Académique Documents
Professionnel Documents
Culture Documents
•••• ••••••••• •
•••••••••••••
•••••••• • • •
I8S58S .asassssssssasssis
MflTHEMflTlCRLrg ..::::::
._••••••
Mass::""""8ma • „ Franck Jedrzejewski
••
••••i
••••••••••••••••I
!••••••••••• • • • • I
m m
!•••••!
!•••••••!
!•••••••!
!••••• !••• !••••••_
!•••••! !•••••! •••••••• _
!••. ••••••••••••
••••••
E CHE R l C fl L E S
•• M U S i Q U E / S C i E N C E S
•••••••••••••••••••••••••••
••••••••••I
••••••I
i
• • • • • • • • • • • • • • • I! » !
II
••••••••••••••••
••• !••••••
•••••• •••••••I
• •••••I •
•••••••I ••••••••!
•••• ••••I
H •••••••••••••••••:
:sssi :gf •••
__•!__
!•••«!
••••••••I ••••••••I
•••I
!••«••!
ircam ElELATOUR • •m •a • • • • •
EDITIONS
Centre F R A N C E
Pompidou www.editlons-delatour.conn
•••••••
MATHEMATICAL THEORY
OF MUSIC
Collection Musique/Sciences
directed by Jean-Michel Bardez & Moreno Andreatta
Published works
Gerard Assayag, Prangois Nicolas, Guerino Mazzola (dir.), Penser la musique avec les
matMmatiques ? (2006)
Andre Riotte, Marcel Mesnage, Formalismes et modules musicaux, 2 vol. (2006)
Carlos Agon, Gerard Assayag, Jean Bresson (ed.), The OM Composer's Book (2006)
Forthcoming
Moreno Andreatta, Jean-Michel Bardez, John Rahn (dir.), Autour de la Set Theory
Guerino Mazzola, La virite du beau dans la musique
MATHEMATICAL THEORY
OF MUSIC
Pranck Jedrzejewski
Collection Musique/Sciences
ircam r r \w? n :i
Centre
•£ -J
bELATOuri
F R A N C E
Pompidou
Editorial Board
Carlos Agon, Ircam/CNRS, Paris
Gerard Assayag, Ircam/CNRS, Paris
Marc Chemillier, University of Caen
Ian Cross, University of Cambridge
Philippe Depalle, McGill University, Montreal
Xavier Hascher, University of Strasbourg
Alain Poirier, National Conservatory of Music and Dancing, Paris
Miller Puckette, University of California, San Diego
Hugues Vinet, Ircam/CNRS, Paris
Editorial Coordination
Claire Marquet
Cover Design
Belleville
Tous droits de traduction, d'adaptation et de reproduction par tous procedes reserves pour tous pays.
Le code de la propriete intellectuelle du ler juillet 1992 n'autorise, aux termes de Particle L. 122-5, 2e
et 3e a), d'une part, « que les copies ou reproductions strictement reservees a I'usage du copiste et non
destinees a une utilisation collective » et, d'autre part, « que les analyses et les courtes citations dans
un but d'exemple et d'illustration ». « Toute representation ou reproduction integrate ou partielle, faite
sans le consentement de I'auteur ou ayants cause, est illicite » (article L. 122-4). Cette representation
ou reproduction par quelque procede que ce soit constituerait done une contrefagon sanctionnee par les
articles L.335-2 et suivants du Code de la propriete intellectuelle.
Preface xi
1 Harmonic Models 1
1.1 Musical Sounds 1
1.2 Musical Intervals 2
1.3 Fourier Analysis 3
1.4 Physical Models 6
1.5 Sonance Models 15
1.6 Geometric Models 17
1.7 Affinity Models 21
4 Neo-Riemannian Investigations 83
Contents
Bibliography 299
Index 331
Preface
The aim of this book is to provide an introduction to the mathematical theory of music,
with emphasis on the fascinating and original connections with musical analysis. Despite
the complexity of the subject, we have tried to make this introduction accessible to
a large audience. Many examples are given for a better understanding. Musicians and
mathematicians are invited to contemplate some nontrivial aspects of the mathematical
construction.
The book is divided into eleven chapters. In the first chapter, we introduce the language
of harmonic models. After a review of some physical models based on the wave equation,
which try to modelize the inharmonicity of the spectrum of certain frequencies, we look
briefly at sonance models. These models are built around the notion of dissonance curves,
which go back to the Hermann Helmholtz's book Sensation of Tones. The next section
is devoted to affinity models introduced by Edmond Costere. Costere's theory could be
generalized to other weight systems, but the crucial question is to know what is the
best weight system for a given context and to give an algebraic demonstration of the
classification obtained. Costere has just given the results for a particular weight system
obtained from a computational software. Today, we have to understand what is behind
these representations and to give the properties of the associated weight systems. In the
last section, we review some geometric methods which lead to the theory of modulation.
Chapter 2 on Taxonomy of chords and scales, is devoted to the classification of pitch-
class sets. We first introduced the notion of a group action, and review Polya's enumer-
ation theory. Four main group actions are considered : the cyclic classification proposed
by Edmond Costere (1954), the dihedral classification proposed by Hanson (1960), Za-
lewski and Forte (1972), the affine classification proposed by the Czech composer Pinos
(1971) and the symmetric classification proposed by the Mexican composer Julio Estrada
(1994). The chapter ends with the enumeration of sets classes.
Chapter 3 investigates the question of determining the limited transposition sets for
a given temperament, and to find a general formula to enumerate them. This problem
is envisaged in the n-tone equal temperament. The enumeration is given in particular
Preface
for the 96-tone equal temperament (1/16 of tones) that is also used by some young
composers.
Chapter 4 is devoted to the Neo-Riemannian investigations. The chapter begins with a
survey of the main results of the Set Theory developed by Allen Forte, briefly illustrated
with short examples. The next section deals with David Lewin's transformation theory
and the Generalized Interval Systems. The legacy of Hugo Riemann is evocated through
the Riemaniann transformations which are used in musical analysis. In this section, we
propose a new interpretation of these transformations using permutations. The group
generated by the new Riemannian transformations is the dihedral group. The permuta-
tions are well-adpated to the design of chords's lattices and isographies.
Chapter 5 on Knots and Braids is one of the most important. It shows how to clas-
sify the 12! twelve tone rows with 554 Gauss diagrams. Schoenberg series are mapped
to singular knots with six double points. From knots or chord diagrams, universal in-
variants are computed. We hope that they will be a precious help in the computational
musical analysis of serial scores. They could also be used as a kind of measurement of
the aesthetic positions when comparing two scores. Braids are introduced to construct
new temperaments. Lastly, in textual analysis, low dimensional topologies are found as
an expression of hidden knots, in the same way that the famous linguist Ferdinand de
Saussure tried to find anagrams under the words.
Chapter 6 on Theory of Temperaments introduces the problem of temperaments, from
Pythagoras to the modern systems. General formulas are established for the division of
the Pythagorean comma and for the division of the syntonic comma. For each choice of a
set of frequencies, the question is to know if one could build an efficient harmonic theory,
which could help us to understand the interactions between notes in chords and musical
progression.
Chapter 7 on Approximations of temperaments begins with a survey on continous
fractions. The approximations of irrational temperaments (i.e. temperaments whose fre-
quency ratios are irrational numbers) by rational temperaments and the converse problem
are investigated. The approximation of the meantone temperament is generally based on
the use of rational approximation of some numbers such as the square root of 5. The
chapter ends with a short description of the Jacobi-Perron algorithm.
Chapter 8 on Tuning groups introduces many new problems. Among them, one difficult
question is to determine the Hellegouarch commas of a given group. Another is to find a
metric well suited to the measurement of harmonic distance.
Chapter 9 on Contemporary Tunings describes modern tunings used by 20 th century
composers, as Harry Partch, Warren Burt, Ben Johnston, Ezra Sims and others. The
chapter ends with a study of the cyclic tunings which have a strange hierarchical organ-
ization.
Chapter 10 deals with Groups and symmetries. We review the mathematical inter-
pretation of symmetry in art and in musical scores. The frieze groups are well known
and have been found in many scores. By substituting the atoms of a crystal lattice by
the frequency ratios derived from the coordinates of the atom, we introduce a new in-
terpretation of crystallographic temperaments. The Bravais temperaments are treated
as an example. Sporadic groups have been used by Olivier Messiaen. Constellations and
cartographic groups end the chapter.
In Chapter 11, Tesselations and Canons, the relationships between tilings and canons
are introduced with many examples. Some musical structures such as chords and twelve
tone rows could be used to tile in two or three dimensions. The classification of canons
xii
Preface
xiii
1
Harmonic Models
The science of sound studies the communication of sound messages created by vibrations
in the propagation medium. It depends on several connex domains. A sound is a physical
phenomenon, but also a perceptual phenomenon. Physical attributes of a sound such as
frequency and amplitude are distinct from perceptual correlates as pitch and loudness.
Sounds are created by compression and expansion of gas molecules in the propagation
medium.
£L = 4TT W / 2
I = pc
where P is the acoustic pressure. The acoustic power is measured in bels or in decibels
(dB). The level is defined from a threshold of hearing I0 = 1012 Watt/m 2 and P0 = 2-10" 5
Pa.
d £ = 10-log-^ = 2 0 - l o g ^ -
h Po
The limit of audibility is 0 dB. A low voice is about 20 dB, a normal voice around 40 dB
and a tutti of a symphonic orchestra about 100 dB.
The sum of two sine waves of closed but distinct frequencies is perceived as a single
sound whose intensity slowly oscillates from large to small values. These vibrations are
called beats. When the amplitude is large, the interference is said to be constructive and
when the amplitude shrinks to zero the interference is destructive. If the frequencies differ
by an amount A / the resulting sound
There are 100 cents in the tempered semitone and 1200 cents in the octave. In the n-tone
equal temperament (n-tet), the octave is divided in n tones. Two consecutive tones are
separated by the interval of 2*/ n or 1200/n cents. The ratios of the just intonation scale
are: unison (1), minor tone (10/9), major tone (9/8), minor third (6/5), major third
(5/4), fourth (4/3) and fifth (3/2). The following intervals will be used in the chapters
about tuning and temperaments.
2
Fourier Analysis
The Pythagorean comma is the difference between twelve just fifths and seven octaves
(3/2P = 3^ = 531441
p
27 219 524288
The syntonic comma is the difference between a major Pythagorean third (81/64) and
a just third (5/4)
_ 81/64 34 81 M
C. = —-.— = —:— = — = 22 cents
5/4 24-5 80
The Holderian comma is the semitone in the 53-tet
CH = 2 1 / 5 3 = 23 cents
DF = 2 1 / 3 1 = 39 cents
The septimal comma is the difference between the Pythagorean minor seventh (16/9)
and the just minor seventh (7/4)
_ 16/9 26 64 _ t
C7 = = 27centS
W = 3^7 63"
The leimma is the difference between three octaves and five just fifths
23 28 256 m
5 = TTF
5 = TTTTT = 90 cents
(3/2) 3 243
The apotome is the difference between seven just fifths and four octaves
. (3/2) 7 37 2187 _ +
A = ' 4 = TTTT
11 = ^TTTTT = 114 cents
2 2 2048
3
1. Harmonic Models
The first sine wave is the fundamental of frequency / . The two others are the second
and third partials with frequencies 2 / and 3 / . More generally, a periodic function f(t) =
f(t + T) of period T and frequency f = 1/T can be written as a trigonometric series
oo
cos
/(*) = y + X > (27r/nt) + bn sm(2irfnt)
n=l
The coefficients are called the Fourier coefficients and are given by the following formulas
a
2 frT
n = ™ / f(x) COS(2TTfnx)dx
1
Jo
and
2 fT
bn= sin 27r nx dx
T f^ ( f )
For non periodic waveforms, we use the Fourier transform, which is a generalization of
the Fourier series. The Fourier transform represents the magnitude of the sound against
the frequency. For a real or complex value function of the Lebesgue space L1, the Fourier
transform is
H-oo
/(«) = / e-2iirvtf(t)dt
-oo
J—o
and the inverse Fourier transform is given by the following formula
r +oo
/(*)= / ezznvtf(v)dv
J—c
Recall that the total energy of a signal is equal to the total energy of its spectrum. So
there is an equality between the L 2 -norm of a function and the L 2 -norm of its Fourier
transform called the ParsevaVs Formula.
P+00 i^ ,2 Z^oo 0
/ \f(v)\ dv= / \f(t)\2dt
/ * 9(v) = f(v)d(v)
Real sound data are usually analyzed by Fast Fourier Transform (FFT). Another kind
of generalization of the Fourier transform is the wavelet transform. Formally, a wavelet
is a real function ^(t) such that the constant
l~ I2
-l-oo
c# = 2TT / -—, ' dx
/ -oo l»l
4
Fourier Analysis
$(v) = v2e~v2l2
a
dt
V |a| Cy J-00 \ J
We discuss now a way of constructing spectral density of a random process X(t) and
we illustrate the use of the FFT on Olivier Messiaen's Mode de valeurs et d'intensites.
In the first voice, Messiaen uses a kind of series. Each note is encoded from 1 to 12. The
time series is composed of the pitch variations X(t) as a function of time, where X(t)
belongs to the set {1, 2, ..., 12}. The data has the following representation
where X is the mean value of pitch variations. The graph of the correlation function is
shown in the next figure.
5
1. Harmonic Models
20 40 60 80 100 120
The corresponding power spectrum is obtained by Fast Fourier Transform (FFT) of the
autocorrelation function.
^ l2 = c2Au + F(t,x)
dt
where t is a positive real (the time) and x an element of R 3 . This equation is subject to
initial conditions
u(0,x) = f(x)
and
-j£(09x)=g(x)
It has been shown that the wave equation admits the solution
6
Physical Models
<^r sin£|MI
\\v\\
and ^
A # t ( t O = cos(t||t/||)
d2u _ 2d2u
dt2 " ° dx2
Suppose that the string of length £ is fixed at both ends. With the boundary conditions
u(0,t) = u(£,t) = 0, and initial conditions u(x,0) = f(x) and du(x,0)/dt = g(x),
D'Alembert shows that the general solution of this equation is given by the sum of
two arbitrary functions representing two waveforms, one traveling to the right and one
traveling to the left, each at the velocity c.
Prom the boundary conditions, we get g(x) = —f(—x). Thus, the solution becomes
where / satisfies
f(x) = f(x + 2l) VxeR
Since / is a periodic function of period 2£9 it admits a Fourier series expansion. If the
fundamental is a sine wave
/ ( z ) = a s i n ( — +tp)
. (nirct\ . /nirxx
It is well-known that the frequency of the n-th partial is given by the formula
fn =
Ye]j^ « = i.2,3,...
where T is the tension of the string and p is its linear density (c = y/T/p).
1) For a string fixed at one end and free at the other, the mode frequencies are given
by
f - -
Inharmonicity models are yet far from reality. At the end of the 19 th century, Lord
Rayleigh proposed to describe the ideal string as a bar of length I
where K2 = I/S is the sectional moment of the bar, p is the density and E the Young's
modulus. This equation can be approximated by the Euler-Bernoulli beam equation which
is a fourth-order differential equation
d2 u 2d
4
u n
with c 2 = EI/pS. If we look for separable solutions u(t,x) = f(x)g(t), it is easy to show
that solutions are given by
with
c
2) If the string is free at both ends, the boundary conditions at x = 0, d2u(t, 0)/dx2 = 0
and d3u(t,0)/dx3 = 0 imply
and
k3(C - A)sm(wt + <p) = 0
that is C = A and D = B.
At x = £, the boundary conditions d2u(t,t)/dx2 = 0 and d3u(t,t)/dxz = 0 yield
and
ks [i4(cosh kt - cos kl) + B(sinh kt - sin kt)] sin(wt + ip) = 0
That is
^4(sinh kt - sin kt) = £(cosh kt - cos kt)
,4(cosh kt - cos kt) = £(sinh kt - sin kt)
8
Physical Models
Eliminating A and B by dividing the first equation by the second, and cross-multipling
yield to
cosh2 kt — 2 cosh kt cos kt + cos2 kt = sinh 2 kt — sin 2 kt
2-2coshAtfcosAtf = 0
or
cos A cosh A = 1 (A)
with A = kt. The allowed frequencies are given by
\lc \l I EI
fn 2 2
2TT£ 27rt ]j pS
where An are solutions of the previous equation (A). Using the identities
o 1 — cos 2x
tan x = - —
1 + cos 2x
and
.o cosh 2x — 1
tanh x = — —
cosh 2x + 1
the equation (A) can be rewritten as
tan
( Y ) = ±tanh(-rr)
where the numbers an tends rapidly to zero when n grows. The interval ratio between
two partials is then
2n + 3 x 2
,2n + l
or [Hueber, 1988]
n + 3 /(n + l ) 2 + l
rn
~ n y ( n + 2) 2 + l
The frequency ratios from the fundamental is the product of each interval ratio (or the
sum when the ratios are expressed in cents) modulo the octave.
n
9
1. Harmonic Models
11n An
4.7300407448
rn (cents)
1755
vn (cents)
555
2 7.8532046240 1165 520
3 10.9956078380 870 190
4 14.1371654912 695 885
5 17.2787596573 578 263
6 20.4203522456 496 759
7 23.5619449020 433 1192
8 26.70353756 385 377
9 29.84513021 347 724
10 32.98672286 315 1039
11 36.12831552 289 128
12 39.26990817 267 395
13 42.41150082 247 642
14 45.55309348 231 873 |
l + cosAcoshA = 0 (B)
tan(-) = ±coth(-)
The frequencies are
where An are solutions of the previous equation (B). These solutions are of the form
[Rayleigh, 1894]
An = i ( 2 n - l ) 7 r - ( - l ) ^ n n = 1,2,3,.-.
The numbers /3 n tend to zero as n tends to infinity,
/2n + l \ 2
10
Physical Models
f(r) = Jn(w-c)
where J n is the n-th Bessel function. The solutions of the wave equation are
T
u(t} r, 0) = AJn(w-) sm(wt + (p) sin(n# + i>)
If the radius of the circular membrane is i?, the solution verifies the boundary condition
f(R) = 0. It follows that w could only takes the values
c
w
— Jnim~^
where j n % m is the m-zero of the n-th Bessel function. The frequency ratios between two
adjacent partials are
Jn,m
^n,m —
Jn—lyTi
and
In vn
Vn = T\T2'"rn mod U vn = 1200
In 2
where the product is taken modulo 2. The numbers vn are the ratios vn are expressed in
cents. The following values show the inharmonicity of the first nominal modes
n rn (cents) vn (cents)
1 806 806
2 507 114
3 376 489
4 300 789
5 251 1040
6 216 56
7 190 246
8 169 415
9 153 568
10 140 708
11
1. Harmonic Models
u(x,y,t) = f(x)g(y)h(t)
dx2
d29(y) +, k. 22g(y)
, = 0
dx2
If the membrane is fixed on the boundary u = 0 for x = 0, x = £x, y = 0 and y = (,y, the
solutions are given by
- ^ I IV lib
/n m = +
' 2y£[ 72-
V 4 / = k4 f
and
g"(t) + w2g(t) = 0
The numbers k and w are related by
k2=w = w /I2p(l-a3)
c h V £
The biharmonic equation is factorized by
( V 2 - * 2 ) ( V 2 + * 2 )« = 0
12
Physical Models
So the solutions can be written uniquely as a sum u\ + ^2, where u\ satisfies V2u = k2u
and U2 satisfies V2u = —k2u. Introducing the hyperbolic Bessel functions
f j4Jn(fca) + B I n ( f c a ) = 0
\ Aj;(fo) + fl/;(fca)=0
Allowed solutions of k depend on integers n, m. They are asymptotically given by
(n + 2ra)7r
£•
"'n,?! 2a
when n tends to infinity. Modal frequencies are
f -—k2
- ch
/o,i « <*-«2
a'
with a « 1.618. The ratios a n>m of the modal frequencies
Jntm — ^n,m/0,1
2) If the plate has a free edge, the fundamental mode is now obtained forn = 2, m = 0
and
13
1. Harmonic Models
i—l
2 11.40 4213
2 1 6.71 3296
2 2 15.97 4797
3 0 2.328 1463
3 1 11.07 4162
4 0 4.11 2447
3) If the plate has a simply supported edge, it can be shown that the fundamental
mode is obtained for n = 0, m = 1 and
j, ch
Jn>m '7n,77i/0,l
are given in the following table for the first values of n and m
This definition proposed by ltd leads to a new differential calculus. For example, it is
well known that in the integral
Ws dWa = \w? - I
Jo
it appears a supplementary term t/2. This equation could also be rewritten in the dif-
ferential form
dWf = 2WtdWt + dt
14
Sonance Models
More generally, for a stochastic process Xt under suitable conditions, we consider the
stochastic differential equation
+b(t,Xt)MhMdWt
dp = ld2(b2(t,y)p) _ d(a(t,y)p)
dt 2 dy2 dy
We have seen in the first chapter, that the governing partial differential equation of a
vibrating membrane is
d2u 2.
If the vibrating membrane is disturbed by a random noise £(£), the equation becomes
and could be treated as a stochastic differential equation. The solutions of the stochastic
equation are new frequencies which tend to the original frequencies when noise goes to
zero. In some cases, inharmonicity could be interpreted as a random perturbation.
15
1. Harmonic Models
simple frequency ratios. The simplest ratios, such as octave (2:1), fifth (3:2), fourth (4:3),
major third (5:4) and minor third (6:5), are consonant intervals. A simple ratio is defined
as a ratio involving small integers or, for some theorists, small prime numbers. In the
Pythagorean meaning, harmonic laws govern also the motion of the planets. The whole
cosmos is embedded in numbers.
Some theorists explain consonance by tonal fusion [Stumpf, 1898] when tones fuse per-
ceptually: Two tones are consonant if they sound like a single tone. Others evoke amp-
litude fluctuations, rapid beating [Helmholtz, 1877], virtual pitches [Terhardt, 1974], ex-
pectation violations [Cazden, 1980] or tonotopic interferences [Greenwood, 1961], [Plomp,
1965], [Kameoka, 1969].
As described in Sensations of Tone, Hermann von Helmholtz [Helmholtz, 1877] in-
troduced the notion of sensory consonance and dissonance by comparing violin tones
sounded together. Helmholtz calculated the resulted dissonance by considering rough-
ness as a function of beats. Fast beats are considered to be rough with a maximum at 32
beats per second. Helmholtz drew dissonance curves by plotting roughness produced by
partials against intervals. Maximal consonances are obtained for the unison, the octave
and the fifth, followed by fourth, major sixth and major third. For each frequency in the
range of an octave, dissonance curves are formed by interactions between a given partial
of the note and the nearest partial of a referenced note labelled c' [see Helmholtz, p. 193].
Like Helmholtz, Harry Partch has drawn a graph of comparative consonance, the One-
Footed Bride [Partch, 1949]. It shows the relative consonance of each interval of his 43
just tone scale. The graph is symmetric around 600 cents and ratios (r, 2/r) are coupled.
"In this exposition 2:1, 3:2 and 4:3 (ratios of the 3-limit) are the Inter-
vals of Power. Those ratios that lie between 4:3 and 3:2 ("tritone" intervals)
-27:30, 11:8, 7:5, 10:7, 16:11, 40:27 are the Intervals of Suspense. The ra-
tios between 21:16 and 7:6, at the threshold of the descent to 1:1, and those
between 32:21 and 12:7, at the threshold of the ascent to 2:1 are the Emotional
Intervals. The ratios starting with 8:7 and descending to 1:1 and those start-
ing with 7:4 and ascending to 2:1, are the Intervals of Approach." [Partch,
1949, p. 156-157].
The tonotopic theory of sensory dissonance was first proposed by Greenwood [Green-
wood, 1961] and extended by Plomp and Levelt [Plomp, 1965], and Kameoka and Kuriy-
agawa [Kameoka, 1969]. In the mid 60's, Plomp and Levelt [Plompt, 1965] asked a large
number of listeners to judge the consonance of a variety of intervals, when sounded by
pairs of pure sine waves. They showed that the interval of maximum sensory dissonance
remains fairly constant relative to the critical band and they observed that in almost all
frequency ranges, the maximal point of roughness occurred at about 1/4 of the critical
bandwidth. The data of the experiment shows an averaged dissonance function of the
form
h(x) = e~^x - e~^x
with /?2 = 3.51 and P1 = 5.75. They also calculated the consonance for two tones, each
with six harmonics. Their results were confirmed in experiments carried out by Kameoka
and Kuriyagawa [Kameoka, 1969] who have proposed a model of calculating the absolute
dissonance of complex tones. A mathematical model is designed in the book of Sethares,
see [Sethares, 1998, p. 300].
Let S be a spectrum with partials at frequencies f\ < $2 < ... < fn and respective
amplitudes ai, <22, ..., an. Sethares defines the dissonance of the sound S by the sum of
16
Geometric Models
with
„, *\rfj-sfi\
13
a2mm(rfj,sfi) + a1
where h(x) = e~^x — e~P^x is the previous function, the constants K = 0.24, a\ = 18.96,
a2 = 0.0207 and n is the number of partials. The dissonance of S at an interval r is given
by
D(r) = H(l, 1) + H (r, r) + H(l, r)
The dissonance of a chord of three notes at the intervals 1, r and s is defined by
Assuming that all amplitudes are equal (a* = 1), it is easy to show that the dissonance
curve D(r) has at most 2n 2 local minima. These minima tend to be located at intervals
r for which r = fa/fj for some partials S. Remark that we could define several models
by changing the function h(x).
Jean-Marc Chouvel [Chouvel, 1998, p. 156] defined the concordance of two partials by
introducing a Gaussian function. For the interval r, the concordance is
CdB{r) = mogl0C(r)
C ( r , 5 ) = C ( r ) + C( 5 ) + C ( r - 5 )
For several sounds, Chouvel obtains dissonance graphs similar to Sethares plots.
17
1. Harmonic Models
Examples. Chords, scales, motives, onsets and most of the usual musical structures are
local compositions. The chords (^4, Z n ) modulo the octave form a local composition over
the n-tone equal temperament R = Zn. For n = 12, we recover the space of classical
harmony.
The following theorem has been established by [Mason, 1970] and generalized in
[Mazzola, 1985]. Pcsets are defined in the chapter 4.1 on Set Theory.
Theorem 3 Let n be a positive integer, and let A, B be two pcsets of same cardinality
m. Then there are two non-negative integers k (called sharps) and I (called flats) with
k-\-l = m — n and k>l>0 and a set of integers called alterations
such that any pc-set A can be translated to any equipollent pcset B by some alterations.
In particular, in the classical temperament (n=12) any scale of seven notes is mapped to
any scale of the seven elements by 3 sharps and 2 flats.
Example. Remark that the definitions of sharps and flats can be exchanged. For example,
A = {0,1,2,3,4,5,6} and B = {0,1,7,8,9,10,11}, the subset {0,1,2,3} of A is mapped
to {10,11,0,1} by the translations of —2, and the subset {4,5,6} is mapped to {7,8,9}
by the translations of 3.
Definition 4 Let (S, M) be a local composition over the ring R. The left module R.S
is generated by the set of {s — x,s G S} for any element x of S. The module is free if it
admits a basis.
Proposition 5 The map F : S —> R.S is a functor from the category of local composition
LOCR to the category of modules ModR.
Mazzola defines also the category of ObLoc of objective local compositions whose
objects are the collection of all local compositions and morphisms are morphisms between
local compositions. He has shown that there is a fully faithful functor from ObLoc to Loc.
