of the
solo, the role of register, how playing multiple saxophones limits Kirks available pitches, or any of
the countless other analytically interesting aspects. It is in many ways a typical jazz solo, and as such
has served primarily as an introduction to the fundamentals of analyzing jazz improvisation, while
at the same time oering a glimpse of how chord-scale transformations might be used in analysis.
137
F7 Bm7 E7 B fm7 Ef7
# J n J J # n n
b b
& b . J
3
Figure 4.19. Two examples of anticipations, from Gene Ammonss solo on Autumn Leaves
(mm. 24, 1:04).
Tenor saxophone duo Gene Ammons and Sonny Stitt recorded Autumn Leaves on the album
Boss Tenors in 1961 (the complete transcription is on p. 206). Autumn Leaves was rst analyzed
in Section 1.5, in connection with diatonic chord spaces. We will return to this subject shortly, but
only after a brief diversion into meter in improvisation (an aspect emphasized in this recording that
did not appear in Kirks solo on Blues for Alice).
Both Ammons and Stitt shift barlines frequently in their solos, while Kirk almost never did.
Meter is not the primary focus here, but it can aect the chord-scale analysis of a passage. The
most common kind of metrical shift is a slight anticipation of the following harmony; Stefan Love
notes that these are so common as to be a clich, and they are usually obvious in analysis.
Figure 4.19 gives two straightforward examples from the beginning of Ammonss solo. The Fs in
m. 2 clearly belongs to the following Bm7 and not to F7, and the Fn in the following bar functions
in the same way.
Sometimes the metric shift is more dramatic; take the passage in Figure 4.20, for example.
Here, Ammons seems to arrive at D7 too late, continuing the E f7 harmony two beats into the next
bar. Jerry Coker refers to these as barline shifts, and says that while they are not intentional,
72. The kind of metrical shifts discussed here are much simpler than those usually discussed in the jazz literature.
The classic work on metrical dissonance in jazz is Cynthia Folio, An Analysis of Polyrhythm in Selected Improvised
Jazz Solos, in Concert Music, Rock, and Jazz since 1945: Essays and Analytical Studies, ed. Elizabeth West Marvin and
Richard Hermann (Rochester: University of Rochester Press, 1995), 10335; Stefan Love provides an overview of other
literature in this area in Subliminal Dissonance or Consonance: Two Views of Jazz Meter, Music Theory Spectrum
35, no. 1 (Spring 2013): 4861.
73. Ibid., 51.
138
Ef7 D7 Gm
b b j b
&b
j
#
given: Cm7 F7
eective: Cm7 F7 Cm7 F7
# n n b
b
&b # n
Figure 4.21. A barline shift from Sonny Stitts solo on Autumn Leaves (mm. 7374, 2:59).
necessarily, they are not errors, either, as they might be in the case of the novice who momentarily
loses his/her place in the progression. In this case it is relatively clear that the arrival of D7 is
delayed, and that the pattern D f CB f G should not be taken as some outgoing scale choice for
D7.
In his solo, Sonny Stitt often uses barline shifts in double-time passages to increase the
eective harmonic rhythm; Figure 4.21 gives one of these passages over a iiV progression. In the
second half of m. 73, Stitt plays a descending F bebop scale, implying the F7 chord a half-bar
early. Instead of continuing to play F7 in the next bar, though, Stitt seems to return to Cm7: he
uses Fs only as a lower neighbor until the last beat, when the (strongly C-minor) gure
GE f DC gure resolves to Fn. Stitt exploits the fact that there is no diatonic shift between ii7
and V7 chords, and alternates between the two freely in a 2f diatonic wash.
139
^ # n .
# - n #
Cm7 F7 Bm7 E7
b
&b J J J
3
RECT, s5, 0
RTF, e, 0
Bm7 anticipation: 2f, Dia. 2f, Dia. 3s, Dia.
Figure 4.22. An ambiguous barline shift in Ammonss solo (mm. 4143, 2:06), with two possible
transformation networks.
Finally, there are some cases when it is unclear whether to read a passage as a barline shift or
as an outgoing scale choice; Figure 4.22 gives a representative example. The Cs and Bn in m. 42
could be heard as anticipating the Bm7 by two beats, or as part of a whole-tone scale over F7, as
shown in the transformation networks. Ammons anticipates the Bm7 in the A sections of the rst
chorus (mm. 3 and 11, the second with the pitches Cs and Bn), giving some credence to the
anticipatory hearing. On the other hand, the rst A section of the second chorus features increased
chromaticismthe D7 in m. 38 uses a diminished scaleand perhaps the second A section
continues the trend. Again, the prismatic approach says that it is less important to decide on a
single interpretation than to realize that both are available for our perception.
Over the course of an improvised solo, a performers choice of scale for a particular chord can
change. Usually, though, some chords are more exible in their chord-scale identity than others.
In the Ammons/Stitt recording, the progression Bm7E7B fm7E f7 in the third and fourth bars
of the A sections is almost always accompanied with the 3s, Dia.4f, Dia. succession. The
77. In this gure and others following, I have omitted the chord symbol from the chord-scale triples to save
space, since it is obvious from the transcription itself.
78. By tallying up all of the appearances of a given chord/scale pairing, it would be possible to develop something
like a snapshot of a performers chord-scale choices. Doing so could perhaps bring some clarity to discussions of jazz
style (as in, for example, David Bakers Giant of Jazz series) by incorporating the chord-scale transformations
developed here. While this is an interesting possibility, we will pursue it no further here.
79. The only time it is not occurs in Stitts rst chorus (3A1 , mm. 6768), where he plays a sustained Dn through
all four chords.
140
Original: C-7 F7 Bbmaj7 Ebmaj7 A-7b5
Figure 4.23. Autumn Leaves, mm. 15, showing substitutions in mm. 34.
ii7 V7 E
ii7 V7 Bb
ii7 V7 A
ii7 V7 E b
ii7 V7 D
ii7 V7 A b
ii7 V7 G
D b
TF T5 T6
ii7 V7
3RD 7TH
home-key dominant D7, on the other hand, eoys a wider range of scalar options, appearing at
various times with the wholehalf diminished scale (see mm. 6, 18, 38), the Lydian diminished
scale (mm. 14, 26, 78), the diatonic collection (m. 82), and the Lydian f7 scale (mm. 102, 110). As
rst noted in Chapter 1 (p. 22n67), the scale choice for the tonic G minor is somewhat exible: it
is always played with En, but sometimes with an upper-neighbor Fn implying a Dorian scale
(mm. 19, 39) and other times with an Fs implying a melodic minor scale (or Gm, f, Lyd. aug., as
in m. 83).
The analysis of Autumn Leaves in Chapter 1 focused on its underlying diatonic nature, but
this diatonicism is attenuated somewhat in the Ammons/Stitt recording. Figure 4.23 gives the rst
ve bars of Autumn Leaves as analyzed in Chapter 1 (top) and as played on Boss Tenors (bottom).
141
Both major seventh chords in the original diatonic succession F7B fmaj7E fmaj7Am7 f5 are
here substituted with chromatically descending iiV progressions; Figure 4.24 shows this set of
substitutions in iiV space. This set of substitutions is relatively easy to understand: rst, the
major seventh chords are turned into dominants (B f7E f7); each is preceded by a ii7 chord
(Fm7B f7B fm7E f7); nally, the rst iiV progression is transposed by a tritone. While these
chords are close together in iiV space, they are far apart in chord-scale space, as noted above;
Bm7E7 is played as 3s, Dia. in this recording, while B fm7E f7 is played as 4f, Dia. a
dierence of ve sharps/seven ats.
The ensemble likely decided to make the substitutions in the solos to prevent the piece from
becoming boring: since almost all of the original chords are diatonic in G minor, playing in the 2f
diatonic collection will work for almost every harmony. The substitutions in mm. 34 of the A
sections provide some variety in the chord-scale options, leaving the diatonic nature of the tune to
manifest in other ways. In both of his choruses, Stitt uses a harmonic generalization from the last
four bars of the bridge through the rst four of the C section, playing the 2f diatonic collection
throughout (Figure 4.25 reproduces this passage from his last chorus). Because of the substitutions
in the A section, this implicit diatonic cycle (rst discussed in connection with Figure 1.8) is the
only one remaining in the solo changes; Stitts choice to use a harmonic generalization highlights
this implicit diatonicism. Because the scale stays the same while the chords move through a
diatonic cycle, every transformation in this passage is the same: Rt3 , e, 0.
Joe Hendersons composition Isotope was the subject of Section 3.1.2, and this section will return
to Hendersons solo on the tune from Inner Urge (1965). Since he is also the composer of the tune,
80. This passage also illustrates the property of signature transformations that f7 is always equivalent to T11 ; see
Hook, Signature Transformations, 14244.
81. The dominant, D7, is not in the 2f collection, since it is not diatonic in G minor. This collection also results
in the major scale (not the Lydian) for the B fmaj7 chords, which is borne out by the E f s in the recording (see
mm. 2324, 55, and 88). The Gm chord itself is usually played with En, placing it in the 1f collection.
82. In fact, Stitts note choices here would also work well if the rhythm section were to play E fmaj7 in the fourth
bar, which is suggested in Figure 1.8.
142
Cm7 F7 Bfmaj7
. -
bb . - j
& #
117 3
Am7f5 D7 Gm (Em7f5)
b b #LN J
&
121
Figure 4.25. Stitts 2 f harmonic generalization, highlighting the implicit diatonic cycle (mm. 11724,
4:11).
b
C7 F7
. b b # j b
& J j n b
#
C Lydian f7 Bf Lydian diminished
Figure 4.26. A scalar shift highlights a change in function in Joe Hendersons solo on Isotope
(mm. 1317, 0:47).
Hendersons fteen improvised choruses (Appendix B, p. 230) can provide some insight into how
he understands its harmony.
As we observed in Section 2.3.2, major-minor seventh chords can function as tonic chords in
the blues. This fact presents something of a problem: the C7 at the beginning of Isotope acts as
tonic at the beginning of its four-bar span, but as a dominant of the following F7 at the end. At
some point during this four bars there must be a pivot fth: a motion from C7 as tonic to C7 as
dominant. This is harmonic information that is not readily available in the chord symbols, but
can be seen in a soloists scale choices.
Figure 4.26 gives a representative example from Hendersons second chorus; here, he uses the
Lydian f7 scale in the rst three bars of the chorus before shifting to the Lydian diminished scale
83. Other blues tunes negotiate this problem by inserting another chord in the rst four bars. Thelonious Monks
Misterioso, to take a typical example, is a blues in B f : the rst four bars contain the progression B f7E f7B f7B f7.
In a case like this, the rst B f7 functions as tonic, while the last two bars function as dominant of the following IV
chord.
143
given: C A7 Af7
eective: C A7 Em7 A7 E fm7 Af7
&J # # b b b
J b
Figure 4.27. Two interpolated ii chords from Hendersons first chorus (mm. 79, 0:39).
in the nal bar. The addition of f9 (D f ) intensies the motion toward F7, since tonic chords do
not usually have this extension. This is a harmonic move that is easily captured using a chord-scale
transformation; we can represent the passage as
Rpivot 5th, f2 , 1
C7, s, Lyd. f7 C7, f, Lyd. dim..
Henderson does not do this in every chorus (he often plays diatonic or blues scales through the
entire four bars), but similar eects can be observed in choruses 1, 7, 9, 14, and 15.
In the two analyses above we saw that substitutions can aect chord-scale choices, but not all
of them do. One of the most common substitutions is actually an interpolation, in which a
performer plays a iiV progression when only a V7 chord is given. Because the motion from ii7 to
V7 does not involve a diatonic shift, these interpolations often do not aect the chord-scale
analysis. Two examples from Hendersons rst chorus are given in Figure 4.27; in each, the
arpeggiation in the rst half of the bar could be seen as diatonic (each involving the ninth of the V7
chord), but they can also be heard as arpeggiations of the ii triad.
The blues oers many opportunities for harmonic generalization, and here I want to focus on
two passages in particular. The rst is in Hendersons fth chorus, the opening of which is shown
in Figure 4.28. In the third bar of the chorus, Henderson begins playing a lower-neighbor gure
84. If a performer chooses non-diatonic scales for the ii7 and V7 chords, the analysis would be aected; in the
diatonic case, it is not.
85. Harmonic generalization is part and parcel of the blues; the blues scale derives its name in part from the fact
that the tonic blues scale can be (and often is) played over an entire blues chorus. As Mark Levine has it, playing the
blues scale over the IIVV chords of a basic blues yields dissonances hardly acceptable in traditional theory. But these
dissonances have been present in jazz since its inception (The Jazz Theory Book, 233).
144
b
C7
j
& . b .
49
t6 t6
F7 Bf7 C A7
b
& #
53
# n
t6 t6 t6 t6 t6 (t6)
2f 2s
s4
Figure 4.28. Signature transformations in the opening of Hendersons fifth chorus (mm. 4956, 1:33).
that continues in sequence through the next four bars. Because this passage is so uniform, we can
draw on the full power of Hooks signature transformations, including the diatonic transposition
operator t. Each pattern moves down a diatonic step (t6 ); the nal arrow is dotted to indicate that
the pattern breaks down at this point, and the lower-neighbor is distorted into a descending minor
third. In mm. 5455, the lower-neighbor pattern is constant, while the underlying collection shifts
from the 2f to the 2s collection, an s4 transformation. Henderson ignores the C chord in m. 55,
anticipating the A7 by a full bar, and also deemphasizes the B f7 (there is no A f present).
Henderson plays a similar passage in the middle of chorus 14, shown in Figure 4.29. Like the
rst, it begins with a diatonic sequence, this time in a stepwise rising motion (a series of t1 s). Here
the sequence breaks down much earlier, but the constant eighth notes recall the previous passage;
both sound like a single line with a shifting diatonic foundation rather than a series of loosely
connected gestures. This passage traverses the same s4 transformation as the rst from the 2f to 2s
86. This operator should be read here as Hooks tk , which operates on oating diatonic forms, not as the tk
operator dened in Section 1.5. See Hook, Signature Transformations, 13941.
87. Incidentally, the diatonic collection of the C7 chord itself is ambiguous. We might hear it as the 1s collection
(as part of a Lydian f7 scale), the 1f collection (hearing C7 as a dominant, as the second mode of of B f Lydian), or
perhaps even the 2f collection (as a harmonic generalization of B f7 in the rst six bars of the chorus). Exactly which
diatonic collection we choose is not important to the analysis here.
145
# b
F7 Bf7 C A7
& b b b # n #
161 3
t1 t1 t1 (t1)
f1 3f s3
2f s2 n s2 2s
s4
Figure 4.29. A similar passage in chorus 14, with signature transformations (mm. 16164, 3:52).
b b b
Gf7 Ef7 C7 Gf7 Ef7 C7 Gf7 Ef7 C7
& b b n b n b b n b j n b b
b n b n
#
b b b n n n
Gf7 Ef7 C7 Gf7 Ef7 C7 Gf7 Ef7 C7
& b b b n b j b b b b
b b b # n # n
Chorus 7 Chorus 8 Chorus 11
diatonic collections, though the route here is more scenic. Henderson does not anticipate the A7
and instead plays the C harmony, splitting the s4 transformation into two s2 s. The rst of these is
further subdivided into an f1 /s3 pair; because the A f comes at the very end of m. 162 as part of an
enclosure gure, it is not as strong as the larger motion from the 2f to the n collection.
In the nal pages of this chapter, I want to take a step back to show that there are aspects of
Hendersons solo that are not easily expressed with chord-scale transformations. In the analysis of
Isotope in Chapter 3, we noted that the turnaround, in which dominant sevenths descend by
minor thirds, was one of the most interesting parts of the tune. The fact that these chords can be
146
related by smooth voice leading was downplayed there, but Henderson often plays these chords
(especially G f7E f7) as arpeggiations that emphasize this voice leading. Figure 4.30 gives several
examples of Hendersons solos on the nal bar of the chorus, along with their resolutions. In all of
these, he emphasizes the G f Gn voice leading from G f7 to E f7, and usually brings out the E f En
over E fC7 as well. The importance of these voice leadings disappears in the chord-scale
transformations, and would be better shown in, for example, one of Tymoczkos voice-leading
spaces.
There are two choruses of Hendersons solo where contrapuntal concerns seem to take
precedence over making the changes; Figure 4.31 reproduces chorus 6 in its entirety, but the same
principle is at work in chorus 12. The particular scale Henderson plays in the rst four bars is
clearly not as important as the chromatic line from E to G and back, taking place over the lower G
pedal point. This chromatic polyphonic melody continues throughout the chorus, culminating in
the arrival at the high C in m. 70. This chorus is one that a Schenkerian voice-leading sketch
would describe particularly well (the rst four bars unfold the tonic interval E5 G5 , and so on), but
that chord-scale transformations do not. Indeed, it is not at all clear which scales are being used,
nor do they even seem particularly important to understanding this chorus.
To the extent that chord-scale transformations describe harmony, though, they are quite
useful, as these three analytical vignettes have shown. Both of the limitations just discussed involve
voice leading, and for jazz musicians, harmony and voice leading do seem to be separate entities.
While for Schenker harmony and voice leading are two aspects of the same phenomenon, George
Russell instead pairs harmony (or chord) and scale. The two theories seem to be quite similar in
their goals; compare Schenkers unication of harmony and voice-leading with Russells repeated
88. The four-voice tesseract in Tymoczko, A Geometry of Music, 106 is his way of describing these motions;
Cohns Four-Cube Trio in Figure 3.13 describes the same voice-leading space.
89. Jerry Coker calls this kind of line cesh, which is short for contrapuntal elaboration of static harmony
(Elements of the Jazz Language, 6167).
90. None of the passages discussing cesh or linear chromaticism in Cokers book, for example, say anything
about harmony, nor do any of the passages that feature discussions of voice-leading (what other authors refer to as
guide tones) mention it as a harmonic phenomenon. In his discussion of the 73 resolution, he says that he is
simply concerned with the smooth connection (voice-leading) of two chords, especially with respect to melodic,
rather than harmonic, implications (Elements of the Jazz Language, 19).
147
b
C7
j # J j J J J j b J j
& #
61
b J j J b j # #
F7 Bf7 C A7
& J
65
b . # b b
Af7 G7 C7 A7 Gf7 Ef7
b b n b
j
& J b J J #
#
J
69
Figure 4.31. Hendersons sixth chorus (mm. 6172, 1:48), featuring contrapuntal motion.
declarations that the Lydian Chromatic Concept unites both the vertical and horizontal aspects of
music. Furthermore, Russells theory of jazz was no less aspirational than Schenkers in its goal to
describe all tonal music; Russell says explicitly that the concept of tonal gravity is the underlying
force of equal tempered music (LCC 223).
Toward the end of the Concept, addressing the claim that his theory can describe all music,
Russell asks why such a theoretical work [should] come from the jazz experience (LCC 223).
This chapter has asked the related question of why Russells theory of the jazz experience seems
not to have been applied in any rigorous theoretical way to jazz itself. If we take seriously the idea
that chords and scales are manifestations of a single phenomenon, then we must look at scales if
we are to fully understand jazz harmony. Transformational theory provides a means of examining
these chord-scales in a systematic way, and the chord-scale transformations developed here have
expanded our musical universe beyond the lead sheet and into jazz performance itself.