C= {0,2,4,5,7,9,11}
18
Geometric Models
VII HI VI II
Example 2. In the diatonic Pythagorean tuning, the frequency ratios are generated by
pure fifths {C (1), D (9/8), E (81/64), F (4/3), G (3/2), A (27/16), B (243/128)}. The
19
1. Harmonic Models
notes are represented by the powers of 3 in the previous ratios: F (-1), C (0), G (1),
D (2), A (3), E (4), B(5). For the major scales, the minimal cadential sets are the five
same sets of the 12-tone equal temperament. For the harmonic minor scale, there are
16 minimal cadential sets, the singleton {III} and all pairs of any two different degrees
different of III. For the melodic minor scale K = {0,2,3,5,7,9,11}, there are 10 minimal
cadential sets Jx = {III}, J2 = {I, II}, h = {I,V}, J 4 = {I,VII}, J5 = {II,IV},
Je = {II, VI}, J7 = {IV, V}, J 8 = {IV, VII}, J9 = {V, VI}, Jio = {VI, VII}.
Example 3. In the Euler Plane, the frequency ratios are given by {C (1), D (9/8), E (5/4),
F (4/3), G (3/2), A (5/3), B(15/8)}. In the basis <3,5>, the notes are represented by the
coordinates determined by the powers of 3 and 5 in the previous ratios, namely C (0,0),
D (2,0), E (0,1), F (-1,0), G (1,0), A (-1,1), B (1,1). The major scale has six minimal
cadential sets Jx = {II}, J2 = {VII}, J 3 = {III,VI}, J4 = {III,IV}, J 5 = {IV,V},
JQ = {V, VI}. The melodic minor scale has also six minimal cadential sets, J\ = {I},
J2 = {II}, J 3 = { / / / } , J 4 = {VI}, J 5 = {VII}, J 6 = {IV,V}. The harmonic minor
scale has 9 minimal cadential sets J\ = {II}, J2 = {HI}, ^3 = {VII}, J4 = {I,IV},
h = {I, V}, J6 = {I, VI}, J7 = {IV, V}, Js = {IV, VI}, J9 = {V, VI}.
Definition 13 Let K be a scale, L = Tn(K) the translated scale and if3, L3 be two tri-
adic sub-interpretations. A modulator is a map defined by the composition of the transla-
tion Tn with an inner symmetry of K which induces an isomorphism between the triadic
sub-interpretation K3 ~ L3.
Example. Let K = {0,2,4,5,7,9,11} be the C-major scale, L the C() major scale L =
Ti(K). The map s(x) = 4 — x defines an inner symmetry of K. Let K3 = {He, Vc} =
{2,5,7,9,11} and L 3 = {IIc$, Vc$} = {0,3,6,8,10} be two sub-interpretations. The map
m(x) = T\ o I(x) = 5 — x defines an isomorphism between K% and L3.
Proposition 16 The setTnQ is covered by triads ofT^y The degrees of the interpret-
ation (T fl Q)(3) of the trace Q inT are called the pivots.
Example. The C-major scale is denoted K = {0,2,14,5,7,9,11} and the C(t major scale
L = {0,1,3,5,6,8,10}. Let J = {II, V} be a minimal cadential set of J in the 12-tone
equal temperament. Let
K3 = {IIC,VC} = {2,5,7,9,11}
and
£3 = {IIct,Vct} = {0,3,6,8,10}
20
Affinity Models
The numbers R = {^(j)}jex define the affinity vector of the setL. A collection of affinity
vectors is called an affinity table.
Definition 19 Let R = {^s(j)}jex be the affinity vector of the set L, the cardinal density
(or tonal gravity,) of a set B G A is
d{B) = 5>(i)
The cardinal density of each perfect chords defines two tables, one for the major triads
and another for the minor triads.
21
1. Harmonic Models
Example. The pentatonic scale (called the Chinese pentatonic scale by Costere) L =(C,
Oji, F&, Gtt, Bb)={0, 3, 6, 8, 10}
'<£ „ J. II- ^ ^
has the following affinity table:
C c« D Dtt E F Ftt G Gtt A Bb B
c 1 I 0 0 0 1 0 1 0 0 0 1
DB 0 0 1 1 1 0 0 0 1 0 1 0
0 1 0 1 1 1 0 0 0 1
n
Gfl 0 1 0
0
1
0
0 0 0 1 1 1 0 0
Bb 0 0 0 1 0 1 0 0 0 1 1 1
R 1 (3 1) 3 (1 3) 1 (3) 2 (2) 2 (3)
The last row is the affinity vector. Parentheses indicate the notes that are not belonging
to the set L. For each major chord, we compute the tonal gravity. The set Mi = [0,4,7]
has the density d(M\) = 1 + 1 + 3 = 5, the set M 2 = [1,5,8] has the density d(M 2 ) =
3 + 3 + 2 = 8, etc. The same is done for the minor chords. The set mi = [5,8,0] has the
density d(rai) = 3 + 2 + 1 = 6, the set m 2 = [6,9,1] has the density d(m2) = 1 + 2 + 3 = 6,
etc. Results are presented in the following table.
[W 8 4 8 6 6 6 7) 6 (6 6 ~W]
\ (6 6 6 8 4 8 5 7 6 6) 6 (7) |
The first row is the density of the major triad from the first chord based on 0 to the
chord based on 11. The second row is the minor chord from the chord finishing by 0 to
the chord finishing by 11. Numbers outside parentheses indicate chords for which all the
notes of the chord belong to the set L. The major chord [8, 0, 3] belongs to L, as well as
the minor chord [3, 6, 10].
Remark that additional weights are given by Costere to reinforce some notes (for
example if a note is repeated or doubled). The affinity table is also applied to microtones
(quarter tones) in [Costere, 1962, p. 35] or to other musical objects [Costere, 1974, p. 184].
It is clear that there are several models according to the definition of the primary weights.
For example, Terhardt considers four relations: the descending fifth, the descending major
third, the ascending and descending major second. In this case, the affinity table of the
pentatonic scale would be
Costere uses the tonal gravity to partition the musical set A in four distinct categories.
Let L be a set of A. Define the following maxima. The first maximum is taken over all
22
Affinity Models
u = max d(t)
v
tcL '
and the second maximum is taken over all perfect chords that are not included in the set
L.
v = max d(t)
Definition 20 A set L is tonally stable or simply stable (S) if the maximum of the
tonal gravity of the constitutive perfect chords is greater than the maximum of the tonal
gravity of the non-constitutive perfect chords, in others words if u> v, i.e.
Definition 21 A set L is balanced (B) if the maximum of the tonal gravity of the con-
stitutive perfect chords is equal to the maximum of the tonal gravity of the non-constitutive
perfect chords, in others words ifu = v, i.e.
Definition 22 A set L is cadential (C) if the maximum of the tonal gravity of the
constitutive perfect chords is less than the maximum of the tonal gravity of the non-
constitutive perfect chords, in others words ifu<v, i.e.
Definition 23 A set L is jarred (J) if the maximum of the tonal gravity of the con-
stitutive perfect chords is less than the maximum of the tonal gravity of the non-constitutive
perfect chords, in others words ifu<v, i.e.
23
1. Harmonic Models
Costere also distinguishes between tonic (T) and non tonic sets (T c ). There are 213
tonic musical sets and 138 non-tonic sets, with the following division, see [Costere, 1954].
s B C J
T 51 65 48 49
rpc 30 37 36 35
T U Tc 81 102 84 84
In the vocabulary of Costere, some properties are structural (not depending on affin-
ity) and others are cardinal properties (depending on affinity). Atonality is a structural
property and polytonality a cardinal property. We investigate first some structural prop-
erties.
Definition 25 A set L has limited intervals if one interval does not belong to L.
For example, the set L = [0, 4, 8, 10] has no fifth. The set [0, 7, 9] has no major third.
Definition 26 A transposition is a map Tn from A to A, defined by
Tn(x) = x + n mod 12
Definition 27 A set L has limited transpositions if two transpositions give the same
set.
For example, the whole-tone scale L = [0, 2, 4, 6, 8, 10] has only two transpositions L
and Ti(L) = [1, 3, 5, 7, 9, 11]. We will explore this concept in details in chapter 3.
Definition 28 A set L has a median symmetry if placed on the unit circle (by the
application k £ L —> (cos(2A;7r/12),sin(2fc7r/12)) a diameter is a symmetry axis.
For example, the set L = [0, 1, 5, 6, 9] has a symmetry around the axis (3, 9).
Definition 29 The inversion of a set L is the map I from A to A, defined by
I(x) = —x mod 12
and the inversion of order n is
In{%) = —x + n mod 12
Definition 30 A set L is reversible if X\L or X\I(L) is a transposition of L. In other
words if it exists m^0 such that
Tm(X\L) =L or Tm(X\I(L)) = L
For example, the set L = [0, 2, 4, 7, 9, 11] is reversible because TQ(L) = X\L. The set
L = [0, 3, 4, 5, 7, 11] is reversible because Ti(X\J(L)) = L.
Theorem 31 There are 94 limited intervals sets, 15 limited transpositions sets (excluded
the chromatic scale), 50 sets with median symmetry and 32 reversible sets with the fol-
lowing division
S B C ~T\
Limited intervals 22 25 20 27
Limited transpositions 3 0 0 12
Median symmetry 16 6 8 20
Reversible 8 3 12 9
where S are the stable sets, B the balanced sets, C the cadential sets and J the jarred sets.
24
Affinity Models
In Costere's theory, cardinal properties are considered in order to investigate the stability
and tonality of sets.
Definition 32 A set Lofn notes is dense or cardinally stable if the density of the set
L is greater or equal to the maximum density of n notes not belonging to the set
1 (2 1 0 1 2) 1 (2 I 1) 1 (1 I 2)
The density of the set L is d(L) = 1 + 1 + 1 = 3. It is less than the density of [1, 5, 7].
The Tristan chord L = {F, B, Dtf, G(t} has the following affinity table
The density of the weights of the set is d(L) = 2 + 1 + 2 + 1 = 6. Let N be the set
N = {C, E, Gb, Bb}, then d(N) = 2 + 3 + 2 + 3 = 10. Thus the Tristan chord is cardinally
unstable or transitive.
Definition 33 A set L has a tonic pole if it exists i G L such that
For example, the set L = [0, 6, 9] of the previous example is cardinally balanced,
because each element of L has the same affinity equal to 1 (see the affinity vector above).
Definition 35 A set L is modulant if there is equality of the affinity for all notes of
at least one perfect chord. In other words, L is modulant if for all i G L, one of the
following relations is true
(1) 3(i) = 3(t + 3) = 9(i + 7)
(2) 3(i) = 9(i + 4) = 9(t - 7)
(3) 9 ( 0 = 3(» - 3) = 3(t + 7)
(4) 3(i) = 3(z - 4) = 9(i - 7)
where the sums are taken modulo 12.
25
1. Harmonic Models
2 (1 2 1) 2 (2 1) 2 (2) 2 (1 2)
One could verify that for each element of L the second relation of the previous definition
is true.
Theorem 36 There are 105 transitive sets, 152 sets with tonic poles, 33 cardinally bal-
anced sets and 159 modulant sets with the following division
S B C J\
Transitive 4 15 42 44
Tonic poles 43 42 39 28
Cardinally balanced 8 4 6 15
Modulant 55 51 31 27
where S are the stable sets, B the balanced sets, C the cadential sets and J the jarred sets.
Investigations around tonality, modality, polytonality and atonality combine structural
and cardinal properties.
For example, the whole-tone scale L = [0, 2, 4, 6, 8, 10], the diminished seventh chord
[0, 3, 6, 9] the cluster [0,1, 2, 3, 4] and the set [0, 1, 2, 4, 10] are atonal sets.
Definition 38 A set L is neutral if it contains only neutral fifth, that is fifth without
associated minor or major thirds.
For example [0, 1, 2, 6, 7, 8], [0, 1, 2, 7, 8, 9] and [0, 1, 2, 3, 4, 5, 6] are neutral sets.
Definition 39 A set L is binary when it contains (at least) one couple formed by the
major and the relative minor triad.
A binary set could contain several couple of major and minor triad. For example
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10] contains six couples. Recall that if [z, i + 4, i + 7] is a major
triad, the relative minor triad is the set [z, i + 3, i + 7] (mod 12). The set [0, 1, 2, 4, 8, 9]
is a binary set. It contains a major chord [9, 1, 4], the relative minor chord [9, 0, 4] and
a minor chord [1, 4, 8].
Definition 40 Let V+ be the set of major perfect triads and V be the set of minor
perfect triads. The inverse minor map opp from V^ to V~ associates for all major triad
t = [i, i + 4, i + 7], the opposite perfect triad opp(t) = [i — 7, i — 4, i\.
26
Affinity Models
Comparision of weight systems are not easy. For example, if we take the chord {G, B,
D, F} we could calculate the affinity table
Gfl Bb
G
C
1
CJ
0
D
1
D«
0
E
0
F
0
ni G
1 1
A
0 0
~B1
0
B 1 0 0 0 1 0 i 0 0 0 1 I
D 0 1 1 1 0 0 0 1 0 1 0 0
F 1 0 0 0 1 1 1 0 0 0 1 0
E (3 1) 2 (1 2) 1 (3) 2 (1 1 2) 1
and deduce the densities of the major and minor perfect chords, namely
Gtf Bb
Maj
C
7
c»
3
D
6
D«
5
E
4
F
5
n6 G
5 5
A
4 5
B
5
Min 5 5 6 3 6 4 6 6 4 4 6 5
We see that the only major chord of highest density is {C, E, G} and that they are 5
minor chords of density 6. Two chords are related to fifth affinity: {C, Eb, G}, {A, C,
E} and three chords are related to chromatic affinity, {Eb, Gb, Bb}, {G, Bb, D} and
{B, D, F§}. We now look at the Terhardt weights, the affinity table is
C c» D Dtt E F Fit G G» A Bb ~B1
G 1 0 0 1 0 1 0 1 0 1 0 0
B 0 1 0 0 1 0 0 1 0 1 0 1
D 1 0 1 0 1 0 0 1 0 0 1 0
F 0 1 0 1 0 1 0 1 0 0 1 0
E (2 2) 1 (2 2) 2 (0) 4 (0 2 2) 1
27
1. Harmonic Models
Two major chords are dominant of weight 8 {C, E, G} and {Eb, G, Bb}. The only minor
chords of the same density is the chord {C, Eb, G}.
Consider for example the first prelude extract from the 24 Preludes of Maurice Ohana.
The first measure is a cluster of eight notes (Bb, C, A, B, D, Ab, Db Eb). The affinity
table calculated with the Costere weights shows an homogenious repartition of the affinity
between 3 and 4 for the pitches of the cluster:
c c» D DJ E F F» G GB A Bb B
Ab 0 1 0 1 0 0 0 1 1 1 0 0
A 0 0 1 0 1 0 0 0 1 1 1 0
Bb 0 0 0 1 0 1 0 0 0 1 1 1
B 1 0 0 0 1 0 1 0 0 0 1 1
C 1 1 0 0 0 1 0 1 0 0 0 1
Db 1 1 1 0 0 0 1 0 1 0 0 0
D 0 1 1 1 0 0 0 1 0 1 0 0
Eb 0 0 1 1 1 0 0 0 1 0 1 0
E 3 4 4 4 (3 2 2 3) 4 4 4 3
28
2
Taxonomy of Chords and Scales
The purpose of this chapter is to propose a general framework of chords classification us-
ing groups action and to rediscover the cyclic classification proposed by Edmond Costere
(1954), the dihedral classification proposed by Howard Hanson (1960), Zalewski and Al-
len Forte (1972), the affine classification proposed by the Czech composer Alois Pinos
(1971), and later by Daniel Starr (1978) and Guerino Mazzola (1985), and the symmetric
classification proposed by the Mexican composer Julio Estrada (1994) (see the review
[Verdi, 2003]). The pitch-class sets theory was extensively studied since the beginning of
the 20 th century, from a theoretical [Lewin, 1960], [Martino,1961], [Howe, 1965], as well as
compositional perspective [Babbitt, I960]. The theory presents the numbering of possible
combinations of the twelve pitches of the equal temperament and was widely studied by
Edmond Costere [Costere, 1954] and by Allen Forte in the 1970,s [Forte, 1973]. Forte's
theory classifies major and minor chords under the same structure and thus is limited
to non-tonal music. In Costere's theory, the inversion used in the dihedral classification
must be excluded. This then leads to a new hierarchy of sounds which no longer has
223 but 351 pitch class sets (corresponding to the cyclic or dihedral group action). We
will illustrate it by examples taken from musical score of the beginning of the century
or from works using enlarged tonality, modality or polytonality [Milhaud, 1923]. The
modes of limited transposition, which were largely used by Olivier Messiaen [Messiaen,
1944], take on a significant structural characteristic for the organization of these classes.
The classification of hexachords is particularly revealing as it shows the tropes of Josef
Matthias Hauer [Hauer, 1926] who put forth at the same time as Schoenberg a different
serial theory. Mazzola's classification in 157 classes and Estrada's classification are also
presented in this chapter. The cyclic classification which includes all the structures of the
dihedral classification, appears to be well-suited to the analysis of pieces in which the
tonal functions are more or less abandoned, like Franz Liszt's last works, Bela Bartok's
Sonates for violin and piano (the first sonata is based on C sharp minor and the second
on C major) or the polytonal works of Frank Bridge and Darius Milhaud, or the modal
works of Olivier Messiaen, Jacques Charpentier, Charles Koechlin, Henri Dutilleux, and
2. Taxonomy of Chords and Scales
even the atonal works. The complementation function is then its implicit definition. It
structures the hexachords and determines the tropes of Joseph Matthias Hauer. The
cyclic classification of Edmond Costere explains analytically and with extreme accuracy
the harmonic mutations of modern and contemporary music up to the advent of serial
techniques. Mathematical investigations and applications of group-theoretic methods in
music theory have been done by D. Halsey, E. Hewitt [Halsey, 1978], H. Pripertinger
[Pripertinger, 1992, 1993, 1997, 1999], M. Andreatta [1995, 2003, 2004] and by Th. Noll
[Noll, 2004].
3 g G G, y = g.x
The orbits are the equivalence classes of this relation. Two elements x, y of X are in the
same orbit if there is an element g of the group G such that y = g.x.
Definition 46 The orbits of G on X are the equivalence classes G(x) of the previous
equivalence relation
G{x) = {g.x \geG}
The set of all orbits is denoted by G\\X = {G(x) | x G X}. The action is transitive (or
G acts transitively,) if there is only one orbit
Definition 47 For each x G X, the stabilizer Gx of x is the set
Gx = {g G G I g.x = x}
Xg = {x G X | g.x = x}
Examples. (1) A subgroup H of G acts by multiplication from the right on G. The orbit
H(g) is the left coset gH. In other words, two elements g, g' G G are in the same left
coset of H if there is an element h G H such that g' = hg. The left coset is denoted by
G/H or gH
G/H = {gh\heH}
The set of the right cosets is denoted by H\G or sometimes Hg.
(2) A group acts on itself by conjugation G x G -» G, (#, h) \—• ghg~l. The orbit
is the conjugacy class of h. The stabilizer of h is the set of its commutative elements
Gh = {g € G, gh = hg}
30
Group Actions
Lemma 48 Let H be a subgroup of G. There is a bisection between the right and the left
cosets. In particular, the sets have the same cardinality
\G/H\ = \H\G\
The map Hg —• g_1H is an embedding from G/H in H\G. It is onto because all classes
gH can be rewritten as (g~1)~1H. •
\(G:1)\ = \G\
Definition 50 A subgroup H ofGis normal if the right cosets are equal to the left cosets
V<?eG, gH = Hg
Theorem 51 (Lagrange) The order and the index of a subgroup H of a finite group
G are divisors of the order of group G
\(G:H)\ = \G\/\H\
Proof The map h —> gh is a bijection from H to gH. All classes gH forms a partition
of distinct subsets of the number of the left cosets. But the number of left cosets is the
index of H in G, so
\G\ = \(G:H)\.\H\
a
Theorem 52 The order of the orbit of x is equal to the index of the stabilizer of one of
its elements
\G(x)\ = \G:Gx\
Proof The map sending the left coset gH to the element g.x G X i s a bijection between
the left cosets of H in G and the orbit of x. •
31
2. Taxonomy of Chords and Scales
\G\\X\ = -±-J2\X9\
Xg = {x € X, gx = x}
Proof
771 El G «
= 1IGI
x€X
- 1
wi^^1
' xex 9eGx
1
' g€G xeXg
1
*geG
D
In order to introduce P6lya's enumeration theorem, we need a generalized version of
this lemma. Let R be a commutative ring such that Q is a subring of R and G be a finite
group acting on a finite set X. A weight function is a function w : X —> R constant on
each G-orbit, i.e.
w{g.x) = w(x) Vg G G, Vx e X
The weight of an orbit G(x) is defined as the weight of any element of this orbit w(G(x)) =
w(x). We recover the original version of this lemma by setting w(x) = 1 for all x G X.
Lemma 55 (Burnside) The sum of weights of G-orbits is the average number of weighted
fixed points
32
Pblya's Enumeration
Proof
1
'x€JY
= ^ E 5>(*)
= i^iE Ew(*)
'^GG xexg
D
(g,f)^fog-1
where p is the permutation representation of g acting on X and the weight function
h : Y —> i? induces a weight function on F x by CJ : Yx -* R
^(/)=ii/i(/(a;))
This function u is constant on the G-orbits on Yx. For each # G G,
1 l
geGk=l
where jfc(p) is the number of cycles of length k of the permutation g in its decomposition
as a product of independent cycles.
33
2. Taxonomy of Chords and Scales
£*(/) = £ n^/W)
fEY* fGY* x€X
£ w (/) = E n w(«))|ui
f&T* feY* u€<g>\\X
n E%)fc
k=l \yeY
D
Since conjugate elements have the same number of A;-cycles in their decomposition, it
is sufficient to compute the cycle index over the conjugacy classes
Example. Let G = S3 be the group of permutation of the set {1,2,3}. S3 has six
elements si=(123)=(l)(2)(3), s 2 =(132)=(l)(23), s 3 =(213)=(12)(3), s 4 =(321)=(13)(2),
s 5 =(231)=(123) and s 6 =(312)=(123). For each element, the number of cycle of length
k is ji(*i) = 3,.72(si) = 0,j 3 (si) = 0, ji(s 2 ) = 1,J2(>2) = 1,J3(«2) = 0, .71(53) =
1,^2(53) = 1,^3(52) = 1, ..., ji(se) = 0, j2(86) = 0,j3(«e) = 1. Thus the cycle index is
the polynomial
P(s3)x)(tut2,t3) = i(«? + 3ti« 2 + 2^3)
Let Y be the set Y = {0,1}, X = {1,2,3} and G = S3. The weight function is defined
on Y by w(0) = 1, w(l) = z. The sum of weights of G-orbits is by Pblya's theorem the
following counting series
34
Classification in Tempered Music
obtained from P(5 , 3,X) by substituting tk by 1 + zk. For the symmetric group Sn we
have the following result.
Definition 59 In the classical temperament, the twelve pitch classes (C, CJJ, D, ..., B)
are identified with Z\i = {0,1,.., 11} according to the code C = 0, CJt = 1, D = 2, ..., B
= 11. A pitch class set (pcsets) is a subset of Z12. A pitch class set of cardinality k is
called a pcset of length k or a k-chord.
Definition 60 Let G be a group acting on the set F(Zn) of pcsets . The elements of the
quotient set F(Zn)/G defines the pcsets classes or simply sets classes relatively to the
action of the group G A pcsets class of cardinality k is a is called a pcsets class of length
k. For the dihedral group Vn, the elements of the quotient set F(Zn)/Vn are called the
set classes under the action of the dihedral group or simply the d-classes. For the cyclic
group Cn, the elements of the quotient set F(Zn)/Cn are called the set classes under the
action of the cyclic group or simply the c-classes or musical assemblies.
Two pcsets are equivalent under the cyclic group Cn if they are reducible to each other
by transposition. A transposition is a mapping
Two pcsets are equivalent under the dihedral group Vn if they are reducible to the same
form by transposition or by inversion followed by transposition. The inversion is the
mapping
i": Zn —> Zn I(x) = —x mod n
35
2. Taxonomy of Chords and Scales
Two pcsets are equivalent under the affine group An if they are reducible to the same
form by affine transformation, i.e. a mapping
Two pcsets are equivalent under the symmetric group Sn if they are reducible to the
same partition of n. For a pcset A = [01, ...,afc], a\ < ... < aki the structure of A is the
vector [A] = [Aai,..., Aa^] where Aa* = a^\ — a; for z = 1,..., k — 1 and Aa& = ai — a/..
The partition of A is the increasing reordering of [A].
The number of A;-chord classes is the number of G-orbits which is the coefficient of zk in
the cycle index for variables
tk = 1 + zk
d\n
Each coefficient of zk is the number of &-chord classes under the Cn group action. There
are, for example, 19 trichords.
Recall that the Euler's totient function ip for integer m is the number of positive
integers not greater than and coprime to m. For m = p prime, ip(p) = p — 1 since all
numbers less than p are relatively prime (= coprime) to p. If m = pa is a power of a
prime p, then <p(p) = pa(l — £). If rai and m2 are coprime, then totient function have
the multiplicative property
<£>(raira2) = <p(mi)ip(m2)
The first values are: p(l) = 1, tp(2) = 1, (p(3) = 2, (p(4) = 2, tp($) = 4, y>(6) = 2,
V (7) = 6, p(8) = 4, y>(9) = 6, ip(10) = 4, y>(ll) = 10, <p(12) = 4, etc.
36
Classification in Tempered Music
Pl , _ / 5P(C„,Zn) + 5 * i 4 n _ 1 ) / 2 if n is odd
{V Zn)
"' " J |P( Cn ,z„) + W r 2 ) l 2 + tT) Hn is even
For the affine group *Ani the result is more complicated. Wan-Di Wei and Ju-Yong (see
[Wei, 1993]) have established the following results.
•r(A2i,%2a) o2
1_ f22(a-l)-lt2a + J2 (2W-V + ^(2i"1)2a"1) #"' +
V t=i
2a~i
2 i11
5>(*) (vtr+2°-H tr - - ) (n t2>)
i=0 \fc=l
Proposition 64 For a prime number p different from 2, the cycle index P = P^ a,z a)
of the affine group An is for n = pa and a > 1
" - ; ^ ^ (!>*<<-•>(-i>r+
Proposition 65 Le£ X, Y be two finite sets, G, H be two finite groups acting on X and
Y. The cycle index of the natural action of G x H on X xY is the product
37
2. Taxonomy of Chords and Scales
P*Q = S E « 0 ) I I I l Q M ) ( M )
a 6
(i) (j) *=u=i
where icd(fc, i) is the smallest integer which is a multiple of k and I, and gcd(k, I) =
kl j lcd(k,l). In particular, for n = p\l...p^T, where pj are prime numbers, the cycle
index of Z n is the product
J U A ) ( * i , * 2 , t s ) = l(t3!+Sht2 + 2t3)
The product
Summarizing the number of assemblies from each class, we get the following table.
Except for the symmetric group, the problem is symmetric: the unison has a com-
plementary 11-note assembly, the intervals have complementary 10-note structures, a
38
Motives and Mosaics
structure of k notes has a complementary structure of (12 — k) notes and for each group
(cyclic or dihedral) the complementary classes have the same number of elements. The
symmetry envelops the hexachords which are symmetrical with respect to themselves,
which explains for now that Matthias Hauer's 44-trope structures do not appear, as
it needs complementation within the hexachords. The structure of the dodecachords is
unique and is equivalent to the chromatic scale whose sounds can be used simultaneously.
These 12-sound chords appear in certain pieces, specially at the conclusion, such as in
Darius Milhaud's Chorfyhores, composed in 1915, or Ivan Wyschnegradsky's Journee de
['Existence for narrator, chorus and orchestra, composed in 1916. When the piece uses
serial techniques, a sub-classification of 12-sound series is necessary. It is possible to clas-
sify the series into 554 serial schemes by structuring the assemblies by tritones within
the series. The unison or single sound has a eleven-note complementary structure, as we
will see in the chapter on knots and braids.