148
Chapter 5
Rhythm Changes
The discussion of chord-scale transformations in the previous chapter concludes the theoretical
portion of this dissertation; this nal chapter will synthesize that theoretical framework in a series
of three longer analyses. All three of the tunes hereThelonious Monks Rhythm-a-ning,
George Colemans Lo-Joe, and Sonny Stitts The Eternal Triangleare instances of a harmonic
archetype known as Rhythm changes, so named for their origin in George Gershwins I Got
Rhythm. Because tunes that use Rhythm changes all share a common harmonic framework, they
are an ideal means to investigate jazz harmony. A complex set of standard substitutions and
harmonic patterns have emerged over the many years jazz musicians have been playing Rhythm
changes; the three analyses in this chapter will allow us to compare these musicians manipulation
of this basic harmonic framework.
It is hard to overestimate the importance of Rhythm changes on jazz practice; along with the blues,
it is one of the most common harmonic types in the bebop era and beyond. David Baker lists
more than 150 Rhythm tunes in his How to Learn Tunes; some of the most well-known of these
are reproduced in Table 5.1. Before moving on to the analyses in the following sections, it will be
useful to examine the form itself, along with some of its more common harmonic substitutions.
1. I Got Rhythm was written in 1930 and rst appeared in the musical Girl Crazy. Because the phrase Rhythm
changes has developed a life beyond its initial meaning, it is rendered throughout this chapter without quotes but with
a capital R. The phrase is normally used as a noun, while adjectival uses drop the changes (as in Oleo is a Rhythm
tune).
2. Thomas Owens, Bebop: The Music and its Players (New York: Oxford University Press, 1995), 13.
3. David Baker, How to Learn Tunes (New Albany, IN: Jamey Aebersold Jazz, 1997), 4244. New tunes that are
based on the chord changes to other tunes are known as contrafacts. Part of the reason for the proliferation of
contrafacts in general (and the genre of Rhythm tunes in particular) is that jazz musicians could avoid paying royalties
to the Gershwins. On contrafacts more generally, see Owens, Bebop, 8 and 1215.
149
Title Composer
Anthropology Charlie Parker/Dizzy Gillespie
Cotton Tail Duke Ellington
52nd Street Theme Thelonious Monk
The Flintstones Hoyt Curtain
Jumpin at the Woodside Count Basie
Moose the Mooche Charlie Parker
Oleo Sonny Rollins
The Serpents Tooth Miles Davis
Tiptoe Thad Jones
Wail Bud Powell
Table 5.1. A selection of Rhythm tunes and their composers.
I Got Rhythm is, like many jazz standards, a 32-bar AABA form; the basic progression is shown
in Figure 5.1. As Andy Jae notes, its changes are not the least bit astonishing; the tune is a
fairly basic set of turnarounds and dominant cycles. Indeed, this feature is one of the reasons for its
popularity: the harmonic framework is something of a blank slate, and allows room for alteration
in a way that more specic sets of changes (like Parkers Blues for Alice, for example) do not.
Another thing that is immediately apparent is the quick harmonic rhythm in the A sections,
which allows soloists the opportunity to show o as they navigate the rapidly moving changes.
Fundamental to the genre of Rhythm tunes is their mix-and-match nature; each part of the
form has many dierent sets of changes, from which the performers may choose freely. Mark
Levine explains this issue succinctly:
When a musician calls a Rhythm tune like Oleo, theres no discussion of which version
of the changes to play. As with the blues, jazz musicians freely mix many versions of
Rhythm changes on the spot, as they improvise. Playing Rhythm changes is a little like
4. The vast majority of Rhythm tunes are in the key of B f . Those in Table 5.1 that are not are Hoyt Curtains
theme to the cartoon The Flintstones and Bud Powells Wail (both in E f ), along with Thad Joness Tiptoe (A f ) and
Thelonious Monks 52nd Street Theme (C).
5. Andy Jae, Jazz Harmony (Tbingen: Advance Music, 1996), 149.
6. Adding to this virtuosity is the fact that many recordings of rhythm changes are quite fast. Of the standard
bebop recordings of tunes in Table 5.1, Parkers recording of Moose the Mooche is 212 bpm, Powells Wail is 270
bpm, while the Parker/Gillespie Anthropology burns along at roughly 305 bpm.
7. The mix-and-match metaphor comes from Jae, Jazz Harmony, 149.
150
A Bf G-7 C-7 F7 Bf G-7 C-7 F7
Bf Bf7 Ef Ef- Bf F7 Bf
B D7 G7
C7 F7
Bf Bf7 Ef Ef- Bf F7 Bf
Figure 5.1. The basic changes to George Gershwins I Got Rhythm (taken from Levine, The Jazz
Theory Book, 238).
knowing several tunes and playing them all at once; thats why Rhythm tunes are
harder to play at rst than a tune with only a single set of changes.
Given this background, the analytical discussion below will proceed in segments: the A sections can
each be broken into four-bar halves, while the bridge is typically treated as a single eight-bar unit.
The rst four bars of the Rhythm A section serve to establish the tonic B f ; Figure 5.2 gives a
number of possible harmonizations of this section. Letter a gives the original Rhythm changes,
while b shows what is by far the most common substitution, replacing B f with Dm7 in the third
bar; this changes the IviiiV turnaround in the last two bars into a iiiviii-V instead. If the B f
harmony is voiced with a major seventh and major ninth (B f DFAC, a very common voicing),
then we can understand the substitution of Dm7 as a simple omission of the root. Letter c goes
8. Mark Levine, The Jazz Theory Book (Petaluma, CA: Sher Music, 1995), 241.
9. These six harmonizations do not begin to constitute a complete set of substitutions for these four bars. Jamey
Aebersold gives 43 harmonizations of the Rhythm A section in I Got Rhythm Changes in All Keys, Jamey Aebersold
Play-A-Long Series, vol. 47 (New Albany, IN: Jamey Aebersold Jazz, 1991), 26.
151
a) Bf G-7 C-7 F7 Bf G-7 C-7 F7
c) Bf G7 C-7 F7 D7 G7 C-7 F7
e) Fs7 B7 E7 A7 D7 G7 C7 F7
further, transforming many of the minor seventh chords into dominant sevenths that serve to lead
more strongly into the following harmonies.
Figure 5.3 shows the relevant portion of iiV space for these rst three harmonizations, along
with a few annotations. The standard harmonization of gure 5.2a can be seen by following the
blue arrows. The substitution of Dm7 in letter b is represented in the space by the red arrows: in
this reading, rst follow the blue arrows until arriving at F7, then follow the red arrows until Gm7
where the blue arrows continue to the tonic B f. The minor-to-dominant substitutions of letter c
are not shown in the space, but are easy enough to imagine: both Gm7 and Dm7 are transformed
by the 3rd 1 operation, and each is replaced by the chord immediately to its north in iiV space (a
EC
substitution which results in the evaded cadence transformation, G7 Cm7, across the bar lines
at the end of mm. 1 and 3).
The harmonization in Figure 5.2d is still more complex. The tritone substitution of D f7 for
G7 in m. 3 is by now familiar, but the G7 in m. 2 has been replaced with a passing diminished
seventh chord. As we rst saw in the analysis of Have You Met Miss Jones? in Section 3.2.3,
fully-diminished sevenths in jazz can often be understood as V7 f9 chords missing their roots. The
152
A-7 D7 G
TF T5
D-7 G7 C
3RD 7TH
G-7 C7 F
C-7 F7 Bb
Figure 5.3. The first four bars of Rhythm changes in iiV space.
B 7 here, then, is a logical substitution for G7 f9, and the C s 7 in the following bar can be
understood as the same substitution of an implied A7 f9 chord (the dominant of the following D
minor), resulting in a chromatically ascending bass line in the rst two bars. Miles Daviss
composition The Serpents Tooth (the opening of which is shown in Figure 5.4) uses a variation
of this progression. Davis also includes a minor-third substitution in m. 4, substituting E fm7A f7
for the diatonic Cm7F7.
The last two harmonizations in Figure 5.2 are somewhat dierent in nature; while any of the
substitutions of letters ad can be swapped in and out at will (the rst two bars of a followed by
the last two bars of d, for example), those in letters ef usually appear intact. Letter e harmonizes
the rst four bars with a cycle of dominant seventh chords (a favorite technique of Thelonious
Monk, and one we will see in the analysis of Rhythm-a-ning below). In contrast to the relatively
compact arrangement of letters ac in iiV space, this cycle traverses nearly the entire space before
153
a) Bf Bf7 Ef Ef- Bf G-7 C-7 F7
arriving at the tonic B f. Letter f is the harmonization from Jimmy Heaths composition
C.T.A., and features a repeated lament-bass pattern from B f down to F7.
The last four bars of the Rhythm A section contain a shift to the subdominant in the rst
two bars, followed by a turnaround in the last two; Figure 5.5 gives several common
harmonizations of this passage. Once again, letter a reproduces the original changes: a seventh is
added to the tonic B f , tipping it towards an E f chord that resolves plagally (via minor iv) back to
tonic before a viiiV turnaround. This plagal motion in the second bar is often substituted with a
backdoor progression, A f7B f , as seen in b (which also precedes the B f7 in the rst bar with a ii7
chord in E f ) and d (which elides the E f and E fm harmonies). Letter c makes the substitution of
Dm7 for B f in the third bar and includes E 7 as a substitution for E f7 f9.
The Rhythm bridge is usually recognizable because of the drastic slowing of the harmonic
rhythm; again, Figure 5.6 gives several common harmonizations, and Figure 5.7 shows them in
iiV space. The standard bridge (letter a) is a simple cycle of dominants, beginning on the III
chord; we will call this the 4-cycle bridge. The most common substitutions here are tritone
substitutions of every other chord, as shown in b and c. The other common option is to insert a ii7
chord before each of the dominants, as shown in d, decomposing each T5 transformation into
10. I have not included another copy of the complete iiV space here, but one can be found in Figure 2.10 (p. 50).
11. This E 7 is functionally ambiguous; it could also stand in for an A7 f9 as the dominant (or C7 f9 as part a
backdoor progression) to the following Dm7. It is spelled as E 7 to produce a smooth bass line from E f in the rst
half of the bar.
154
a) D7 G7 C7 F7
b) D7 Df7 C7 B7
c) Af7 G7 Gf7 F7
Ef-7 Af7 Df
c
A- 7
D 7
b
d
a
Df7 Gf
D-7 G7
Gf7 B
G-7 C7
B7 E
C-7 F7 B b
Figure 5.7. The four Rhythm bridge harmonizations of Figure 5.6 in iiV space.
155
TF 3rd. (This procedure could of course be combined with the tritone-substituted versions in b
and c as well.) Other less conventional harmonizations are also possible; The Eternal Triangle
and Lo-Joe both use specialized bridges which we will see in later sections.
It should be apparent from this discussion that Rhythm tunes can vary widely in their
harmonic particulars. The mix-and-match nature of their construction means that the chords used
by an ensemble can change even over the course of a single performance: a rhythm section might
prefer one harmonization of the bridge during a saxophone solo and opt for another during a piano
solo, for example. The harmonizations given in Figures 5.2, 5.5, and 5.6 have only begun to scratch
the surface; because most of the tune consists of turnarounds, any of the countless possible
turnarounds could be used instead. It is easy to imagine a Rhythm tune that makes use of the
descending minor-third turnaround of Hendersons Isotope in the A sections or fast-moving
Coltrane changes over the bridge.
Still, though, the many harmonizations of Rhythm changes all share certain attributes:
32-bar AABA form, half-note harmonic rhythm, a move to the subdominant in the sixth bar of
the A sections, and so on. These aspects, combined with the popularity of the form, means that
Rhythm tunes are usually apparent to performing jazz musicians, even though the changes
themselves might be quite removed from Gershwins original (as in the case of Figure 5.2e above, or
in George Colemans Lo-Joe below). To say that a tune is a Rhythm tune is akin to saying a
piece is in sonata form: as listeners we can expect certain general things to be true, but the
particulars of the instantiation will vary from piece to piece (or performance to performance, or
even chorus to chorus).
12. Given the modular nature of Rhythm tunes, the genre seems particularly ripe for a schema theory approach,
following in the footsteps of Robert Gjerdingen, Music in the Galant Style (New York: Oxford University Press, 2007).
The two- and four-bar units here (turnarounds, dominant cycles) are not unlike the stock phrases used in the galant
style, and jazz pedagogical materials like the previously-cited Aebersold and Jae texts could easily serve as analogs to
the 18th-century Italian partimenti often used by schema theorists.
156
5.1.2 Harmonic Substitution vs. Chord-Scale Elaboration
Before moving on to the three analyses proper, it will be helpful to return to an issue rst
mentioned in the last chapter in connection with Rahsaan Roland Kirks solo on Blues for Alice.
In many cases, it is not clear whether a particular improvised passage should be heard as a
harmonic substitution or as an outgoing chord-scale choice over a more basic harmony. In the case
of non-Rhythm tunes, we can usually rely on the head to provide the authoritative changes for the
tune, and it is likely that we choose to hear that particular set of changes throughout the
performance. Rhythm changes, though, bring this problem to the fore, since we cannot depend on
a single set of canonical changes.
By way of a short illustration, consider again the melodic passage that opens Miles Daviss
Serpents Tooth (rst shown with Daviss original progression in Figure 5.4). If this were an
improvised passage, it seems likely that the rst choice of harmonies would not be those used by
Davis, given the clear outlines of both G7 and A7 chords in the second halves of mm. 12. It is
also possible to hear this passage as a series of outgoing scale choices over a standard diatonic
progression, hearing the CsEG fragment as part of a diminished scale over F7. Three possible
hearings of these rst two bars are shown in Figure 5.8, which gives locations in chord-scale space
for each harmony. They are shown here in ingoing-to-outgoing order: a uses only diatonic scales,
b uses the same collections but hears the Lydian diminished scales over the diminished seventh
chords, while c emphasizes more widely shifting diatonic collections and scale choices.
While this prismatic approach to analysis may have seemed excessive for the relatively
insignicant passages in the last chapter where it was used, it will take a central role in our study of
Rhythm changes. Because the harmonic structure of the tunes is so uid, it is impossible to claim
with any certainty that a particular set of changes constitutes some Platonic tune, in the same way
that we might be able to for Autumn Leaves or All the Things You Are. To x a set of
denitive changes for a particular passage is to misrepresent the fundamental nature of Rhythm
tunes in jazz practice; the changes are often ill-dened even among the players themselves (as the
above quotation from Levine attests). Engaging with a single Rhythm tune, then, constitutes an
157
b
&b # n
n
a) Bf, f, Dia. G7, n, Dia. Cm7, 2f, Dia. A7, 2s, Dia. Dm7, n, Dia.
b) Bf, f, Dia. B7, n, Lyd. dim. Cm7, 2f, Dia. Cs7, 2s, Lyd. dim. Dm7, n, Dia.
c) Bf, f, Dia. Gm7, 2f, Blues Cm7, 2f, Dia. F7, 6s, WH dim. Bf, f, Dia.
Figure 5.8. Three possible hearings of the opening of The Serpents Tooth.
engagement with an entire genre of tunes, with all their attendant history. Transformational
theory, with its ability to refract a passage into many possible interpretations, oers us a way into
this rich network of harmonic possibilities inherent to the genre.
5.2.1 Head
Thelonious Monks Rhythm-a-ning is a basic Rhythm tune, and as such will be an illustrative
rst example. The head of the tune is shown in Figure 5.9 as it appears in the Thelonious Monk
Fake Book. The source recording for this lead sheet is from Monks album Criss-Cross (1963); we
will analyze a dierent performance below, but the dierences in the head are insignicant. What
13. A lack of understanding about the recorded history of jazz is often seen as something of a social mistake,
especially in the case of Rhythm changes. Mark Levine tells a story of playing a Rhythm tune with Sonny Stitt, when
Stitt began to play the cycle progression of Figure 5.2e over the A sections. Levine recounts: after a couple of
choruses, glares from Sonny, and a growing sense of feeling smaller and smaller, I nally strolled, or stopped playing.
After the set, I asked him what were the changes he was playing, and he growled just listen, man. This story is
accompanied by a footnote that (based on the word of saxophonist Don Byas) attributes the cycle progression to
pianist Art Tatum. Levine, The Jazz Theory Book, 242.
Phil Ford notes that this kind of secret knowledge is fundamental to understanding jazz as a part of an emerging
hip culture in the 1940s50s, which coincides with the time that Rhythm tunes began to proliferate. As he puts it,
knowing the score is what practically denes the hipster: hip, in its original meaning, means to be aware
(Somewhere/Nowhere: Hipness as an Aesthetic, The Musical Quarterly 86, no. 1 [April 2002]: 54). We might
instead read knowing the changes, since to really know Rhythm changes involves knowing about the wide variety of
harmonic possibilities that denes the genre.
14. Don Sickler, ed., Thelonious Monk Fake Book, Steve Cardenas, transcriber (Milwaukee: Hal Leonard, 2002).
158
A1/A2
b
& b 44 .. b n . . j
b
. . b - . .
.
b j j j j b j b ..
& b b . j
b - - b - . b
Bass -
walks: Bf Ef Bf
bb
D7 G7
& n # - j . - jn # j n n
- .
j
n -
F7
bb .
C7
& j j # j b n
n # - . - #
n - n . . . n -
A3
b j
&b . b b
b . .
. -
. n .
.
b j j j j b j
& b b . . b
Bass
b - - b - b -
walks: B f Ef Bf
solos : Bf rhythm changes
is noteworthy about this lead sheet is that there are no changes given in the A sections; there is
only an indication that the solos are to be played over Rhythm changes. This speaks not only to
the ubiquity of the form, but also to its uidity, since a single denitive version is not given.
Nevertheless, there are a few aspects of the head that might have an impact on a soloists
harmonic choices. The rst is the arpeggiation of an E f major triad in the second bar of the A
sections. None of the common sets of changes in Figure 5.2 use E f in the second bar, but this
15. The second volume of the old Real Book does give changes during the A sections, but they are somewhat
inaccurate; in any case, it also includes the indication to solo over Rhythm changes.
159
plagal motion is essential to the tune. The other important feature of the tune is the whole-tone
ascent at the end of the bridge. Monk is well-known for his propensity towards the whole-tone
scale, and we will see this manifest below in his solo on the tune.
Instead of the Criss-Cross recording, we will instead focus our analytical attention on a live
recording made in 1958, on the album Thelonious in Action. This recording is attractive for a
number of reasons. First, tenor saxophonist Johnny Grin takes eleven full choruses on the tune,
allowing the opportunity to analyze a somewhat longer selection of music than we did in the
previous chapter. Second, after Grins second chorus, Monk does not play at all, leaving only the
bass and drums to accompany the tenor saxophone. This, combined with the ambiguity of
Rhythm changes, provides something of a blank harmonic slate, leaving Grins improvised lines
to do the bulk of the harmonic work.
When approached with the fast-moving harmonies in the Rhythm A sections, Johnny Grins
preferred strategy seems to be to ignore them: he frequently uses harmonic generalizations in the
A sections. Often these generalizations are diatonic, using the 2f collection; Figure 5.10 gives a
representative example from chorus 8A2 . While we could perhaps imply a diatonic set of chord
changes like those in Figure 5.2ab, the rising arpeggios harmonizing the top line F5 F6 seem to
take precedence over any particular harmonization. Similar rising diatonic patterns can be found in
mm. 14 (chorus 1A1 ) and mm. 16972 (6A2 ).
Other times, Grin plays passages that are nearly diatonic, but altered somewhat to t an
underlying harmony. Figure 5.11 gives an example from chorus 9A1 (chorus 11A1 is similar). In
16. The harmonization Cm7C s 7 ts the melody, but does not appear in Monks recordings, where the bassist
consistently arrives on E f on the downbeat of the second bar.
17. Robert Hodson provides a similar analysis of the Criss-Cross recording (though not from a transformational
perspective) in Interaction, Improvisation, and Interplay in Jazz (New York: Routledge, 2007), 6674. His analysis
focuses more strongly on the interactive elements of the performance than the harmonic ones.
18. Again, this is a common occurrence for Monk. During particularly good solos, he would rise from the piano
and dance around the stage (an aspect of his performance on display throughout Charlotte Zwerins documentary of
Monk, Thelonious Monk: Straight No Chaser [Warner Bros., 1988], VHS). In the live recording here, he can
occasionally be heard shouting words of encouragement to Grin.
160
n
bb . J
&
233
b b n
& J J b J
257
Rs1, 2 R f1, -2
Figure 5.11. Altered diatonic generalization in Griins solo (mm. 25760, 4:22), with two possible
transformation networks.
this passage, the line is mostly diatonic, with the exception of the Bn and A f in m. 259, implying a
G7 f9 harmony. This passage, unlike the diatonic ascent in Figure 5.10, ts better with a diatonic
chord progression, as shown in the upper transformation network. While it is certainly possible to
hear a 2f diatonic swath throughout these four bars (represented in the lower network), hearing
the half-note harmonic rhythm brings out the contrast between G minor in the rst bar and the
altered G dominant seventh in the third.
The most common harmonic generalization Grin uses is the B f blues scale, which often
appears in the last A section of a chorus. The clearest example of this is also the rst, at the end of
his third chorus; this passage is reproduced in Figure 5.12, and similar clear statements of the blues
scale can be found in 4A3 , 8A3 , 10A3 , and 11A3 . Because Grin generalizes the A sections so
frequently, the two-bar harmonic rhythm of the bridge often sounds like an acceleration of
161
> b
b
&b # j
89
n
j
.
j
#
b # J
&b
93
Figure 5.12. Griins blues generalization in a final A section (mm. 8996, 1:55).
b b b b .
b b b b b b b
&b J J
25
n n b b b
b n n b b
&b
29 3 3
Figure 5.13. A D f m triad as a B f blues subset in Griins solo (mm. 2532, 0:58).
harmonic activity rather than its usual role as a relaxation of the half-note harmonic rhythm of the
A sections. This blues generalization in the last A section of a chorus, then, helps to increase the
contrast to the dominant cycle of the bridge.
The blues scale also provides an explanation for Grins seemingly unusual implication of
D fm at the end of the rst chorus, shown in Figure 5.13. It is not immediately apparent how to
understand the passage in mm. 2528 (see Figure 5.14): Grin could be superimposing D fm over a
B f diatonic progression, implying B fm7 f5, or using a B f halfwhole diminished scale
generalization. Given his inclination for the blues scale in the last A section of the tune, though,
my own hearing leans towards this D fm triad as a subset of the B f blues scale. Grin also
emphasizes the pitches D f and A f at the end of his second and tenth choruses; the former implies
D f major, while the latter leans more clearly toward B f .
162
n b b n b
a) b) c) d)
b b b b b b b b
b b n b n
b n
&
Figure 5.14. Four possible harmonic contexts for a D f m triad: a pure triad; the upper tones of a
B fm7 f5; a diminished scale subset; and a blues scale subset.
bb n n b b
& J
201
Figure 5.15. Half-note harmonic rhythm in an A section of Griins solo (mm. 2014, 3:33).
Grin does not always generalize the A sections; he sometimes plays the half-note harmonic
rhythm of the tune itself. The clearest example of this occurs in chorus 7A2 , which is reproduced
in Figure 5.15. While it seems clear that Grin hears the half-note harmonic changes here, the
melodic patterns he plays are harmonically ambiguous, owing to their limited range. While the
progression shown in a is most likelyit combines the passing diminished seventh in the second
bar with the applied dominant of C in the thirdwe might hear the harmonization at b instead,
with A7 in the place of C s 7 and a tritone substitution for G7 f9. Still other, less conventional,
hearings are possible; letter c shows a hearing that moves to E f , by rst moving to E f7 in the
second bar (a nod towards the heads tilt towards the subdominant in the same formal location),
and then via an applied dominant to a modied iiVI in E f .
Admittedly, this last hearing may be dicult to discern, not least because it is so far from the
typical rst four bars of Rhythm changes. Another important reason, and one we have yet to
consider, is the role of the other band members in shaping harmony. Indeed, the principal role of
the rhythm section is to provide the harmonic framework for the soloists. Since Monk does not
163
bb n n b b
Tenor
Sax & J
b
? b
b n
Bass
Bf Af7? Bf Cm7 F7
Figure 5.16. Griins improvised solo line, along with Ahmed Abdul-Maliks bass line (mm. 2014,
3:33). Bass sounds as written.
play during Grins solo, we might look to bassist Ahmed Abdul-Maliks line during these four
bars, which is shown in Figure 5.16. Abdul-Malik does not seem to use the half-note harmonic
rhythm here, and instead plays a generalization in the rst two bars, walking up the B f major scale
(or Lydian, depending on whether the E f or En is heard as the chromatic pitch). The strong
tonicdominant motion from A f to E f in the third bar seems to imply some A f harmony, perhaps
as a backdoor substitution to the downbeat B f in the fourth bar. He does gesture towards the
home-key iiV in the last bar: we might hear the B f G on beats 23 as a weak arpeggiation of
Cm7, and the nal C as a representative of F7 (which resolves to B f in the next bar). While this
bass line certainly provides insight into Abdul-Maliks conception of the harmony of these four
bars, it does not necessarily tell us anything more about Grins harmonic understanding; it is
entirely possible (and common, as here) that all band members do not share exactly the same
harmonic framework, especially in a Rhythm tune.
As we have noted before, our job as listeners and analysts is not necessarily to decide on a set
of denitive changes. This ambiguity is a critical part of understanding exactly what jazz harmony
is, and carries with it important epistemological questionsquestions, incidentally, which relate to
my own suspicion of the Schenkerian analysis of jazz rst sketched in Section 1.2. If we take
19. A detailed account of exactly how walking bass lines project harmony is beyond the scope of this project; for a
good overview, see Todd Coolman, The Bottom Line: The Ultimate Bass Line Book (New Albany, IN: Jamey Aebersold
Jazz, 1990).
20. There is much more to be said about the role of interaction in negotiating harmony in Rhythm tunes; we will
return to this idea in more detail in the analysis of The Eternal Triangle in Section 5.4.
164
Abdul-Maliks bass line as the harmony, are we then to understand some of Grins note choices as
incorrect? Or vice versa, if Grins solo line represents the true version of the harmony, why does
Abdul-Malik choose to ignore it? In Rhythm-a-ning, do the changes for the head (determined
by whom?) hold through all of the solos, or is the harmonic framework considered anew in every
chorus? In order to make a Schenkerian voice-leading sketch of the passage in Figure 5.15, we
would be forced to contend with these issues, since determining what pitches are consonant (or
more structural) is dependent on being able to identify the underlying harmony unambiguously.
My own contention is that the realities of jazz performance necessitate a more uid theoretical
conception of harmony; the prismatic transformational approach allows us to make these kinds of
distinctions by presenting multiple transformation networks (representing multiple harmonic
hearings) of a single passage.
Compared to his strategies for the A sections, Grins bridges are much less varied. In most
choruses, he uses the standard 4-cycle bridge; Figure 5.17 gives an example from his nal chorus.
In this passage, Grin repeats the rising arpeggio, altering it in each two-bar phrase to t with the
descending-fths harmonic pattern. He often provides additional harmonic interest by changing
the scale to lead more strongly to the following harmony (not unlike the technique Joe Henderson
used in the rst four bars of his solo on Isotope, examined in Section 4.3.3). Figure 5.18 gives a
passage from Grins second chorus; here, both the D7 and G7 gain a f9 in their last two beats,
while the F7 gets a s5 (or f13) in its nal bar, which acts as a common-tone connection with the B f
blues scale that follows in chorus 2A3 . The other common alteration Grin makes is the tritone
substitution, as shown in Figure 5.19.
21. This diculty is perhaps one of the reason that Steve Larsons book on the subject (Analyzing Jazz: A
Schenkerian Approach [Hillsdale, NY: Pendragon Press, 2009]) focuses primarily on solo piano recordings of Round
Midnight, a piece in which the harmonies are well-denedunlike Rhythm changesand there are no other band
members to muddy the waters. (Larson does include a live recording from Bill Evans and a partial transcription of a
Bud Powell recording, each of which uses a piano trio, plus an ensemble recording from Evanss Conversations with
Myself, a multi-track recording in which Evans acts as all three members of the ensemble.) Garrett Michaelsen critiques
Larson on this same point, and also suggests that Larson overemphasizes the importance of harmony in general
(Analyzing Musical Interaction in Jazz Improvisations of the 1960s [PhD diss., Indiana University, 2013], 910).
22. The main interest in this particular passage is metric: Grin superimposes a three-beat pattern over the
quadruple meter starting in m. 341.
165
D7 G7
n n n .
bb n # n n # n n J
&
337
C7 F7
bb n b n b J J
&
341
Figure 5.17. A 4-cycle bridge from Griins last chorus (mm. 33744, 5:30).
# n n
D7 G7
b b # n # n n n n n b
& #
49
C7
n b
F7
b n n n b
&b b J J
53 3
Figure 5.18. A 4-cycle bridge, with chord-scale elaborations that lead more strongly toward the
following harmony (mm. 4956, 1:19).
. n b b
D7 G7 Df7
bb J # n b b # n n b
& J
241
n # # #
C7 B7
bb n n # # n #
& # # # b n
245
Figure 5.19. The bridge from Griins eighth chorus, with tritone substitutions shown in green
(mm. 24148, 4:08).
166
5.2.3 Monks Solo Harmony
While Grins tenor saxophone solo displays a number of interesting harmonic formations,
Thelonious Monks own solo on Rhythm-a-ning exhibits a few more, and is worth a brief visit
here. Monk only plays three choruses on the tune, and his rst is characteristically sparse.
Throughout all three A sections of this rst chorus (chorus 12 in the transcription), he plays only
pitches from the B f pentatonic collectionthe same collection he uses to comp behind Grins
rst three choruses. The recurring rhythmic motive is altered slightly so that it ts the
harmonies of the standard 4-cycle bridge in mm. 36976, before returning to the B f pentatonic
collection for the nal eight bars of the chorus.
Beginning in chorus 13, Monk consistently plays an 8-chord dominant cycle in the A sections
(the harmonization rst seen in Figure 5.2e). Because this harmonization is so distinct from the
ordinary Rhythm A section, Monk simply arpeggiates each chord to avoid blurring the overall
progression. Playing a winding bebop line through the dominant cycle might risk the coherence of
the substitution, especially if the bass player did not pick up on this harmonization and played a B f
diatonic bass line. Figure 5.20 reproduces chorus 13A1 , showing the dominant cycle in the rst
four bars, followed by a B f blues harmonic generalization in the next four.
These dominant-cycle A sections are always paired with bridges that use the whole-tone
scales. The head of Rhythm-a-ning uses the whole-tone scale in the bridge (clearly over the F7,
and implied over C7 as well), and its use in Monks solo helps to provide coherence to the
performance as a whole. The bridge from chorus 14 is shown in Figure 5.21; though the
whole-tone collection shifts between adjacent dominant seventh chords, the passage is
harmonically consistent. This uniformity is easy to see in the chord-scale analysis: every change of
harmony is represented by the transformation RT5 , f1 , 0.
23. Comping is what jazz musicians call the act of accompanying (or complementing) a soloist; see
Mark Levine, The Jazz Piano Book (Petaluma, CA: Sher Music, 1989), 22334.
24. Thelonious in Action is a live recording made at the end of an eight-week run at the Five Spot Cafe in New
York (Robin D. G. Kelley, Thelonious Monk: The Life and Times of an American Original [New York: Free Press,
2009], 24243). Ahmed Abdul-Malik was used to playing with Monk by the time of the recording, and catches the
dominant-cycle A section almost immediately; Monks strong left-hand entrance on Fs at m. 385 removes any doubt as
to the progression that will follow.
167
Fs7 B7 E7 A7 D7 G7 C7 F7
b n b b
& b # # n # n # # n n # n n # n n
n .
? bb
# n n
385
Bf
b
& b b n b j b n b b b n b
(blues)
? bb
389
Figure 5.20. Monks dominant-cycle A section and blues generalization from chorus 13A1
(mm. 38592, 6:12).
D7s5 G7s5
bb # # n # b b n
(anticipation)
n # .. # n
&
433
s, WT n, WT
RT5, f1, 0 RT5, f1, 0
bb b b b
C7s5 F7s5
# j
b n b b # j n n
J n
b
&
437
f, WT 2f, WT
RT5, f1, 0
Figure 5.21. Monks bridge from chorus 14, using the whole-tone scale (mm. 43340, 6:53).
168
This brief analysis of Rhythm-a-ning has illustrated that no single set of changes can
adequately describe this tune. While in a more standard tune like Autumn Leaves, individual
chords might change slightly (adding an extension or using a tritone substitution), the basic
progression remains intact; rarely do we encounter a situation like that of Monks 8-cycle A
section, where an entire set is replaced with another. This mix-and-match approach to harmony is
an essential element of Rhythm tunes. Though the dozens of Rhythm contrafacts are all based on
a single chord progression, the wide range of harmonic approaches means that no two Rhythm
tunes sound exactly alike.
To this point in this study, the analyses have focused on harmony as reected in a lead sheet, or
how particular performances conrm (or contradict) these given lead-sheet harmonies. This
approach naturally requires a lead sheet to exist in the rst place, which is not always the case.
Even in cases where one does exist, there are several reasons a jazz musician might want to create a
lead sheet anew: it may be inaccurate (often the case with fake books); it may not reect a
particular recording the musician wants to emulate (John Coltranes recording of Body and Soul,
for example, does not use the standard changes); or the musician may simply want to practice ear
training.
In these cases, the transformational approach to harmony can be used in reverse, so to
speak; rather than analyzing how a soloist elaborates on a given set of changes, we can take the raw
material of a recording and deduce a likely set of changes. This section will do just that, using
George Colemans composition Lo-Joe, recorded on the album Amsterdam After Dark (1979).
Lo-Joe was recorded somewhat later than the other tunes analyzed here, and as such is somewhat
more harmonically adventurous. It is recognizably a Rhythm tune, though with a highly altered
bridge, and in the key of D f rather than the usual B f .
25. I am grateful to Prof. Tom Walsh for bringing this tune to my attention.
169
Before beginning with the analysis, a few disclaimers are in order. While a transcription of
the head (complete with piano and bass parts) can be found in Appendix B on p. 236, I want to
emphasize the fact that a full transcription is in general not necessary to create a lead sheet, and is
provided here only as an expedient to writing about the process. A skilled jazz musician would
likely transcribe only the melody, and determine the harmonies simply by ear, without necessarily
writing anything down. Next, there is the question of whether or not Coleman and his bandmates
ever played from a lead sheet at all; might we be manufacturing a somehow false lead sheet rather
than reconstructing one? This question is not important for our purposes here, since lead sheets
are such a common way of conveying jazz tunes. Even if Coleman did not give his bandmates a lead
sheet, he must have had some means of communicating the harmonic progression of the tune, and
a lead sheet is the canonical way to notate this kind of progression in jazz.
The opening of melody of Lo-Joe, shown in Figure 5.22, appears to be straightforward,
outlining mostly major triads and major seventh chords. The resulting succession of harmonies,
though, does not seem to reect any of the usual Rhythm openings, or indeed any ordinary jazz
progression at all. When combined with the ensemble, though, it becomes clear that the melody
consists primarily of upper extensions to harmonies. Figure 5.23 gives the same passage from the
second A section along with the piano and bass parts. On the downbeats of the rst and third
bars, bassist Sam Jones plays a D f , and pianist Hilton Ruiz plays an identical voicing. Combined
with the knowledge of standard Rhythm A sections, we can be relatively condent that the
harmony here is the tonic D fmaj7; Colemans opening gure in ascending fourths is then
understood as outlining the 13th, 9th, and 5th of this chord.
26. This focus on melodic upper extensions becomes more common after the bebop era; Robert Hodson
discusses this (in connection with how players create individual melodic proles) in Interaction, Improvisation, and
Interplay, 4246.
27. The second A section is given here only because there is more activity in the piano, providing more harmonic
information. Because there are three nearly identical A sections in Rhythm tunes, they can generally be exchanged
freely. In this analysis, only one A section is usually given, but the reader is encouraged to compare the corresponding
locations in the other two A sections.
28. One of the hallmarks of later (post-1960) jazz is the increase of the use of perfect fourths, especially in piano
voicings; see Dmitri Tymoczko, A Geometry of Music: Harmony and Counterpoint in the Extended Common Practice
(New York: Oxford University Press, 2011), 35760 for an overview of this practice.
170
bb j
& b b b n n n b b b n n n b b # n # n b n b n n #
arpeggios: GM GfM Am Af Dmaj7 Dfmaj7
b j
& b bbb n n n b b b n n n b b # n # n b n b n n #
bb
& b b b b n # n j # n n
b n b n n b n n
n
? bb b n n
bb J
? bb b b J
bb
n b n n n b b
9
Figure 5.23. The first four bars of the A section of Lo-Joe, with ensemble (mm. 912).
171
The third-bar tonic is preceded, as usual, by a iiV progression, but the melody over the A f7
contains the pitches En, An, and Bn. These pitches can be understood as the s5 (or f13), f9, and s9,
of the chord, which are all representative extensions of the altered chord (hereafter, 7alt).
Depending on our analytical priorities, we might analyze this harmony as A f7alt (in which case the
extended tones are rst-class chord members) or as a diatonic A f7 in which the melody notes were
part of an outgoing scale choice: A f7, 4s, Lyd. dim.. In either case, understanding the A f7
helps to understand the harmony in the previous bar, which is a B f7, again with s5 in Ruizs
voicing and f9 in the melody (and, in section A3 , with s9 in the voicing). The rst two bars of the
A section, then, contain a standard IVIiiV progression, with both dominant sevenths played as
7alt chords.
The next two bars are perhaps the most dicult in the entire piece; Figure 5.24 provides these
two bars from all three A sections of the head. While the rhythm section pitches are mostly
consistent, the harmony is not so clear. The starting and ending points are stable, and correspond
with ordinary Rhythm changes, with D fmaj7 in bar 3 of the A section moving to D f7 in bar 5.