Al = {vA,b\ AeGL2(Zn),beZ2n}
where GL,2(Zn) is the general linear group. The mapping o~A,b is defined by
Applying P61ya's enumeration theorem, the computation of cycle indices of the action of
the group J?n on Z\ leads to formulas such as for n = 2
39
2. Taxonomy of Chords and Scales
The G-mosaics are the G-orbits on the set of all partitions II n and the GA:-mosaics or
G-mosaics of size k are the G-orbits on Hnik-
Theorem 66 Let Sm be the symmetric group on the set {0,1, 2, ..., m-1} and let Mk
be the number
1
' ' *W)€GxS fc i=l
where ji(g) and Ji(cr) are the numbers of i-cycles in the decomposition of g € G and
a e Sk The numbers of G-mosaics in Zn is Mn and the number of Gk-mosaics is
Mk - Mk-\ (with M0 = 0).
Example. For n = 12, the previous formula gives the following results [Fripertinger, 1998].
40
Cyclic Chords
higher fifth is equivalent to adding 7 (a fifth is worth 7 half-tones) to each number [7,
10, 13, 16] and to keep only the balance of the division by 12 [7, 10, 1, 4] i.e. [1, 4, 7,
10] (Cfl, E, G, Bb), which is a structure similar to the initial chord. Inversion, which is
one of the essential functions of the serial theory, and even of the fugues of J.-S. Bach,
cannot be integrated esily to the classification. To consider that a major and a minor
chord have the same structure is impossible in the analysis of a tonal piece and even in
the case of modal or atonal works, the distinction between major and minor structures
should be made in most cases [Eimert, 1924], [Perle, 1962]. In Alban Berg's Sonata for
piano, the first phrase finishes on the fourth beat by a B minor chord which colors the
whole analysis. The first of the Vingt regards sur Venfant Jesus by Olivier Messiaen,
which is a work written in Mode 2, is colored by the tonality of F sharp major ; limited
transposition sets are, according to Messiaen, "in the atmosphere of several tonalities at
once, without poly tonality" [Messiaen, 1954, p. 57]. Another problem linked to inversion
is the complementation function which is in the dihedral classification slightly modified
as compared to its traditional form with respect to conventional sets. For example, for
sets, the structure [0, 1, 2, 3, 4, 6 ] is complemented by the notes which are not in this
set, i.e., [5, 7, 8, 9, 10, 11] which, transposed to the lower sixth minor, is a new structure
of the initial form [0, 2, 3, 4, 5, 6]. On the other hand, if the inversion is integrated, the
initial set becomes the set [11, 9, 8, 7, 6, 5] which, according to dihedral classification,
is equivalent to itself, and therefore is its own complement, which is no longer patterned
after sets.
41
2. Taxonomy of Chords and Scales
Definition 69 A cyclic chord is a pitch class set for which the free monoid generated
by the affine mapping r
< r > (x) = {x, r(x), r 2 (#),...}
is finite. The cycle index v is the period of the cycle r p + v = rp.
Definition 70 The transition matrix Taib of a cyclic chord is the matrix defined by the
generator r = Ma$ as follows
and 0 otherwise. The characteristic of the chord is the characteristic polynomial of the
transition matrix
P(x) = det(Ta>6 - xl)
Definition 71 Let fin be the cyclic group of prime roots £ such that £ n = 1. The cyclo-
tomic polynomial is defined by
<i>n(x)= n ^ - o
<e/i n
where the product is done over all generators of the cyclic group \xn.
Proposition 72 (1) The coefficients of the cyclotomic polynomials are integers
4>n{x) e Z[x]
(2) For all integer n > 1
xn -1=n^M
d\n
M ) = Y<xk
x
k=0
where [i is the Moebius function defined by fi(l) = 1, /x(n) = 0ifnisa square factor and
fj,(n) = (—l) r if n is a product of r distinct prime numbers.
Theorem 73 A class set is a cyclic chord for the action of the affine group if and only
if its characteristic polynomial is different from </>|2, </>i</>2, </>?</>2, 0o^i> 01 $$^2 ^3 where
(j)j are the cyclotomic polynomials.
Recall that the first cyclotomic polynomials are </>Q(X) = a;, (/>i(x) = x—1, <p2(x) = x+1,
(j)s(x) = # 2 + x + l, 04(x) = z 2 + l, </>6(z) = x2-x + l, (j>i2(x) = x4-x2 + l. For different
values of a and 6, the characteristic polynomials are
Po,b(x) = P6ib(x) = xu(x - 1) = 0 l V i ( a ) for all b = 0,.., 11.
Pifi{x) = {x-l)™ = <ff{x)
P\fik+l{x) = P\$k+s{x) = X12 - 1 = <f>i<l>2<l>3<l>4</>e<f>i2(x)
4 2
PIA(X) = Plt8(x) = (x- l) (x + x + l ) 4 = </>i4(x)
42
Enumerations of Pcsets Classes
is the basic scale for Scriabin's Prelude opus 74 no.l. The final cadence is written in this
scale.
43
2. Taxonomy of Chords and Scales
Paul Hindemith [Hindemith, 1940] classified the intervals by order of harmonic affinity.
He obtained a series (C, G, F, A, E, E\>, A\>, D, Bb, D\>, B, Pji) which he used in the
piano piece Ludus Tonalis as a structural element. The series is even printed in spiral
form on the cover page. The intervals lead to a complete redefinition of tonalities. In
the following table, the first column is the representative of the sets of the dihedral or
cyclic classification, the second column is the number of the class set, the third gives the
prime form of this pcset, the fourth column is the interval vector and the last give the
complement of the set.
In the affine classification, there are also six classes, but two classes have the same
symmetry group, reducing the number to only five classes. In the following table, the
first column is the representative of the set, the second column gives the number of the
class set under the action of the cyclic group, the third gives the class names according
to Mazzola's classification table, the fourth column provides a representative and the last
column is the symmetry group of the class.
In the symmetric classification, there are six classes of 2-chords which are the same as
in the dihedral classification. But there are only two classes of 10-chords represented by
the following representatives.
Representatives
**********QQ
B Sets
S10-1 0,1,2,3,4,5,6,7,8,9
*********Q*Q S10-2 0,1,2,3,4,5,6,7,8,10
44
Enumerations of Pcsets Classes
no. 10, opus 70 opens with an augmented fifth chord (Gb, Bb, D) and a diminished fifth
chord (E\>, G\>, B\>\>). The inflection of the fifth of classical chords results in inflected
chords, the most famous of which is the major chord with descending inflection of the
fifth (C, E, G\>). Scriabin used the transpostion of the enneaphonic scale
In the affine, classification, there are only 9 classes (12 classes if we include Mazzola's
extension ".1" see [Mazzola, 2002]). In the following table, the first column is a repres-
entative of the set. The second column is the number in the cyclic classification. The
prime form is given in the third column. The interval vector is in the fourth column.
The columns 5 and 6 give respectively the name of the chord in Forte's and in Mazzola's
classification. The last column is the symmetry group of the set.
45
2. Taxonomy of Chords and Scales
iJ={l,Mij0,MM,Mi>8 with z ^ 3 , 6 , 9 }
In the symmetric classification, there are 12 classes of trichords and 3 classes of ennea-
chords. The last column of the following table gives the 3-partitions of twelve.
Representatives
****#****QQQ
t Sets Partitions
S9-1 0,1,2,3,4,5,6,7,8 1,1,1,1,1,1,1,1,4
********Q*QQ S9-2 0,1,2,3,4,5,6,7,9 1,1,1,1,1,1,1,2,3
*******Q*Q*Q S9-3 0,1,2,3,4,5,6,8,10 1,1,1,1,1,1,2,2,2 |
Tetrachords and Octachords. In the classical theory, chords of four sounds are
reduced to the seventh chords. In the harmony by superimposition of fourths, the chords
found in Wagner's Tristan, Liszt's Mephistovalses, and in works by Milhaud, Honegger
and Stravinsky, are important structural chords. By superimposing fourths, five chords
of main fourths are obtained ([0, 2, 5, 7], [0, 2, 6, 7], [0, 1, 5, 6], [0, 1, 5, 7] and
[0, 1, 6, 7]) which create harmony by steps of fourths. The structures of four sounds
are also very present in modal compositions. The chords theme in Olivier Messiaen's
Vingts regards sur Venfant Jesus is made up of three chords of four sounds (of structure
[0, 2, 5, 7], [0, 1, 5, 6], [0, 3, 4, 9], [0, 4, 5, 6]). The atonal pieces also use chords of four
sounds. Each of Aaron Copland's Piano Variations, written in 1930 explores a motif of
46
Enumerations of Pcsets Classes
four notes. Tetrachords are widely used in Serge Prokofiev's Visions fugitives, opus 22.
In his Second sonata, Nikolai' Roslavetz uses a scale of eight notes (E, F, F$, G |J, A$} B,
Q,D)
In the cyclic and dihedral classification, the following table shows the representative,
the number of the set in the cyclic classification, the prime form of the set (third column),
the interval vector (fourth column), the name of the set in Forte's classification and the
prime form of the complement set.
| Representative Sets Iv Forte | Complements
| ****00000000
1*C4-1 0,1,2,3 9111 4-1/8-1 0,1,2,3,4,5,6,7
1 ***o*ooooooo C4-2 0,1,2,4 8112 4-2/- 0,2,3,4,5,6,7,8
| **o**ooooooo C4-3 0,1,3,4 8121 4-3/8-3 0,1,2,3,4,5,6,9
| *0***0000000 C4-4 0,2,3,4 8211 78-2 0,1,2,3,4,5,6,8
1 ***oo*oooooo C4-5 0,1,2,5 7113 4-4/- 0,1,3,4,5,6,7,8
| **0*0*000000 C4-6 0,1,3,5 7122 4-11/- 0,2,3,4,5,6,7,8
1 **oo**oooooo C4-7 0,1,4,5 7131 4-7/8-7 0,1,2,3,4,5,8,9
*0**0*000000 C4-8 0,2,3,5 7212 4-10/8-10 0,2,3,4,5,6,7,9
1 *o*o**oooooo C4-9 0,2,4,5 7221 -/8-11 0,1,2,3,4,5,7,9
*00***000000 C4-10 0,3,4,5 7311 -/8-4 0,1,2,3,4,5,7,8
***000*00000 C4-11 0,1,2,6 6114 4-5/- 0,1,2,4,5,6,7,8
| **o*oo*ooooo C4-12 0,1,3,6 6123 4-13/- 0,2,3,5,6,7,8,9
1 **oo*o*ooooo C4-13 0,1,4,6 6132 4Z15/- 0,1,3,4,5,6,7,8
| **000**00000 C4-14 0,1,5,6 6141 4-8/8-8 0,1,2,3,4,7,8,9
1 *o**oo*ooooo C4-15 0,2,3,6 6213 4-12/8-12 0,1,3,4,5,6,7,9
1 *o*o*o*ooooo C4-16 0,2,4,6 6222 4-21/8-21 0,1,2,3,4,6,8,10
| *0*00**00000 C4-17 0,2,5,6 6231 -/8Z15 0,1,2,3,4,6,8,9
1 *oo**o*ooooo C4-18 0,3,4,6 6312 V- 0,2,3,4,5,6,8,9
1 *oo*o**ooooo C4-19 0,3,5,6 6321 -/8-13 0,1,2,3,4,6,7,9
| *000***00000 C4-20 0,4,5,6 6411 -/8-5 0,1,2,3,4,6,7,8
1 ***oooo*oooo C4-21 0,1,2,7 5115 4-6/8-6 0,1,2,3,5,6,7,8
1 **o*ooo*oooo C4-22 0,1,3,7 5124 4Z29/- 0,2,3,4,6,7,8,9
1 **oo*oo*oooo C4-23 0,1,4,7 5133 4-18/- 0,1,3,4,6,7,8,9
1 **ooo*o*oooo C4-24 0,1,5,7 5142 4-16/- 0,1,2,4,6,7,8,9
| **0000**0000 C4-25 0,1,6,7 5151 4-9/8-9 0,1,2,3,6,7,8,9
| *0**000*0000 C4-26 0,2,3,7 5214 4-14/8-14 0,1,2,4,5,6,7,9
| *o*o*oo*oooo C4-27 0,2,4,7 5223 4-22/- 0,1,2,3,5,7,9,10
| *o*oo*o*oooo C4-28 0,2,5,7 5232 4-23/8-23 0,1,2,3,5,7,8,10
| *0*000**0000 C4-29 0,2,6,7 5241 -/8-16 0,1,2,3,5,7,8,9
| *00**00*0000 C4-30 0,3,4,7 5313 4-17/8-17 0,1,3,4,5,6,8,9
1 *oo*o*o*oooo C4-31 0,3,5,7 5322 -/8-22 0,1,2,3,5,6,8,10
| *00*00**0000 C4-32 0,3,6,7 5331 -/8-18 0,1,2,3,5,6,8,9
*000**0*0000 C4-33 0,4,5,7 5412 7- 0,2,3,4,5,7,8,9
1 *ooo*o**oooo C4-34 0,4,6,7 5421 -/8Z29 0,1,2,3,5,6,7,9
*000*000*000 C4-35 0,1,4,8 4134 4-19/- 0,1,2,4,5,6,8,9,10
| **000*00*000 C4-36 0,1,5,8 4143 4-20/8-20 0,1,2,4,5,7,8,9
1 *o*o*ooo*ooo C4-37 0,2,4,8 4224 4-24/8-24 0,1,2,4,5,6,8,10
*o*oo*oo*ooo C4-38 0,2,5,8 4233 4-27/- 0,1,3,4,6,7,8,10
*0*000*0*000 C4-39 0,2,6,8 4242 4-25/8-25 0,1,2,4,6,7,8,10
*00**000*000 C4-40 0,3,4,8 4314 -/8-19 0,1,2,4,5,6,8,9
*00*0*00*000 C4-41 0,3,5,8 4323 4-26/8-26 0,1,2,4,5,7,9,10
*00*00*0*000 C4-42 0,3,6,8 4332 -/8-27 0,1,2,4,5,7,8,10
*00*00*00*00 C4-43 0,3,6,9 3333 4-28/8-28 0,1,3,4,6,7,9,10 |
47
2. Taxonomy of Chords and Scales
In the affine classification, there are only 21 classes (29 extended classes) of tetrachords
and octachords, obtained by complementation. Remark that there are also ten classes
that have a symmetry group reduced to {1}, that is ten class sets with no symmetry. Some
authors have seen in this fact a foundation or a contribution to the atonality problem.
The following table shows the representatives of the class, the number of the set in the
cyclic classification (second column), Mazzola's number in the affine classification. The
last column gives the symmetry groups of the set.
In the symmetric classification, there are only fifteen classes of tetrachords given in
the following table. The last column is the 4-partition of twelve.
48
Enumerations of Pcsets Classes
Representatives
*******QQQQQ
tJ Sets Partitions
S8-1 0,1,2,3,4,5,6,7 1,1,1,1,1,1,1,5
******Q*QQQQ S8-2 0,1,2,3,4,5,6,8 1,1,1,1,1,1,2,4
******QQ*QQQ S8-3 0,1,2,3,4,5,6,9 1,1,1,1,1,1,3,3
*****0*o*ooo S8-4 0,1,2,3,4,5,7,9 1,1,1,1,1,2,2,3
*****Q*Q*Q*Q S8-5 0,1,2,3,4,6,8,10 1,1,1,1,2,2,2,2 |
Pentachords and Heptachords. The first atonal piece of the 20th century is a
work from the composer Abel Decaux, who settled in the United States in 1923 where
he taught at the Eastman School of Music of Rochester. Called the French Schoenberg,
Decaux composed four pieces for piano called Clairs de lune between 1900 and 1907.
The first piece, Minuit passe, develops a cell of three notes in a mysterious atmosphere
and stops on a chord of five sounds (B\>, D, Gtf, B, F$). The second piece, La ruelle,
begins with a chord of structure [0, 2, 4, 6, 8] and ends with a chord of the first piece.
Franz Liszt's Nuage gris also ends on an atonal chord of five sounds (A, E\>, G, B, F$).
Elliott Carter's Concerto for orchestra is built on 38 different possibilities of five-sound
chords. The pentatonic modes and in particular the major pentatonic (C, D, E, F, A)
and the minor pentatonic (C, D, E\>, G, A\>) were exploited by A. Tcherepnin (Etudes
de piano sur la gamme pentatonique, opus 51, Cinq etudes de concert chinoises, opus 52,
Etudes techniques sur la gamme pentatonique, opus 53). As for structures of seven sounds,
they are found in transition works used as scales. For example, Bartok uses a scale from
Roman folklore (C, D, E, F, Fj{, A, B), Scriabine uses a heptaphonic scale made of a
synthetic chord completed by the fundamental fifth (D, D\>, E, JFJJ, G, A, B\>) and many
Hispanic compositions use the melodic Andalusian minor. This scale (A, B\>, CJf, D, E,
Fj{, G|J) gives Hispanic elements found in Albeniz, Debussy (Iberia, La puerta del vino),
Granados (Goyescas), Lalo (Symphonie espagnole), Ravel (Alborada del gracioso), etc.
Dutilleux's Metaboles on the chord (D, F$, D$, B\>, E). The low E played by the double
basses gives the tenor of the whole piece.
The 72 Carnatic modes of Indian music only offer 36 different structures, as differ-
ent modes may have the same structure. This is the case for the Dheragankarabhama
49
2. Taxonomy of Chords and Scales
which is only the scale in C major and which can be found in five transpositions: Hanu-
matodi (C Phrygian), Narabhairavi (C Hypodorian), Harikambogi (C Mixolydian) and
Matsyakaliani (C Lydian). The Indian modes have also been used in Western music.
For example, August Chapuis in his Suite for piano uses the Eastern scale (D, E, F,
GU, A, Bb, C)J) which is the Srimendra mode, also used by Albert Roussel in his opera
Padm&vati, in the transposed form Mayamalavagaula. But the systematic exploitation of
these modes is due to composer Jacques Charpentier (72 Etudes kamatiques) who wrote
a work for each mode between 1956 and 1961. The classification of five-sound structures
shows the classical scales within the cyclic classification: major scale, minor scale, as-
cending melodic minor scale and descending melodic minor scale but cannot distinguish
the tenors of the ecclesiastic modes which are structurally different only in the starting
note, whether it is tonic or final. Jolivet's Etude sur les modes antiques exploits the A,
F and E modes. In Charles Koechlin's Heures Persannes opus 65, pentachords are used
for example in the fourth piece, entitled Matin frais dans la haute vallee.
50
Enumerations of Pcsets Classes
In the affine classification, there are 25 classes (36 extended classes) instead of the
38 classes of the dihedral classification. The following table shows the representative,
the number of the set in the cyclic classification, the prime form of the set, the interval
vector, the name in Mazolla's classification and the symmetry groups of the sets.
51
2. Taxonomy of Chords and Scales
In the symmetric classification, there are only 13 classes of pentachords given in the
following table.
53
2. Taxonomy of Chords and Scales
In the affine classification, there are only 34 classes (36 extended classes) instead of 80
in the cyclic classification and 50 in the dihedral classification. The following table gives
the representative, the number in the cyclic classification, the prime form of the set, the
interval vector (fourth column), the number in the affine classification (fifth column) and
the symmetry groups of the sets.
54
Enumerations of Pcsets Classes
#1 = {1,M7JO,M5,2,MI1J2,MI>6,M7,6,M5,8,MII)8}
iJ 2 = {l,M7,o,M5,4,Mii j4 }
H3 = {l,T5,M4 a ,M8,i,M4,5,M 8 j 5,Mii, 1 i,M3,ii,M7,ii,Mi 1 j,M3,7,M 7 ,7}
H4 = {l.Tj.Mij with i = 5,7,11 and j = 2,4,6,8,10}
In the symmetric classification, there are only 11 classes of hexachords. They are given
in the following table.
| Representatives B Sets Partitions
| ******oooooo S6-1 0,1,2,3,4,5 1,1,1,1,1,7
*****o*ooooo S6-2 0,1,2,3,4,6 1,1,1,1,2,6
*****oo*oooo S6-3 0,1,2,3,4,7 1,1,1,1,3,5
*****000*000 S6-4 0,1,2,3,4,8 1,1,1,1,4,4
****0*0*0000 S6-5 0,1,2,3,5,7 1,1,1,2,2,5
****0*00*000 S6-6 0,1,2,3,5,8 1,1,1,2,3,4
****oo*oo*oo S6-7 0,1,2,3,6,9 1,1,1,3,3,3
***o*o*o*ooo S6-8 0,1,2,4,6,8 1,1,2,2,2,4
***0*0*00*00 S6-9 0,1,2,4,6,9 1,1,2,2,3,3
**0*0*0*0*00 S6-10 0,1,3,5,7,9 1,2,2,2,2,3
*0*0*0*0*0*0 S6-11 0,2,4,6,8,10 2,2,2,2,2,2 |
3
Limited Transpositions Sets
Limited transpositions sets are of first importance in the combinatorial study of the n-
tone equal temperament (n-tet). They structure chords, modes and the whole tuning
assemblies by their singular properties. The problem of finding the subsets of the chro-
matic scale in the n-tet which are globally invariant by transposition has been studied
by Messiaen since 1929. Let us recall that in tempered music, a combination of notes has
limited transpositions when it gives back the same notes after less than twelve transpos-
itions. For example, the tritone (diabolus in musica) has only six transpositions, instead
of other intervals which have twelve transpositions. This chapter gives the mathematical
background for studying limited transpositions sets in n-tone equal temperament and
enumerates these sets for different values of n.
58
General Theory
Algorithm. The last theorem leads to the following algorithm. Recall that a partition of
n is a way of writing the integer n as a sum of positive integers. For each partition of n,
we have to determine if a reordering of this partition has a d-symmetry. If it is the case,
we must add it to the set of solutions only if it (or a permutation of it) has not been
added before.
Examples. Let's discuss some examples for different values of n (2 < n < 10).
Lt-sets of Z2.The 2-tet case is trivial. It is reduced to one musical interval, the triton of
frequency ratio \f2 (600 cents).
Lt-sets of Z3. The 3-tet, identified with Z^ {0,1,2}, is based on the interval of frequency
ratio 2 1 / 3 (400 cents). Since 3 is prime, there is only one limited transpositions set
£ = {0,1,2}.
Lt-sets of Z4. The 4-tet, identified with Z4= {0,1,2,3}, is based on the interval of fre-
quency ratio 2 1 / 4 (300 cents). Prom the relation k x \HA\ = 4, we see that the order of
the isotropy group HA is 2 or 4. If \HA\ = 4, the isotropy group is Z 4 . There is only
one limited transpositions set which is Z4 itself. If \HA\ = 2, then the isotropy group is
HA = {0,2}. If 1 is in A, then, by the the previous theorem, A contains also 3. In this
case, there are two sets of the form Tj. (A), {0,2} and {1,3}. By union, we construct the
set {0,2} U {1,3} = {0,1,2,3} = Z 4 , but it is already obtained. Thus there is one limited
transpositions set, the set {0,2}. To sum up, there are two limited transpositions sets in
the 4-tet, E\ = {0,2} and E2 = {0,1,2,3}. The set E\ has two transpositions E\ and
Ti{E\) = {1,3} and the set E2 has only one transposition.
Lt-sets of Z 5 . The 5-tet has only one limited transpositions set because 5 is prime.
Lt-sets of ZQ. The 6-tet is the whole tone scale. There are 4 limited transpositions sets.
The set {0,3} has 3 transpositions. The set {0,2,4} has 2 transpositions. The set {0,1,3,4}
has 3 transpositions and the set {0, 1, 2, 3, 4, 5} has only one transposition.
59
3. Limited Transpositions Sets
Lt-sets of Z7. The 7-tet has only one limited transpositions set because 7 is prime.
Lt-sets of Zs- The 8-tet has five limited transpositions sets. The set {0,4} has 4 transpos-
itions. The set {0,2,4,6} has 2 transpositions. The set {0,1,4,5} has 4 transpositions. The
set of 6 notes {0,1,2,4,5,6} has 4 transpositions and the chromatic set {0,1,2,3,4,5,6,7}
has only one transposition.
Lt-sets of Zg. The 9-tet has three limited transpositions sets L = {£?!, E2, -£"3}. If \HA\ =
9, then the isotropy group is a limited transpositions set E\ = {0,1,2,3,4,5,6,7,8}. If
\HA\ = 3, then the isotropy group is HA = {0,3,6} = E2. The transpositions are
Ti{HA) = {1,4,7} and T2(HA) = {2,5,8}. The union HAVT1(HA)\JT2(HA) is equal
to Ei. The set E3 = HAUTi(HA) = {0,1,3,4,6,7} is a limited transpositions set. The
union HA U T2(HA) = {0,2,3,5,6,8} is the transposition by 2 of the set £3.
Lt-sets of Z10. The 10-tet identified with Z10 is based on the interval of frequency ratio
2i/io ^ 2 0 cents). It has 8 limited transpositions sets. The set {0,5} has 5 transposi-
tions. Its complementary set {0,1,2,3,5,6,7,8} has also 5 transpositions. The set {0,2,5,7}
and {0,1,5,6} has 5 transpositions as well as the complementary sets {0,1,3,5,6,8} and
{0,1,2,5,6,7}. The chromatic set {0,1,2,3,4,5,6,7,8,9} has only one transposition.
60
Messiaen Modes
The five first sets are classified or altered chords, the seven following modes (mode 1
to mode 7) show Messiaen's modes and the last three (modes A, B and C) are defective
because they are included in Messiaen's modes: mode A (or mode 8) in mode 3, mode
B (or mode 9) in mode 2 and mode C (or mode 10) is included in mode 6. Remark that
the mode 6 is included in the mode 7 which is not really used by Messiaen. Lastly, the
chromatic total has only one transposition and contains all the other modes. The following
figure shows the distribution of the limited transpositions sets in tempered music. The
rc-axis represents the number of notes and the y-axis the number of transpositions. The
lefts side of the tree corresponds to the chords and the right side corresponds to the
Messiaen modes.
2 »o
I
(4 transp.)
(6 transp.)
(9 notes)
(8 notes)
Mode 3
Mode 4
J
(6 transp.)
(8 notes)
ModeS
61
3. Limited Transpositions Sets
Let's introduce Messiaean's modes in the classical notation and in the molecular
presentation. In the molecular presentation, the x-axis represents the major thirds (trans-
lation T4) and the y-axis represents the tritonus (translation TQ). In this frame {T^TQ),
the fifths (or fourths) are placed along the diagonal axis.
The whole tone scale is the first Messiaen's modes. It has only two transpositions.
$ . . .il'iM'
$ !»!>»'*
The mode 2 (also called the Bertha's mode) has three transpositions.
i >^.M«'^
22
^ ^
$
23
I , ^ 1»fl»^ m »
The molecular presentation shows the limited transpositions subsets: the modes B
{0, 1, 3, 6, 7, 9} and C {0, 1, 4, 6, 7, 10}, the chord of diminished seventh {0, 3, 6,
9}, the chord of altered seventh {1, 3, 7, 9}, the double fifth {0, 1, 6, 7}, various trans-
positions and four tritonus represented by the vertical segments.
This mode is widely used in the first regard of Messiaen's Vingt regards sur Venfant
Jesus.
62
Messiaen Modes
» . I>» \* 0^ ^
The molecular presentation shows the limited transpositions subsets of mode 3 the
mode 1 {0, 2, 4, 6, 8, 10}, the mode A {0, 1, 4, 5, 8, 9}, the chord of altered seventh
{0, 2, 6, 8} and the chord of augmented fifth {0, 4, 8}.
I ^ i » 'i*«» ^ ^
It has several limited transpositions subsets: the mode 5 {0, 1, 2, 6, 7, 8}, the mode B
{0, 1, 3, 6, 7, 9}, the mode C {0, 2, 3, 6, 8, 9}, the chords of diminished seventh {0, 3,
6, 9}, of altered seventh {0, 2, 6, 8}, the double fifth {0, 1, 6, 7}.
ipppp^s^
It includes the chord of altered seventh {0, 2, 6, 8} and the double fifth {0, 1, 6, 7}.
1,
$ » . « » ll» **
63
3. Limited Transpositions Sets
The molecular presentation shows the limited transpositions subsets of mode 6: the mode
1 {0, 2, 4, 6, 8, 10}, the mode 5 {0, 1, 2, 6, 7, 8}, the mode B {1, 2, 4, 7, 8, 10}, the mode
C {0, 1, 4, 6, 7, 10}, the chord of diminished seventh {1, 4, 7, 10}, the chord of altered
seventh {0, 2, 6, 8}, the double fifth {0, 1, 6, 7} and the chord of augmented fifth {0, 4,
8}.
i» ii» i«
$ • i * ii«
The mode 7 includes all the sets of limited transpositions except the modes 3 and A. It
is a good complement of these modes.
The mode B is included in the mode 2. This mode is cited by Messiaen in his Technique
de mon langage musical (example 342).
The mode C is included in mode 6 3 (third transposition of mode 6). This mode is also
cited in Messiaen's Technique de mon langage musical (example 341).
64
Messiaen Modes
Theorem 84 In the 12-tet, there are only 16 limited transpositions sets which can be
classified with the number of transpositions k.
1) If k = 1, there is only one limited transpositions set,
2) If k = 2, there is only one limited transpositions set,
3) If k = 3, there are two limited transpositions sets,
4) Ifk = 4, there are three limited transpositions sets,
5) If k = 6, there are nine limited transpositions sets.
Proof The 12-tet is identified with Z12. Since the order of the isotropy group is a divisor
of 12, we have to study five cases.
1) If l-ff^l = 12, then the chromatic scale
^ = {0,1,2,3,4,5,6,7,8,9,10,11}
HA= E2 = {0,2,4,6,8,10}
E2 U T\ (E2) = Ei
HA=E3 = {0,3,6,9}
By making the unions of E$ with T\{E3) and T2(E^)i three new sets are obtained. The
set
£3 ^ ( £ 3 ) = £4 = {0,1,3,4,6,7,9,10}
is Messiaen's mode 2. The set
£3 U T2(E3) = {0,2,3,5,6,8,9,11} = T 2 (£ 4 )
65
3. Limited Transpositions Sets
HA=E5 = {0,4,8}
E 5 UT 1 ( J B 5 ) = £ 6 = {0,1,4,5,8,9}
-E?5 U T2(E§) = E2
is Messiaen's mode 3. The other combinations are two transpositions of this mode, namely
and
£5 U T2(E5) U T3(E5) = T2{E7)
^ = E 8 = {0,6}
The set
tfgUTj^ura^UTaC^UTsCEs) = T5(E9)
T
£8UT1(£8)UT2(£'8)ur4(£:8)ur5(£8) = 4(£g)
EsUT1(Es)UT3(E8)UT4(E8)UT5(Es) = T3(E9)
E8l)T2(E8)UT3(E8)\JT4(Es)uT5(E8) = T2(E9)
The set
E8l)T1(E8)UT2(E8)L)T3(E8) = Ew = {0,1,2,3,6,7,8,9}
is Messiaen's mode 4. The following unions are transposition sets.
E8UT1(E8)l)T2(E8)UT5(E8) = T5(E10)
E 8 UT 1 (E 8 )UT 4 (£' 8 )UT5(£ , 8 ) = T4{E10)
E8l)T3(E8)UT4(E8)UT5(E8) = T3(E10)
The set
Es U T^Es) U T2(E8) U T4(E8) = Eu = {0,1,2,4,6,7,8,10}
66
Messiaen Modes
is a new limited transpositions set (Messiaen's mode 6). The other sets are transpositions
of mode 6.
ESUT1(ES)UT3(ES)UT5(ES) = T5(En)
E8UT2(E8)UT4(E8)l)T5(E8) = T4(En)
E8UT2(E8)liT3(E8)uT4(E8) = T2(En)
The sets
E8UT1(E8)UT3(E8)\JT4(E8) = E4
E8UT2(E8)UT3(E8)UT5(E8) = T2(E4)
are transpositions of the mode 2 already obtained. The study of the combinations of the
form E8 U Ti(E8) U Tj(E8) reveals three new limited transpositions sets. The set
E8 UT^Eg) UT2(E8) = E12 = {0,1,2,6,7,8}
is the mode 5. The following sets are transposition sets of mode 5.
EsUT1(Es)UT6(Es) = T5(E12)
E8UT4(E8)UT5(E8) = T4(E12)
The set
E8UT1(E8)L)T3(E8) = E13 = {0,1,3,6,7,9}
is the mode B. The following sets are transposition sets of mode B
E8UT2(E8)UT5(E8) = T5(E13)
E8UT3(E8)UT4(E8) = T3(E13)
The set
E8 U I\(£ 8 ) U T4(E8) = E14 = {0,1,4,6,7,10}
is the mode C. The following sets are transposition of the preceding set.
E8UT3(E8)UT5(E8) = T5(E14)
E8UT2(E8)UT3(E8) = T2(E14)
The set
E8\JT2(E8)l)T4(E8) = E2
is the whole tone scale already obtained. The next step is the study of the sets E8\JTj(E8).
This leads to two new limited transpositions sets
E8 U T^Es) = E8U T5(E8) = E15 = {0,1,6,7}
and
E8 U T2(E8) = E8U T4(E8) = E16 = {0,2,6,8}
The set
£$ U T3(E8) — E3
is already obtained. D
67
3. Limited Transpositions Sets
The following proposition shows that the limited transpositions set L is a structural
element of A in the sense that L connects the transpositions of A and thus makes easier
the modulations.
Proposition 86 Let d be a divisor of n. Denote Ld the set
L d = {0,fc,2fc,...,(d-l)fc}
where n = kd. Let A be a pcset class with exactly n transpositions (A is not a limited
transpositions set) and containing a set L of pcset class in Ld. Then L is included in
exactly d pcset classes Tm(A), with m = 0,1,..., (d — 1).
Definition 87 A set A of m pitch classes is neighbor of a set B of the same cardinality
if A and B have m — 1 pitch classes in common
\AnB\=m-l
The following result shows that for n = 12, there is only one major scale. It also means
that there is only one dichotomy of the 12-tone in 7 white keys and 5 black keys.
Theorem 88 Let m,n,u be non-negative numbers with 0 < m <n, 1 <u < n, and u is
prime with n. There is only one pcset class A verifying
(1) The cardinality of A is m, \A\ = m
(2) A has exactly n transpositions, A £ C
(3) A is in the neighborhood ofTu(A), \AnTu(A)\ =m-l
Let x be a pitch class ofZn, a representative of A is given by the set,
A= {x,Tu{x),Tl{x),...,TZ-l{x)}
Example. For n = 12, m = u = 7 and x = 5, the set A is the C major scale
A = {5,0,7,2,9,4,11}, that is the white keys. For m = 5 and x = 6, A is the black
keys, A = {10,3,8,1,6}.
With the three criteria given in the previous theorem, one could define a diatonic scale.
Thus the theorem says that there is only one diatonic scale for each m. Balzano [1980]
68
Diatonic Scales and Microtonality
and Gould [2000] defined diatonic scales from the diagonal sets emerging from the tiling
of the plane. In the following figure, the z-axis represents the translations T4 and the
y-axis the translations T3.
2-
1 11
11-
8' 8
VL
IV
5
I
2
The limited transpositions sets are also structural elements of the minor scale.
Proposition 89 In the 12-tone equal temperament, there is only one pcset class such
that each transposition A has six elements and A is the union of an element of L3 and
an element of L4. A is the minor scale without its tonic.
Example. Let's consider A minor. Without its tonic (i.e. the pitch class 9), A is the set
A = {0,2,4,5,8,11}. This set is the union of an element of L3, {0, 4, 8} and an element
of L4 {2, 5, 8, 11}. Adding a pitch class to A leads to form a set of 7 notes. There are
six choices, but only two are in the neighborhood of the major scale.
Theorem 90 In the 12-tone equal temperament, there are only two pcset classes in the
neighborhood of the major scale: the minor scale A, and its transposition at the third of
the inversion of A, 14(A) = T4l(A).
Example. In the neighborhood of the C major scale {0, 2, 4, 5, 7, 9, 11}, there are only
two scales A ={9, 11, 0, 2, 3, 5, 8}, A minor and the transformation /4(A) = {11, 0, 2,
4, 5, 7, 8}.
Definition 91 Let m,n,u be non-negative numbers with 0 < m < n, 1 < u < n, and u,
n are coprimes. A pcset class A with cardinality m is a diatonic scale if
(1) A does not contain three consecutive pitch classes
Vx e Z n , {x, x + 1, x + 2} £ A
(2) A contains one pcset of Ld, with d divisor of n
3x € Z n , Ts({0,fc,2fc,..., (k - l)d}) e A
(3) A is in the neighborhood ofTu(A)
\AnTu(A)\=m-l
Theorem 92 In the n-tone equal temperament, for a given u as above, there is only one
diatonic scale A of cardinal m, given by
69
3. Limited Transpositions Sets
Examples. (1) For n = 18. The integer u is one of 1, 5, 7, 11, 13, 17 (11 = —1, 13 = —5
and 17 = —7). For u = 5, the set of divisor of n is {1, 2, 3, 6, 9}. The sequence 0, 5, 10,
15, 15, 2, 7, 13, 17, 4, 9, 14, ... contains two elements of L2 (0, 9) and (5, 14). Thus, in
the field of the thirds of tone (i.e. 18-tet), the diatonic scale of length m = 10 is the set
whose elements are located before 14 in the previous sequence. A = {0, 2, 4, 5, 7, 9, 10,
12, 15, 17}. For u = 7, we get A = {0, 2, 3, 6, 7, 9, 10, 13, 14, 17}.
(2) For n = 24. The cycle of the multiples of u = 7 ends with the second element of L2 (7,
19). Thus in the field of the quarter tones (i.e. 24-tet), the diatonic scale of length m = 13
is the set A = {0, 1, 4, 5, 7, 8, 11, 12, 14, 15, 18, 21, 22}. It is composed of G major scale
{0, 4, 8, 12, 14, 18, 22} = 2 x {0, 2, 4, 6, 7, 9, 11} and the transposition a quarter tone
up of B\> major scale without triton 2 x {0, 2, 3, 5, 7, 10}, to avoid ambiguity.
Remark. Our definition of diatonic scales are different from the definitions of G. Balzano
and M. Gould. However in some cases, the Bolzano-Gould scale is a subset of the unique
diatonic scale. For example, in the case Z ^ + i ) , with k = 4, the tiling of the plane with
the translations T4 vertically and T§ horizontally leads to the following figure.
9 14 19 4 9
1
1
5 10 15 0 -- 5
1
1
1 6 11 -- 16 1
1
1
17 2 -- 7 12 17
1
1
13 -- 18 3 8 13
1
1
9 14 19 4 9
The diagonal set U = {0, 2, 5, 7, 9, 11, 13, 18} (which is considered as a diatonic scale by
M. Gould) is included in the set A = U U {3,14} which is precisely the diatonic scale for
n = 18, u = 9, and m = 11. The cycle of the multiple of 9 is not stopped at 9, although
9 appears two times in the diagonal. Thus a limited transpositions set of L2 (3, 13) is
included in the diatonic scale A in order to characterize the transpositions of the set A.
Theorem 93 In the 24-tet, there are only 381 limited transpositions sets which can be
classified by the number of transpositions k.
1) If k = 1, there is only one limited transpositions set,
2) If k = 2, there is only one limited transpositions set,
3) If k = 3, there are two limited transpositions sets,
4) If k = 4, there are three limited transpositions sets,
5) If k = 6, there are nine limited transpositions sets,
6) Ifk = 8, there are 30 limited transpositions sets,
7) If k = 12, there are 335 limited transpositions sets.
70
Quarter-Tone Limited Transpositions Sets
Proof Let H be the isotropy group. Since Card(H) is a divisor of 24, the number k
takes its value in the set {1,2,3,4,6,8,12}. It is then easy to construct in each case,
the isotropy group H containing the note {0}. By union of separate sets Ti(if), one
constructs the set of all limited transpositions sets. Since the subgroups of Z24 are not
separate, we could find, during the analysis, some sets already obtained. The proof can
be simplified by symmetry arguments, but we will not do that, because the complete
demonstration gives also the explicit forms of the limited transpositions sets, and not
only their enumeration. Seven cases have to be considered:
D2 = HUT1(H)
and its transposition sets
A set of 18 notes
£3= | J Tj{H)
j=l,2,3
and its transpositions
.7=1,2,3
T H
T2(Ds) = U ^ )
.7=0,2,3
T3(D3) = | J Tj(H)
j=0,l,3
71
3. Limited Transpositions Sets
which are already obtained. The two other sets are transpositions of E2 and E3
HUT4(H) = T4(E2)
and
HUTb(H) = n(Es)
c) Three sets of twelve notes. Prom the six sets Ti(H) , i = 0,1, ...5, one makes (g) = 10
sets of twelve notes containing {0}, but only three are generator sets. The set
E4= | J Tj(H)
i=o,i,2
and
T H
T5(£4) = (J ^ )
j=0,l,5
The set
E5= | J Tj(H)
j=0,l,3
and
T5(E5)= | J ?}(#)
i=0,2,5
The set
E 6= |J ^(ff)
i=0,l,4
and
T5(E4)= (J ^(tf)
J=0,3,5
72
Quarter-Tone Limited Transpositions Sets
d) Two sets of 16 notes. Prom the six sets Ti(H) with i = 0,1,2, ..,5, one constructs
(5) = 10 sets of 16 notes containing {0}, but two only are generator sets. The set
E7 = |J Tj(H)
J'=0,l,2,3
E8 = |J TjiH)
i=0,l,2,4
has also four equivalent forms Tj(Es) with j = 0,2,4,5. The set
C2= [J T,(ff)
j=0,l,3,4
e) A set of 20 notes
E9= |J Tj(H)
j=0,l,2,3,4
b) Three sets of six notes. Prom the seven sets Ti(H) with i = 1,2,..., 7, we get
D2 = HUT4(H)
c) Seven sets of nine notes. From the eight sets Ti(H) with i = 0,1,.., 7, one can build
(2) = 21 sets of the form H U Ti(H) U Tj(H) with i ^ j . However, each of those sets
has three equivalent forms (including itself), so there are only 21/3 = 7 sets, denoted by
*6, ,Fn.
d) Eight sets of twelve notes. There are (3) = 35 sets of the following form.
[" 3 1
J' = 1
73
3. Limited Transpositions Sets
D 2= |J Tj(H)
i=0,l,4,5
Consequently, there are three forms to subtract to the 35 sets, so 35 — 3 = 32. Each
of these 32 sets has four equivalent forms (including itself), so we have 32/4 = 8 sets,
denoted by F12,...., F\g
e) Seven sets of 15 notes, denoted by F20, ....,^26- There are Q = 35 sets of the form
X^k) with k = 5 and
ffc-i
X$k)=H\J
Each of these sets has five equivalent forms, so it remains only 35/5 = 7 sets of 5x \H\ = 15
notes.
f) Three sets of 18 notes. There are Q = 21 sets of the form X\ ' with k = 6. But the
set
D 2= |J Tj(H)
i=0,2,3,4,6,7
is already obtained. Since it has three equivalent forms, there are only 21 — 3 = 18 sets
which has, each, six equivalent forms. So, there are 18/6 = 3 sets of 18 notes.
g) A set of 21 notes
F30= |J Tj(H)
j=0,l,...,5
7) If k = 12 then H = 12Z 24 = {0,12}. One constructs all the unions from the sets Ti(H)
with i = 0,1,..., 11. We get
b) Five sets of 4 notes. There are 11 sets of the kind X\ ' with k = 2. Since the set
E1=HUT6(H)
is already mentioned, it remains 11 — 1 = 10 sets, each with two equivalent forms. So,
there are 10/2 = 5 sets of 4 notes.
c) 18 sets of 6 notes. There are ^) = 55 sets of the kind X\ ' with k = 3. Since the set
DX = (J Tj(H)
.7=0,4,8
74
Quarter-Tone Limited Transpositions Sets
d) 40 sets of 8 notes. There are (^J = 165 sets of the form X\ ' with k = 4. Since the
sets
Ci= (J Tj{H) = Ti{C1)
j=0,3,6,9
fori = 0,3,6,9.
E2 = (J Tj(H)=T6(H)
j=0,l,6,7
E3= |J Tj(H)=T6(E3)
j=0,2,6,8
and its transposition T^Es) = Tio(^3), have been already mentioned, there are only
(165 - 5)/4 = 40 sets of 8 notes.
e) 66 sets of 10 notes. There are ^) = 330 sets of the kind X\ ' with k = 5. Consequently,
we have 330/5 = 66 sets of 10 notes.
f) 75 sets of 12 notes. There are (g1) = 462 sets of the form x\k) with k = 6. But the
sets
B= |J TiiJS)
j=0,2,4,6,8,10
and
D 2= (J Tj(H)
.7=0,1,4,5,8,9
E4= |J T^iJ)
j=0,l,2,6,7,8
£5 = (J Tj(H)
j=0,l,3,6,7,9
E6= |J T,(ff)
j=0,l,4,6,7,10
and its equivalents T2(EQ) and T5(i?6) are already mentioned. Amongst the sets of the
form X\ ' some sets can be identified with Fi for i = 12 to 19. We will prove by reductio
ad absurdum that it is impossible. If so, the sets Fi will verify the identity
T12(Fi) = Fi
75
3. Limited Transpositions Sets
But the notes {16,17} belongs to all the sets Fi, because the sets Fi are of the form
Fi= |J ^({0,8,16})
j=0,l,p,q
So, the notes {4,5} belongs to all the sets Fi , it means that
Fi= |J ^({0,8,16})
5=0,1,4,5
for all i. But this set is the set D2 which is already mentioned. The hypothesis Tyz^Fi) =
F{ is absurd. The seeking sets are not of the form Fi. Thus there are 12 sets already
mentioned, and consequently, there are (465 — 12)/6 = 75 sets of 12 notes.
g) 66 sets of 14 notes. There are ft) = 462 sets of the kind X\ ' for k = 7 and so
462/7 = 66 sets of 14 notes.
h) 40 sets of 16 notes. There are ft) = 330 sets of the kind x\k) with k = 8. Since the
sets C2 and its transposition, the sets E7 and E$ with their four equivalent sets, that is
a total of 10 sets, have been already mentioned, there are (330 — 10)/8 = 40 sets of 16
notes.
i) 18 sets of 18 notes. There are (g1) = 165 sets of the kind X\ ' with k = 9. Since the
three sets equivalent to D3 are already mentioned, there are (165 — 3)/9 = 18 sets of 18
notes.
j) 5 sets of 20 notes. There are ft) = 55 sets of the kind x\ ' with k = 10. Since the five
sets equivalent to Eg are already mentioned, there are (55 — 5)/10 = 5 sets of 20 notes.
k) A set of 24 notes. There are 11 sets of the kind X\ ' with k = 11. Consequently, there
are 11/11 = 1 set of 24 notes. •
3.5 Enumerations
The goal of this section is to find the number of limited transpositions sets for a given
equal temperament of n notes. P61ya theorem gives a generator polynomial in which the
coefficient of xk is exactly the number A;-chord in n-tone equal temperament. For the
action of the cyclic group, the polynomial is given by the formula
d\n
where ip is the Euler's function. For x = 0, this formula gives the number of pcsets classes
n d
l^
d\n
The following result has been established in [Ballon, 2004]. We denote Ln the number of
limited transpositions sets, Mn the number of modes, Kn the number of modes associated
with limited transpositions sets. A pcset class of k degrees is associated with at most k
modes.
76
Enumerations
Let Ln>fc the number of limited transpositions set of k notes, and KU}k the number of
modes associated with the limited transpositions sets of k notes. A pcset class of k notes
which is not a limited transpositions set has exactly k modes, thus the numbers of the
respective quantities verify the relation
As the number of pcsets classes of k notes and the number of pcsets classes oin — k notes
are equal, the sum over k leads to
Pn-Ln = -(Mn-Kn)
n
Prom this relation, we get the expression of Ln. On the other hand, the number of modes
of k notes is
(n-l)!
M, l k
' ~ [k - 1 j ~ (n • *)!(fc - 1 ) !
This expresison leads to the total number of modes:
n-l
k=i fc=ov K J
The relation Kn is more difficult to prove. Let Mn be the set of modes and /Cn the set
of modes associated with limited transpositions sets. Let n be n = Pilp22'~Prr a n d m be
a divisor of n such that m ^ 1. The elements of /Cn are constructed from the elements
of the sets Mn/p. with j = 1,..., r. For each divisor m of n, the injection
maps a mode M of n/m notes with a limited transpositions modes of /Cn by repeating
n/m times the structure of M. The set /Cn is the union of the sets
r
K,n = \JfPj {Mn/pj)
3. Limited Transpositions Sets
By applying the following result to a collection of r sets Ai,..., Ar and by denoting \A\
the cardinal of the set A,
r r r—k+1 r—fc+2 r
[JAj = ]T(-i)*+1 £ £ .. £ |^nA i2 n...n^ t |
j=l fc=l ii=l i2=ii+l i fc =ifc_i+l
we get
r r-A:+l r-fc+2
w
^ = i/t;i=B-i) E E -
k=l ii=l i2=ii+l
r
/ y |/pi1--Pifc(^n/pi1..pifc)|
*fc=ifc-i+l
/ m i (,«A/ln/mi
and then
Theorem 95 Let n be the number of notes per octave in the equal temperament, and k
an integer less than n. LeVs denote m the number of common prime factors of n and k
and qi>q2>~">Qm these factors. The number of modes Mnj/- is given by
m m—fe+lm—/c+2 m
fc+1 Qilqi2-"(lik
Mn,k = £(-D E E •••• E k ;)
The number of pcsets classes (Pcsc) in the n-tet and the number of limited transpos-
itions sets classes (Ltsc) are given in the following table for different values of n, in the
cyclic classification.
78
Enumerations
13 n Ltsc
1
Pcsc
3
114n Ltsc
20
Pcsc
1 181
4 2 5 15 9 2 191
5 1 7 16 35 4 115
6 4 13 17 1 7 711
7 1 19 18 69 14 601
8 5 35 19 1 27 595
9 3 59 20 110 52 487
10 8 107 21 21 99 879
11 1 187 22 188 190 745
12 16 351 23 1 364 723
13 1 631 | 24 381 699 251 |
For divisions of ra-th of tone (m > 5, m = n/6), we get large values
T h e o r e m 96 Let H he the set of divisors of n strictly less than n and Lk he the set of all
values I G H such that Lk = 0 mod n. The numbers x^^ ofk-chords with £ transpositions
verify for each k the equation
(n-l)l ^f l\
k\(n-k)\ ^ \ n)
where Xkye are positive integers and Pn^ is the coefficient of xk in the Pdlya generator
polynomial of k-chords.
The previous formula gives the number of limited transpositions sets. But in some
cases, there are several solutions for the Xkyi and another argument to solve the system
is needed.
Examples. Lt-sets of Z30. For the 30-tone equal temperament (Fifths of tone), we have
to solve several equations.
For fc = 6, the equation
5rc6)5 + 4z 6j io + 3z 6) i5 = 111
has solutions
#6,5 = n\
#6,10 = ^1 + 3 n 2
#6,15 = 37 - 3ni - 4n 2
79
3. Limited Transpositions Sets
where n i , n^ are positive integers. To determine these integers, we need another argu-
ment. Let a be a limited transpositions set of 6 notes which has 5 transpositions. T h e
structure of a is of t h e form [a] = [ai,..., as] with a\ 4-... 4- CLQ = 30. Since a must have
5 transpositions, the only solution is a\ = ... = a^ = 5. Thus n\ = 1. To determine n^
we have to look at the limited transpositions sets with 6 notes and 10 transpositions.
Since they are 4 partitions of 10 of length 2, there are only 4 limited transpositions sets
and thus n^ = 1. T h e last equation gives t h e number of limited transpositions sets of 65
notes which have 15 transpositions, #6,15 = 30. For k = 10, the equation
Examples. Lt-sets of ZyQ. For the 36-tone equal temperament (sixths of tone), we have
to solve several equations. For k = 4, the equation
3#4,9 + 2X4}i8 = 19
80
Enumerations
has solutions
f x 4)9 = 1 + 2ni
\ #4,18 = 8 - 3ni
As they are only one limited transpositions set of 4 notes with 9 transpositions, n\ = 0.
For k = 6, the equation
5x 6j6 + 4z 6) i2 + 3x 6) i 8 = 160
has several solutions
#6,6 = ^1
#6,12 = 1 + n i + 3 n 2
#6,18 = 52 - 3ni - 4n 2
Let a be a limited transpositions set of 6 notes which has 6 transpositions. The structure
of a is of the form [a] = [ai, ...,ae] with ai + ... + a$ = 36. Since a must have 6 trans-
positions, the only solution is a\ = ... = a^ = 6. Thus n\ = 1. In the same way, if a is a
limited transpositions set of 6 notes which has now 12 transpositions. The structure of
a must be of the form [A, A, A] with A = [ai, a2] and ai + a 2 = 12. The partition of 12
leads to 5 solutions [1, 11], [2, 10], [3, 9] [4, 8] and [5, 7]. Thus z 6 ,i 2 = 5 and n 2 = 1. To
sum up, we have one set with 6 transpositions, 5 sets with 12 transpositions and 45 sets
with 18 transpositions. For k = 8, the equation
4#9,4 + 3x9 j i 2 = 58
has solutions
( #12,3 = ^1
#12,6 = 1*2
#12,9 = ni+ 2?23
#12,12 = 2ni + n 2 + 3n4
[ #12,18 = 1098 - 6ni - 3n 2 - 3n 3 - 4n 4
Let a be a limited transpositions set of 12 notes. The structure of a is of the form
[a] = [ai,...,ai 2 ] with a\ + ... + ai 2 = 36. If a has 3 transpositions, the only solution
is a,j = 3 for all j . Thus n\ = 1. If a has 6 transpositions, the structure of a is of the
form [A, A, A, A, A, A], where A = [ai,a 2 ] and a\ + a 2 = 6. There are two solutions
81
3. Limited Transpositions Sets
82
4
Neo-Riemannian Investigations
This chapter starts with a review of the main results of American Set Theory (AST). It
is not very easy to know when the AST was really born. Perhaps the most appropriate
introduction is to examine the papers of Milton Babbitt, who gave fruitful intuitions
since his paper on Set Structure as a Compositional Determinant published in 1961, and
other works on the studies of musical sets and partitions [Martino, 1961], [Howe, 1965],
[Gamer, 1967]. Set Theory was set out by Allen Forte who published The Structure of
Atonal Music in 1973. The pcsets classes were the first expression of the dihedral group
action on the twelve notes. In the 80's, Georg Perle's Serial Composition and Atonality
(1981) devoted to the atonal music of Schoenberg, Webern and Berg, and John Rahn's
Basic Atonal Theory (1980) become standard references in the field of contemporary
music theory. The group-theoretic properties of musical systems were pointed out by M.
Babbitt (1965) and G. Balzano (1980).
The Generalized Interval System (GIS) introduced by D. Lewin in musical theory
provides a new abstract construction. Its book Generalized Musical Intervals and Trans-
formation (1987) redefines the transformations introduced in its previous papers leading
to a neo-Riemannian theory arising in response to analytical problems posed by chro-
matic and extended tonal music. J. Clough, G. Myerson (1985) and E. Agmon (1989)
explore diatonic systems. John Clough and Jack Douthett investigate abstracts properties
of musical groups. J. Kovachi, D. Clampitt, R. Cohn, R. Peck and Ian Quinn also use
general mathematical tools to understand the conceptual transformations emphasized
by neo-Riemannian theory. Henry Klumpenhouver proposed new analytical investiga-
tions based on isographies and networks to make clear harmonic and melodic transitions
through contextual transformations, see for example [Klumpenhouwer, 1991], [Lewin,
1990], [Hascher, 2005], [Lambert, 2002] and [Stoeker, 2002].