Beat three following the D fmaj7 seems to be Bm7 (arpeggiated in the melody), but at this point
Ruiz seems to double the harmonic rhythm, playing four chords in the last bar of this passage.
Joness bass line is also very consistent here, but it is unclear whether this acceleration of the
harmonic rhythm is real or only a surface elaboration.
Figure 5.25 gives several interpretations of this passage. All three of the interpretations begin
with D fmaj7 and end with D f7, with Bm7 on beat 3 of the rst bar. Letter a conforms most
strongly with the melody, but the progression is unusual: there is a chromatic slipping eect from
Bm7 to B fm7 that pregures the chromatic motion to the tonic via a tritone-substituted
29. Recall that the altered chord has a specic meaning in jazz, as a dominant that includes both s9 and f9
(along with other possible extensions). To avoid confusion, and because they are prevalent throughout Lo-Joe, in
this analysis the altered chord will be called 7alt.
30. The unusual signature designation here is an artifact of Russells categorizations: normally dominant chords
have the parent Lydian tonic a whole-step below their roots (the parent Lydian tonic of G7 is F), but for 7alt chords
the Lydian tonic is a half-step above (the parent of G7alt is A f ). Here, the chord is eectively A f7alt (given the En,
An, and Bn in the melody), which Russell would analyze as part of the A Lydian diminished scale
(ABCnDsEFsGs, or enharmonically starting on A f : A f AnBnCDsEnGf).
31. During the solos, there there are consistently only two chords in this bar, but then again, Lo-Joe could be a
tune in which there are separate sets of head and solo changes.
172
Dfmaj7 Df7
b j
& b bbb b b # n # n b n b
n n
# b
b n
b
& b b b b b # n #
n
n bn n n
n
b bb
? bb #
A1 bbb
? bb n n b n b n
bbb n n b
b j
& b b b b b # n
b n
n n
n b bb
n
? bb n
A2 bbb J
? b b b
bbb n n n b
b J
b
& b b b b # #
n
n
n bn n n
n
b bb
? bb
A3 bbb
? b b b n b
bbb n n b J
Figure 5.24. The third and fourth bars of all three A sections of Lo-Joe. The melody is shown on the
top sta, and each grouping of three staves below contains piano and bass parts. Each group ends
when Joness bass reaches the tonic Df.
173
bb j # n n
b b b n # n
& b b b n b # b
a) Dfmaj7 Bm7 Bfm7 D7 Df7
dominant, D7D fmaj7. Letter b is a hearing with doubled harmonic rhythm that follows the bass
line. The progression here makes more harmonic sense, featuring mostly iiV progressions,
though the melodic support for some of the chordsthe A7 and both E7sis weak at best.
Letter c focuses on the piano line, returning to two chords per bar. In this hearing the top notes of
the voicings in the second bar are heard in both cases as s9 moving to f9: a logical hearing, but one
that is not strongly supported by the melody or bass line. None of these hearings seem to t the
music perfectly, but each does t some aspect of it. As we have seen before, it is of little analytical
use to decide on a single true analysis, though it does have consequences for our imaginary lead
sheet author (who must put a set of changes with this melody). This may be a passage in which a
lead sheet is not sucient; no single xed interpretation can adequately capture the essence of this
harmonic motion. Only in their interaction (and in the recording itself ) does a full picture of the
harmony emerge.
The second half of the A section is much simpler, though the rst A section ends dierently
than the other two. (This is typical of Rhythm tunes, and would probably be notated on a lead
sheet as rst and second endings, as in Rhythm-a-ning in Figure 5.9.) The fth bar of the A
section contains a D f7 chord, which moves to the subdominant G fmaj7 in the following measure.
The second half of m. 6 moves to some kind of C f chord, though Ruizs piano voicings in the head
are unhelpful in determining its quality (in the solos it is usually played as a dominant seventh).
174
Fm7 E7 Efm7 D7
b n n b
A1 & b bbb n # b n n
n
Df7 Gfmaj7 Cf7
b
& b bbb
n b A
b
Fm7 Bf7 Efm7 Af7 Dfmaj7
b n b
A2, A3 & b bbb n b n
The rst A section then moves to a tritone-substituted turnaround, while the other two double the
harmonic rhythm to arrive on tonic in the eighth bar (see Figure 5.26).
The bridge of Lo-Joe is its most distinctive feature, and is given with chord changes in
Figure 5.27. This bridge is clearly inspired by the last half of the bridge of Eternal Triangle
(discussed in the next section), and its sequential nature is helpful to our imagined lead-sheet
author: once a single bar is determined, it can simply be transposed to all of the others. Here, each
bar contains a single iiV progression, made explicit in the bass and with basic three- and four-note
voicings in the piano.
This bridge is phenomenologically rich: the iiV progressions themselves are clear, but the
connections between them admit of multiple possibilities (as a listener, or for an improvising
musician). Figure 5.28 gives several possible transformation networks for the rst half of bridge of
Lo-Joe. The analysis at a is a horizontal one: do a iiV, move up a tritone and do another,
move down a half-step and do it again, and so on. Letter b emphasizes the descending fth
motion: play a iiV and its tritone transposition, then move down a fth and repeat. This hearing
respects the descending fths present in the standard Rhythm bridge, and also reects the
organization of iiV space, shown (in letter c). Letter d highlights the tritone relationships
between bars, and also encourages hearing the 3rd transformation connection between F7Fm7
32. This progression is given (in the key of B f ) in Aebersold, I Got Rhythm Changes in All Keys, 27.
175
b b n
Cm7 F7 Fsm7 B7 Fm7 Bf7 Bm7 E7
bb
&bbb n n n n # # n n J n # n
#
17
J #
b b
Bfm7 Ef7 Em7 A7 Efm7 Af7 Am7 D7
bb b b b n j b n J n n
&bbb n n # n
n
21 n
a) a horizontal reading
T6 T11 T6
Cm7 F7 Fsm7 B7 Fm7 Bb7 Bm7 E7
Figure 5.28. Several possible transformation networks of the bridge (first four bars). Unlabeled
arrows indicate the TF transformation.
176
and B7Bm7 chords. Hearing the bridge this way allows hearing as if the music is bouncing back
and forth between two normal Rhythm bridges, one in D f , the other in G. Network e emphasizes
the half-step relationships, and encourages a connection between the rst and last pairs of chords
and the central two.
It may seem as though transformations themselves have not played an important role in
reconstructing a lead sheet for Lo-Joe. This reconstruction, though, has taken place against the
background of the musical spaces developed in the earlier chapters of this study (all of which have
transformations as their logical basis). These musical spaces provide a mostly-unseen structuring
principle to the analytical work in this section. Because the spaces were developed to demonstrate
functional jazz harmony, to recognize that Lo-Joe uses functional harmony is to recognize that it
likely reects an orderly representation in (say) iiV space. This, combined with the knowledge of
Rhythm changes in general, means that we could easily reject the triadic analysis in Figure 5.22 as
a nonsensical jazz progression. This notion of syntax is one that is often implicit in the
construction of the spaces, but comes to the fore when used for the kinds of harmonic
determination done here.
The nal Rhythm tune of this chapter was the inspiration for the bridge of Lo-Joe: Sonny Stitts
The Eternal Triangle. The canonical recording appears on Dizzy Gillespies album Sonny Side Up
(1957), featuring Stitt along with Sonny Rollins, both on tenor saxophone. The album is widely
regarded as one of the best jam session albums in jazz, and Eternal Triangle is often singled out
as the standout performance of the record. This two-tenor format will allow us the opportunity
to explore more deeply the role of interaction between players in shaping harmony.
33. It is also a concept that is absent from other prevalent theories of jazz harmony: it is easy to imagine an
analysis in which the arpeggiated triads and seventh chords of the melody of Lo-Joe form coherent voice-leading
strands to and from structural tonic chord members, for example.
34. See, for example, Stephen Cooks review on AllMusic, where he notes that on The Eternal Triangle, in
particular, Stitt and Rollins impress in their roles as tenor titans . . . an embarrassment of solo riches comes tumbling
177
First, though, a brief analysis of the tune itself is in order. The head of Eternal Triangle is
shown in Figure 5.29; the A sections are standard Rhythm changes, featuring fast-moving bebop
melodic lines. The B section, though, is unique to this tune, and features iiV progressions
descending by half-step. We might imagine this bridge as being derived from the standard Rhythm
bridge, as shown in Figure 5.30. In the rst step, the typical IIIVIIIV is compressed into the
second half of the bridge. To preserve the correct length, Stitt extends the fths cycle backward by
two chords to E7, maintaining the original harmonic rhythm of one chord every two bars,
increasing the harmonic work done by the bridge (and consequently, the area of iiV space it
traverses). In the next step of the derivation, each dominant seventh is replaced by a iiV
progression; nally, every other iiV progression is replaced with its tritone substitute, resulting in
the chromatic descent of the bridge itself.
Because the A sections of Eternal Triangle use typical harmonies, both Rollinss and Stitts
solos display many of the same solo approaches we saw in Grins solo on Rhythm-a-ning above.
Both players use harmonic generalizations of various types: diatonic (choruses 2A3 , 5A3 , 8A2 , and
11A1 , for example); blues (4A1 , 9A2 , 11A3 ); and other scales (the half-whole diminished scale in
9A1 and 12A2 ). Stitt in particular emphasizes the half-note harmonic rhythm of the A sections,
often playing bebop lines that change accidentals frequently to highlight the harmonic shifts (see
choruses 6A3 and 10A2 , for example). Many other common harmonic devices can also be found,
including tritone substitutions (mm. 133) and cesh (mm. 3335, 193, and 39596), in which a
chromatic melodic line embellishes a single unchanging harmony (see p. 147n89).
One harmonic aspect of Rollinss solo does deserve special mention, as it has not yet appeared
in the analytical examples. Figure 5.31 shows the fth and sixth bars of an A section, where we
would normally expect the harmonies B f7E f7. Here, Rollins clearly arpeggiates a Bm7 instead;
this is a feature which is often called side-slipping or side-stepping. The overall harmony of
this bar is B f7 (sometimes with its preceding ii7 ), but here Rollins plays a harmony a half-step
out of both these mens horns (Stephen Cook, Review of Dizzy Gillespie, Sonny Side Up, AllMusic.com, accessed
July 13, 2015, http://www.allmusic.com/album/sonny-side-up-mw0000188698).
35. Jerry Coker denes side-slipping as deliberately playing out-of-the-key for the sake of creating tension
(Elements of the Jazz Language for the Developing Improvisor [Miami: Belwin, 1991], 83).
178
A1/A2 Bf Gm7 Cm7 F7 Dm7 G7 Cm7 F7
b 4 . j n
b
& 4 .
3
b b # b n b b
Fm7 Bf7 Ef7 Dm7 G7 Cm7 F7 Cm7 F7 Bf
& b .. n b .. .. b
J
B Bm7 E7 Bfm7 E f7
b n #. b b b
&b n
J
J
b b # b n b b
Fm7 Bf7 Ef7 Dm7 G7 Cm7 F7 Bf
& b .. n b
179
standard bridge D7 G7 C7 F7
compressed,
extended backward E7 A7 D7 G7 C7 F7
tritone
substitutions Bm7 E7 Bbm7 Eb7 Am7 D7 Abm7 Db7 Gm7 C7 Fsm7 B7
Figure 5.30. Derivation of the bridge of Eternal Triangle from a standard Rhythm bridge.
bb # n b b n
&
Bm7
180
b n b
Bm7 E7 Bbm7 Eb7
b # # j b J b
&b # # n
113
3
3s, Dia. 4f, Dia.
Rf 7, 0 Rf 7, 0
n
Am7 D7 Abm7 D b7 Gm7 C7 Fsm7 B7
b n n b b b b N n # # n # # n # n
(f9)
& b
117
n
s, Dia. 6f, Dia. f, Dia. 4s, Dia. 4s, Lyd. dim.
Rf 7, 0 Rf 7, 0 Rf 7, 0
Re, 2
Figure 5.32. A non-sequential bridge from Sonny Rollinss solo (mm. 11320, 2:12).
away. The motion from B f7 to Bm7 is distant in iiV space, though is closely related to the slide7
7th 1 slide7
transformation (B f7 Bm7). It is also a convincing way of playing outside, which
is what jazz musicians call improvised lines that do not seem to connect with the underlying
harmony. Outside playing becomes an important feature of more modern jazz improvisations, but
it also features prominently in this recording when Rollins and Stitt begin trading (a section to
which we will return below).
Given that the bridge of Eternal Triangle is its most interesting feature, it will be
worthwhile to examine the solo strategies of Rollins and Stitt separately before moving on to
discuss how the two interact with each other. These strategies can be broken down into two basic
types: sequential and non-sequential solo approaches. The non-sequential solo approach to this
bridge is less common, and a single example should suce. Figure 5.32 gives the bridge from
Rollinss fourth chorus; because the harmonies are so fast-moving, he uses mostly diatonic scales.
Though the melodic line is not sequential, Rollinss shifting diatonic palette works in counction
36. Dmitri Tymoczko notes that side-slipping usually demonstrates ecient voice-leading; he also emphasizes its
role in creating polychordal or polytonal eects; see A Geometry of Music, 37478.
37. I suspect that the outside terminology is related to Russells outgoing scale choices, but I have not been
able to corroborate this suspicion.
181
Am7 D7 Abm7 D b7 Gm7 C7 Fsm7 B7
b n n b. b b #. j
a) &b J
N # #
85
bb . n # b. b b n. n n # . # n #
J
b) &
149
Figure 5.33. Two sequential bridges from Rollinss solo (mm. 8588, 1:48 and mm. 14952, 2:42).
with the chromatically descending iiV progressions; the chord-scale analysis highlights that the f7
signature transformation is equivalent to T11 .
Because the bridge is made up of these chromatically descending iiVs, the most obvious
approach for an improviser is to play sequentially, repeating a single pattern over each iiV
progression. This approach is seen most often in the second half of the bridge, where the harmonic
rhythm doubles; Figure 5.33 gives two examples of this strategy from Sonny Rollinss solo. In a,
from the third chorus, Rollins plays the same quarter-note pattern in all four bars, highlighting the
T11 s of the progression itself. Letter b, from the fth chorus, is more complicated, but the basic
idea is the same. Here, Rollins begins on the ninth of the minor seventh chord and descends to the
third of the dominant seventh (diatonically, in the key of the dominant). The rhythm here is more
varied, and linking material is inserted in the third bar, but the sequential pattern is still clear.
38. That f7 = T11 is one of the properties of signature transformations; see Julian Hook, Signature
Transformations, in Music Theory and Mathematics: Chords, Collections, and Transformations, ed. Jack Douthett,
Martha M. Hyde, and Charles J. Smith (Rochester: University of Rochester Press, 2008), 14244.
39. This repeating pattern is one of what are sometimes called digital patterns. Though these patterns are more
common at faster tempos (and in eighth notes), this is a standard 5321 pattern, where the numbers indicate scale
degrees of the minor seventh chord. Jerry Coker gives the passage at Figure 5.33a as an example of digital patterns in
Elements of the Jazz Language, 11.
182
metric structure: 3 3 3... +2 5
# n n b b b n n b n
bb n n # n # # n # # n
&
213
D bebop Df bebop C bebop B bebop
Figure 5.34. A sequential bridge with metric shift, from Sonny Stitts solo (mm. 21316, 3:34).
Stitt seems to prefer sequential bridges more than Rollins does, but Stitt often introduces a
metric shift as well; Figure 5.34 gives a representative example from chorus 7 (similar passages can
be found in choruses 9, 10, and 12). Here, Stitt plays a descending bebop scale over every dominant
seventh chord. The rst two of these are three-beat patterns, creating a metrical grouping
dissonance (G3/4). When Stitt repeats the pattern on the C bebop scale, it should end on the
downbeat of the third bar. As if realizing he has arrived too earlyafter all, the C7 chord does not
begin until the third barStitt extends the scale another two beats, resulting in a ve-beat scale
that descends an entire octave. With ve beats remaining in the bridge after the end of this
pattern, he repeats it using the B bebop scale; a nal An in the last beat acts as an enclosure to the
tonic B f that begins the next A section. While the harmonic rhythm of Eternal Triangle
normally doubles in the last half of the bridge, Stitts frequent metric shifts accelerate it even
further, giving his solos even more momentum into the nal A section of each chorus.
Though the strictly sequential patterns are mostly restricted to the last half of the bridge, the
rst half can also support sequential patterns; Figure 5.35 shows an example from the eighth
chorus. Because the harmonic rhythm is slower at the beginning of the bridge, the sequences are
usually somewhat looser. Here, Stitt plays a decorated Bm7 arpeggio, followed by the same pattern
a bar later over B fm7. (While in the passage in Figure 5.34 Stitt ignored the ii7 chords, here he
seems to ignore the E7 instead.) The pattern breaks in the middle of the third bar, where Stitt
moves toward a standard diatonic pattern for the nal E f7 chord.
40. On metrical dissonance in jazz, see especially Stefan Love, Subliminal Dissonance or Consonance: Two
Views of Jazz Meter, Music Theory Spectrum 35, no. 1 (Spring 2013): 5254 and throughout.
183
# # b n b b b n b
Bm7 (E7) Bbm7 Eb7
b
&b n # b
241 3 3
Figure 5.35. A loosely sequential bridge from Stitts solo (mm. 24144, 3:57).
With a few exceptions (including Lo-Joe in the previous section), the analyses so far have been
interested primarily in a soloists improvised line, paying only passing attention to the fact that
these improvisations occur within a framework of group improvisation. Sonny Rollinss solo on
Eternal Triangle, for example, does not take place in isolation; he is supported by the rhythm
section, and in later choruses he and Stitt trade, alternating 4- or 8-bar segments of improvised
melodies. Understanding this interaction is crucial to understanding the performance as a whole.
This section will focus on these moments of interaction in Eternal Triangle, acknowledging the
role that interaction plays in harmony, and vice versa.
One of the rst models of interaction in jazz is found in Robert Hodsons Interaction,
Improvisation, and Interplay in Jazz. He uses a semiotic model borrowed from Jean-Jacques
Nattiez, in which a work of music is both a product (a score/sound) and a set of processes: both
the poietic process of composition and the esthesic process of a listener. Hodson adapts this
model for jazz performance, since a jazz musician simultaneously creates the sound and listens to
the other band members sounds. The two separate components of poiesis and esthesis in Nattiezs
model form something of a feedback loop for jazz musicians, since their musical utterances are
often shaped by those of their fellow musicians (while at the same time, potentially inuencing
those musicians themselves).
184
Garrett Michaelsen, in his dissertation on interaction in jazz, critiques Hodson on this model.
He argues that while Hodsons focus on the ensemble as a whole (via the esthesic process) is
valuable, Hodson does not go far enough. In his analyses, Hodson places himself in an improvisers
perspective, and as Michaelsen notes, it is . . . unclear how this vantage point enables musical
analysis because it leaves no room for the outside observers non-poietic perspective. Instead,
Michaelsen oers a listener-based approach, in which a particular auditory stream (a bass line, for
example) might be heard as inuencing another (like an improvised saxophone line). As with
harmony, there is often not a single correct analysis of a given interaction, and so Michaelsens
listener-based approach ts nicely with the prismatic approach to harmony taken here.