4. NeoRiemannian Investigations
Definition 97 A pitch class (pc) is one of the twelve notes designated by the integers
{0, 1, ..., 11} identifying with Z12. C is labeled by 0, Oft is labeled by 1, ..., and B is labeled
by 11. Sometimes, the pitch classes are coded with the hexadecimal representation by the
integers {0, 1, 2, .., 8, 9, a, b). The letters a and b stands for 11 and 12 respectively.
Enharmonic equivalents (such as Oj[ and Db) are coded with the same symbol.
Definition 98 A pitch class set fpcset,) is a subset 0/Z12, i>£- a set of pitch classes.
A serie (or tone row,) is an ordered pcset. A fc-serie (or a k-row^ is an ordered pcset
of cardinality k. A pcset is said to be in normal order if arranged in ascending or-
der, the set is put as compactly as possible by a cyclic permutation. More precisely,
let A = [i4o,^4i,...,Ak-i] be a pcset arranged in ascending order Ao < ... < Ak-\.
For each cyclic permutation a, the index vector is defined from the permuted pcset
a(A) = [AT(0), Ar(i),---,47(fc-i)] by the vector (uu...,uk-\) with
The normal order is obtained for a given permutation ao if the number built on the
coordinates of the index vector, called the compact number
Example. The set [0,3,6,8,9] is not in normal order. Its normal order is [3,6,8,9,0]. The
compact numbers of the five cyclic permutations are: iV([0, 3, 6, 8, 9]) = 9863, iV([3, 6, 8,
9, 0]) = 9653, iV([6, 8, 9, 0, 3]) = 9632, AT([8,9,0,3,6]) = 10741, AT([9,0,3,6,8]) = 11963.
Definition 99 The dihedral group D\2 acts on the set of all pcsets. Two pcsets are equi-
valent (or D\2 -equivalent) if "they are reducible to the same prime form by transposition
or by inversion followed by transposition". Pc-sets classes (also called simply set classes)
are identified with the dihedral classification. The set is in prime form (dihedral or Forte's
prime form) if it is transposed so that the first integer is 0 and it is the most compact
form among its inversion.
Example. The set A = [2,3,5,7,11] is not in prime form. The normal order of A is
[11,2,3,5,7], and transposed one semitone up is [0,3,4,6,8] and AT([0, 3, 4, 6, 8]) = 8643.
The inversion of A, 1(A) = [1,5,7,9,10] is not in normal order. The normal order is
[5,7,9,10,1] and transposed a fourth down [0,2,4,5,8]. As N([0, 2, 4, 5, 8]) = 8542 < 7V([0,
3, 4, 6, 8]) then the prime form of A is [0,2,4,5,8]. The determination of the prime form
can be done easily with the circle representation. In this representation, the pitches are
placed on a circle as drawn below. Transpositions are obtained by rotating the circle.
The inversion is just a mirroring of the figure.
84
Set Theory
Definition 100 The set matrix of the prime form A = [AQ, A\, ...,Ak-i] is the k x k
matrix U defined by
Uij = Ai-i + Aj-i mod 12
Proposition 101 Let A be a pcset of length k, A= [Ao, A\,..., -Afc-i]. If any number m
appears k times in the set matrix U, or in other words, if there exists a number m such
that
k
= J2 kUij=m)
Im(A) = TmI(A) = A
A=[Ao9Au..nAk-i\
Proposition 103 The comparison matrix is skew-symmetric, i.e. CT = —C. The com-
parison matrix of the inverse set 1(A) is the transposition matrix of C.
C(I(A)) = CT
C(R(A)) = CR
defined by
C ij = Ck+l-j,k+l-i
85
4. Neo-Riemannian Investigations
Example. Let A be the 4-serie BACH, that is the pcset A = [10,9,0,11]. The comparison
matrix is
/ 0 -1 -1 1 \
C(A) = 1 0 - 1 1
1 1 0 1
V - l -1 -1 0/
The inverse set 1(A) = [1,2,11,0] has a comparison matrix CT. The retrograde set
R(A) = [11,0,9,10] has a comparison of the form
( 0 -1 -1 - 1 \
1 0 1 1
C(R{A))
1 -1 0 1
V l -1 -1
and the retrograde-inverse RI(A) = [0,11,2,1] has the following comparison matrix
1
C(RI(A)) =
Thus C(RI(A)) = C(R(A))T. But remark that C(IR(A)) is different from C(RI(A)).
Definition 104 The complement Ac of a pcset A is the set of all elements of Zu not
contained in A, i.e. Ac = Zyi\A. The set (of six pitch class sets) is self-complement if
the set A and its complement Ac are reductible to the same prime form.
Example. The set A = [0, 1, 2, 4, 5, 8] and its complement Ac = [3, 6, 7, 9, 10, 11] are
reducible to the same prime form A. A is self-complement. The set B = [0,2,4,5,7,9] is
also self-complement. But the set C = [0,2,3,5,6,8] has a complement reducible to the
prime form Cc = [0,2,3,4,6,9] different from the prime form C.
Definition 105 The interval class (\c) of two pitch classes x,y is the mapping d :
Z12 x Z12 - • {0,1,2,..., 6} defined by
mod 12 if \x - y\ < 6
d(xyy)
y\ mod 12 if \x — y\ > 6
There are only seven interval classes or types of intervals (ascending or descending):
unison (0), semitone (1), tone or dyad (2), minor third (3), major third (4) fourth or
fifth (5) and tritone (6).
Definition 106 The derivation of a serie S = [So,..., Sk] of length k + 1 with So < ... <
Sk is the pcset D(S) = [D0,...Dk] defined by Dj = d(Sj,Sj+i) = Sj+i — Sj mod 12
for all j = 0,..., k — 1, and Dk = Sk — So- The iterated derivation is defined recursively
Dm(S).
Definition 107 The interval vector (iv) of a pcset A is a 6-tuple that represents the
number of each interval classes included in A. The first coordinate is the number of the
smallest interval (a semi-tone in the twelve tone music), the second is the number of
seconds, etc. until the number of tritones.
86
Set Theory
Example. The whole tone scale A = [0,2,4,6,8,10] has an interval vector iv(A) =
0600603. The Prometheus hexachord B = [0,1,3,5,7,9] has iv(B) = 142422 which
means that B is composed of 1 semitone, 4 tones, 2 minor thirds etc. Remark that,
sometimes, the interval vector is defined as a 12-tuple for the twelve interval classes.
Definition 108 The interval class content vector (icv) of two pcsets A and B is the
7-tuple icv(A, B) = [vo, •••^6] defined by v0 = \A D B\ and for i = 1,.., 6, Vi = \AC\kB\ /2
where \X\ is the cardinality of the set X and Adk B is the set of all pairs of interval
class k :
A fife B = {(x,y) GAxB, d{x,y) = k}
Definition 109 A Z-related pair is a pair of sets A, B with the same interval vector
iv(A) = iv{B), but not reducible to the same prime form.
Example. The sets A = [0,1,2,4,5,7] and B = [0,1,2,3,5,8] have the same interval
vector iv(A) = iv(B) = 333231, but have different prime form.
Definition 110 Two sets A and B of same cardinality m are in similarity relation Rp
if two representatives A and B of the pcsets classes of A and B contain at least one
common subset of cardinality m—1.
Definition 111 Two sets A and B of same cardinality m are in similarity relation i?o
if they have no corresponding interval vector coordinates in common
Definition 112 Two sets A and B of same cardinality m are in similarity relation Ri
if they have four out of their six interval vector digits equivalent and the two dissimilar
digits are switched in position.
Definition 113 Two sets A and B of same cardinality m are in similarity relation R2
if they have four out of their six interval vector digits equivalent and the two dissimilar
digits are not equivalent.
Definition 114 Two sets classes A, B of distinct cardinalities (2 < \A\ < 10, 2 < \B\ <
10,) are in subset relation if one is included in the other, that is
ADCB «=> A C B or B C A
87
4. Neo-Riemannian Investigations
Definition 115 The set complex of a set class A, denoted K(A, Ac) or simply K(A) is
the set of all set classes B in subset relation with A or Ac, in other words, the sets B
such that B contains A or is contained in A or B contains Ac or is contained in Ac.
BeK(A) ^=> BDCA or B DC Ac
Example. The subcomplex of the set A = [0,2,4,6,8,10] is the set Kh{A) = {i?i, £ 2 ,
.., B7} with Bx = [0,2,4], B2 = [0,2,6], Bz = [0,4,8], B4 = [0,2,4,6], B5 = [0,2,4,8],
S 6 = [0,2,6,8],B 7 = [0,2,4,6,8].
in other words, the interval function counts how many times the note k of X has its
i-transpose in Y. The interval vector (also denoted by \X,Y)) is the 12-tuple whose
coordinates are the interval functions h(x,Y)(i) f°r i = 0,..., 11.
Example. The interval function of the sets X = {4,10} and Y = {1,5,9} is
88
Generalized Interval Systems
Definition 119 Let X, Y be two non-null pcsets. The injection function INJ(X, Y)(f)
relatively to a transformation f is the number of elements x of X such that f(x) £Y
INJ(X,Y)(f)=y£l{fix)eY)
xex
Theorem 120 Let T{ be the translation ofi, X and Y two non-null pcsets, then
INJ{X,Y){Ti) = IFUNC{X,Y){i)
Proof
D
C
Theorem 121 Let A be an hexachord \A\ = \A \. For all bijectionf, we have
INJ(A,A)(f) = INJ(Ac,Ac)(f)
and in particular
IFUNC(A,A) = IFUNC(AC,AC)
Proof This theorem is a consequence of the properties of the characteristic functions.
11
INJ(A,A)(f) = £ l(Hx)€A) = J2 lA(k) lf-HA)(k)
xeA k=0
= £(i-i A .(fc))(i-i/-W*))
k
= ^-E^w-E1/-1^)^)
k
= 12-6-6+ ^ l(f(x)€Ae)
c
xeA
= INJ(Ac,Ac)(f)
a
Theorem 122 Let A be an hexachord with \A\ = \AC\ = 6. For all bisection f, the
injection function verifies
In particular,
IFUNC(A, Ac) = 6 - IFUNC{A, A)
89
4. Neo-Riemannian Investigations
Proof
\A\ = 22 l(xeA) = ^2 1
(®Gi4)-l(/(a;)€i4uA«)
=
S1(/(*)eA) + ^^meA*)
xEA XEA
= INJ(A,A)(f) + INJ(A,Ac)(f)
and
IFUNC(AC, A) + IFUNC(AC, Ac) = 6
•
If we define the Fourier transform of the characteristic function by
then it is well known that the Fourier transform of the convolution product is the point-
wise product of their Fourier transforms. So the Fourier transform of the interval function
is the pointwise multiplication of the complex conjugate of F(\x) with «F(ly) that is
J T ( X , r ) = ^ ( b F ) .F(ly)
Theorem 123 The interval function of two non-null pcsets X and Y is constant if and
only if its Fourier transform Th{X, Y)(i) vanishes for i = 1,..., 11.
Lewin has shown that it is the case if the set S = X or S = Y satisfies the following
properties:
(FP1) S is the disjoint union of tritone sets (#i+i — Xi = 6) and (or) augmented triad
sets (#i + i — Xi = 4).
(FP2) | 5 n T 3 ( i 4 ) | = \SnA\ for all set A of the class A = {0,1,6,7} i.e. the number
of notes in common between A and S is the same as the number of notes in common
between S and the 3-transpose of A.
D = {0,2,4,6,8,10}
90
Riemannian Transformations
Theorem 124 The interval function of two non-null pcsets X and Y is alternate h(X, Y)
= (p>#>P><7>P><Z>P><Z>P>tf>P><z) if and only if its Fourier transform T'h{X,Y){i) vanishes
for all i different ofi = 0 and i = 6. In other words, if and only if either X or Y satisfies
the first four previous properties (FP1) to (FP4).
Theorem 125 Let X, Y, Z be three non-null pcsets. The interval function /i(X, Y) is
equal to h(X, Z) if and only if the Fourier transform of lx or the Fourier transform
of ly — l z vanishes everywhere. In other words, if and only if Y and Z have the same
cardinal number \Y\ = \Z\ and either X verifies the following (FPi) properties or Y and
Z verify the (FR*) properties (for i = 1,..., 5).
(FR5) \Y H D\ - \Y flTi(D)| = \ZnD\-\ZD TX{D)\, where £> is the whole tone scale
D = {0,2,4,6,8,10}.
[a,6,6 + 3] 3-> [ a - 3 , a , b ]
[a - 4, a, b] <—> [a, 6, b + 4]
91
4. Neo-Riemannian Investigations
The composition of these transformations has been defined by Lewin. The suhdominant
transformation S is the composition
S = RoL
(denoted by S = RL). S transforms a major triad to a major triad transposed a fifth
up, and a minor triad to a minor triad transposed a fifth down.
Examples. S([0,4,7]) = [7,11,2] and S([0,3,7]) = [5,8,0]. Remark that S is not an
involution.
7 •1
k J5L iw I» J 5 L I,
•+6
[l.o
92
Isographies and K-nets
Xavier Hascher [Hascher, 2002] has shown how the 4-chords of Franz Liszt's Ladilaus
Teleki are organized in isographies. Robert Peck [Peck, 2003] analyzed the opening canon
of no. 8, Nacht from Arnold Schoenberg's Pierrot Lunaire, op. 21. The trichords of the
first three measures display a triangular network of the same pcset 3-3 [0,1,4]. Moreover,
theses chords are also the vertices of a supernetwork, in which edges represent hyper-
operators [Tn] and [In\.
Definition 126 Hyper-operators between two K-nets are positive isographies if
(1) The two K-nets have the same disposition of T and I relations,
(2) The values j of the translations Tj are the same in the two networks,
(3) The values j of the inversions Ij in the second network minus the values i of the
correspondant U arrows in the first network are equal to m (with m > 1).
j—i=m mod 12
If m = 0, the graphs are equivalent and the two networks are said to be in strong
isography.
Definition 127 Hyper-opertors between two K-nets are negative isographies if
(1) The two K-nets have the same disposition of T and I relations,
(2) The values j of the translations Tj in the second network are the inverse of the
values i of the correspondant T{ arrows in the first network,
(3) The values j of the inversions Ij in the second network minus the inverse values i
of the correspondant Ii arrows in the first network are equal to m, with m > 1.
j — I(i) = m mod 12
The following definition of (n, m)-isographies generalizes the previous ones. For n = 1,
we recover the definition of positive isographies and for n = 11, we recover the definition
of negative isographies.
Definition 128 Hyper-opertors between two K-nets are (n,m)-isographies if
(1) The two K-nets have the same disposition of T and I relations,
(2) The values j of the translations Tj in the second network are the translations Tn
(with n € {1,2,.., \\}) of the values i of the correspondant T{ arrows in the first network
3=Tn(i)
(3) The values j of the inversions Ij in the second network minus the values Tn(i) of
the correspondant Ii arrows in the first network are equal to m, with m > 1.
j — Tn(i) = m mod 12
In the two following examples, we show that some graphs could preserve the same
structure, even if there are no inversions in these graphs. To avoid confusion, we replace
hyper-operators by t-operators which act on the notes of the isographies. In the Second
Sonate for piano of Nikolai Roslavetz, the same structure is kept. The sets include a
minor third and a fifth transposed a semitone up. The set is completed by a tritone,
which is drawn on the bottom of the graph by the translation TQ. The first measures
evolve by simple translations, denoted by t-operators [T3] or [T7], at the minor third or
at the fifth. These t-operators are composed of the same translations as the horizontal
operators (T3 and Tj) of the initial structure. Here are the structure of the first measures:
93
4. Neo-Riemannian Investigations
M«,4
8 « - ^ - 1 - ^ 4 | U^i-4^7 I
4 | l- 11 |
2 |
1«JL.6_!i*9
l- I-
4
I- 7 J
9
I
MM. 9 Met. 10 MM. 11-13
1 6 4-3L11-JJU2 I ,JL,_S., I
m I , I , I-
O4J1-5-X.8
-M. I, I, I- Uu I, I, I'' 2 4JL 7-2*10
I"
2
l- I- 9 4
I I I I
The Prelude opus 74 n° 4 of Scriabin is difficult to analyze using isographies. The first
measures are analyzed in the following nets:
A • Mode 3 transp. 2 mes. 1 £-Mode 3 transp 2mes 2 A-Mode 3 transp. 2
1o*JL7.Ji>11 I
[4.tl IV, ja
I"
•
!'•
4 10
2 |
[JiT,ol [T„l p lU
1
I''
0
1"
11 |
Td<B) • Mode 3 transp. 4 B - Mode 3 transp. 2 A-Mode 3 transp 2
I 8 4-Il-5J!l*o,
IT„] g | 5 «JL.2-iU6 [T,l, jJY^
l"
10 |_
1"
1 |
I" 4 |
94
Parsimonious Graphs
This Prelude is written in the Messiaen's mode 3. This mode has four transpositions.
The first measures begin with the second transposition of the mode. Three types of
chords are used: the chord A {0, 1, 4, 9}, the chord B {1, 5, 8, 9} and the chord C
{0, 5, 8, 11}. Each chord has a common horizontal structure composed of one minor
third and one major third (denoted by T3 and T 4 ).
The t-operators have two parts. A part in which the notes are simply transposed and
denoted by [Tn]. And a second part which modifies only the vertical components of the
graph, denoted by [Jn]. This part transforms the vertical arrows of translation T; by
=
Jn\J-i) J-i+n-
For example, the passage of the first graph to the second transforms a chord i to a
chord C. The action of the translation [Tn] translates the notes {0, 9, 1} of the chord
A to the notes {11, 8, 0}. The second part of the t-operator [J2] transforms the vertical
arrow T3 into the translation T3+2 = T5. Consequently, the note {4} of the chord A is
first translated by Tn to give the note {3}, and then modified by [J2] to give the note
3 + 2 = 5 of the chord C.
The t-operator [JmTn] acts on the notes of the horizontal arrows by the translation
[Tn] and on the notes of the vertical arrows by the translation [T n + m ].
Definition 129 Let X and Y be two pcsets of same cardinality. The pcsets X and Y
are Pij related if, in the notes that are not common to both X and Y, i notes move by
interval class 1, and j notes move by interval class 2. More precisely, if the non-common
tones are denoted X\Y = {x\,X2> ...,£i+j}, there exists a bisection f : X —» V verifying
m
t( \ _ ( ^fc i 1 ° d 12 if 1 < k < i
^Xk'~{ xk±2 mod 12 if i+l<k<i+j
This definition covers the definition of Cohn (Pi,o and Po,i-stated) and the definition
of Childs (P2,o-related). The contextual Riemannian transformations P and L are P\}Q-
related and the R transformation is i^i-related. In the Chicken-Wire Torus represented
in the following figure, the edges represents the contextual transformations P, L, R. The
figure depicts the parsimony of triads in the sense of Richard Cohn.
95
4. Neo-Riemannian Investigations
The hexatonic mode of limited transposition [0,3,4,7,8,11] is organized in six triads (C,
Cm, E, Em, Ab and Abm) which defined the hexatonic cycle. These triads themselves
are P^o-related. They are placed at the vertices of a cube. The two remaining vertices
of the cube are occupied by two augmented triads, which have the greatest potential to
couple hexatonic cycles via P^o-relations. The graph obtained is called the Cube Dance.
Another method induced by P^o and Po,i-relations is depicted by a graph called Power
Towers. We invite the reader to refer to the paper of Douthett and Steinbach [Douthett,
1998].
96
Permutational Transformations
/ 0 1 2 3 4 5 6 7 8 9 10 11 \
^ 7 6 5 4 3 2 1 0 11 10 9 8 j
or in cycle notation
P = (0,7)(l,6)(2 ) 5)(3 ) 4)(8 ) ll)(9,10)
It transforms a major triad based on the pitch n into a minor triad based on the pitch
(12 — n) mod 12. This transformation is the inversion h(x) = — x + 7 mod 12.
The Relative transformation R defined by
/ 0 1 2 3 4 5 6 7 8 9 10 11 \
^ " ^ 4 3 2 1 0 1110 9 8 7 6 b )
It exchanges a major triad based on the pitch n with a minor triad based on the pitch
(9 — n) mod 12. This transformation is the inversion U(x) = — x + 4 mod 12.
The Leittonwechsel transformation L is defined by
/0 1 2 3 4 5 6 7 8 9 10 11 \
\ 11 10 9 8 7 6 5 4 3 2 1 0 j
and in cycle notation
transforms a major triad based on the pitch n into a minor triad based on the pitch
(4 — n) mod 12. This transformation is the inversion I\\{x) = — x + 11 mod 12.
The new transformations lead to three important groups: the hexatonic, octatonic and
the PLR or dihedral group.
The Hexatonic group is the group generated by the transformations P and L. It has
the presentation:
97
4. Neo-Riemannian Investigations
where In is the inversion of order n (x —> —x + n mod 12) and Tn is the transposition
of order n (x —> x + n mod 12). The group is represented by an hexagon. The vertices
are the elements of the group and the edges are alternatively the transformations P and
L. Let A be a pcset of k notes, the graph of G\ represents a pitch class, given by the
action of the group G\ on A. For example, the atonal triad A = [0,6,11] gives five others
triads P(A) = [7,8,1], LP(A) = [4,3,10], PLP(A) = [3,4,9], (LP)2(A) = [8,7,2] and
P(LP)2(A) = [11,0,5]. Remark that each triad is the same pitch class set (3-5 in the
Forte's classification).
The Octatonic group is generated by the two permutations P and R.
The group has eight elements. With the previous notation of transpositions and inver-
sions, we get
<?2 = {To,/7,T9,/io,T6,/i,r3,/4}
The graph is an octagon, whose edges are alternatively the transformations P and R.
As in the case of the hexatonic group, one vertex of the graph accepts one pcset, which
is propagated through other vertices via the P and R permutations.
The group PLR is generated by the three permutations. It has the presentation
r _ I P , L , R | P2 = L2 = R2 = (PL) 3 \
u
- \ = (PRL)2 = (PR)4 = 1 /
Substituting P, L, R by the inversions J7, 1^ 7n, one verifies that G is the dihedral
group. The elements of G are drawn on the following figure.
-x+6
The interior octogone is rotated by 37r/4 and identified with the exterior octogone. The
figure is then a torus on which three octagons are drawn in the meridian plane and four
hexagons, based on the succession PLPLPLP, lie on the transversal planes of the torus.
98
Permutational Transformations
Because of the planar representation, these hexagons are not easily seen. Starting from
a point situated on the interior octogone, the path PLPLPLP leads to a point located
on the exterior octogone which is exactly the same as the starting point.
As G is the dihedral group, the figure is a tesselation of the transpositions and inver-
sions of a pitch-class set. In the tonal music context, one chord, for example a major triad
[C, E, G] (but it could also be a chord of four notes or more), is placed on a vertex of the
figure and propagated with the P, L, R transformations. Prom the figure obtained, the
consonant triads are connected with the transformation LR, when the base pitch moves
a fifth up.
LR LR LR LR
D E B F#
Example. We understand why isographies could easily be drawn with this representation.
But this representation does not concern only consonant triads, and our scheme can be
filled with any pcset, for example, with the atonal triad [0, 1, 6]. In the piano piece of
George Crumb entitled Gargoyles (Makrocosmos vol. 77), the right hand moves
PR
[2,3,8] - ^ - + [5,6,11] - ^ - + [0,1,6]
PL
[3,4,9] [7,8,1] - ^ t [1,2,7]
LR PR
[1,0,7] - ^ - > [10,3,4] [11,10,5]
PL (RP)2
[1,2,8] [0,5,6] [6,11,0]
The same transformations are used in both hands, except in the last triads of our example,
in which the reverse transformation is used (PRPR on the right hand, and RPRP on the
left hand). But if we allow hands crossings, the P-transform brings the set [7, 8, 1] to the
set [6, 11, 0] and [0, 5, 6] to [1, 2, 7]. In the last chord, the triads are played together.
Example. Thomas Noll has studied afRne transformations Ma$ : x —» ax + fi mod 12. He
has shown that some of them generate monoids. The following graph depicts a monoid
based on two afRne mappings u and v. On this picture, a, b and c are constants, and the
functions u and v verify the relations u3 = u, v3 = v and u2v = vu2.
-• u
If u and v are of the form f(x) = ax + (3 mod 12 and satisfy f(A) C A for the major
chord A = {0,4,7}, then the functions u and v are 0, 4, 7, x, 3a; + 7, 4a:, 8a; + 4 or 9a; + 4.
If we choose for example ^(a;) = 3a; + 7 and v(x) = 8x + 4, then the monoid has three
99
4. Neo-Riemannian Investigations
constants a = 7, b = 0 and c = 4. In the same way, if we want to keep the minor triad
unchanged A = {0,3,7} the functions u and v are 0, 3, 7, a;, 3a; + 3, 4a; + 3, 8a; + 7 and
9a;. The choice u(x) = 3x + 3 and v(a;) = 8a; + 7 leads to the constants a = 0, b = 7 et
c = 3.
But others monoids are possible. Hexacycles can be found using the affine transform-
ations
fa (x) = 7x + 1 h2(x) = 7x + 4
/^(z) = 7a;+ 7 /i 4 (z) = 7a; + 10
In the following figure, the choice of three indices z, j , k G {1,2,34} leads to the hexacycle.
h,(x+4) h,(x)
k
*
x+4 x+8
K
h/x+8)
7,11,2,5,8
7x+10^ .Jx+1
3,7,4,1,10 2,6,3,0,9
7x+10l 7x+1
11,3,6,9,0 3,7,10,1,4
7x+4 7x+4
1,5,2,11,8
100
Permutational Transformations
Example. The mode 2 of Messiaen (C, CJ|, E\>, E, F$, G, A, B\>). which is a limited
transpositions mode, has only three transpositions. Trying to find affine transformations,
we obtained two permutations closely related to mode 2. The following permutation
/ 0 1 2 3 4 5 6 7 8 9 10 11 \
a
\^16 1 1 4 9 2 7 0 5 10 3 8 J
a =(0,1,6,7X2,11,8,5X3,4,9,10)
/ 0 1 2 3 4 5 6 7 8 9 10 11 \
1 ^ 3 8 1 6 11 4 9 2 7 0 5 10 )
a=(0,3,6,9)(l,8,7,2)(4,ll,10,5)
is the affine transformation x —• 5x + 3 mod 12. The group H is generated by the two
permutations and has the presentation
The group has only twelve elements and appears to be well adapted to the three trans-
positions of mode 2. Consider a characteristic chord of mode 2 (C, E, G, Fjj), that is
a chord that only belongs to one of the three transpositions, the group H acts on this
chord and the action produces twelve chords. Remark that the major triad is included
in this chord.
In the previous figure, there are only four chords of mode 2 1 (first transposition), four
chords of mode 2 2 (second transposition), and four chords of mode 2 3 (third transposi-
tion). The group facilitates the transpositions inside mode 2. It is clear that it could not
101
4. Neo-Riemannian Investigations
Example. He de feu 2 is a piece for piano composed in 1949-50 with three other pieces in
Quatre etudes de rythme. It is based on various inversions of two series of twelve notes,
which is interpreted by two permutations. The first permutation is given below
0 1 2 3 4 5 6 7 8 9 10 11 \
6 5 7 4 8 3 9 2 10 1 11 0 J
or expressed in the cycle notation
a =(0,6,9,1,5,3,4,8,10,11X2,7)
/ 0 1 2 3 4 5 6 7 8 9 10 1l\
^ 5 6 4 7 3 8 2 9 1 1 0 0 11/
6 =(0,5,8,1,6,2,4,3,7,9,10)(11)
These two permutations a and b generate the Mathieu group Mi 2. It has the following
presentation.