Michaelsen discusses harmony only in passing (usually in connection with a particular musical
example), while Hodson dedicates an entire chapter to the role of interaction in harmony. Many
of the questions that concern Hodson are the same as those we have confronted during the course
of this study:
How can one reconcile the disparity between dierent versions, both written down and
performed, of the same harmonic progression? Does it even need to be reconciled?
Some scholars criticize the eort to reconcile these variants as an attempt to force a
Western ideology of coherenceand a modernist ontology of the pieceonto a music to
which it doesnt really apply. But, if this kind of [harmonic] coherence is not a part of
jazz, then why do jazz musicians talk about a soloist making the changes, or an
improvised line as either making sense or not? There must be some criteria for musical
coherence.
Hodson answers these questions by borrowing a linguistic metaphor from Noam Chomsky: he
argues that jazz musicians play the deep structure of a tune, which might be realized in any
number of waysand can be revealed by analyzing musical interaction. Rhythm tunes, he argues,
can be generated from the deep structure of I Got Rhythm: its A sections consist of
43. Michaelsen, Analyzing Musical Interaction, 2425.
44. Ibid., 3638 and throughout. Michaelsens work draws on Albert Bregmans schema-based stream segregation,
while acknowledging that auditory streams in jazz are more exible than Bregmans own; see Albert S. Bregman,
Auditory Scene Analysis (Cambridge, MA: MIT Press, 1990).
45. The closest Michaelsen gets to an outright discussion of harmony comes in his discussion of interaction with
referents: he notes that chord changes are inherently open-ended expressions of harmony that permit a wide variety
of possible chord voicings, and that dierent referents will motivate projections of varying specicity (Analyzing
Musical Interaction, 9091).
46. Hodson, Interaction, Improvisation, and Interplay, 53.
185
prolongations of B f , while the bridge begins o tonic and contains a motion back towards it. This
appeal to linguistics does provide some way of understanding the myriad of Rhythm
harmonizations, but it leaves something to be desired: it is not as though any set of chord changes
can appear in the Rhythm A section, provided it starts and ends with B f . Again, the
transformational approach to harmony developed here allows us a means to specify the ways in
which this deep structure is modied, and an interactional analysis of Eternal Triangle seems a
natural way of exploring these modications.
As a rst step in that direction, consider Figure 5.36, from the end of chorus 11. In the nal A
section of the chorus, Stitt plays a very strong blues generalization. Though this is a common
choice (especially in A3 sections), this particular occurrence is marked by the strong emphasis on
the blues in the rhythm section parts as well. After the fast-moving harmonies of the bridge (the
end of which is given in the transcription), the groups convergence on eight bars of blues has a
striking eect.
The instigating factor for this blues generalization might well have been Tommy Bryants
decision to play a tonic pedal in m. 345. In a Rhythm tune like Eternal Triangle, the constant
half-note harmonic rhythm can become tedious, and a pedal point is one of the most eective ways
a bass player can counteract this tendency. This is the rst time in the 5 elapsed minutes of the
recording Bryant uses such a pedal, and Stitt and Bryants brother Ray on piano are very likely to
have noticed.
Stitt responds to this tonic pedal by playing an emphatic blues lick with a prominent A f .
This pitch, f7, is unlikely to occur over any of the common harmonies of the rst bar of the
Rhythm A section (except perhaps as a f9 over G7), and its repetition as a long note in the second
bar cements its status as a member of the B f blues scale. Pianist Ray Bryant, hearing the bass pedal
47. Charlie Persips drums are not included in the transcription, since they do not contribute to the harmony.
48. Throughout these interactional analyses, I follow Michaelsens own use of language, which relocates a
listeners perception into the player. Thus, Stitt responds to this tonic pedal should be understood as I, the
listener/analyst, hear Stitts musical utterance as a response to the tonic pedal of the bass. This language, though
potentially confusing, is in general clearer and easier to read. See Michaelsen, Analyzing Musical Interaction, 4647.
186
Am7 D7 Abm7 D b7 Gm7 C7 Fsm7 B7
b n n b b b n # #
n b #
Stitt
(tenor)
&b b
3
b j
& b n j # b b b
jn j
n
# n #
Ray Bryant
? b b J b j
J #
(piano)
b n
J
Tommy
? bb
b
b # n
Bryant
(bass)
341
b b. . b b n b #
11A3
& bb J
3
b
&b
b n ww # n b
? b b b ww b
b n
? b
b b
345
bb b b b J .
12A1
.
&
b
&b
#
? b b w
b
>
? b n
b
349
Figure 5.36. The end of chorus 11, including rhythm section (mm. 34153, 5:19; bass sounds as
written).
187
along with Stitts blues lick, then launches into a series of blues voicings. Combined with the
tonic pedal, Bryants diminished chords and neighboring 64 chord of m. 348 give the section a
strong blues feel that both reinforces and is reinforced by Stitts solo line.
This group interaction is what Michaelsen would call strongly convergent, in that all three
of the members involved play music that supports the others. The convergence continues in the
last four bars of the section, as Tommy Bryant moves away from the tonic pedal to outline the
blues-inected progression B fB f7/DE fE 7F7. In m. 351, he arrives on a dominant pedal at
the same moment that Stitt also concludes his line on F, all reinforced by Ray Bryants strong
left-hand F in the piano. This convergence on the dominant (combined with Charlie Persips
drum ll) provides a strong push into the following chorus, where the three fall back into their
usual, less convergent, roles.
The most clearly interactional moments of Eternal Triangle come between the two
saxophonists themselves after the trading begins in chorus 14. Stitt and Rollins trade fours for
three choruses, then trade eights for three more to end the saxophone soloing. The fours supply
many good examples of harmonic interaction; chorus 15A2 (shown in Figure 5.37) gives a
representative example. Here, Rollins plays a diatonic gure in the rst two bars, then side-slips in
the next two to play a gure that outlines B major rather than B f . Stitts response is at once both
convergent and divergent: he enters on the same pitch that Rollins did and plays a very similar
gure in his rst bar (marked with an x in Figure 5.37), but his starting pitch is Fn, dissonant with
the B major triad Rollins is playing at the same time. The last A section of the same chorus
(shown in Figure 5.38) illustrates a more convergent interaction. Here, Rollins begins with a
motive outlining the pitches GG f FD. Stitt seamlessly picks up this line on the downbeat of
49. At the same time, Bryants piano voicings project some continuity with the end of the bridge, since he
continues moving his right hand in parallel thirds.
50. In Michaelsens terminology, the three auditory streams project similar futures (Analyzing Musical
Interaction, 60). The particulars of auditory streams and implications need not concern us too much here, since it is
relatively clear that all three members project the blues.
51. Incidentally, the arrival on F in the piano is perhaps unusually contrapuntal: it seems as though Bryants voicing
in m. 350 is dictated by the lower-voice motion GG f F, while the upper voice Cs demands resolution to D, fullled
on the downbeat of the next chorus. These contrapuntal motions all combine with the bass En to produce a German
seventh chord in the second half of m. 350a fairly typical predominant in classical music, but extremely rare in jazz.
188
m. 477, continuing the motive for another four bars, at which point Rollins reclaims it for a bar
before moving on to new material in the beginning of the next chorus.
bb # n # # n . #
x
j b n #.
[S.S.]
& J J
457 [S.R.] 5
bb n
3
&
461
Figure 5.37. A harmonic interaction between Rollins (red notes) and Stitt (blue notes) from chorus 15
(mm. 45763, 6:52).
bb b # . b J # # b n n b # b
[S.R.]
&
473
b b b b b
[S.S.] [S.R.]
& b b b j b
477
b
b
& b # b
481
Figure 5.38. A convergent interaction from later in chorus 15 (mm. 47381, 7:05).
189
5.4.3 Extended Analysis: Trading Eights, Harmony, and Interaction
For the nal analysis in this dissertation, I want to take a more detailed look at the way harmony
functions over a longer period of time, using the nal three choruses of saxophone soloing on
Eternal Triangle. These are the choruses where Stitt and Rollins trade eights; since each soloist
gets an entire section of a chorus, he has more harmonic leeway than in the rapid four-bar
segments of the preceding three choruses. Since these are the nal three choruses of their solos,
this portion is also where the saxophone dueling comes to a head. In it, the suggestions of
outside playing initially suggested in Rollinss very rst chorus (recall Figure 5.31) reach a nal
realization, before winding down again in the nal chorus.
The eight-bar trading begins at the beginning of chorus 17, after a non-musical interaction:
someone (probably Stitt) can be heard on the recording asking keep going? Stitts rst eight bars
are a typical diatonic A section, with only incidental chromaticism. Rollinss response in chorus
17A2 is more outside, and is shown in Figure 5.39. The most blatant chromaticism here is in the
second bar, where we would normally expect Cm7F7 or Cm7C s 7. As is usually the case with
outside playing, it is not clear how we should interpret Rollinss note choices here, though the
chord-scale triples below the sta give a few suggestions. After this second bar he returns to more
inside playing, though with a few more outgoing scale choices than usual: he plays both s5 and f9
over the F7 in m. 524.
Bf Gm7 D7f9? G7 C7 F7
b n # n #
&b # n J # J n n b # b b
521
Figure 5.39. Rollinss outside playing in the second A section of chorus 17 (mm. 52124, 7:44).
190
bb # n n # n n b # n
E7 A7? Dm7 G7 Cm7 F7
& J # #
537
A7, 2s, Dia. side-slipping
b b
Bf7
local
4
&b J J
541
f13 n13
chord-scale ambiguity
Figure 5.40. Rollinss increased chromaticism in the final A section of chorus 17 (mm. 53742, 7:56).
Stitts ensuing bridge is again typical, using only the expected diatonic collections. That he
chooses not to play outside on the bridge illustrates a basic principle of jazz harmony, in that
progressions that are less commonplace are generally played more inside. The bridge of Eternal
Triangle is its most distinctive feature, so soloists typically play improvisations that highlight these
harmonies, whereas the A sections of the tune are more typical and admit of greater elaboration.
Playing far outside the changes on the bridge risks obscuring the harmonic progression that is an
essential feature of this particular Rhythm tune (and perhaps, in an amateur performance, giving
the impression that outside playing will be mistaken for not knowing the changes).
In the nal A section of the chorus, Rollins increases the chromaticism even further, as shown
in Figure 5.40. His rst bar seems to imply a motion from E7 to A; while we might understand
the E harmony as a tritone substitution for the tonic B f , the A chord in the second half of the rst
bar does not make sense as an ordinary substitution in any of the usual harmonizations of the A
section. Given that the A7 harmony seems to continue into the second bar, we might instead hear
this outside playing as a downward side-slipping, substituting an A diatonic collection for the tonic
B f . Side-slipping is one of Rollinss preferred methods of playing outside: we observed it in his
rst chorus, and he repeats the technique in chorus 14A3 (mm. 44142). Rollins returns to more
52. A similar phenomenon occurs in the nal turnaround of Joe Hendersons Isotope, discussed in Section 4.3.3.
Because the turnaround is so distinct, Henderson simply arpeggiates the harmonies in most of his improvised choruses
to make them as clear as possible.
191
typical harmonies after these rst two bars, though he still emphasizes dissonant tones: m. 540, for
example, features strongly accented dissonances, with the local 4 (Levines avoid note) appearing
on beats 1 and 3. All of this chromaticism combines to form an improvised line even more outside
than Rollinss rst eight bars of trading, as though his own sense of harmony is being slowly
detached from that of both his own rhythm section and from the A section of the tune itself.
In the beginning of the next chorus, Stitt seems to take the outside-playing bait, launching
into a dominant-cycle A section very much like Monks solo on Rhythm-a-ning. This harmonic
move seems to take the rhythm section by surprise, and draws our attention to the interaction not
only between the two soloists, but between them and the rhythm section as well (a transcription
including the piano and bass parts is given in Figure 5.41). Faced with this unexpected dominant
cycle, pianist Ray Bryants solution is simply to stop playing, while bassist Tommy Bryant instead
plays a B f pedal, as if to stress the tonic in the midst of Stitts cycle. Since the 8-chord dominant
cycle lasts only four bars, all three band members all return to a typical A-section harmonic
structure in m. 549.
Perhaps anticipating another outside response from Rollins, Ray Bryant does not immediately
begin playing in the next A section, and Tommy Bryant opts this time for a dominant pedal. In
the face of harmonic uncertainty, this approach from the rhythm section makes sense: a dominant
pedal in the bass will work to build tension no matter what Rollins decides to play, and Ray
Bryants wait-and-see approach prevents any harmonic clashes. As it turns out, Rollins does play a
non-diatonic line, and again it is not entirely clear what harmonic framework he has in mind. His
line in the rst four bars of chorus 18A2 is a loose, descending, motivic repetition that recalls the
C.T.A. harmonization of Figure 5.2f (with its whole-step descent). While the motive and
direction of the line are relatively clear, we might also hear this line, in counction with Ray
Bryants comping in the last two bars, as a series of outgoing scale choices on a more standard
progression, as shown in Figure 5.42. Like Stitt, Rollins returns to a more diatonic approach in
the last four bars of the section, setting up the way for Stitts bridge.
192
b n # # n n
18A1 8:03
Tenor Sax b
& # # # # # n # n n n b
[S.S.]
b
&b
? bb
Piano
? b
Bass b
545
bb n b b b
& b b
b
& b w
w
? b
b n
? b
b n
549
j n # n
3
bb A # b
18A2 [S.R.]
n. b .
& J J
3
b
&b j j
? bb J b J n b
? b
b
553
Figure 5.41. Transcription of chorus 18, including rhythm section (mm. 54576, 8:03).
193
bb N
&
b
&b
# n
j
.
? b b ..
b
n
? b
b
557
b
bb nn # n #
18B
# n # n b n b
&
[S.S.]
b j
&b # ww b b n www
b
# w
? b
b
n.
? bb n n b #
561 n b
b
n b b b
# n n b
&b
b
& b n
b b #
? b # b n n
b n
? b b
b # b b n # # n n
565 b
194
j
3
n b b
18A3 [S.R.]
b J J J J n b # b
&b
b
& b n # n
j
# n # # n #
? b # n b
b # n n b
n J
? b
b b n n # n
569
b b
&b
b n b
b
&b
b
w
b
? b
b
.
? bb
573 n
Bf C7 F7 Bf7
Re, -2 R f1, 1
n # A n n. # b b
bb J .
f, Whole-tone f, Lyd. dim. 2f, Lyd. f7
Rollins & J J
b
T10 T11 T11
R. Bryant
? b
b J J n b b ...
Figure 5.42. One possible interpretation of Rollinss descending line (mm. 55356, 8:09).
195
expected: Bf Cm7 F7 Bf
side-slipping
Figure 5.43. Stitts side-slipping at the beginning of chorus 19 (mm. 57779, 8:29).
The bridge of chorus 18 proceeds normally in all three instruments: Stitts scale choices are
almost completely diatonic, Ray Bryant uses standard 3-note voicings, and Tommy Bryant plays a
bass line emphasizing the T11 root motion. While the bridge of Eternal Triangle usually
functions as the locus of harmonic activity in a chorus, the bridge here has an almost calming
eect after the harmonic disruption in the rst half of the chorus. The tension of all of the outside
playing seems to disappear as all of the band members (and we as listeners) relax into the bridge,
relatively condent that it will progress as expected.
This harmonic tranquility does not last long, though, as Ray Bryant instigates another
dominant cycle at the beginning of chorus 18A3 . He enters emphatically on an F s7 chord, and
includes the bass note (typically omitted by pianists) in his left hand, as if to demonstrate that he
understood Stitts cycle in 18A1 and is willing to support it for this A section. His brother Tommy
catches the cycle almost immediately: after playing a B f on the downbeat of m. 569, he makes his
way to a Bn by beat 3, and continues with the cycle all the way through the rst four bars. This
time, though, Rollins does not follow along, playing the B f Lydian scale for the rst two bars. Ray
Bryants enthusiasm for the dominant cycle fades quickly, hearing that Rollins does not follow
along; by the third bar the piano voicings are nearly inaudible, returning only in the nal bar
(where C7F7 is a characteristic choice regardless of the particular harmonization used).
After a more common diatonic ending in the last four bars of chorus 18, Stitt begins the next
chorus with a side-slipping gesture, arpeggiating a C sm7 chord before the expected Cm7 in the
196
b J . b j j .
19A2 [S.R.]
[S.S.]
& b J J .
Tenor
J
b ! Bmaj7
& b
n # # n # # n
Piano
? bb
Bass
n # # # n b b
583
2f, Dia. 5s, Dia. 2f, Dia.
Rs7, 0 R f 7, 0
Figure 5.44. Ensemble arrival on Bmaj7 at chorus 19A2 (mm. 58388, 8:33).
second bar of the chorus (see Figure 5.43). After this brief moment, though, Stitt plays more or
less diatonically until the end of this A section.
At the beginning of chorus 19A2 , the side-slipping occurring sporadically throughout the
saxophone solos reaches its culmination in a remarkable moment of ensemble convergence. Stitt
ends his eight bars with a tonic triadic descent, ending on B f . At the same moment, both Ray and
Tommy Bryant land on a Bmaj7 chord, turning Stitts tonic B f into a chordal major seventh (see
the transcription in Figure 5.44). This side-slipped B major lasts only two bars, and by m. 587
the rhythm section moves back to B f . For his part, Rollins plays a repeating motive over all eight
bars of this section consisting of the pitches E f , D, and B f . Over the Bmaj7, this lick emphasizes
the major third and seventh of the chord, but when the rhythm section slips back to B f , Rollins is
left emphasizing the dissonant 4. This approach is certainly not as outside as Stitts dominant cycle
of chorus 18A1 , but the dissonance still lends a sense of outsidedness to the section as a whole.
In the nal bar of his A section, Rollins resolves the E f to D, after which the bridge and nal A
section proceed almost completely diatonically.
53. Given the convergence of all three members of the ensemble on this Bmaj7 chord, it seems likely that this
harmonic move was planned in advance. Though it is impossible to say for certain, the fact that it occurs on the last
chorus of trading provides some support; it is easy to imagine a situation in which the band decides (before recording)
that a move to B major in the second A section will mark the last chorus of saxophone solos.
197
Heard as a unit, the nal three choruses form a progression from inside to outside playing and
back again. Stitts diatonic playing in chorus 17 is challenged by Rollinss suggestions of outside
playing, weakly at rst in 17A2 but more strongly in 17A3 . Stitt takes the suggestion in the
opening of chorus 18, playing a dominant cycle that takes the rhythm section by surprise. Rollins
responds not with a dominant cycle, but with a motivic response that seems almost completely
disconnected from typical Rhythm harmony. The rhythm section tries to anticipate a dominant
cycle in the choruss nal A section, only to be disregarded by Rollins. Amid the harmonic
confusion, Stitt begins chorus 19 with some mild side-slipping before returning to more inside
playing. This side-slipping is answered strongly by the rhythm section, which begins the following
A section a half-step o on Bmaj7. After a mildly dissonant eight bars from Rollins, the bridge
and nal A section return inside to conclude the saxophone trading and make space for Dizzy
Gillespies trumpet solo that follows. The harmony in these choruses is far more nuanced than the
chord changes of the lead sheet might suggest; it is a dynamic, ever-changing attribute, revealed
not only in the interaction among the musicians themselves, but also between the musicians and us
as listeners.
As the end of this study approaches, it may be useful to return to the fundamental question: what
is jazz harmony, really? Or, as Hodson states the problem, what exactly are you analyzing when
you analyze jazz harmony? Does the harmony exist in the chord symbols on a lead sheet? In the
voicings of a pianist or walking line of a bassist? Does harmony exist in a tune itself, independent
of any particular performance of it? If a performers harmony conicts with the chord symbol,
which one is correct?
These are fundamental questions, and questions that do not often arise in the study of notated
music. The answer, I think, is that harmony in jazz is all of these things, and more. A single
harmony can be captured by a chord symbol, but this chord symbol is only part of the story. A
198
transcription of a piano voicing for that symbol does not represent a nal solution either, since in
jazz a repeating harmonic framework forms the basis of a performance, and pianists do not usually
play identical voicings for a single chord throughout. We may be tempted to turn towards the
soloist as the arbiter of harmonic identity, but as we have seen, even a seemingly clear progression
in a solo instrument can be obscured by the choices of a rhythm section.
It is my hope that the transformational approach developed throughout this dissertation is
better equipped to deal with this chimerical nature of jazz harmony. Chapter 1 began by treating
harmony in something of a clean-room fashion, grouping chord symbols into diatonic sets
without worrying too much about what the chord symbols actually represented. Chapters 2 and 3
took steps towards clarifying this nature, representing chord symbols in their most basic form of
root, third, and seventh. This abstraction allowed us to explore connections between harmonies
that do not share a diatonic collection, using variations on a basic musical space of iiVI
progressions. In Chapter 4, our conception of harmony expanded from three-note chords into
many-note scales, as we drew on the work of George Russell to create a chord-scale space. This
work permitted a change of focus from chord symbols and abstract tunes to jazz performance itself,
investigating the way in which harmony can function for improvising performers. This nal
chapter has continued this work by looking through the lens of Rhythm changes, itself a harmonic
archetype, variously instantiated in countless Rhythm contrafacts, each of which is elaborated in
individual performances.
In this last chapter, the transformations themselves receded into the background somewhat,
used only as a tool for discussing harmony in service of other analytical points. This is intentional,
and as Julian Hook notes in his review of Lewins GMIT, transformation theory is a large and
varied toolbox; there are only some minimal instructions for using the tools, and no designs at all
for what one can build with it. For all the focus on harmony, it is often the case that harmony is
not the most interesting aspect of a particular passage. Just as we tell our students that a Roman
numeral analysis does not mark the end of the analytical process, neither does a completed
55. Julian Hook, David Lewin and the Complexity of the Beautiful, Intgral 21 (2007): 166.
199
chord-scale labeling mean that we can check o a passage as analyzed and move on to the next
chorus. This dissertation has developed a exible set of tools for analyzing harmonytools that
can be used as necessary whenever a need to discuss harmony arises.
Given that harmony touches nearly every aspect of jazz performance and analysis, the
foundational work here might be applied or extended in any number of ways. It is easy to imagine
other transformations that might appear more commonly in other (especially later) jazz repertoires,
where either the ordered-triple or chord-scale approaches might be fruitfully applied. Dmitri
Tymoczko has argued that jazz is a modernist synthesis, and that jazz musicians act as custodians
of a tradition of advanced tonal thinking. Though I tend to disagree with this historical view of
jazz, it may well be the case that the techniques developed here might lend insight into the musics
Tycmozko identies on either historical end of common-practice jazz: the tonality of the
impressionists on one side and that of the minimalists on the other.
Even within the common-practice jazz era of this dissertation, there is room for expansion.
The solo analyses done here only begin to scratch the surface, and a detailed investigation of
chord-scale choices among dierent performers might well lead to some meaningful distinction
between, for example, a Johnny Grin solo and a Joe Henderson solo. Certainly every soloist has
their own style, and the way in which they interpret harmony is often an essential component of
that style. It is also easy to imagine that saxophonists in general have a dierent kind of harmonic
language than do trumpeters, trombonists, pianists, or bao players; chord-scale transformations
could help to tease out these distinctions.
The analyses here have focused exclusively on small-group jazz, since there are fewer moving
parts to manage, but harmony is of course present in (nearly) all jazz performance. One of my own
interests lies in the implications of harmony for composers and arrangers of big-band music. When
an arranger like Thad Jones or Jim McNeely sits down to arrange a jazz standard, they typically do
so with a knowledge of harmony earned from experience as a player. The fact that large ensemble
music is typically composed and written down means that the author maintains a tighter control
200
over the harmony than in an improvised setting; this allows the opportunity for a detailed shaping
of harmony over the course of a tune. These standards are often reharmonized in interesting ways,
and the transformational approach is ideally suited to discussing these arrangements in dialogue
with their original sources.
One of the aims of this dissertation has been to take seriously the manner in which jazz
musicians themselves discuss harmony. Since these musicians do not often speak technically about
harmony, I have used pedagogical materials as a way of getting at this insiders view. This focus
has often been implicit: there has been no extended literature review of pedagogical texts, nor was
there a need to use these texts to the exclusion of other, more academic, treatments of jazz
harmony. When a choice arose, however, I usually opted to cite Jerry Coker, Mark Levine, or
Jamey Aebersold rather than Henry Martin, Dmitri Tymoczko, or Steve Larson. This choice was
made not to disparage the important work of other theorists working on jazz harmony, but rather
as a means to acknowledge the real work on harmony in the jazz community (in addition to the
developments of the theory community).
At the outset of this study, I suggested that a transformational approach to jazz harmony
might constitute a set of analytical values dierent from the Eurological values of Schenkerian
analysis (as often applied to jazz). The prismatic style of transformational analysis, borrowed from
Steven Rings, has helped to enable this alternate value system. Nearly all of the analyses here have
avoided taking a single synthetic view of a passage, opting instead for a perspective in which
multiple, sometimes conicting, analyses can be considered individually in turn. This multifaceted
approach is intended to reect the nature of jazz harmony itself, and its application can enable us to
narrow the gap between bring jazz theory and jazz practice.
201
Appendix A
List of Recordings
This appendix provides a list of all of the musical examples used in this dissertation, along with
recordings where they can be found. For each tune, I have tried to provide at least two reference
recordings; they have been chosen because they present clear statements of the standard changes
analyzed in the text. Recordings are listed here only by the artist and album title; full details can be
found in the discography that follows. In cases where a tune has a recording that is considered
denitive, it is marked with a star. (This list reects my own proclivity towards pianists and
saxophonists, and is by no means exhaustive.)
Chapter 1
Jazz discography
. in the 21st century can be frustrating, given the many dierent formats and reissues of original
sources. I have tried to be as clear as possible in the discography, listing both original release information and more
readily available reissues, when available. Three online sources have been invaluable in compiling this information: The
Jazz Discography Project (http://www.jazzdisco.org; Nobuaki Togashi, Kohji Shaolin Matsubayashi, and Masayuki
Hatta, maintainers); Discogs (http://www.discogs.com; created by Kevin Lewandowski); and AllMusic
(http://www.allmusic.com; All Media Network).
202
Chapter 2
Chapter 3
203
Chapter 5
204
Appendix B
Transcriptions
Transcription is a notoriously dicult problem in jazz. As Steve Larson (among others) has noted,
any transcription is also, in some sense, an analysis. In the transcriptions that follow, I have (in
general) notated only the solo line, focusing primarily on the pitches and rhythms. This means
that many of the so-called secondary parameters are absent from the notation: dynamics,
intonation, phrasing, issues of timing, etc. To that end, the transcriptions should be seen as
companions to the source recordings, not replacements for them.
In the transcriptions of the Rhythm tunes, no chord symbols are given. One of the principal
arguments of Chapter 5 is that the chord symbols are somewhat uid through the course of a
performance; I have simply omitted the chord symbols rather than assigning them based on the
solo line. In the transcriptions, the notes remain more-or-less uninterpreted (though the notation
of accidentals is a question of interpretation); examples in the text refer to a given passages
harmonic context.
List of Transcriptions
1. Steve Larson, Analyzing Jazz: A Schenkerian Approach (Hillsdale, NY: Pendragon Press, 2009), 2.
205
Autumn Leaves
As played on Boss Tenors (1961) Joseph Kosma/Jacques Prvert
Gene Ammons/Sonny Stitt, tenor saxophones trans. Michael McClimon
B b m7 E b7
1A1 1:02
b 4 # n # n b
Ammons F7 B m7 E7
& b 4 j . . J J J J
C m7
Tenor Sax
3
A m7b 5
bb n # n b j nm j
D7 Gm
5
& n # J n .
B b m7 E b7
b # n # # n n b b
1A2 C m7 F7 B m7 E7
b n # #
& b n J # # n
9 3 3
A m7b 5
b n b n j .
D7 Gm
&b J . #
m
13
A m7b 5
1B
b # n j
D7 Gm
b
& j .
17 n
b
b n
C m7 F7
& b # n n b #
j
21
B b maj7
b b
&
23
1C A m7b 5 E m7b 5
b
D7 G m7
& b # n
3
25 3
The head
. changes for this recording are as given in Figure 1.6, but the solo changes shown here are slightly dierent.
The third and fourth bars of each A section consistently substitute Bm7E7B fm7E f7 for B fmaj7E fmaj7, and the
progression Gm7C7Fm7B f7 in the third and fourth bars of the C sections appear here as Gm7Em7 f5.
206
E b7
b n # -
D7 Gm
&b
j b j b n .
#
29 3 k .
B b m7 E b7
^
2A1 1:53 C m7
b # n # n b .
F7 B m7 E7
& b . n J
k
~~
~~~
33 3
A m7b 5
n n .
b #
D7 Gm
&b
37 3
. b
B b m7 E b7
^ # n . # - n #
bb J J
C m7 F7 B m7 E7
2A2
#
& J 3
41
A m7b 5
b n b # b
n # j .
D7 Gm
& b j
#
J J
45 3
A m7b 5
# . -
2B
b .
D7 Gm
&b #
j
49 3
b
b
C m7
-
.
B maj7 F7
&b j .
j
53
A m7b 5 E m7b 5
b
bb # n .
2C
D7 G m7
n .
& J
57 3
E b7
D7 Gm Stitt
b b n n
5
&b # n J n
61
(rhythm unclear)
207
3A1 2:46 C m7 B b m7 E b7
b w
F7 B m7 E7
&b j
65
(out)
A m7b 5
b # . n
j n
D7 Gm
& b #
j
J J .
69
3A2
b
# n n b
C m7 F7
&b #
73
n .
B b m7 E b7
# n . b
n # #
B m7 E7
b n b
& b # n #
75 5 3
n
A m7b 5
.
# #
D7 Gm
b n # n n .
&b # n #
77 2
A m7b 5
#
3B
bb j .
D7 Gm
j j
81
& n # n #
B b maj7
.
C m7 F7
b
&b J
85 5 5
5
3C A m7b 5 E m7b 5
b .
D7 G m7
& b
89
E b7
b D 7
J
Gm
b n
&b
93
208
4A1 3:38 C m7
b n # n b
F7
& b # n .
97
B b m7 E b7
b # # n n n b n b b n b
B m7 E7
& b
99 3
3
A m7b 5
b b n . j j
D7 6 Gm
&b n # n # n
101 3
n # n.
B b m7 E b7
# n n b
4A2 C m7 F7 B m7 E7
b # #
&b
105 3
A m7b 5
bb n b #
J n .
J #
D7 Gm
&
-
109 3
. -
A m7b 5 j
bb
4B
#
D7
Gm
&
. -
113
F 7 b
. -
b
C m7 B maj7
b j
& #
117 3
4C A m7b 5
E m7b 5
b #
D7 G m7
& b J
121
E b7
b j j .
D7 Gm
& b j # b .
#
125
5A1 4:31 C m7
b
&b (to piano solo)
129
209
Blues for Alice
As played on We Free Kings (1961) Charlie Parker
Rahsaan Roland Kirk, saxophones trans. Michael McClimon
1 (0:40)
D m7 b G 7 C m7b F7
4 .
F maj7 E m7 A7
b j
& 4 J J
Tenor sax
B b7 B b m7 E b7 A b m7 D b7
b b n b # b n
A m7 D7
j
& b #
5
j j b b b # b - b . - .
G m7 C7 F maj D m7 G m7 C7
& b # j n
9
C m7
2 (0:55) F maj7
# n b
b #
E m7D m7 G7 A7 F7
& b j # # n # J
13
B b7 B b m7 E b7 A b m7 D b7
b b b b
b
A m7 D7
& b # b n b
17
# n b C 7 b
G m7 F maj D m7 G m7 C7
&b
>
&b b
21 Soprano sax (Manzello)
The album
. gives Kirks name only as Roland Kirk; he added Rahsaan to his name in 1969. Kirk often played multiple
instruments simultaneously; the transcription tries to make this clear, providing a separate sta for each instrument.
The manzello is a modied soprano saxophone (both names are given here).
210
w
3 (1:09) F maj7
b b - n . b
E m7 A7 D m7 G7 C m7 F7
&b
3
& b b. b b - n .
25
B b7 b b A b m7 D b7
3
N . # > .
B m7 E 7 A m7 D7
b b b
&b #
j J J
29 (tenor) 3
G m7
C7
F maj
.
D m7 G m7 C7
& b n b n b #
33
k
4 (1:24) F maj7
#
E m7
A7
D m7 G7
b C m7 F7
&b J J J
&b j
37 (soprano)
J J
B b7
b b
&b
3
(to piano, drum solos; choruses 58)
& b b b
41
3
9 (2:34) F maj7
j
E m7 A7 D m7 G7 C m7 F7
&b b n j b
97 Soprano sax (Manzello)
B b7 B b m7 E b7 A b m7 D b7
b b j b
A m7 D7
&b j
101
#
211
j J J
G m7 C7 F maj D m7 G m7 C7
& b J
.
105
10 (2:48) F maj7
# # b
E m7 A7 D m7 G7 C m7 F7
&b # n b
109
B b7 B b m7 E b7 A b m7 D b7
b b # n b n # j # .
A m7 D7
& b b b
3 j
113
G m7
.
C7
b n
F maj D m7 G m7 C7
^
j J J
&b
117
# .
b -
11 (3:02) F maj7
n # # #
E m7 A7 D m7 G7 C m7 F7
&b J
121
B b7 B b m7 E b7 A b m7 D b7
b j J J
A m7 D7
&b w #
125
n .
b m b
G m7 C7 F maj D m7 G m7 C7
&b # . b b # b
129
F maj7
&b
(to bass solo)
133
212
The Eternal Triangle
As played on Sonny Side Up (1958) Sonny Stitt
Sonny Rollins/Sonny Stitt, tenor saxophones trans. Michael McClimon
1A1 0:38
b
& b 44 j b b
[S.R.]
Tenor Sax
b
&b .
J
5
3
n
1A2
b n
&b J . b #
9
b j
. # b n
&b J
13
3
# b b
1B
bb n
&
17
b n j # . b b n n n b n # #
&b b #
21
1A3
b n j
&b
25
3
b
& b J # j
29
Sonny. Rollins is indicated in the transcription with [S.R], and Sonny Stitt with [S.S.].
213
2A1 1:05
b n n b b
&b J
33
bb b
J
&
37
bb . .
2A2
& J J
41
b # b
b n n b b n
&b n
45
2B
# n # # b b b
b n b
&b
49
b b n n b n # # # b n # #
& n # n
53
2A3
bb n
& n
57
b b b b # .
&b
J
61
b
3A1 1:32
b n n
&b J J
65
214
b b b . b
& b N b b
69
3A2
b
& bb J J J
73
b b
&b J . J
77
# n # # N b . b n
3B
b n b
&b J
81
b n b. b b #. j
&b J N # #
85
3A3
b
&b n b
n
b
89
b # n. b
& b n b J
93 3
3
4A1 1:59
b b b b
&b
b J J J J J J J J
97
b
&b
b J b J n J J n J J J J J j
101
J J J
215
b j J J
4A2
& b J b b n
105
J J J
b b b
b - .
&b n
109
4B
b n b
b
&b # # j b J b
# # n
113
3
b b n n b b b b N n # n # n # # n # n
& n
117
.
4A3
b
&b J
121
- -
b b b b b . .
& . J k
125
-
. . .
5A1 2:25
b J #
j
&b
129
# n # # b b . n
& bb n J
J
133
b
5A2
b
&b j J
137
3
216
b b
&b b n b
141
3
5B
b # # n jn n b b
& b # # . # # n #
145 3
bb . n # b. b b n. n n # . # n #
J
&
149
.
5A3
b
&b
153
b b
b n b n b
&b # b
157
6A1 2:52
b
&b
161
b j
&b
. n b b
[S.S.]
165
J
3
6A2
n
& bb n # n b
169
b b #
& b b
n b J .
173
217
b n # n # n # n # # n b b
6B
& b b #
177 3
# b b b b n b
bb n . J J J
n # n # # n
# #
&
181
6A3
b b n b # b n b # n
&
185 3 3
b . b
&b J #
189
- . n # # b
7A1 3:18
b b #
&
193
b n #
&b
j
#
197 3 3
b #
7A2
bb
&
201 3
bb N b J .
& b n
3
205 #
# n # b . b .
7B
b # n # n # n b n b J J
&b
209 3 3
218
> # n n . b > b . n > n b n > # n #
b b n .
n # n # # n
&b
213
. n
7A3
b b b # J # n #
&
217 3
bb b b b J .
.
&
221
n # n b
8A1 3:44
& bb
225 3
b b b N w
& #
229
3
>
bb
8A2
& J
233
.
bb b
& b # J
3
237 3
# # > b n b b b n b
8B
b
&b n # b
241 3 3
b j
& b # . # b . n #
245
219
8A3
b n b # n n b # n
& b n
249 3 3 3 3 3 3
3 3
b b b b >
>
b.
&
253
b
9A1 4:10
# b n
& bb # b
n #
257
b b # b # n
J J
&
261 3
w b .
9A2
b >
b J
&b
265 3
b b
b
&b b # n b j .
269 3
9B
b # J n # n n N b
b #
3
& b # #
273
b > > b
& b # n n b b b n n b n
277 >
bb b n n .
9A3
& n n b
281 3
220
b b
& b #
285
3
bb b n b n n # n #
10A1 4:37
&
289 3
bb b b b n J .
&
293 3
10A2
& bb n n n
297
bb b b b J
& b b b b n n
3
301 3
10B
# n # # b . b
b
&b
n # n # n b n b J
305 3
b n # b b b n n # # n # b n b
&b
309
b
10A3
b n b # n
&b
313 3 3
b b .
&b b
317 #
221
11A1 5:03
b n b
&b J
321
& bb #
325 3
w # #
11A2
b J n #
&b
329
bb b b n n # J .
&
333 3 3
b
bb n # n # # n # n # n b n b b
11B
& b
337 3
bb n n n b b b b b n # # #
&
341 3
b b. . b b n b #
11A3
b
&b J
345 3
b b b b b J . .