This group has 95040 elements and, of course, only some of them are used by Messiaen.
However, it induces a real potential of combinations, which maybe charges the music of
inner perspectives.
Ma : x —» a.x mod 12
_ / 0 1 2 3 4 5 6 7 8 9 10 11 \
5
V 0 5 10 3 8 1 6 11 4 9 2 7 J
The notes {0, 3, 6, 9} are fixed points. The others are exchanged in four cycles
MB = (1,5)(2,10)(4,8)(7,11)
102
Morris Groups
_ / 0 1 2 3 4 5 6 7 8 9 10 11 \
7
" \ 0 7 2 M 11 6 1 8 3 10 5 J
which let the notes {0, 2, 4, 6, 8, 10} fixed and exchanges the others in three cycles
M 7 = (1,7X3,9X5,11)
/ = (1,11)(2,10)(3,9)(4,8)(5,7)
T1 = (0,1,2,3,4,5,6,7,8,9,10,11)
The group generated by the transpositions and the multiplication by 5 or 7 has 24
elements
\(TUM<>)\ = \(T1,M7)\ = 2A
If we add the inversion, the group has 48 elements
1 ( ^ , 7 ^ 5 ) 1 = 1 ( ^ , 7 ^ 7 ) 1 = 48
In the alpha familly, Morris considers the permutations that exchange the neighbor notes.
The prototype of this familly is the permutation
a = (0,1)(2,3)(4,5)(6,7)(8,9)(10,11)
The group generated by the permutations a, Ti, M5, M7 and J defines some new trans-
formations which can act on pcsets or on twelve-tone series. The smallest group is the
group
Ga = (Tua)
generated by the transpositions T\ and the permutation a of neighbor notes. It has 72
elements and has the presentation
In the beta familly, Morris considers the permutations of 2-cycles which exchange the
notes distant of one tone. The prototype of this familly who has 8 permutations is the
beta permutation
/3 = (0,2)(1,3)(4,6)(5,7)(8,10)(9,11)
The smallest group generated by this permutation (Ti,/J) has 108 elements and the
largest (Ti,/3, J,M 5 ) has 432 elements.
103
4. Neo-Riemannian Investigations
In the gamma familly, Morris considers the permutations of 2-cycles which exchange
the notes having the distance equal to one minor third. The prototype of this famille
who has 8 permutations is the element
7 = (0 > 3)(l,4)(2,5)(6,9)(7,10)(8,ll)
The smallest group generated by this permutation (Ti, 7) has 96 elements and the biggest
(Ti,7,7,M 5 ) has 192 elements.
In the delta familly, Morris considers the permutations which exchange the notes dis-
tant of one major third. The prototype is
5 = (4,8)(5,9)(6,10)(7,11)
The group (Ti,<$,M5) has 648 elements and the group (Ti,<S,/,M) has 1296 elements.
The following table gives the order of the group indicated in the first column when x is
one of the permutations a, /?, 7 or S.
G a P 7 5 1
(Tx,x) 72 108 96 648
(Ti,I,x) 144 216 192 1296
(T!,M5,x) 144 216 192 648
(TuM7,x) 72 216 96 1296
(TuM5,I,x) 144 432 192 1296
{TltM7tI,x) 144 432 192 1296
The same study has been done with the permutations which exchange the notes of a
limited transpositions set and leave the others unchanged. One can show that the group
(Ti,M 5 ,a;) extends the possibility of schoenbergian transformations, since the biggest
group has 46080 elements. The following table gives the order of the group (Ti,Ms,x)
for each limited transpositions set.
| Names X \{TuM5,x)\
| Tritone (0,6) 768
Aug. Fifth. (0, 4, 8) 648
| Dim. Seventh (0, 3, 6, 9) 384
| Alt. Seventh (0, 2, 6, 8) 2304
Double Fifths (0, 1 , 6 , 7 ) 23040
| Mode 1 (0, 2, 4, 6, 8, 10) 144
Mode 2 (0, 1, 3, 4, 6, 7, 9, 10) 384
Mode 3 (0, 1, 2, 4, 5, 6, 8, 9, 10) 3888
| Mode 4 (0, 1, 2, 3, 6, 7, 8, 9) 23040
| Mode 5 (0, 1, 2, 6, 7, 8) 46080
Mode 6 (0, 1, 2, 4, 6, 7, 8, 10) 3840
Mode 7 (0, 1, 2, 3, 4, 6, 7, 8, 9, 10) 46080
Mode A (0, 1, 4, 5, 8, 9) 3888
Mode B (0, 1, 3, 6, 7, 9) 46080
Mode C (0, 1, 4, 6, 7, 10) 46080
Chromatic Scale (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11) 24 |
104
5
Knots and Braids
This chapter shows three applications of knots theory to the musical field. The first ap-
plication deals with dodecaphonic series, which can be classified with Gauss diagrams and
so leads to a better understanding of their structural meanings. The second application
concerns tuning and temperaments and shows how to build a new temperament from a
knot. The last application is more general as it could be used in any text. It suggests
how to highlight knots in literary or musical texts, and shows new topological structures
of low dimension.
diagrams. Two knots are equivalent (ambient isotonic) in space if their planar diagrams
are equivalent under the three Reidmeister moves.
One defines the number of crossings c(K) of a knot K as the minimal number of
crossing points in the planar representation of the knot K. A crossing of an oriented
knot is said to be positive or right (resp. negative or left) if somebody moving on the
higher branch sees the lower cord passing from right-hand side to the left (resp. from the
left-hand side to the right). The index of a right crossing is +1 and the index of a left
crossing is —1. For a knot K, the linking number is the sum of all the index crossings,
and for a link K with Kn components (strands), the linking number is the half of the
sum of the linking number of each component
lk{K) = ±5>0K„)
n
One way of classifying knots is to consider the minimal number of crossings. The sum of
two oriented knots is obtained by removing a small arc from each knot and to glue the
four endpoints respecting the orientation of each knot. The unknot O is the untangled
circle S1.
Definition 130 A knot K, different from the unknot, is called a prime knot if the sum
K = K\ + if2 implies that K\ or K<i is the unknot.
Knots are designated by their crossing number and an order number. The knot 3i
is the trefoil knot, the knot 4i the eight knot, etc. Prime knots have been tabulating
until 16 crossings. The first work was done by Thomas K. Kirkman in the ISSO's. The
classification was studied by C. N. Little and reviewed by Kurt Reidmeister up to 9
crossings in 1932. John Conway, Hugh Dowker, Jim Hoste and Morwen Thistlethwaite
contributed to the classification which is now known up to 16 crossings. The following
table gives for each number of crossings (first column) the number of prime knots (second
column).
Cross. Primes || Cross. Primes
3 1 10 165
4 1 11 552
5 2 12 2176
6 3 13 9988
7 7 14 46972
8 21 1516 253293
9 49 |1 1388705
To characterize knots, mathematicians try to find new invariants. These invariants are
numbers (like the linking number), polynomials or groups. For a general introduction,
see [Adams, 1994], [Lickorish, 1997], [Sossinsky, 1997], etc.
106
Dodecaphonic Knots
the first paragraph, we show how to build the Gauss diagram of a twelve-tone row. Let
us choose a series, for example
According to the identification of Z12 with the twelve notes, this series could also be
written as
0,8,6,7,2,4,9,11,1,5,3,10
Put the notes of the series on a circle
G
Gb
" \ 8 Ab
\
0 C
\ /
*°Bb
Bit
V Eb
Db
Rub out notes. What is left is an abstract structure called a Gauss or a chord diagram.
It is easy to verify that the 48 derived forms of the 12-tone row are mapped on the same
Gauss diagram. That is why Gauss diagrams are used to classify dodecaphonic series.
All the transpositions of the series have the same chord diagram (you have eventually to
rotate the diagram). The retrogradation of the series is placed on the diagram which is
the mirror symmetry of the original diagram. The inversion of the series uses the same
107
5. Knots and Braids
chord diagram and the retrogradation of the inversion is mapped by the mirror diagram.
All the 48 derived serial forms use the same chord diagram (viewed in the space, possibly
mirroring or rotating).
The couples are (0, 2), (1, 4), (3, 8), (5, 11), (6, 10) and (7,9). Put the numbers on the
circle from 0 to 11 and join the numbers of the couples by a chord, we obtain the Gauss
diagram associated with the singular knot
108
Gauss Diagrams
Remark that in general several knots are associated with the same Gauss diagram
(change for example one of the undercrossings to an overcrossing). In the same way,
several twelve-tone rows are associated with the same Gauss diagram. By labelling each
chord by a letter and by reading the word starting from 0 one obtains the following
diagram
1 ]
geG
Since G has 48 elements, we have to determine the fixed points for each element. But
the elements Tn (1 < n < 11), TnR (n ^ 6), T n J , TnIR (for n even) have no fixed
points and the identity has 12! fixed points. It remains to look at seven elements of G,
the operation T6R and TnIR for n odd. The fixed tone rows by the operation T6R are
the tone rows in which the six last entries are the reverse of the six first transposed by
a tritone. So there are
£ = 1 2 x l 0 x 8 x 6 x 4 x 2 = 2 6 -6!
ways of doing this. The same can be done for the transformation TnIR: the first six
entries determine the number of fixed tone rows. So they are also 2 6 • 6! series, for each
n odd. Finally there are
12! + 7 • 2 6 • 6!
109
5. Knots and Braids
fixed tone rows. Thus we get | 5 | = (12! - t)/48 = 9985920 tone rows classes.
More generally, for the n-tet, we have the following result.
•
Theorem 132 In the n-tone equal temperament (n > 3), there are exactly n\ series of
notes. Under the equivalence of the derived forms (i. e. under the action of the dihedral
group), there are
Under the action of the dihedral group, the number of Gauss diagrams in the 2n-tet is
given by
with
L»J
n\
E k\(n-2k)\
k=0
where \x\ indicates the greatest integer less than or equal to x.
The proof of this theorem is given in the paper of Khruzin [Khruzin, 2000].
The following table gives the number of Gauss diagrams (dn) for the 2n tone equal
temperaments. For the usual equal temperament (n = 6), there are exactly 554 Gauss
diagrams which classify dodecaphonic series with respect to their structural properties.
n Cn dn Temp.
3 5 5 6-tet
4 18 17 8-tet
5 105 79 10-tet
6 902 554 12-tet
7 9749 5283 14-tet
8 127072 65346 16-tet
9 1915951 966156 18-tet
10 32743182 16411700 20-tet
11 625002933 312702217 22-tet
110
Gauss Diagrams
A complete catalogue of the 554 Gauss diagrams has been set up. The diagrams are
classified by the Gauss word order. Each of the six chords are labelled by the numbers 1,
2,.., 6 and the Gauss word is a permutation of 112233445566. For two Gauss diagrams D
and L, D < L if and only if w(D) < w(L) where w(D) is the Gauss word of D. We can
also write the Gauss word in a form slightly different by locating the chords of different
types: the chord a connects two consecutive points of the Gauss diagram, b connects
the two extreme points of three consecutive points of the circle, and so on, until the
type / which connects the two extreme points of seven consecutive points of the circle.
Moreover, one uses the opposite (e.g. e _ 1 is a chord linking the two extreme points of
eight consecutive points). For example, the series of Jean Barraque Au-deld du hasard
(0, 8, 7, 1, 4, 2, 10, 3, 11, 5, 6, 9) is represented by the word
X = aba1-1 dab
and corresponds to the diagram D m in the classification of the Gauss diagrams. The
structural vector of a Gauss diagram is made of six components which count the number
of chords of each type a, 6, c, d, e, / . For example, the structural vector of D m is 220200.
This means that this diagram includes two chords of the type a, two chords of the type
6, none of the type c, two chords of the type d and none of the types e and / . In the
classification, the first Gauss diagram D\ is used by B.A. Zimmermann in his opera Die
Soldaten.
Di X = a6
Gauss Word 112233445566
Structural Vector 600000
Permutation
(0 1) (2 3) (4 5) (6 7) (8 9) (10 11)
B.A. Zimmermann, Die Soldaten, Acte I
Another example is the following diagram D349 which is used by Karel Goeyvaerts in
his Sonata for two pianos.
^349 X = afd~xe2a
Gauss Word 112345662453
Structural vector 200121
Permutation
(0 1) (2 8) (3 11) (4 9) (5 10) (6 7)
Karel Goeyvaerts, Sonata for two pianos.
The symmetric diagram D358 has been used by Anton Webern in his Symphony, opus
m
21.
Z/35g X = ac~1e~1eca
Gauss Word 112345665432
Structural Vector 202020
Permutation
(0 1) (2 11) (3 10) (4 9) (5 8) (6 7)
A. Webern, Symphony, opus 21
in
5. Knots and Braids
The last diagram of the dodecaphonic classification is used by B.A. Zimmerman in the
Interludes of his opera Die Soldaten.
£>554 X= f
Gauss Word 123456123456
Structural Vector 000006
Permutation
(0 6) (1 7) (2 8) (3 9) (4 10) (5 11)
B.A. Zimmerman, Interludes.
Ecce Ancilla Domini of Jean Etienne Marie is built on a series of 24 notes in the
space of quarter-tones (F %+, Ftf, G, B, C\+, A, D+, C, F, G|t, A$+, Ctf, B+, E, G$+,
Aft, G+, D%, F+, E-h, A-h, D, D$+, C+). The chord diagrams in this space can be
represented by a Chinese diagram linking two Gauss diagrams.
5,4,0,9,7,2,8,1,3,6,10,11
by Luigi Nono in Canto Sospeso (D^s)
9,10,8,11,7,0,6,1,5,2,4,3
by Ernst Krenek in the Studies in Counterpoint (Di3$)
3,6,1,7,0,2,11,10,8,4,5,9
by Andre Riotte in many pieces and in particular in Multiple for string quartet composed
in 1962 and in his Inventions for piano solo (D50)
0,1,5,2,7,6,8,11,9,3,10,4
and by Milton Babbitt in Three Compositions for piano (£353)
10,3,5,2,0,1,7,11,6,9,8,4
112
Serial Algebra
The Gauss diagram of these series includes two (and only two) chords of type a, because
the first and the last note of the series is a tritone. Remark that all the type of chords
are not necessarily present in an all-interval series. For example, in the Gauss diagram
D14 which is denoted by X = a2bc2b, there are no chords of type d, e and / .
Definition 134 A transpose series is a series in which there is a transposition from the
set A to the set Ac. If the set of the transpose series is denoted by T, we have
SeT&3i Ti(A) = Ac
Definition 135 A reverse series is a series for which it exists a transposition from the
set 1(A) = —A to the set Ac. If we note TZ the set of reverse series, we have
SeTl^3i Ti(-A) = AC
Definition 137 A semi-reverse series is a series for which it exists a non-identity trans-
position which sends the set —A to the set A. If TV denotes the set of semi-reverse series,
we have
S e TV <# 3 t ^ 0, Ti(-A) = A
Examples. The series of Luigi Nono's Canto Sospeso (9, 10, 8, 11, 7, 0, 6, 1, 5, 2, 4, 3)
which has a Gauss diagram D358 is a semi-reverse serie, because Tj(—A) = A. The series
113
5. Knots and Braids
A of Jean Barraque's Concerto (7, 4, 9, 0, 6, 1, 2, 10, 5, 11, 3, 8) which has the Gauss
diagram D271 is a semi-reverse serie.
Many series do not belong to the four previous types. For example, the series of Pierre
Boulez's Sonatine for flute and piano (0, 11, 7, 1, 8, 4, 3, 9, 2, 10, 5, 6), the Cycle de
Partisan furieux of Le Marteau sans Maitre (3, 1, 4, 5, 9, 6, 10, 7, 8, 11, 2, 0), Karel
Goeyvaerts's Opus 2 for 13 instruments (11, 6, 5, 4, 7, 8, 3, 2, 9,10,1, 0), Jacques Lenot's
Third Sonate for piano (5, 8, 7, 2, 6, 0, 3, 9, 10, 1, 4, 11), Peter Schat's Polonaise opus
2 (6, 8, 9, 11, 2, 10, 0, 1, 3, 4, 5, 7), Jan van Vlijmen's Costruzione per due pianoforti
(0, 1, 5, 4, 7, 3, 2, 9, 8, 10, 6, 11) not belong to the set T U V U Tl U W. The following
figure shows the different relationships between the serial sets.
18 % (84)
A. Schoenberg Op. 30
Remark that the classification of Costere-Parsysz works with non well-ordered sets
and does not preserve the structural properties of the series. This classification is fully
compatible with the Gauss diagrams classification.
5.6 Combinatoriality
The combinatoriality is a notion first introduced by Milton Babbitt. Since combinatori-
ality applies to any set, we only consider the combinatoriality of hexachords in order to
compare the two classifications of Costere and Babbitt. The basic transformations and
their associated rows are denoted by P (basic row), I (inversion), R (retrograde) and RI
(retrograde of the inversion).
Definition 138 A semi-combinatorial row is a twelve-tone row such that any of its
transformations other than the retrograde can be transposed so that the first six notes are
equivalent to the last six notes of the original set, not necessarily in the same order.
114
Combinatoriality
115
5. Knots and Braids
S = AL\T3(A) = AuT9(A)
S =[0,11,7,8,3,1,2,10,6,5,4,9]
Iv(S) = [3,2,2,4,3,1]
Pij=¥(Xn = j\Xn.1=i)
which means that the Markov chain doesn't take into account the whole history of the
process but depends only on the last state (at time n — 1). Under suitable conditions, it
exists a unique stationary distribution of the chain A satisfying the condition \P = A.
More precisely for all states, we have
116
Markov Chains
n J
k=o
S = - J2XiPi* lo
&Pii
Markov chains have been used by Iannis Xenakis for example in his piece Analogique B.
The composer defines some Markovian random processes according to range of pitch
classes, intensity and density. He uses two transition matrices
/ 1/5
^ 4/5
4/5
1/5 ) <K'")
The entropies of the two matrices are quite similar: S(P) = 0.707 and S(Q) = 0.722.
In Archorripsis, Iannis Xenakis used a discrete distribution of independent events.
On the finite set (#i, #2, > •••> xn) the Poisson distribution depending on the parameter A
attributes the probability
One shows that the mean and the variance of the Poisson distribution is equal to the
A parameter. In Archorripsis, the unfolding of the pitch classes, speeds and durations
are also organized according to the Poisson distribution. For each musical parameter,
random points are controlled by the parameter A. The matrix used by Xenakis is given
in the following figure. The number of events of an element of the matrix is proportional
to the darkness of each cell. Dark cells represent 4 events by measure, and white cells
represent no event in the measure.
Time
Markovian analysis could also be used to measure some gaps in serial compositions.
For example, consider the transitions matrices of Anton Webern's opus 17, n° 1 defined
as follows. The theoretical transition matrix Qij represents the transitions from the pitch
117
5. Knots and Braids
class j to the pitch class i in the 48 forms of the dodecaphonic serie. Write the original
series and its 11 transpositions. Do the same for the retrogradation series, the inverse
and the retrogradation of the inverse: we obtain 48 series of 12 tones. Now count the
transitions from j to i and divide by the total number of transitions, we obtain the
probability transition Qij which is a 12x12 matrix.
Qi Q* \
Q
= Y4 Q 2 Qi J
with
/0 0 1 \
7 7 2
0 0 2
Qi = 2 7 0
2 0 7
\1 2 0/
and
/0 2 1 0 7 \
1 1 0 2 0
2 0 1 2 2
Q% = 2 1 2 1 2
0 2
2 1 0 1
V7 0 2 2 1 0 /
The transition matrix Py represents in the score the empirical probability of transition
from the pitch class j to the pitch class i, that is the number of j followed by pitch class
i divided by the total number of transitions.
p= Pl Pi\
\ * Pj
with
°
I 3/283/17
0
7/34
2/7
3/34 0 1/17 \
1/14 0 1/14
0 0 0 1/26 1/13 2/13
Pi = 2/27 2/27 1/9 0 5/27 1/27
6/25 3/25 1/25 2/25 0 1/25
V 1/26 0 1/26 3/26 2/13 o J
The submatrix Pi is
118
Groups and Diagrams
The submatrix P3 is
°
( 1/10 1/6
0
7/30 1/30 1/10
1/6 1/6 1/30
1/30 \
1/30
1/29 2/29 0 6/29 0 2/29
P4 =
1/33 1/33 2/33 0 2/33 4/33
2/25 1/25 3/25 3/25 0 4/25
^ 7/26 1/13 0 0 3/13 0 /
Using the 2-norm of a matrix A, defined as the supremum of the eigenvalues of A.A1
we find for the Webern's piece
||P - Q | | 2 « 0.65
The 2-norm between P and Q defines the discrepancy of a serial composition, a kind of
distance between the musical piece and the raw materials. It is another method to map
the serial pieces.
119
5. Knots and Braids
(0,1)(2,11)(3,10)(4,9)(5,8)(6,7)
These pairs define the permutation P358 associated with the diagram Ds^s-
/ O l 2 3 4 5 6 7 8 9 10 11 \
358
^ 1 0 1110 9 8 7 6 5 4 3 2 )
Gn = (P n ,Ti,7)
The order of this group is called the generating power of the diagram Dn.
Examples. 1) The group associated with the diagram D358 is the group
G358 = CP358»?i,I)
The GAP software gives the presentation of this group in terms of generators and rela-
tions
G358 = (a, 6, c I a 2 = b12 = c2 = ab^c = l )
The group has
|G 3 5 8 | = 24
elements. This group and the dihedral group have the same number of elements.
2) The series D554 used by B.A. Zimmermann in his opera Die Soldaten is also used
by Pierre Boulez in the Cycle des bourreaux de solitude (Piece 6) of Le Marteau sans
Maitre.
2,3,4,5,6,7,8,9,10,11,0,1
The associated group
has 24 elements.
3) The series used by Anton Webern in his Concerto for nine instruments, opus 24
11,10,2,3,7,6,8,4,5,0,1,9
120
Tuning Braids
is associated with the diagram -D538. The associated group generated by the three oper-
ators
G538 = (Pb38,Ti,I)
has 7680 elements.
5) The series used by Dallapiccola in Quademo Musicale di Annalibera
10,11,3,6,8,2,7,1,5,9,0,4
is associated with the diagram D$Q. The group generated by the three operators
I I X I
1 2 i i+1 n
I I X I
The set Bn of braids with n strands has a natural structure of group, called the Artin
group. The product of two braids a and b is the braid ab obtained by putting the two
braids end to end. The inverse of a braid b is a mirror image of b.
121
5. Knots and Braids
Theorem 145 (Artin) The braid group Bn is isomorphic to the group whose generators
are G\, ...,cr n _i and verifying the following relations
GiGj1 = O^Oi = 1
aiaj=o-jai if\j-i\>2 for ij = 1,2, ...,n - 1
OiGi+\Oi = <7i + i<7i<7i + i % = 1,2,..., n — 2
These relations can be used to verify that two braids are equivalent. For example, the
braid a^1 (c^o^c^cr^"1 is equivalent to cr2
Some braids are used to describe tuning and temperaments. Meantone temperaments
are based on two different elementary intervals u\ and cr2. The first interval G\ is the
distance between the first two notes (C-CJJ) and the second interval cr2 is the distance
between the second notes (C(t-D). In the meantone temperaments, two consecutive notes
are separated by either o\ or <J2. The first interval concerns the notes (C-Cfl), (Eb-E),
(F-F(t), (G-GjJ) and (Bb-B). The other intervals are of type cr2. Meantone temperaments
are described by the word
Another example is the Marpurg H temperament built on two basic intervals denoted
a and b: a is equal to 96 cents and b to 102 cents. The structure of the temperament
repeats (abb) four times.
(abb)(abb)(abb)(abb)
122
Tuning Braids
a
P"^^c 1 1
b
b
^^^^.J.. a
h.J
b
a
,, J G
b.
b
IsTS a
b ••'• Bb
1 1 b f'" B
L_ fi
In the same way used by Harry Partch, one can build new temperaments. Let's choose
a knot of twelve crossings and a law of crossing. For example, just multiply the interval
of the crossed branch. To have a consistent law of crossing, the intervals must verify the
two closure relations
c 2 = ab2
and
a2bcs = 1
By choosing different values for a, b and c verifying the closure relations, one can do the
computation of all intervals modulo the octave, possibly by multiplying or dividing by a
power of 2 in order to get each interval in the same octave. For example, for a = 16/9,
b = (20) 1 / 4 /a and c = 56/4 we get a = 773 cents, 6 = 524 cents and c = 789 cents. This
new temperament is represented in the following figure.
123
5. Knots and Braids
[ [ black shirt ;
red noise
(( black towel ]
; red) shadow
red ; red
One can ask to find all the word white. Each occurrence will be denoted by A, the
occurrence of the inverse word black will be denoted by small a. The same can be done
for other colors: the occurrences of the word red are denoted by B, and the occurrences
of its inverse blue are denoted by small 6. The word obtained is
aBaBBB
One can check if this word has a corresponding braid or not. The word aBaBBB cor-
responds to the knot 62. In this way, we hope to find many hidden braids in a given text
and we suppose that this will show topological relationships. All parameters are involved.
If we look at the punctuation marks, and if we assign the bracket [ to the letter A, the
inverse bracket ] to the letter small a, the sign ; to the letter B and the parenthesis ( t o
C and its inverse ) to the letter c, we find the word
AABCCaBcB
which represents the knot 74. Other occurrences could form a knot such as lexical units
or classes, isotopic relations, semantic units, as well as contextual representations. In a
musical text, in the same way, one could choose a pattern A and its inverse a, for example
the retrograde notes or rhythms, or other parameters. One can choose also other patterns
B, C, etc. and try to find a word that corresponds to a braid. This method can be used
to find invariant knots. For example, in the evolution between languages in linguistics
or in musicology, we could find a knot or a set of knots based on structures of a given
124
Textual Analysis
language L\ and the same knots in a language Li which is more recent than Li, This
could be an argument to maintain that the language L2 has relationships with L\. In
that case, the relation is supposed to be relevant and through the analysis of many texts
leads to some new categorical classifications.
125
6
Theory of Temperaments
The tuning of musical instruments has been widely studied in the world by all civiliz-
ations. The disciples of Pythagoras, Aristoxenus, Ptolemy and others have set out the
Greek tuning theories. The Pythagorean system based on the natural fifth and on the
octave was used in Europe in the fifteenth century. Later, musicians tried to complete the
cycle of fifths. But it is well known that there is no integer solution to this problem. Many
approximation solutions appear, consisting of tempering the fifths, that is diminishing
the ratio of the natural fifth to adjust successive fifths in an entire number of octaves.
Other musicians and theorists gave up with the idea of diminishing the fifths and tried
with meantone temperaments to set a tuning system with natural thirds. Others with
just intonation tried to use small and whole numbered ratios, and others tried to establish
a coherent set of ratios using irregular divisions. As there are great implications from the
esthetical point of view, the problem has passed through the centuries and is still active.
The coefficients verify the relations (in what follows, all sums are considered modulo 12)
12 12
i=l i=l
The partial sums are defined by
and
^ = 1^2 E, | i + ¥ ^
where a is defined by
1200 37
a= In
ln(2) V213/12
that is about 14 cents.
128
Pythagorean Tunings
If all the fifths are tempered by a fraction ai of Pythagorean comma (i.e. if all 6^ = 0),
the table of the general temperament of twelve tone can be written as follows.
129
6. Theory of Temperaments
130
Historical Temperaments
a
=2*k
and
2 ,5r
35
The general form of the meantone temperaments is given in the following table.
©'
has no integer solutions (except the trivial case (0,0)). Pythgoras has proposed an approx-
imate solution (12, 7). Musicians and theorists have tried to get approximate solutions,
usually by diminishing the pure fifths with a fraction of comma. The system thus obtained
is called a temperament.
131
6. Theory of Temperaments
1
a = a0-\ =
ai +
1
a2 +
or in a compact abbreviated notation
a = [a 0 ;ai,a 2 ,...]
with a semicolon to separate the integer part. When the integer part is null (ao = 0), the
notation a = [a\, a 2 ,...] is also used. For typographic convenience, the continued fraction
is also written as
1 1 1
a0 + ; ; "...
It is well known that every real number a has a unique continued fraction expansion.
This expansion is finite if a is rational (a € Q) and infinite if a is irrational (a € '.
7. Approximations of Temperaments
Let [a] be the integer part of a, the largest integer less or equal a. Define the sequence
xn by the relations x$ = a and
1
Xn+l =
Pn r , 1
c n = — = [ao;ai,...,a n J = a 0 H =j
*n a2 +
1
+—
Example 1. The computation of the n-th convergent of
Proposition 146 Continued fractions of a real number a verifies the following proper-
ties:
Proof (1) It is easy to see by induction that (qn) is a strictly increasing sequence, since
qo = 1 < qi = a\ and if qn-\ < qn we have
<?n+l n , Qn-1 . „ . -,
134
Continued Fractions
Pn+1 = Pn^n+1 + P n - 1
Qn+1 = qnQ>n+l + Qn-1
leads to
Pn+ltfn " P n t f n + l = - ( P n t f n - 1 ~ Pn-lQn-l)
and since
Piqo -poqi = l
we get
Pn+l?n -Pnqn+1 = (-1)*
(3) Dividing the previous equation by qnqn+\> we get
Pn+1 Pn _ (~l)n+1
tfn+1 qn qnqn+1
or
(-l)n+1
Cn+l ^n —
#n<7n+l
Thus, the distance between cn and c n _2 is
Pn Pn-2
Cn - C n - 2 =
<Zn qn-1
Pn-1 Pn-2 ("I)71"1
<Zn-l <7n-2 qn-iqn
n 2
(-l) ~ (-l)^1
qn-2qn-i qn-iqn
T h e even convergents C2& form an increasing sequence and t h e odd convergents C2k+i
form a decreasing sequence. •
T h e o r e m 1 4 7 For n € N, we have
1 Pn 1
< a <
tfn(tfn+l +tfn) tfn qnqn+\
Pn\ Pn+1 _ Pn
a— <
Qn\ | #n+l tfn | tfntfn+1
135
7. Approximations of Temperaments
thus
1 1 1
tfntfn+i
< ^
2gn_ign
<
4qn-iqn-2
<...
T h e number a is t h e sum of t h e alternate serie
oo
OO / \
^°°tfn qo ^ V ? n Qn-lJ
n
Po 1 (-l)
- ^
40 + tfotfl + ... + tfntfn+l
- — + ...
•
T h e latest property leads to t h e following theorem which gives the best approximation
of a real number.
T h e o r e m 1 4 8 Among the fractions p/q with q<qn, the closest to the real number a is
the convergent pn/qn.
V Pn
a —q <^ a - qn
and suppose, in a first time, t h a t t h e fraction p/q is between cn = pn/qn and c n + i =
Pn+i/<?n+i- T h e sequence (..., Pn/qn,p/q,Pn+i/qn+i, ~) is increasing or decreasing. For a
non-negative integer N,
Pn _P N Pn+l 1
<
qn q qnq qn+i qn #n<Zn+l
t h a t is q > Nqn+\ > qn+i. If t h e sequence is increasing, using the property (2), we get
Consequently p > p n + i . Now if p/q does not belong to the interval [c n , c n +i] since
1 1
qnqn+i < ^ 2qn-1qn < ..
136
Continued Fractions
the length of the interval [c n ,c n + i] is at most the half of the length of [c n _i,c n ]. Thus
p/q belongs to [c n _i, c n ]. Prom the first part of the proof, we see that
V Pn
a - < a - q
Q n
we have q > qn and p > pn. The approximation of the real a by a convergent is better
than the approximation by other fractions simpler than the convergent. •
If the rational p/q is a best approximation to a, then p/q is a convergent. But it is not
true that if the rational p/q satisfies
1
a
q2
then p/q is a convergent of a. However a theorem of Hurwitz states that if a rational
number satisfies
a <
q 2q2
then it is a convergent in the continued fraction expansion of a.
Examples. To find q just fifths (3/2) lying in p octaves, one has to solve the equation
This equation has no integer solutions. So, we have to find the best approximation
f-^|) = 0
representing the frequency of the individual intervals of the 1200-tet, for each k =
1,..., 1200. In the sequence of the convergents c£\ which are expressed in cents by the
formula
1 (n)
» 1200 log Ck
log 2
we keep only the first convergent (denoted simply c&) such that the approximated ratio
is less than 0.5 cents
» k\ < l / 2
137
7. Approximations of Temperaments
0.3
0.25
0.2
0.15:
0.1:
0.05 :
~ d ' " " '260' ' " 400 ' ' 6do" " ' 800' " ' iobd ' ' 1200
Superpartial ratios (i.e. ratios of the form (fc + l)/k) have an important property:
superpartial ratios are Hellegouarch commas. They are ratios of the first just intervals
(2/1, 3/2, 4/3, 5/4, 6/5, etc) and also they are the best approximations of microtones.
In the sequence of convergents of the approximation of a microtone of frequency 21/71 in
the n-tet, there is always a superpartial ratio. The following table shows the superpartial
ratios in the computation by continued fractions of the microtones (second column), the
number of degrees n (n-tet, first column) and the values in cents of the microtones (third
column).
110n Ratios
15/14
Cents
120
136n Ratios
52/51
Cents
33
12 17/16 105 | 53 77/76 23
17 25/24 71 54 79/78 22
18 26/25 68 60 87/86 20
19 28/27 63 66 95/94 18
24 35/34 50 72 104/103 17
31 45/44 39 | 96 138/137 13 |
138
Meantone Approximations
V = log(2)
q log(3/(2ft»))
The following table gives the approximations for the classical meantone temperaments.
The number of degrees m of the approximated m-tet is the denominator of the conver-
gents. The two first convergents (1 and 1/2) are omitted. The last column gives the usual
approximation.
1 0r Authors
Pythagoras
Convergents
3/5, 7/12, 24/41, 31/53, 179/306
Us.
53
1/9 Romieu 3/5, 4/7, 7/12, 130/223, 137/235 223
1/8 Romieu 3/5, 4/7, 7/12, 74/127, 81/139 139
1/7 Romieu 3/5, 4/7, 7/12, 46/79, 53/91, 364/625 91
1/6 Silbermann 3/5, 4/7, 7/12, 32/55, 39/67, 71/122 55
3/17 Riccati 3/5, 4/7, 7/12, 25/43, 32/55, 569/798 55
1/5 Sauveur 3/5, 4/7, 7/12, 18/31, 25/43, 768/1321 43
3/14 Riccati 3/5, 4/7, 7/12, 18/31, 25/43, 43/74, 74
5/23 Smith 3/5, 4/7, 7/12, 18/31, 43/74, 534/919 74
2/9 Rossi 3/5, 4/7, 7/12, 18/31, 43/74, 61/105 74
1/4 Aaron 3/5, 4/7, 7/12, 11/19, 18/31, 101/174 31
5/18 Smith 3/5, 4/7, 7/12, 11/19, 29/50, 707/1219 50
2/7 Zarlino 3/5, 4/7, 7/12, 11/19, 29/50, 69/119 50
3/10 Harrisson 3/5, 4/7, 7/12, 11/19, 40/69, 51/88 69
1/3 Salinas 3/5, 4/7, 7/12, 11/19, 733/1266 19 1
The classical meantone temperament has irrational ratios (C, 1, Db, 5 7 / 4 /16, D >/5/2,
Eb 4/5 3 / 4 , E, 5/4, F 2/5 1 / 4 , Ffl, 5>/5/8, G 5 1 / 4 , G(J, 25/16, A 5 3 / 4 /2, Bb, 4/y/E, B, 5 5 / 4 /4).
To get a rational system that approximates this tuning, one has to find fractions to be
used instead of powers of 5. If we carefully look at the system proposed by Verheijen, as
it has been stated by Simon Stevin in Van de Spiegeling der Singeonst (1884), one could
show that if we use the following approximations
^ m v^ieo 4^i95
559 107 128
all the ratios of the Verheijen system are described (C, 1, D\>, 40/39, D, 625/559, E\>,
125/107, E, 5/4, F, 107/80, Fj, 3125/2236, G, 160/107, Ab, 25/16, A, 107/64, B\>,
1000/559, B, 200/107).
139
7. Approximations of Temperaments
Let aij be the number of steps giving the closest approximation to Cj in the i-tone equal
temperament and the error e^ of this approximation
e =
ij \cj ~~ 0i s
ij j\
Suppose that the target interval size is a random variable, then the probability of getting
an error is
Hall has done a systematic evaluation of equal temperaments through n = 612. It appears
that for the sequence of just intervals
F = {1,5/4,4/3,7/5,3/2,8/5,5/3,7/4}
the remarkability function exceeds 2.5 for 12,19, 31,41, 53, 62, 68, 72, 84, 87, 99,103,109,
etc. We have done the computation of the remarkability function for other temperaments.
Some results are given in the following table.
| Farey 12, 19, 31, 41, 53, 62, 68, 72, 84, 87, 99, 103, 109, 115, 118
| Partch 7-limit 19, 27, 31, 41, 53, 62, 68, 72, 84, 87, 99, 103, 109, 118
Partcli 11-limit 31, 41, 46, 58, 72, 80, 94, 118, 130, 152, 159, 183, 190, 198
Partch 13-limit 58, 66, 72, 87, 111, 130, 183, 190, 198, 224, 270, 296, 311
Zarlinian diatonic 12, 19, 22, 24, 31, 34, 41, 46, 53, 65, 75, 77, 84, 87, 94
Pythagoras 12, 24, 29, 41, 53, 65, 77, 82, 94, 106, 118, 130, 135, 147, 159
Kepler I 12, 41, 53, 65, 77, 94, 106, 118, 130, 159, 171, 183, 195, 205
al-Kindi 12, 24, 29, 41, 53, 65, 82, 94, 106, 118, 135, 147, 159, 171, 188
Mayumi Reinhardt 10, 66, 74, 87, 103, 113, 130, 140, 147, 177, 187, 198, 206, 207
The remarkability of the temperament of Mayumi Reinhardt {1, 14/13, 13/12, 16/13,
13/10, 18/13, 13/9, 20/13, 13/8, 22/13, 13/7, 208/105} based on the prime number
140
Multiple Continued Fractions
13 is very different from the other scales. Other functions such as the Paul Erlich's
harmonic entropy have tried to understand the relationship between non-equal and equal
temperaments, and to provide better approximations.
David C. Keenan has proposed the following definition of the harmonic error (limited
to three prime numbers 3, 5, 7) to measure the discrepancy of equal tunings in a way
very close to the remarkability.
Definition 149 Let R be the round function. For each prime number p > 3, we define
the p-th harmonic error of the n-tet by
1200
ep(n) log
nlog2 2#(nlogP/log2)
M§#(5?)
The following plot shows the variations of the inverse of the harmonic error for n from
5 to 400.
2-\
141
7. Approximations of Temperaments
XJI Oi Xn Oi
142
Multiple Continued Fractions
starts with the definition of the sequence (a n ,6 n ). The initial values are ao = a and
bo = p. The sequence is defined by the 2-dimensional Gauss transformation
T{x y)= m
' { ^(wVx)'mia(wl))
where {x} denotes the fractional part of x and
vn = v(Tn-1(a,/3))
where the function v{x^y) denotes the position of l/{x} in T(x,y), i.e. v(x)y) = i when
the z-th coordinate of T(x,y) is !/{#}. The matrices A(un, vn) are given by
i( un 1 0
A
(UnA) = 0 0
\° 0 1
and
j( un 1 0
A
(un,2) = \ 0 0 1
0 0
It has been shown that the product
Cn = A(univn)A(un-uvn-i)...A(uuVi)
we find the approximations (7/12, 4/7, 2/3) and (1/3, 2/7, 1/3), (11/19, 7/12, 2/3) and
(6/19, 1/3, 1/3), (18/31, 2/3, 11/19) and (10/31, 1/3, 6/19), (10/17, 11/19, 18/31) and
(11/34, 6/19, 10/31), etc. The modified Jacobi-Perron algorithm has been generalized to
d-dimensions.
143
8
Tuning Groups
The twelve tone equal temperament is generated by the repetition of the semitone inter-
val. Rational temperaments are also generated by the repetition of one or several intervals,
but some relations must be added, such as the approximation of the pythagorean comma
by the unit ratio, in order to close the cycle of fifths. The various methods to consider a
temperament are explained in this chapter.
8.1 Definitions
Equal temperaments are cyclic groups. They are generated by the repetition of a single
ratio { l , r , r 2 , . . . . , r n _ 1 } . This set is closed by the octavian relation rn = 2, which induces
an isomorphism of this set with the cyclic group. For the 12-tone equal temperament,
the ratio r is the semitone r = 2 1 / 1 2 . Unfortunately, most of temperaments are not cyclic
groups. As we would like to have a complete description of the frequency ratios by the
powers of r, we introduce the notion of a pseudo-cyclic group. Consider a temperament
T of n pitch-classes, i.e. an increasing sequence of frequency ratios xo = 1 and Xj G [1,^[
for j = 1,.., n — 1 (in general, 77 = 2 is the octave)
T = {l,xiyX2,...,xn-i}
T={l,r,r2c2,r3c3,...,rn-1cn_i}
r r
8. Tuning Groups
Example. The Pythagorean temperament T = {1, 256/243, 9/8, 32/27, 81/64, 4/3,
1024/729, 3/2, 128/81, 27/16, 16/9, 243/128} is a pseudo-cyclic group whose gener-
ator is the leimma r = 256/243 and the regulators are c<i = cz = p, C4 = C5 = CQ = p 2 ,
c7 = c$ = eg = p 4 and c\\ = c\2 = pb. The ratio p = 531441/524288 is the Pythagorean
comma. The set
describes each ratio and ends with the closure relation r12p5 = 2. The regulators define
a unique relation
P = i
which means that in order to close the cycle of fifths, the Pythagorean comma must be
a unit approximation or a unit like.
Definition 150 Two pseudo-cyclic groups T\ and T2 of same period 77 and same gener-
ator r are equivalent T\ 1Z T2 if they have the same units like.
Definition 151 Let r be a generator and ui, ..,u m be m units like. The tuning group G
generated by r modulo the relations Uj = 1, for j = 1, ..., m
is the class of all pseudo-cyclic groups for the equivalence relation 1Z.
Examples. (1) The set
H = | l , r , ra, r a^r a , r a , r a , r a , ,r a , r a ,r a }
is a system built with a generator r and only one regulator a. The closure relation is
r6a6 = 2. This system corresponds to the tuning group
G =< r, - = 1>
r
Meantone temperaments are pseudo-cyclic groups with one regulator.
(2) The set
is a pseudo-cyclic group with two regulators and the closure relation r 4 a 4 6 4 = 2. This
pseudo-cyclic group is associated with the tuning group
a
^ •. b
G =< r, - = 1, - = 1>
r r
For example, the group
9 3 ? 3 4
n - i - 1
G-< = 1 = 1 >
8' 2^ ' 215
describes the diatonic Zarlinian temperament.
146
Definitions
the smallest height. For example, let's consider the following presentation of the tuning
group
9 3?
-,
G = <
8 ' 2 * s l >
This group has one generator r = 9/8 and one relation u = 1, with u = 3 7 /2 1 1 . The
group G has elements of the form l , r , r 2 , r 3 , . . . . , r n modulo the relation given by u.
The ratios X{ = rluk (where i = 1,2, ...,n and k G Z) form an increasing sequence
#i < #2 < #3 < .... < xn limited by the octave. As there are several possibilities to
choose the number fc, there are several ratios. One choose the ratio Pijq% which have the
smallest height
ftmin = infmax(pi,gi)
The first note of our tuning system is C with ratio 1. If we multiply this ratio by the
generator r, we get x\ = r = 9/8. If we multiply again this value by the generator, we
have r2 = 81/64 or 408 cents. But this ratio has not the smallest height, because it can
be divided by u
_ r 2 _ 3 * 2 ^ _ 2* _ 32
X2
" "JT " ¥ ~W " 33 " 27
If we multiply this new ratio by the generator, we get the new value
2532 4
^3 = ^ = 3 3 2 3 = 3
which has the smallest height. Multiplying again the ratio #3 by the generator r leads to
X4 = x^r = 3/2. We build the temperament by the powers of the generator, modulo the
smallest height. We obtain: C (1), D (9/8), E (32/27), F (4/3), G (3/2), A (27/16), B
(16/9) and C (2).
Remark. One could also define a tuning group with several generators. Let R = {ri, ...r n }
be the set of generators and U = {^1, ...yum} the set of relations, the tuning group
describes the class of the pseudo-cyclic groups whose elements are of the form
r r
l 2 * ' * rn u
l '"um
with r = 10/9 and s = 9/8 is composed of all elements of the form r V , that is r, 5, r 2 ,
rs, s 2 , r 3 , r2s, rs2, s3, etc. modulo the relation. One can verify that this group leads to
a temperament of 26 notes.
T h e o r e m 152 Letpi, p2, P3,..., pn be prime numbers in increasing order and let r\ be
the number
n =Pl Vl P'S -'Pn'
147
8. Tuning Groups
Let G be the group generated by the prime numbers pi, P2, P3,---> Pn and H be the group
generated by the ratios
n = Pi P2 -Pnn with i = 1 , . . . , n
The number r\ is a generator of the group G/H if and only if the determinant of the
matrix A = (a^) is ±1
•^n,™
gcd(ai)n+i,a2,n+l,---)^n,n+l) = 1
__ 28 31* 3*__
G
-<35> 2i9 = 1 > 24 = >
148
Formal Languages
8 -19 -4
8 -19
detA = -5 12 4 = — -5 = -(96-95) = - !
12
0 0 -1
If we note p the Pythagorean comma, k the syntonic comma, the usual tuning groups
are
G = < 21/71 >, the n-tone equal temperament
G = < r , p = 1 >, the Pythagorean temperaments
G = < r , p = 1, fc = 1 >, the syntonic temperaments
G =<r, k12n = p >, the meantone temperaments
G = < (3/2)*/ n >, the non-octaviant temperaments (n = 7, TEQJ of Serge Cordier,
n = 9, Wendy Carlos's alpha, etc.)
G =<p, 2 65 = 3 4 1 >, the Janko temperament of 41 notes
G = < p, 2 84 = 3 5 3 >, the Mercator temperament of 53 notes
X = (ab)(ab)(ab)(ab) = (ab)4
describes a system in which the first interval is a, the second 6, the third a, etc. With
this convention, the tempered system is denoted by a 12 . The problem underlying the
description of the tuning systems is to understand the relationships of the chords and
scales in non-tempered systems which are modified under transpositions. For example
the major scale have a structure of the form: tone, tone, semi-tone, tone, tone, tone,
semi-tone (ttsttts). In the meantone temperaments which are described by the word
X = (abbab)ab(abbab)
U = {ab)2{b){abfb
In this expression, parenthesis represent the intervals of the scale: the first interval is
(ab), the second (ab) the third 6 etc. One can see easily that the major scale D, F, G,
A, B\> has the same structure. But Cft major has a complete different structure
(b2)(ab)(a)(ab)(b2)(ab)(a)
The study of these relationships and their topography could explain the predominance
of some modulations.
149
8. Tuning Groups
Example 1. The tempered system is the simplest group. It is built by the repetition of
the semitone interval, which is the generator of the tuning group
G =< 2 1 / 1 2 >
It is described by the concatenation of the twelve repetition of the interval a of 100 cents.
X = a12
All transpositions have the same structure.
Example 2. The Pythagorean system is a tuning group in which the leimma (£) is the
generator and the Pythagorean comma (a/£) is a unit like
X = {iataU)p{iHataH)
in which p is the Pythagorean comma between G\> and F§, I is a leimma of 90 cents
(256/243 = 2 8 /3 5 ) and a denotes the apotome of 114 cents. Leimma, apotome, and
Pythagorean comma satisfy the relation
a = £ . p = 3 7 /2 1 1
Example 3. The meantone temperament of Pietro Aaron is a tuning group built with the
semitone generator and in which the enharmonic comma is a unit like
57/4 27 ,
X = (abbab)ab(abbab)
where a is the semitone of 76 cents (5 7 / 4 /16) and 6 is a semitone of 117 cents (8/5 5 / 4 ).
X = (abbab)ab(abbab)
7
3 28
7 n anc sanotner
where a is the semitone ( ^ j ^ ~ ) * ^* semitone (-^ fc5n). Let's denote c the
difference between a and 6, the meantone temperaments are described by a pseudo-cyclic
group of the form
H =< l , r , r 2 c , r 3 c 2 , r V , r V , r V , r 7 c 4 , r 8 c 4 , r V , r 1 V , r 1 1 c 6 >
The generator r is the semitone a and the regulator is c = b/a. The closure relation is
given by
a567 = 2
150
Graphical Representations
ih
D
l-lll-V
C-E-G +1
-1 viV *
Deltafirstthirds
A Yy (Syntonic comma)
XF
viA
B -1 X
Kimberge r II (C Major)
151
8. Tuning Groups
db(Z) = logbh(?)
where h(p/q) = max(p,q). For a k-chord of rational frequencies x = (1, fi, ..., fk-i) the
harmonic metric is defined by
db(x) = log6 max(A, A/i,.., Xfk-i)
where fi = pi/qi and A = gcdfa).
Remark that the map db is a metric on Q + \ { 0 } . When the frequencies are not rational,
the harmonic metric is extended using the property
h(xn) = [h(x))n VneN
Examples. In the Pythagorean tuning, the major chord x = (C\^E^G^C2) has the har-
monic distance equal to
34 3
d2(x) = log2max(l,26,-,2)
= log 2 max(2 6 ,3 4 ,2 5 .3,2 7 )
= log 2 (2 7 ) = 7.
In the 12-tone temperament, for the same chord x = (1, 2 1 / 3 , 2 7 / 1 2 , 2) we have
x12 = (1, 2 4 , 2 7 , 2 12 ) and
h(x12) = 2 12 = (h(x))12
So h(x) = 2 and
d2(x) = log2(2) = 1
Definition 154 The harmonic gap of a rational frequency p/q (p, q relatively prime and
P, Q>1) is the logarithm sum of its numerator and denominator
H{p/q) = logp + logg
The harmonic distance of two ratios r\ = p\jq\ and r2 = P2/Q2 is defined by
d(r1,r2) = \H(r2)-H(r1)\
The harmonic gap verifies the following properties.
152
Harmonic Metrics
Definition 155 The harmonic complexity of a frequency ratio p/q whose decomposition
in prime factors has the form p/q = 2a°p^1 ...p%n where pi,..., pn <we prime numbers
greater than 2 is the sum of the prime factors different of 2
C(p/q)=Pi + ...+pn
Definition 156 The indigestibility of a prime number p is
. , 2(p-l)2 . .
X\P) = ¥ P %s prime
V
The indigestibility of a product of two prime numbers is (by definition) the sum of the
indigestibility of its factors
x(pq) = x(p) + x(q)
The harmonicity of the ratio p/q is
sgn(x(p) - x(q))
hip/q) =
x(p) + x(q)
where sgn{x) = \x\ /x is the sign of x ^ 0.
Example. The indigestibility of the first numbers is x(2) = 1, x(3) = 8/3, x(5) = 32/5.
The harmonicity of the pure fifth 3/2 is h{3/2) = 3/11. The harmonicity of the syntonic
comma /i(81/80) = 52/5. The harmonicity of the Pythagorean comma is /i(3 12 /2 19 ) =
1/51.
Definition 157 The p-adic norm of the rational frequency ratio x = par/s where p is
a prime number, r and s are integers not divisible by p, is defined by
\x\D=p-a
with the convention that |0| = 0.
Example. For the syntonic comma, 81/80 = 3 4 /(2 4 • 5), the p-adic norms are
81 81 1 81
= 16,
80 2 80 3 "81' 80
- + ....+ \q\
\q\ "Pi
or
- = sup
\q
l-l l-l
\q\ p\
153
8. Tuning Groups
Theorem 160 Let p, q be two prime numbers with p < q. The commas of the group
G =< p,q> are the convergents of the irrational number a = log pj log q.
In the general case G =< piy ...,p n >, commas are given by the Chemillier-Duchamp
algorithm, see [Chemillier, 1999].
Proof We must show that all comma is a convergent. Let a = pxqy be a comma. Then x
and y have opposite signs and \x/y\ is a best approximation of a. Thus a is a convergent.
Conversely, let's define the sequence (an) by
q rn-2
r _ i = p, 7*0 = — , rn = r—
p°o {rn-irn
Since p < q, the number a$ exists and r 0 is the first convergent of a. There exits some
numbers an > 1 such that
C-i < rn-2 < C-l 1
because r n _ i < r n _2 and (log(r n _i), log(r n _2)) are linearly independent on Q. The
number rn is a comma. Let b = pxqy such that
|log p (6)|<|log p (r n )|
that is
\x + V logp(g) | < \xn + yn log p (g)|
and
(x„,J/„) = ( ( - i r - V n , ( - l ) n 9 n )
Since pn/qn is a best approximation of log p (g), we have \y\ < qn and \x\ > pn. Con-
sequently, x and y have opposite signs and then h(b) > h(rn). D
154
Consistency and Completeness
Definition 161 Let G be the group G =< p,q > and rn a comma of G. The group
G/ <rn> is isomorphic to Z and defines natural scales. For p = 2 and q = 3, the scales
are called Pythagorean scales.
- n = 4 / 3 , # i = {1,2}
-r2 = 9/8, H2 = {1,3/2,2}
-r3 = 2 8 /3 5 , H3 = {1,9/8,4/3,3/2,16/9,2}
- r\ = 3 1 2 /2 1 9 , H4 is the twelve Pythagorean temperament
- 7"5 = 2 6 5 /3 4 1 , H& is the Janko temperament of 41-degrees
- TQ = 3 5 3 /2 8 4 , HQ is the Mercator temperament of 53 degrees.
Example. The Partch 3-limit H3 = {1, 4/3, 3/2, 2} is 12-consistent. Since A12(l) = 0,
^12(4/3) = 5, A12(3/2) = 7, A12(2) = 12. We have for example
^ 1 2 ( 5 X 5 ) = 5 + 7 = ^12(2)
The Partch 5-limit H5 = {1, 6/5, 5/4, 4/3, 3/2, 8/5, 5/3, 2} is not 11-consistent. The
approximated indices in the 11-tet are respectively (0, 3, 4, 5, 6, 7, 8, 11) and
A 1 1 (|) + X 1 1 (J)-» + 4 ^ 1 ( J x f ) - A l ( | ) . .
155
8. Tuning Groups
there exists an integer r smallest as possible such that the ratios a\, a2, ..., ar € X\{1}
verify
An(ar) - An{ar-i) + ... + ( - l ) r _ 1 j 4 n ( a i ) = j (modn)
For all degrees j , the maximum value taken by r is called the diameter of X and is
denoted by 5n(X).