&
349
12A1 5:29
b b . n n #
b
& J
353 3 3 3 3
222
b # n b j # j
&b b b #
357
# n # b n #
12A2
bb b n # #
& #
361 3
.
bb b J
& b # J
3
365
J 3
-
# # # J n b
12B
b
&b # .
J
A J b
369 3 3
# n n . b b b n n b . n # #
b # n # # n
&b
373
. n b
12A3
b b b # J #
&
377
b b b .
& b # b
381
13A1 5:55
b
& b b b b b
b
385
bb b n n # #
& # b
389
223
bb .
j n
13A2
#
&
393
b . # n -
bb
# #
&
397 3
13B
# n #
b # n # n # n b n b b
&b b b b
3
401 3
b
&b # b b
n n
405
13A3
b n b
&b #
409
b b j # . n J .
& b b # . J
413
bb
14A1 6:20 (begin trading 4s)
. .
[S.R.]
& J
417
b
&b J
[S.S.]
421
. # # n j j
14A2
b
[S.S.]
& b J . n b # b
[S.R.]
425
224
bb # n b n
&
429
14B
j # # # b b
bb [S.R.]
# n n b
&
433 3 3
bb n # n b b b n n n # # n # n J
[S.S.] [S.R.]
&
437
n # # # # A n n n n b
14A3
&b
b n b n b
441
bb b b n n
3
[S.S.]
& .
J
445 3
[S.S.] j
[S.R.]
15A1 6:46
b b
# n n b
& b b
n
449
.
bb b n # j
J
&
453 3
3
b
15A2 [S.R.]
j b n #. # n # # n . #
[S.S.]
& b J
457 J
5
bb n
3
n #
&
461
225
15B
# n # #
b n # n # n n b b b
&b
[S.R.]
465 3 3
bb n j # j b b b b n n
[S.S.] 3
# # n .
& # n #
469
3 3 3
bb b b # . b J # # b n n b # b
15A3 [S.R.]
&
473
b b b b b
[S.S.] [S.R.]
&b b b j
b
b
477
b
16A1 7:12
b b
# n
[S.S.]
b
& # b #
481
3 3
bb b b J .
& J J
485
b J J . . . .
16A2
b
[S.R.]
&b J J b
489 3
b b b
[S.S.]
b
& b b J J
493
# n # # N A b
16B
bb #
[S.R.]
& J J J J J J J J b J n
497
226
bb n n # b b b > . # n # n j
[S.R.]
[S.S.]
& n #
501
j j
bb j n # n
16A3
&
505
b .
[S.S.]
j
&b b n
n
17A1 7:37 (S.S. continues, trading 8s)
b b n b
&b # n
513
b b b
j .
& # #
517 3
3
17A2
b n # n #
bb [S.R.]
& # n J # J n n # b b
521
bb n j b b # n # b n #
& b n [S.S.] #
525
3
17B
# # # n # # b b #
b n n # n # n b
&b
529 3 3 3
n b n.
bb # n b j b b n n b J
&
533 3
227
b n n # n
17A3
b # # n #
& b n b #
537 [S.R.]
J
b b . .
b
& J J J
541
bb # n # # n n #
18A1 8:03
& # # # n # n n n b
545
[S.S.]
b
bb n b b b
549
& b
j n #
n
3
n . # J b
18A2
.
[S.R.]
b A b
&b J
553 3
& bb N
557
bb nn # n # b
18B
# n #
& n b n b
561
[S.S.]
b n b b b # n n b
&b
565
j
3
bb n b
18A3 [S.R.]
JJJ b n b
& J # b
569
228
b
3
[S.S.]
b b b n b
&
573
19A1 8:29
bb n # n # n b #
&
577
bb n b # n J .
& J
581
b [S.R.]j
19A2
j
b
& J . .
585
J
# n n
3
b
&b
[S.S.]
589
n >
# # b b
19B
b b j n # n b #
& # #
593 3
b # # n # #[S.R.]
bb n n b b
n
n n
&
597 3 3 3
19A3
b
&b J
J J J J J J # j
601
b b n b b #
&
605
3
20A1 8:54
b .
&b (trumpet solo)
609
229
Isotope
As played on Inner Urge (1965) Joe Henderson
Joe Henderson, tenor sax trans. Michael McClimon
b b N
1 (0:31) C 7
j j b b
& 44 J b b
B b7
b
F7 C A7
& J J # #
5
A b7 G b7 E b7
b J b b b n b
G7 C7 A7
& b b b b b
9
2 (0:47) C 7
. b b j b
& J # b
13
B b7
F7
b b b
C
# A7
n b b b
& #
j
17 3
A b7 G b7 E b7
b # G7
n b b b n b
C7 A7
& J # # b
21
3 (1:02) C 7
. b b
j j
& # J #
25
B b7
b b b
F7 C A7
& J b j
29
230
A b7 G7 C7 A7 G b7 E b7
b n
& b b b b b J b b b b b n b
33
4 (1:17) C 7
>
b b .
& n
~~~
37
B b7b ~.~~
~~ b . #
F7 C A7
& #
41
#
A b7
j b b b b # b
G7
& # j
45
5
G b7 E b7
b # b n b b b b
C7 A7
# b b b b b
& # # #
5 (1:32) C 7
b
. j
& b .
49
B b7
b
F7 C A7
& #
53
# n
A b7 G b7 E b7
. j b b
b n b
G7 C7 A7
j b
b j
3
& j b b # 3 J
57 b
231
6 (1:48) C 7
# b
J j J J J j b J j
j
& #
61
B b7
# #
F7 C A7
& J b J j J b j
65
A b7 G b7 E b7
b . # b b
G7 C7 A7
b b n b
j
& J b J J
#
J #
69
7 (2:03) C 7
b b b
j n J # b # n b b
& #
J #
73 3
B b7
F7 C
& b b # # n b
77
A b7
b b b n
A7
# b b
& # # b b b b n n
80
G b7 E b7
# # n b n b # #
G7 C7 A7
& b # b b b b n b j
82 3
5
8 (2:18) C 7
b
j
&
85 (false ngerings)
232
b
B 7 b b C A 7
# b J
F7
&
89 (Alternate A
signicantly at)
A b7 G b7 E b7
b
# b b b b b n n
G7 C7 A7
& J J
93
N 3
b b # #
9 (2:33) C 7
n J
&
97 3 3 3 3 3
B b7
b m b # b A 7# n b
b
F7 C
& b b
n b
101 3
A b7 b b
b b b
G7 C7
# A 7 # b G 7b b n E 7n
b b
&
105 3
10 (2:48) C 7
n b b b
&
109
B b7
b # b
F7 C A7
& b b j b b
# #
113 3
A b7 G7 C7 A7 G b7 E b7
b j
& J b b
117
233
. b .
11 (3:03) C 7
. b
& J J J
B b7
121
b
F7
j
b
b b # b # # # #
C A7
& b
125
A b7 G b7 E b7
b b
G7 C7 A7
& b b j # # b b b # n
129 3 3
b
12 (3:18) C 7
j
& # # # . b
133
F7 B b7 C A7
& j b j b n
J #
#
j
137
A b7 G b7 E b7
b n .
G7 C7 A7
& b b n b b
141
# k
.
b b b
13 (3:33) C 7
b
F7
& J J
145 3 3
B b7 A b7
# # b b b b
C A7
b # #
& # # n
150
J J J
5
234
G b7 E b7
b # n # b # # b n b b n
G7 C7 A7
& b b
154
J
5
6
14 (3:48) C 7
b . b b
& J b J n b
b b
157 3
9
B b7
# # b
F7 C A7
& b b b # n
161 3
A b7 G b7 E b7
b b
G7 C7 A7
b
& b n b . j
J .
165
b
b b
15 (4:02) C 7
b
& J J n b b b
b b b
169
B b7
# #
F7 C A7
b n # b j
& b b #
173 3
A b7 G b7 E b7
G7 C7 A7
n b b J J J # n b b n b
& b
177
16 (4:17) C 7
& n. b
181
J (piano solo)
235
Lo-Joe
As played on Amsterdam After Dark (1979)
Coleman, tenor sax; Hilton Ruiz, piano; Sam Jones, bass
George Coleman
trans. Michael McClimon
A1
bb 4 j
& b b b 4 n n n bbb nnn b b # n # n bnb nn#
Tenor Sax
bb 4 j
& b b b 4 j b # n # n bn n n
n n n
Piano
? bb 4 J #
bbb 4 J
b
Bass
? bb 4
bbb 4 n n
n n nn b
b n
bb n n b
&bbb n b A n # b n n n
b
j j
bb j #
& b b b b b j
n b
n n
b n b
? bb J J
J # b # J
bbb
? b b b n b n n b n
bbb
5
This .transcription contains only the head of the tune; the bass line sounds down an octave. In many cases, the piano
voicings are dicult to hear in the recording. I have tried to keep the transcription as faithful as possible, but the
disclaimer on using the transcription as a supplement to (not a replacement for) the recording applies more strongly
than usual in this case.
236
A2
bb j
& b b b n n n b b b n n n b b # n # n b n b n n #
bb j
& b b b b n # b n
n b #
b n
n n n
n n
n
? b b n
bbb J n
? b b b b
bbb n b n n n b J
9
b b n n b
& b bbb n b A
n
b
b
& b b b b b b j
n b
j
j
b
? bb
bbb J J J
n
? bb b
bbb n b
13
b b n
B
bb
&bbb n n n n # # n n J n # n
#
J #
b b n j
& b b b n
# n j # #
# n n
# # n n
? bb b n n n b n #
bb # b J
J
? bb b
n
# n n n. n n
bb n J
17
237
bb b b b n j b b
&bbb n n # n b n J n n
n
n
bb j j
& b b b b b n j n n n
n # # b
? b b b J n b
bbb J n b J n #
? b b b
n n n b b n b
n
bbb
21
A3
b j
& b bbb b b n n n b b b n n n b b # n # n b n b n n #
b
b n j
& b bbb
n # # n
n
n
n b n n
n
? bb
bbb
? b b b b
bbb b n b n n n b
25
bb b n n b
& b b b b n b A n
bb j j
& b b b b
b n b # b n n
? bb b
bb J J
? b b b J n
bb b J
29
238
Rhythm-a-ning
As played on Thelonious in Action (1958) Thelonious Monk
Johnny Grin, tenor; Monk, piano trans. Michael McClimon
bb 4 b
1A1 0:36
b b # b
&b
b J J
5
> # b b n # n
1A2
b b n
& b n #
9 3
n
bb b b b J
& #
13 3
3
1B
n b b n
b # n n n
&b # n n b n
17
b # n n
b
# # b n
&b
3
21 3 3
b b b .
b b b
1A3
b
&b b J b b b
J
25
n b b b
bb n n n b b
&
29 3 3
239
. #
b n J n
2A1 1:05
& b # n
33
b j
& b b n b j .
37
n b
2A2
bb # j b n
&
41 # 3 3
b n > # n
b b J J J
&b J
45
bb # n # n n n n # n n n b
2B
& #
49
bb n n n b n b
b J J
&
53 3
.
b b b
2A3
b
& b b J
57
bb n b b b
& #
61 3
b b b
3A1 1:34
b
&b J J J
65
240
b b b
& bb J n #
69
b b > b b
b b b b
3A2
& b b
73
b
b b N b n b b b n n #
77
&
3B
n n n
b
b # n
&b
81
bb b # n
b
3
&
85
b
3A3
b >
&b
j
#
89
n
j
.
j
b # #
J
&b
93
b b
bb J J J J J b n b b
4A1 2:02
& #
97 3
b
bb b b n
&
101
241
bb b n # # # n b n
4A2
&
105
b b n
b
& b J # J
~~~
109 3 3
bb # n n b A b n b #
4B
& n b n
113
b
b
&b # b n n b # b
117
J
3
bb b b b b
4A3
& b
J J
121 3
.
b .
& b J . . # J J
125
b b b
2:30
n b # n n # n b
5A1
b
&b #
129 3
b b
& b n b b #
133
5A2
b b b n n # n j
& b #
137 3
242
bb J b J b
& J J J
141
5B
n . # b n #
b n #
J # n n n #
&b
145 9
n b b # n
b n n b
&b n b # b
149
5A3
b # b
&b
b
# n # n n b
153
b b
& b # j # #
157
3
> >
# n
6A1 2:58
b
&b
161
b b
bb n b J b j # n
&
165 3
6A2
b b J
&
169
b b
bb n b b b b
b
&
173
243
b # n
6B
b n
& b # n n b # n
177
b . n n . # n
b b n b
3
& # n # n J
181 3
- - - - # - n - # - b b
- - -
6A3
b #
&b
185
bb b J
b j
J
b J N
&
189
. # j
b b b
7A1 3:26
b b n J
&
193
b b
& b b # b b
#
197
3
bb n n b b
7A2
& J
201
b b b # J
&b # b
205
3
.
bb # n b b
7B
n
& n n n
209 3 3
244
# . # n # # #
bb n n J # # n
& # # # # n
213 3
#
7A3
b
&b J
217
.
b b n b b
& J
221
w b b
3:54
8A1
b b n b n # n
&b #
225 3
bb b n b b b . b J b J J
& J
229
n
bb . J
8A2
&
233
b n
b b J J J b J
&
237
8B
. n # n b b # n n b b
bb J J b
&
241
n # # #
bb n n # # n
& # # #
245 # b n
245
.
8A3
>
b
&b # J
249
.
b
&b #
J
253
b n
9A1 4:22
& b b
257
J J J
>
bb > > # > > >
&
261
bb #
9A2
& J
265 3 3
b b
b b J
&b # J
~~~
269 3
bb # n # n n n # n b n
9B
&
273
b #
b b n n b n
&b # b
277 3
9A3
b b
&b b b # n b
#
281
3
246
bb .
& J
285
n # n # # b b n b b
10A1 4:49
& bb
289 5 3 3 5 5 3 3 5 5
b n b b # b b b n # n n b #
bb b n
3
&
293 3 3 3
5 5 5 5 3 3
3 3
bb
10A2
j n . # n
& #
297 3 3
7 7
b n b
& bb b n b #
J
301 3 3 3
# # # n n n
10B
b b # n # n n n # n
& # #
305
b 3 # n n n b
b
& n # n n b b J J
309
b b b
10A3
b .
& b b J J J J
J
b
J
313
b b . # n # j .
&b J
317
247
11A1 5:17
b n
b JJ J
&b
321 3
b b b b b n n b
& b b # j J J
325
3
bb b b
11A2
& b b n # n b n
329
b # j
& b b
333
3
n n n .
bb n # n n # n
11B
& J
337
b n n b J J
&b
341
~ b n b
11A3
bb
&
345
b .
bb
J . . w
&
349 3
248
12A1 5:44
b
&b w
b
&b
www www
Piano
? bb
353
w w
b
&b
www www www
? bb
357
w w w
12A2
b
&b j j
? bb
361
b
&b j
? bb
365
249
12B
b
& b n
j j
n
? bb w
369
ww w
b
&b j j j j
b n
? bb n w w bw w
373 w
12A3
b
&b j
? bb
377
j j
b .
&b b .
LH
J J
? bb
381
b n b b
13A1 6:12
& b # # n # n # # n n # n n # n n
n .
? b
b # n n
385 (very faint)
250
b b n b j b n b b b n b
&b
? bb
389
13A2
b
& b # # n # n # # n n # # n n b
n n n
v
? bb n
# n
393
b b b
&b b b
b
? bb b
397
bb b # n
13B
#
&
? bb
ww ww
401
#
bb n # # n n #
& b
? b
b
405
251
13A3
b n b b
& b # # n # n # # n # n
n n # n n
? bb
409
j
b b b
.
b . .
b J
& J
.
J
LH
? bb
413
b n # # n N n # # n n n # n n b n b
14A1 6:39
& b
? bb
417
# n n
b J
& b b b
? bb # N
421
b n # # n N # # n n # n n b n
14A2
& b
? b
b # n n
425
252
b
& b b bw w
? bb bw w
429
b # # n #
14B
b b b n n # .. # n
&
? bb
433
bb b b b J # n b n b b # j n n
j
b
&
? bb
437
14A3
b
& b n # # n N # # n n # n
n b n b
? bb n
# n
441
15A1 7:07
b
&b
b b www
? b
(to bass solo)
b
w
445
253
Bibliography
Adams, Kyle. When Does the Present Become the Past? A Re-examination of Presentism and
Historicism. Paper presented at the annual meeting of the Society for Music Theory,
Charlotte, NC, November 2013.
Aebersold, Jamey, ed. Charlie Parker Omnibook. Atlantic Music Corp., 1978.
. I Got Rhythm Changes in All Keys. Jamey Aebersold Play-A-Long Series, vol. 47. New
Albany, IN: Jamey Aebersold Jazz, 1991.
. The IIVI Progression. Jamey Aebersold Play-A-Long Series, vol. 3. New Albany, IN:
Jamey Aebersold Jazz, 1974.
Arthurs, Daniel. Reconstructing Tonal Principles in the Music of Brad Mehldau. PhD diss.,
Indiana University, 2011.
Baker, David. How to Learn Tunes. New Albany, IN: Jamey Aebersold Jazz, 1997.
. How to Play Bebop, Vol. 1: The Bebop Scales and Other Scales in Common Use. Van Nuys,
CA: Alfred, 1985.
Berendt, Joachim-Ernst, and Gnther Huesmann. The Jazz Book: From Ragtime to the 21st
Century. 7th ed. Translated by H. and B. Bredigkeit, Dan Morgenstern, Tim Nevill, and
Jeb Bishop. Chicago: Lawrence Hill, 2009.
Berliner, Paul. Thinking in Jazz: The Innite Art of Improvisation. Chicago: University of Chicago
Press, 1994.
Biamonte, Nicole. Augmented-Sixth Chords vs. Tritone Substitutes. Music Theory Online 14,
no. 2 (June 2008).
Bishop, John. A Permutational Triadic Approach to Jazz Harmony and the Chord/Scale
Relationship. PhD diss., Louisiana State University, 2012.
Bregman, Albert S. Auditory Scene Analysis. Cambridge, MA: MIT Press, 1990.
Bribitzer-Stull, Matthew. The AfCE Complex: The Origin and Function of Chromatic Major
Third Collections in Nineteenth-Century Music. Music Theory Spectrum 28, no. 2 (Fall
2006): 16790.
Burt, Peter. Takemitsu and the Lydian Chromatic Concept of George Russell. Contemporary
Music Review 21, no. 4 (2002): 79109.
254
Callender, Clifton. Voice-Leading Parsimony in the Music of Alexander Scriabin. Journal of
Music Theory 42, no. 2 (1998): 21933.
Callender, Clifton, Ian Quinn, and Dmitri Tymoczko. Generalized Voice-Leading Spaces. Science
320, no. 5874 (April 2008): 34648.
Capuzzo, Guy. Neo-Riemannian Theory and the Analysis of Pop-Rock Music. Music Theory
Spectrum 26, no. 2 (October 2004): 17799.
. Pat Martinos The Nature of the Guitar : An Intersection of Jazz Theory and
Neo-Riemannian Theory. Music Theory Online 12, no. 1 (February 2006).
Carter, Nathan. Visual Group Theory. Washington, D.C.: Mathematical Association of America,
2009.
Clement, Brett. A New Lydian Theory for Frank Zappas Modal Music. Music Theory Spectrum
36, no. 1 (Spring 2014).
Clough, John, Nora Engebretsen, and Jonathan Kochavi. Scales, Sets, and Interval Cycles: A
Taxonomy. Music Theory Spectrum 21, no. 1 (Spring 1999): 47101.
Clough, John, and Gerald Myerson. Variety and Multiplicity in Diatonic Systems. Journal of
Music Theory 29, no. 2 (Autumn 1985): 24970.
Cohn, Richard. Audacious Euphony: Chromatic Harmony and the Triads Second Nature. New York:
Oxford University Press, 2012.
Coker, Jerry. Elements of the Jazz Language for the Developing Improvisor. Miami: Belwin, 1991.
Cook, Nicholas. Schenkers Theory of Music as Ethics. Journal of Musicology 7, no. 4 (October
1989): 41539.
Cook, Stephen. Review of Dizzy Gillespie, Sonny Side Up. AllMusic.com. Accessed July 13, 2015.
http://www.allmusic.com/album/sonny-side-up-mw0000188698.
Coolman, Todd. The Bottom Line: The Ultimate Bass Line Book. New Albany, IN: Jamey
Aebersold Jazz, 1990.
De Cosmo, Emile. The Diatonic Cycle. North Bergen, N.J.: EDC Publications, 1970.
255
Demsey, David. Chromatic Third Relations in the Music of John Coltrane. Annual Review of
Jazz Studies 5 (1991): 14580.
Deveaux, Scott. The Birth of Bebop: A Social and Musical History. Berkeley: University of California
Press, 1997.
Douthett, Jack, and Peter Steinbach. Parsimonious Graphs: A Study in Parsimony, Contextual
Transformations, and Modes of Limited Transposition. Journal of Music Theory 42, no. 2
(1998): 24163.
Folio, Cynthia. An Analysis of Polyrhythm in Selected Improvised Jazz Solos. In Concert Music,
Rock, and Jazz since 1945: Essays and Analytical Studies, edited by Elizabeth West Marvin and
Richard Hermann, 10335. Rochester: University of Rochester Press, 1995.
Ford, Philip. Somewhere/Nowhere: Hipness as an Aesthetic. The Musical Quarterly 86, no. 1
(April 2002): 4981.
Forte, Allen. The American Popular Ballad of the Golden Era, 19241950. Princeton University
Press, 1995.
Gjerdingen, Robert. Music in the Galant Style. New York: Oxford University Press, 2007.
Gollin, Edward. Some Aspects of Three-Dimensional Tonnetze. Journal of Music Theory 42,
no. 2 (1998): 195206.
Goodheart, Matthew. The Giant Steps Fragment. Perspectives of New Music 39, no. 2 (July
2001): 6395.
Graf, Richard, and Barrie Nettles. The Chord Scale Theory and Jazz Harmony. Advance Music,
1997.
Grossman, Israel, and Wilhelm Magnus. Groups and Their Graphs. Washington, D.C.:
Mathematical Association of America, 1964.
Hall, Rachel Wells. Review of David Lewin, Generalized Musical Intervals and Transformations and
Musical Form and Transformation: Four Analytic Essays. Journal of the American Musicological
Society 62, no. 1 (Spring 2009): 20522.
Hodson, Robert. Interaction, Improvisation, and Interplay in Jazz. New York: Routledge, 2007.
256
Hook, Julian. Contemporary Methods in Mathematical Music Theory: A Comparative Case
Study. Journal of Mathematics and Music 7, no. 2 (2013): 89102.
. David Lewin and the Complexity of the Beautiful. Intgral 21 (2007): 15590.
Hyer, Brian. What Is a Function? In The Oxford Handbook of Neo-Riemannian Music Theories,
edited by Edward Gollin and Alexander Rehding, 92139. New York: Oxford University Press,
2011.
Julien, Patricia. The Structural Function of Harmonic Relations in Wayne Shorters Early
Compositions: 19591963. PhD diss., University of Maryland, 2003.
Kelley, Robin D. G. Thelonious Monk: The Life and Times of an American Original. New York: Free
Press, 2009.
Kernfeld, Barry. The Story of Fake Books: Bootlegging Songs to Musicians. Lanham, MD: Scarecrow
Press, 2006.
Klumpenhouwer, Henry. In Order to Stay Asleep as Observers: The Nature and Origins of
Anti-Cartesianism in Lewins Generalized Musical Intervals and Transformations. Music
Theory Spectrum 28, no. 2 (Fall 2006): 27789.
Larson, Steve. Analyzing Jazz: A Schenkerian Approach. Hillsdale, NY: Pendragon Press, 2009.
257
Larson, Steve. Composition versus Improvisation? Journal of Music Theory 49, no. 2 (October
2005): 24175.
. Schenkerian Analysis of Modern Jazz: Questions about Method. Music Theory Spectrum
20, no. 2 (October 1998): 20941.
. Strict Use of Analytic Notation. Journal of Music Theory Pedagogy 10 (1996): 3171.
Lerdahl, Fred. Tonal Pitch Space. New York: Oxford University Press, 2004.
Lester, Joel. Compositional Theory in the Eighteenth Century. Cambridge, MA: Harvard University
Press, 1992.
Levine, Mark. The Jazz Piano Book. Petaluma, CA: Sher Music, 1989.
Lewin, David. Generalized Musical Intervals and Transformations. New York: Oxford University
Press, [1987] 2007.
. Musical Form and Transformation: Four Analytic Essays. New York: Oxford University
Press, [1993] 2007.
Lewis, George. Improvised Music after 1950: Afrological and Eurological Perspectives. Black
Music Research Journal 16, no. 1 (Spring 1996): 91122.
Love, Stefan. A Model of Common-Tone Connections Among Jazz Scales. Journal of Music
Theory Pedagogy 23 (2009): 15569.
Martin, Henry. Charlie Parker and Thematic Improvisation. Lanham, MD: Scarecrow Press, 1996.
. Jazz Harmony: A Syntactic Background. Annual Review of Jazz Studies 4 (1988): 930.
. Schenker and the Tonal Jazz Repertory. Tdschrift voor Muziektheorie 16, no. 1 (2011):
120.
McFarland, Mark. Schenker and the Tonal Jazz Repertory: A Response to Martin. Music Theory
Online 18, no. 3 (September 2012).
258
McGowan, James. Dynamic Consonance in Selected Piano Performances of Tonal Jazz.
PhD diss., Eastman School of Music, 2005.
. Chord-Scale Networks in the Music of Wayne Shorter. Paper presented at the West
Coast Conference of Music Theory and Analysis, Eugene, OR, March 2012.
Mulholland, Joe, and Tom Hojnacki. The Berklee Book of Jazz Harmony. Boston: Berklee Press,
2013.
Owens, Thomas. Bebop: The Music and its Players. New York: Oxford University Press, 1995.
Phlert, Werner. Basic Harmony. Translated by Jrgen Krohn and Norman Bowie. Werner Phlert
Publications, 1989.
Porter, Lewis. John Coltrane: His Life and Music. Ann Arbor: University of Michigan Press, 1998.
Rings, Steven. Riemannian Analytical Values, Paleo- and Neo-. In The Oxford Handbook of
Neo-Riemannian Music Theories, edited by Edward Gollin and Alexander Rehding, 487511.
New York: Oxford University Press, 2011.
Russell, George. The Lydian Chromatic Concept of Tonal Organization. New York: Concept, 1959.
. The Lydian Chromatic Concept of Tonal Organization. 4th ed. The Art and Science of
Tonal Gravity. Brookline, MA: Concept, 2001.
Samarotto, Frank. Treading the Limits of Tonal Coherence: Transformation vs. Prolongation in
Selected Works by Brahms. Paper presented at the annual meeting of the Society for Music
Theory, Madison, WI, November 2003.
Sanguinetti, Giorgio. The Art of Partimento: History, Theory, and Practice. New York: Oxford
University Press, 2012.
Santa, Matthew. Nonatonic Progressions in the Music of John Coltrane. Annual Review of Jazz
Studies 13 (2003): 1325.
259
Satyendra, Ramon. An Informal Introduction to Some Formal Concepts from Lewins
Transformational Theory. Journal of Music Theory 48, no. 1 (Spring 2004): 99141.
Schenker, Heinrich. Free Composition [Der freie Satz]. Edited and translated by Ernst Oster. New
York: Longman, [1935] 1979.
Sickler, Don, ed. Thelonious Monk Fake Book. Steve Cardenas, transcriber. Milwaukee: Hal
Leonard, 2002.
Straus, Joseph. Uniformity, Balance, and Smoothness in Atonal Voice Leading. Music Theory
Spectrum 25, no. 2 (October 2003): 30552.
Strunk, Steven. The Harmony of Early Bop: A Layered Approach. Journal of Jazz Studies 6
(1979): 453.
. Linear Intervallic Patterns in Jazz Repertory. Annual Review of Jazz Studies 8 (1996):
63115.
Swift, Richard G. Letter to the editor. Journal of Music Theory 4, no. 1 (April 1960): 128.
Tufte, Edward. The Visual Display of Quantitative Information. 2nd ed. Cheshire, CT: Graphics
Press, 2001.
. The Generalized Tonnetz. Journal of Music Theory 56, no. 1 (2012): 152.
. Generalizing Musical Intervals. Journal of Music Theory 53, no. 2 (Fall 2009): 22754.
. A Geometry of Music: Harmony and Counterpoint in the Extended Common Practice. New
York: Oxford University Press, 2011.
. The Geometry of Musical Chords. Science 313, no. 5783 (July 2006): 7274.
. Scale Networks and Debussy. Journal of Music Theory 48, no. 2 (October 2004): 21994.
Voglewede, Matthew. Metrically Dissonant Layers of Swing: Double Time in Two of Louis
Armstrongs Performances of Lazy River. Paper presented at the annual meeting of the
Music Theory Society of the Mid-Atlantic, Philadelphia, PA, March 2013.
260
Waters, Keith. Chick Corea, Postbop Harmony, and Jazzs Second Practice. Paper presented at
the annual meeting of the Society for Music Theory, Charlotte, NC, November 2013.
Waters, Keith, and J. Kent Williams. Modeling Diatonic, Acoustic, Hexatonic, and Octatonic
Harmonies and Progressions in Two- and Three-Dimensional Pitch Spaces; or Jazz Harmony
after 1960. Music Theory Online 16, no. 3 (August 2010).
Zwerin, Charlotte. Thelonious Monk: Straight No Chaser. Warner Bros., 1988. VHS.
261
Discography
Ammons, Gene, and Sonny Stitt. Boss Tenors. Verve v6 8426, 1961. Reissued, Verve 837 440-2, 1991,
cd.
Baker, Chet. Cool Baker, Vol. 2. Pacic Jazz k18p-9260, 1953. Available on Chet Baker Quartet
Featuring Russ Freeman. Blue Note 4931642, 1998, cd.
. Smokin with the Chet Baker Quintet. Prestige pr 7449, 1965. Reissued, Prestige
9020470372, 2001, cd.
Blakey, Art, and the Jazz Messengers. At the Cafe Bohemia, Vol. 1. Blue Note blp 1507, 1956.
Reissued, Blue Note cdp 7 46521 2, 1987, cd.
Coleman, George. Amsterdam After Dark. Timeless sjp 129, 1979. Reissued, Timeless Jazz Legacy
tjl 74509, 2010, cd.
Coltrane, John. Coltranes Sound. Atlantic lp 1419, 1960. Reissued, Rhino/Atlantic 7567813582,
1988, cd.
. Giant Steps. Atlantic lp 1311, 1959. Reissued, Rhino/Atlantic 8122752032, 1998, cd.
Davis, Miles. Collectors Items. Prestige prlp 7044, 1956. Reissued, Fantasy 0712, 1990, cd.
. Cookin with the Miles Davis Quintet. Prestige 7094, 1957. Reissued, Prestige 61046, 2003,
cd.
. Miles Davis, Vol. 2. Blue Note blp 5022, 1953. Reissued, Blue Note bnz 240, 1985, cd.
[Miles Davis All Stars]. Walkin. Prestige prlp 7076, 1957. Reissued, Original Jazz Classics
ojccd 213, 1991, cd.
DeFrancesco, Joey. Live: The Authorized Bootleg. Concord Jazz 7230123, 2007, cd.
Evans, Bill. How My Heart Sings! Riverside rlp 473, 1962. Reissued, Original Jazz Classics ojccd
369-2, 1995, cd.
. Portrait in Jazz. Riverside rlp 12-315, 1960. Reissued, Original Jazz Classics ojccd 088-2,
1992, cd.
. Sunday at the Village Vanguard. Riverside rlp 376, 1961. Available on The Complete Village
Vanguard Recordings 1961. Riverside 4443, 2005, cd.
262
Flanagan, Tommy. Giant Steps: In Memory of John Coltrane. Ea 4022, 1982. Reissued, Ea cd
4022-41, 1988, cd.
Gillespie, Dizzy. Dizzy Gillespie. RCA Victor lpv 530, 1966. Available on Dizzy Gillespie: The
Complete RCA Victor Recordings. Bluebird 07863 66528-2, 1995, cd.
Gillespie, Dizzy, Sonny Stitt, and Sonny Rollins. Sonny Side Up. Verve mgv 8262, 1957. Reissued,
Verve 5214262, 1992, cd.
Gordon, Dexter. More Power! Prestige pr 7680, 1969. Reissued, Prestige/Original Jazz Classics
ojccd 815-2, 1994, cd.
Guaraldi, Vince. A Flower is a Lovesome Thing. Fantasy 3-257, 1958. Reissued, Original Jazz Classics
ojccd 235-2, 1995, cd.
Harris, Barry. At the Jazz Workshop. Riverside rlp 326, 1960. Reissued, Original Jazz Classics
ojccd 208-2, 1992, cd.
. Newer than New. Riverside rlp 413, 1961. Reissued, Riverside vicj 60691, 2000, cd.
. Inner Urge. Blue Note blp 4189, 1965. Reissued, Blue Note bst 84189, 1985, cd.
Jamal, Ahmad. Jamal at the Pershing, Vol. 2. Argo lp-667, 1961. Available on Complete Live at the
Pershing Lounge 1958. Gambit 69264, 2007, cd.
Jones, Hank. Bop Redux. Muse mr 5123, 1977. Reissued, Pony Canyon Records 30059, 2002, cd.
Kelly, Wynton. Someday My Prince Will Come. Vee Jay vjs 3038, 1961. Reissued, Vee Jay nvj2-902,
1992, cd.
Kirk, [Rahsaan] Roland. We Free Kings. Mercury 20679, 1961. Reissued, Mercury 826455-2, 1991,
cd.
Konitz, Lee, and Hal Galper. Windows. Inner City ic 2057, 1977. Reissued, Steeplechase sccd
31057, 1987, cd.
Monk, Thelonious. Criss-Cross. Columbia cl 2038, 1963. Reissued, BMG/Sony 770514, 1993, cd.
. Thelonious in Action. Riverside rlp 12-262, 1958. Reissued, Original Jazz Classics ojccd
1861032, 1992, cd.
Morgan, Lee. Candy. Blue Note blp 1590, 1957. Reissued, Blue Note bnz 67, 1987, cd.
263
Morgan, Lee. Cornbread. Blue Note blp 4222, 1965. Reissued, Blue Note bnz 109, 1988, cd.
Navarro, Fats. The Fabulous Fats Navarro, Vol. 2. Blue Note blp 1532, 1956. Reissued, Blue Note
cdp 781532 2, 1989, cd.
Parker, Charlie. Charlie Parker, Vol. 1. Dial lp 201, 1946. Available on Charlie Parker: The
Complete Savoy and Dial Master Takes. Savoy svy 17149, 2002, cd.
. Summit Meeting at Birdland. Columbia jc 34831, 1977. Reissued, TriStar 80914, 1994, cd.
. The Magnicent Charlie Parker. Clef mg c-646, 1955. Blues for Alice is available on
Conrmation: Best of the Verve Years. Verve 314 527 815-2, 1995, cd.
Pass, Joe. Virtuoso. Pablo 2310 708, 1974. Reissued, Original Jazz Classics ojccd 31990-02, 2010, cd.
Peterson, Oscar. The Way I Really Play. MPS 15180, 1968. Reissued, Verve B0002233-16, 2003, cd.
. We Get Requests. Verve v/v6 8606, 1964. Reissued, Verve/PolyGram 5214422, 1997, cd.
Powell, Bud. The Amazing Bud Powell, Vol. 1. Blue Note blp 1503, 1955. Reissued, Blue Note cdp
7815032, 1989, cd.
The Quintet [Charlie Parker, Dizzy Gillespie, Bud Powell, Charlie Mingus, Max Roach]. Jazz at
Massey Hall. Debut deb 124, 1953. Reissued, Original Jazz Classics ojccd 044, 1990, cd.
Rodney, Red. One for Bird. Steeplechase sccd 31238, 1989, cd.
Rollins, Sonny, and Coleman Hawkins. Sonny Meets Hawk! RCA/Victor lsp-2712, 1963. Reissued,
RCA/Victor 09026 63479-2, 1999, cd.
Shearing, George. I Hear a Rhapsody: Live at the Blue Note. Telarc cd 83310, 1992, cd.
Shorter, Wayne. Speak No Evil. Blue Note blp 4194, 1964. Reissued, Blue Note cdp 7465092, 1987,
cd.
Taylor, Art. Taylors Wailers. Prestige 7117, 1957. Reissued, Original Jazz Classics ojccd 094-2,
1992, cd.
Woods, Phil, and Gordon Beck. Live at Wigmore Hall, London: The Complete Concert. JMS
186862, 1996, cd.
264
[email protected]
Michael McClimon www.mcclimon.org
Education
Indiana University, Bloomington, IN
PhD, Music eory
Dissertation: A Transformational Approach to Jazz Harmony
Minor fields: Music History and Literature, Jazz Studies
Teaching Experience
present Lecturer in Music, Furman University
Duties: Teach both written and aural components of sophomore music
theory (chromatic harmony, form, post-tonal techniques). Direct
independent studies with undergraduate music theory majors (topics have
included Schenkerian analysis, transformational theory, and late Romantic
symphonies). Team-teach a weekly composition seminar with the rest of the
theory/composition faculty.
Conference Papers
Reconsidering the Lydian Chromatic Concept: George Russell as Historical eorist.
Society for Music eory (St. Louis, MO, October )
Professional Activity
present Associate Webmaster, Society for Music eory
present Senior Editorial Assistant, Music eory Online
present Member, SMT Jazz Interest Group Awards Committee
Member-at-large, Society for Music eory Networking Committee
Digital Projects Specialist, Center for the History of Music eory and
Literature
Editor, Indiana eory Review
Soware Development Contractor, Indiana University, Music
Fundamentals Online
Editorial Assistant, Music eory Online
Various positions, Indiana University Graduate eory Association
Treasurer (), Symposium Program Committee (,
)
Various positions, Indiana eory Review
Circulation Manager (), Editorial Sta ()