156
9
Contemporary Tunings
The rediscovery of microtones at the beginning of the 20 th century and some new in-
vestigations on just intonation have led to a set of new theories which have been applied
in many musical works. Whereas Julian Carrillo, Alois H&ba, Ivan Wyschnegradsky and
others have used the equal temperament based on the equal division of tone, Adriaan
Fokker has developed a theory of 31-tone equal temperament. Many new tunings have
been explored in the field open by Harry Partch, Lou Harrison, Ezra Sims, Ervin Wilson
and Ben Johnston.
describes the set of all products of the form 3a • 56 • n where a and b take all integer values
a = 0, 1, 2 and b = 0, 1, 2, 3. So, one gets the set of possibilities, namely: n, 3n, 5n,
9n, 15n, 25n, 45n, 75n, 125n, 225n, 375n, 1125n. If one gives the value of n, for example
n = 1/5, the sequence becomes 1/5, 3/5, 1, 9/5, 3, 5, 9, 15, 25, 45, 75, 225. Introducing
the power of 2 in order to put all values in the octave [1, 2[, we get a complete tuning
system: C (1), D (9/8, 204 cents), D+ (75/64, 275 cents), D|f (6/5, 316 cents), E (5/4,
386 cents), Pji (45/32, 590 cents), G (3/2, 702 cents), G+ (25/16, 773 cents), Gfl (8/5,
814 cents), A+ (225/128, 977 cents), A% (9/5, 1018 cents), B (15/8, 1088 cents). The
9. Contemporary Tunings
3 a • 56 • V • • • n
They could describe many well-known temperaments. The genus 3 6 -n is the Eratosthene
diatonic system described by Ptolemaus. The genus 3 1 1 • n is the Pythagorean tempera-
ment built on the fifths ascending (if n = 311) and descending (if n = 1/3 11 ). The genus
3 3 • 5 2 • n corresponds to the zarlinian temperament. In the 20 th century, this procedure
has been widely studied by Adriaan Fokker (1889-1972) in relation with lattices proper-
ties. From the Euler expression 3 P • 5 9 • 7 5 , Fokker draws in a multi-dimensional space, p
fifths in the x direction, q thirds in the y direction over the fifths and s sevenths in the z
direction. For example, the genus 3 4 • 5 in the Fokker representation is described by the
lattice
E - B - F I - Q - G j l
I I I I I
C — G — D — A — E
All notes separated by a horizontal line are fifths, and all notes separated by vertical
line are major thirds. As in a crystallographic description, we get a lattice representation
where atoms are replaced by notes. So many results of crystallography, and in particular
results based on symmetry, can be applied to Fokker representation.
Another great contribution of Adriaan Fokker concerns the 31-tone equal temperament
(31-tet), also called tricesimoprimal or trentunisonic system. This system had been stud-
ied for the first time by Christiaan Huygens (1629-1695) who published in 1691 Novus
cyclus harmonious. In this temperament, the octave is divided in 31 equal parts called
diesis. The notes are named by prefixation. Let re be a note, xi is the note x plus one
diesis, xis is the note x plus two diesis, xisi is the note plus three diesis and xisis is the
note plus four diesis. For example, ais is A augmented by two diesis. Similar prefixes are
used for the descending order, xeh is the note x minus one diesis, xes is x minus two
diesis, xeseh is x minus three diesis and xeses is x minus four diesis. The tone is divided
in five diesis, and semitone has three diesis. In the 31-tet, enharmonic notes are different
from the 12-tet. In the next figure, the genus 3 2 • 5 • 7 is represented.
158
Euler-Fokker Genera
The drawback of the Euler-Fokker genera is that some genera are equivalent. For
example, the genus 3 • 7 2 built on C has the same notes as the genus 7 5 built on G. Then,
it is natural to ask: how many different genera can be built ? We have the following
result.
Theorem 165 Let (p, q, r) be the description in diesis of three constituent intervals of
the Euler-Fokker genus 3 a • 5 6 • 7C.
(1) There are exactly 156 different genera with (p, q,r) = (18,10,25), i.e. with a fifth of
18 diesis, a third of 10 diesis and a seventh of 25 diesis.
(2) If we substitute the major third by a minor third of 8 diesis, there are only 136
different genera, with (p,#,r) = (18,8,25).
(3) There are 125 different genera with (p,q,r) = (20,10,30).
Adriaan Fokker composed some musical pieces with the Euler-Fokker genera under the
pseudonym of Arie de Klein, for the tricesimoprimal organ: Canzone (3 2 • 7), Preludium
chromaticum (3-5-7), Improvisando (5 2 • 7), Tenacita and Canzonetta (3 • 7 2 ), Inventio
(5 • 72) and Capriccio (7 3 ). He has also commanded works to other dutch composers
through the Huygens-Fokker Foundation (Stichting Nauwtuisterendheid).
Henk Badings (1907-1987) has composed many 31-tet pieces. The Suite van kleine
stukken was written in 1954 with the Euler genus 3 3 • 5 2 Reeks van kleine klankstukken
for organ used 3 3 • 5 2 in Preludio, 3 3 • 5 3 (Siciliano), 5 3 • 7 2 (Perpetuum mobile), 5 2 • 73,
3 3 • 7 2 , 3 2 • 5 • 7 (Passepied) and 3 • 5 2 • 7.
Hans Kox (1930) also used Euler genera. Drei Stukken for violin is written in the 3 2 • 7 3
genus. Vue des Anges on a poem of Rilke (1959) is built on the 3 3 • 72 genus. The serenade
(1968) for two violins in five parts is written in the genus 3 • 7 2 in different transpositions.
The canon of the first part used two transpositions 3 • 7 2 (di -f£h) and 3 • 7 2 (ges-a£h)
included in the genus 3 • 5 • 7 2 (di-feh).
Jan van Dijk (1918) used many different genera in Musica per organo trentunisono.
Each of the twelve pieces is written in different genera: Prelude and fugue for oboe and
organ uses 5 • 3 4 (g, dis) 7 • 3 3 • 5 (ai, gis) 7 • 3 4 (ai, 6), the second piece Fugue for violin,
alto and organ uses: 7 3 • 3 3 (ai, d£h), 7 3 • 3 2 (ai, ei), 7 4 • 3 (ai, g), 7 • 3 4 (di, e), the third
piece Canon for organ: 7 5 (ai, ais), the fourth piece Fugue for flute, oboe and organ:
7 • 3 2 • 5 2 (ai, feh), The fifth piece 5a, Toccata for organ solo: 7 2 • 3 2 • 5 (ai, b£h), and the
fifth piece 5b, Fugue for oboe, violin and organ: 7- 3- 5 3 (ai, dih), the sixth piece Prelude
and fugue for organ solo: 7 2 • 3 • 5 2 (ci, b£h), the seventh piece, Corrente alia fuga for
organ solo uses 7 3 • 3 • 5 (ai, dis), the eighth 8a, Capriccio for flute in G and organ uses
7 • 5 4 (ai, b£h), the eighth part 8b, Fugue for flute in G and organ 7 2 • 5 3 (ai, ei), the
ninth piece, Duettino e Ciaconna for two flutes and organ: 7 3 • 5 2 (ai, bis) 7 4 -5 (ai, c)
and the tenth Basso solo for pedalier uses 7 5 (ai, ais).
The british composer Alan Ridout (1934-1996) also used Euler genera in his String
Trio (1961). The first movement Chants uses the genus 3 3 • 7 2 , the second movement
Dances uses the genera 5 2 • 7 3 and 3 • 5 2 • 7, the third and final movement takes back the
elements of the first part.
Many other composers have been involved in the 31-tet history, such as Anton de Beer
(1924), Jon Catler (1957), Ivor Darreg (1917-1994), Harold Fortuin, Eugen Frischnecht,
Jaap Geraedts (1924), Joop van Goozen (1960), Alois Haba (1893-1973), Anthon van der
Horst (1899-1965), Martinus J. Lttrsen, Joel Mandelbaum (1932), Peter Schat (1935), Al-
phonse Stallaert (1920), Paul Christian van Westering (1911) and Ivan Wyschnegradsky
(1893-1979), who wrote a piece for organ with the technique of non-octaviant spaces.
159
9. Contemporary Tunings
Definition 166 The Farey sequence Fn of order n is the increasing sequence of irre-
ductible fractions p/q (i. e. gcd(p) q) = 1) between 0 and 1 whose denominators do not
exceed n. The Farey temperament TFn is the shifted Farey sequence. It is defined by
adding one to each ratio of the Farey sequence, namely TFn = Fn + 1.
Fi = {0,l}
by inserting mediants of each neighboring pairs of rational numbers. If p/q and p'/q' are
such pairs, the mediant is denned by
P + rf
q + q'
The first Farey sequences are
1 X
F -1° \
,r_r0 1 1 1 2 3 1
4
MU's^'sU'i'
Remark that the Farey temperaments TFn are not the set of all rational numbers p/q
such that p and q are less or equal n. For example, in the Farey temperament TF5 we
have the number 7/4 which does not verify this property. However, each number x of TFn
is the mediant of two consecutive numbers of TFn-\. The Partch n-limit temperament
Hn verifies this property, if p/q G Hn then q <p <n. The main properties of the Farey
sequences are
Proposition 167 (1) Ifpi/qi, Pv/qi are two consecutive ratios p\/q\ <Pi/q2 then
\piq2~qiP2\ = 1
(2) Ifpi/qi, P2/92, P3/93 are three consecutive ratios p\/q\ <p^/q2 <Pz/qz then
P2 _ Pi +P3
q2 q\ + qz
(3) The number of terms in the Farey sequence of order n (i.e. the cardinal of the set
Fn) is
n
|FB| = l + 5>(fc)
fe=l
where ip(k) is the Euler totient function.
160
Farey Temperaments
The fact that the sequence of the denominators of a Farey sequence is palindromic, is
not obvious. It is proved by the embedding of the Farey sequences into the Stern-Brocot
tree. The Stern-Brocot tree is a binary tree obtained by iteratively inserting the mediant
of two adjacent fractions and starting with points 0/1 and 1/0. The relations between the
Stern-Brocot Tree and Well-formed Scales are described in [Clampitt, 1997] and [Carey,
1989].
The ratios on the left side of the tree are less than 1 and the ratios on the right side are
greater than 1. For the ratio on the left side, the place of each ratio p/q can be encoded
by a word describing the way to get the number starting from 1/1. Denoting L the left
movement and R the right movement, the word LLRL or simply L2RL represents the
ratio 3/8.
1
0
4 5 5 4 3 3 2 1
More generally, movements are associated with continued fractions. The function given
by the exponents rj of the dj (= L or R) components
f(a\\o%. ..<") = [ 0 ; r i , r 2 , . . . , r n _ i , l + r n ]
describes the continued fraction of the point associated with the word
a a
l 2 ~'an
1 +
2
The sum of the exponents minus 1 (fc = ri +... + r n — 1) is the row number in which the
fraction is located. For example 3/8 is located in the fourth row (2 + 1 + 2 — 1 = 4) of
the Stern-Brocot tree.
The Farey sequences (and then the Farey temperaments) are related to Ford circles.
The Ford circle C(p,q) is the circle centered at (p/q)±l/(2p2)) and of radius l/(2q2).
If we put the Farey ratios Pi/qi on the real line, we could draw the circles C(pi,<fc). If
161
9. Contemporary Tunings
Pi/Qi) P2/92, P3/93 are three consecutive terms of the Farey sequence, then the circles
C(pi,gi) and C(p2>#2) are tangent at the point a\ given by
„ _ (V2 gi l__\
and the circles C(p2>#2) and C(ps,qs) are tangent at the point 0,2 given by
U2 92(92+9*)'92+9$)
The point a\ lies on the circumference of the circle built on the two points of the real
line (pi/#i,0) and (p2/</2»0). The point a2 belongs to the circle built on (p2/<72,0) and
(^3/^3,0). If we plot all the fractions on the z-axis and their corresponding Ford circles
generated from a Farey series of order n, they form a set of tangent circles [Ford, 1938].
Remark that Farey sequences have been used by Ivan Wyschnegradsky in his research
on rhythmical figures of ultrachromatism [Wyschnegradsky, ed. 1996]. Another important
point is that phase transitions occur in Farey temperaments.
Phase transitions are singularities of variables such as free energy. For a dynamical
system characterized by a Hamiltonian H, we define the Gibbs measure by the probability
of density
e-PH
/(*) = Z
where /? is linked to the temperature T of the system, namely /? = 1/kT. The partition
function is
Z = I e-^*) dx
The internal energy of the system is the expectation of the Hamiltonian, considered as a
random variable
B
"-•*-- -W
The free energy is defined by
F = -kTlogZ = U-TS
The entropy is given by the following equations
f)F f)F
s = = fc(log z + m = km2
~ ar W
162
Harry Partch Tonality Diamond
*-(SX-^(W
Usually, phase transitions are singularities of the calorific capacity. For the Farey tem-
peraments, we will see that the free energy has a discontinuity atft= 2. Recall that the
Farey sequences are the set Fm of all rational fractions of [0,1] whose denominators are
less or equal m. The set Fm-\ is obtained from the set Fm by inserting between two
consecutive fractions pi/qi and p2/q2 the number
Pi +P2
tfl+42
For each m, let Fm be the Farey set
=
I'm l^lj^r-j^mj
and for each Xi = Pi/qi and Xi+i = pi+i/c^+i define the number
v% = Pi+i + Qi
i=l
FW)= lim Z l l o g ^ )
m->oo p m
Theorem 170 The free energy of the Farey temperaments F(fi) exhibits a unique phase
transition atft= 2.
Proof See the paper of P. Kleban and A.E. Ozluk [Kleban, 1998]. •
163
9. Contemporary Tunings
1/3
1/5 5/3
1/1 5/5 3/3
5/1 3/5
3/1
X = A • d • A1 with A = (abc)(aba)
where a, b} c, and d denote the intervals of the temperament and A1 the palindome of
A, A1 = (aba)(cba). More generally, we have the following definition.
Definition 171 Let p be a prime number. A p-limit system is a temperament in which
each ratio is a rational number whose denominator and numerator factors are products
of prime numbers less or equal p.
The Partch 7-limit temperament is built in the same manner on the numbers <
3,5,7 > and leads to the system of 13 degrees: {1, 8/7, 7/6, 6/5, 5/4, 4/3, 7/5, 10/7,
3/2, 8/5, 5/3, 12/7, 7/4}. The system is described by the palindromic word
The Partch 9-limit is built on the numbers < 3,5,7,9 > and leads to the 19 degrees
temperament {1, 10/9, 9/8, 8/7, 7/6, 6/5, 5/4, 9/7, 4/3, 7/5, 10/7, 3/2, 14/9, 8/5, 5/3,
12/7, 7/4, 16/9, 9/5}. If we denote the constitutive intervals by a = 10/9, b = 81/80,
c = 64/63, d = 49/48, e = 36/35, / = 25/24, g = 28/27, h = 21/20, i = 50/49, the
9-limit is the palindromic word
X = A-n-Ai
with
A = abcdef ghijklm
164
Ben Johnston Lattices
The Partch 13-limit is the temperament of 41 degrees {1, 14/13, 13/12, 12/11, 11/10,
10/9, 9/8, 8/7, 7/6,13/11, 6/5,11/9,16/13, 5/4,14/11, 9/7,13/10,4/3,11/8,18/13, 7/5,
10/7, 16/11, 3/2, 20/13, 14/9, 11/7, 8/5, 18/11, 5/3, 22/13, 12/7, 7/4, 16/9, 9/5, 20/11,
11/16, 24/13, 13/7}. If we denote the constituent intervals a = 14/13, b = 169/168,
c = 144/143, d = 121/120, e = 100/99, / = 81/80, g = 64/63, h = 49/48, i = 78/77,
j = 66/65, k = 55/54, / = 65/64, m = 56/55, n = 99/98, o = 91/90, p = 40/39,
q = 33/32, r = 50/49, the 13-limit is the palindromic word
X = A • r • A1
with
A = (abcdefghijk)c(lmnopq)c(o)
The Partch monophonic system is the 43 tone temperament that has been used by
Harry Partch for tuning the chromelodeon an instrument built in 1941. The ratios are
{1, 81/80, 33/32, 21/20, 16/15, 12/11, 11/10, 10/9, 9/8, 8/7, 7/6, 32/27, 6/5, 11/9, 5/4,
14/11, 9/7, 21/16, 4/3, 27/20, 11/8, 7/5, 10/7, 16/11, 40/27, 3/2, 32/21, 14/9, 11/7,
8/5, 18/11, 5/3, 27/16, 12/7, 7/4, 16/9, 9/5, 20/11, 11/6, 15/8, 40/21, 64/33, 160/81}.
The temperament is symmetric around 7/5 and is described by the word
X = A-i>A*
with
A= (abc)(jdef)(ajgj)(abdc)(hgj)(abc)
25 1 312 K 34
G = < r = - , 1= ^ , 5 E E - >
The lattice (3 • 7) leads to (C 1, D\> 49/48, D 9/8, Eb 7/6, E 9/7, F 4/3, Pfl 72/49, G
3/2, Ab 14/9, A 12/7, Bb 7/4, B 27/14), which is the tuning group
49 1 _ 312 ^ _ 26
G = < r = 1 = 7 = >
4 8 ' 2^' 3 ^
165
9. Contemporary Tunings
The lattice ( 3 - 5 - 7 ) leads to the 22 degrees temperament (C 1, D\>' 28/27, D\>" 16/15,
D> 10/9, Db"' 8/7, E\>' 7/6, £b 6/5, E 5/4, £ b " 9/7, F' 21/16, F " 27/20, i^|' 45/32,
F&" 81/56, G 3/2, A\>' 14/9, 4b 8/5, 4 5/3, AV 12/7, £ b ' 7/4, 5b 9/5, 5 15/8, Bb"
37/14, C 2), which corresponds to the tuning group
2 35
G
28
1 = 5
_3^_ 7
f_
= < r = -
27' 322' 212' - 32 >
This lattice is used in One Man (1967) for trombone solo, Rose (1971) for choir, Mass
(1972) for choir, ten trombones, double-basse and percussions, and in the Fourth String
Quartet (1973). Works such as The Song of Innocence (1975) for soprano and instruments
and Two Sonnets of Shakespeare (1979) use the bidimensional lattice (3 • 11). Diversion
(1979) for 11 instruments and the Sixth String Quartet is built on the lattice (3• 5-7-11).
Duo for two violins (1978), the Fifth String Quartet (1979) use the five prime numbers
(3 • 5 • 7 • 11 • 13). The Microtonal Suite for piano (1978) and Twelve Partials (1980) use
the lattice (3 • 5 • 7 • 11 • 13 • 17 • 19) and the Ninth String Quartet, written in 1987, uses
a 10-dimensionnal lattice (3 • 5 • 7 • 11 • 13 • 17 • 19 • 23 • 29 • 31).
V n [#i>£2,-»,0n]
2 4 [1,3,5,7]
166
Non-octaviant Tunings
Eliminating the doubles, we get a 19 degrees temperament. Many CPS have a great
number of degrees. But the CPS
3 9 [1,2,..., 9]
4 8 [1,2,..., 8]
(3/2) 6 / 7 602
Notes
C
Freq.
1
Cents |
0 \nG 3/2 702
Ctf (3/2)^7 100 G« (3/2) 8 / 7 802
D (3/2) 2 / 7 201 A (3/2) 9 / 7 903
Dfl (3/2) 3 / 7 301 AJJ (3/2) 10 / 7 1003
E (3/2)4/7 401 B (3/2)n/7 1103
F (3/2)5/7 501 | C (3/2) 12 / 7 1203
Wendy Carlos has proposed equal temperaments which are good approximations to
the just fifth (3/2) and the just major third (5/4). The fifth is divided in 9,11 or 20 equal
parts leading respectively to the alpha, beta and gamma temperaments. These numbers
come from the expansion in continued fraction of the ratio
log(3/2)
= [1; 1,4,2,6,1,10,135,...]
log(5/4)
whose numerators and denominators of convergents (1, 2, 9/5, 20/11, 129/71, ..) lead to
the sequence (1, 2, 5, 9, 11, 20, 71, ...).
The alpha temperament is the system of 18 degrees built on the repetition of the interval
of ratio (3/2) a / 9 , about 78 cents. In this system, nine intervals (702 cents) approximate
167
9. Contemporary Tunings
the just fifth (3/2, 702 cents), five intervals (390 cents) approximate the just major third
(5/4, 386 cents) and four intervals (312 cents) approximate the just minor third (6/5,
316 cents). The ratio x verifying this approximation is given by the formula
The beta temperament is the system of 22 degrees built on the repetition of the interval
of ratio (3/2) 1 / 1 1 , about 64 cents. In this system, 11 intervals (702 cents) approximate
the just fifth (3/2, 702 cents), six intervals (383 cents) approximate the just major third
(5/4, 386 cents)) and five intervals (319 cents) approximate the just minor third (6/5,
316 cents). The ratio x verifying this approximation is given by the formula
The gamma temperament is the system of 35 degrees built on the repetition of the
interval of ratio (3/2) 1 / 2 0 , about 35 cents. In this system, 20 intervals (702 cents) ap-
proximate the just fifth (3/2, 702 cents), 11 intervals (386 cents) approximate the just
major third (5/4, 386 cents)) and nine intervals (316 cents) approximate the just minor
third (6/5, 316 cents). The ratio x verifying this approximation is given by the formula
Refinements of these tunings are obtained by dividing each interval into two new
intervals. This leads to alpha prime, beta prime and gamma prime presented in the
following table.
168
Cyclic Tunings
X = (adabcb)a(bcbada)
with a = 27/25 (133 cents), b = 49/45 (147 cents), c = 375/343 (154 cents) and
d = 625/567 (169 cents).
and normalize the ratios by a power of 2 to put them in the same octave [1, 2[. This leads
to new temperaments which have many common properties. One of them is the existence
of prime scales generated by only two intervals. The most famous examples of the cyclic
tunings are the Pythagorean tunings based on the pure fifth (w = 3/2) and the golden
scale based on the golden fifth (a fifth which depends on the golden ratio, solution of the
equation <>/ =(/)-{-1). But here, we will consider new tunings given by irrational number
w such as 7T = 3.14159..., e = 2.71828..., the Euler constant 7 = 0.5772156649..., the
Catalan number C = 0.9159655942, and the Feigenbaum number F = 4.6692106...
^ 3 = C1C2C1
where c\ = 4/3 = 498 cents and c<i = co/ci = 9/8 = 204 cents. The next system
(k = 2) is made from the fifths {B\>, F, C, G, D} and composed of two intervals c<i and
c
3 = C1/C2 = 32/27 = 294 cents. The scale well ordered is made of 5 notes {C, D, F, G,
B\>} and represented by the word
P 5 = C2C3C2C3C2 = (C2C3)2C2
169
9. Contemporary Tunings
For k = 3, we get the seven degrees Pythagorean scale {C, D, E\>, F, G, A, B\>} repres-
ented by the word
Pi = c2c4(c2fc4c2
in which the two constituent intervals are related to the previous one by the relation
c4 = C3/C2 = 256/243. This system has only two constituent intervals c2 and c 4 . If we
continue the algorithm with k = 4, we construct the new scale {C 1, D 9/8, E\> 32/27,
E 81/84, F 4/3, G 3/2, A\> 128/81, A 27/16, B\> 16/9} which is described by the word
P9 = a(bcb)a(bcb)a
where the intervals a, b and c are given by a = 9/8, b = 256/243 and c = 2187/2048. We
see that this scale has three constituent intervals instead of two. It is not a prime scale.
The first scale which contains all the previous scales and has only two intervals is the
scale P12, involving the two intervals c4 = c$/c2 = 256/243 and C5 = C2/C4 = 2187/2048.
The following table shows the next steps. The first column gives the number of degrees
in the scale, the second and third columns are the numbers of constituent intervals. The
fourth column gives the relation of the constituent interval and the last column is its
value in cents.
Golden Tunings. Golden cyclic scales are based on the golden fifth of 696 cents
_ 15-N/5
w= 2^—
The constituents of the prime scales are presented in the following table whose columns
have the same meaning as the previous one. The first constituent interval is c\ = 2/w =
504 cents.
170
Cyclic Tunings
Exponential Tunings. Exponential cyclic scales are based on the powers of the num-
ber e, the basis of the Neperian logarithms e = 2.71828... The constituents of the prime
scales are given in the following table (same meaning as previous). The first constituent
interval is c\ = e/2 = 531 cents.
P i Tunings. John Harrison (1693-1776) is a British watchmaker who has studied the
7r-systems based on the ratio 21/271" of 191 cents. Here we studied the tunings built with
the powers of the number w. The constituents of the prime scales are presented in the
following table. The first constituent interval is c\ = 4/n = 418 cents.
7 = 0.5772156649...
the Eulerian cyclic tunings are presented in the following table. The first constituent
intervals are c\ = 2j = 249 cents and c<i = I/27 2 = 703 cents.
171
9. Contemporary Tunings
C a t a l a n Tunings. Catalan tunings are made from the powers of the Catalan number
C = 0.9159655942... The constituents of the prime scales are presented in the following
table. The first constituent interval is c\ = 1/C = 152 cents.
Feigenbaum Tunings. Feigenbaum tunings which occurred in fractal music, are built
on the Feigenbaum constant F = 4.6692106.... The constituents of the prime scales are
presented in the following table. The first constituent interval is c\ = F/A = 268 cents.
3 2ci 1C2 c2 =25/^2 664
4 3ci 1C3 c3 =c2/c1 = 27/F3 397
5 4ci 1C4 C4 =C3/Cl =29/F4 129
9 5C4 4C5 c5 =ci/c4 = F 5 / 2 n 139
13 9c 4 4C6 ce =c5/c4 =F9/220 10
22 13ce 9c 7 c7 =c4/c6 =229/F13 118
31 22ce 9c 8 c8 =c7/c6 = 2 4 9 / ^ 2 2 108
40 31ce 9c 9 C9 = C 8 / C 6 =269/F31 98
49 40ce 9cio cio = c 9 / c 6 = 2 8 9 / F 4 0 87
58 49ce 9cn en =c10/c6 =2109/F49 77
172
10
Groups and Symmetries
The symmetries that appear in music are related to group theory. Symmetries of periodic
patterns and big lattices such as tunings and temperaments are well described by multi-
dimensional crystallographic groups. The symmetry of a geometric object (i.e. a subset
of a metric space) is the set of all transformations which brings the object into the same
figure as the original or its mirror image. The set of all transformations that leaves an
object unchanged is called the symmetry group (or point group) for that object. These
transformations are generated by elementary symmetry operations which are reflection,
rotation and translation. They can be applied to any musical object. This is why each
musical object has a symmetry group, possibly reduced to the identity element.
Definition 172 Let X be a set, F(X) be the free group and R be a set of elements of
F(X). Each element of R can be written as a finite product X\ ...XJI where each Xj for
j = 1, ...,n is an element of X or an inverse of an element of X. Let N be the smallest
10. Groups and Symmetries
normal subgroup of F(X) containing R. Then the factor G = F(X)/N is called the
presentation of the group G =< X \ R = 1 > determined by the generators X and the
relations R.
Coo=<t>
is generated by one element without relation. It represents the repetition of the same
pattern
p p p p p p p p p
Cn=<r\rn = l>
This group is generated by the rotation r about a fixed axis of angle 2kir/n where
fc = 0 , l , 2 , . . . ( n - l ) .
is the group of (proper) rotation (det(r) = +1) which leaves a regular dihedral n-gon
invariant.
The derived group G' of a group G is the group generated by the set of commutators
[x, y] = x~1y~1xy in G. If G = < X \ R > and H = < Y \ S > are two groups generated
by respectively the sets X and Y", then the direct product has the presentation
GxH=<X,Y | R,S,[X,Y]>
Definition 173 A frieze group is a discrete subgroup of the euclidean plane whose trans-
lation subgroup is infinite cyclic.
174
Frieze Groups
Frieze groups are symmetry groups of infinite strip patterns. More precisely, frieze
groups are infinite subgroups of isometric transformations of the euclidian plane M2,
which map a strip (i.e. Rx[—1, +1]) and are discrete. All translations are parallel.
T h e o r e m 174 In the euclidean plane, there are exactly seven frieze groups.
We now describe these groups. The crystallographers notation uses two dimensional
symbols. The first symbol represents the element of symmetry perpendicular to the direc-
tion of the translation. The second denotes the element of symmetry parallel or perpen-
dicular to the direction of the translation. A number n means a n-fold axis. Sometimes
the notation starts with p for periodic, but generally is omitted. Symbol m (mirror) de-
notes a reflection and g (glide) represents a glide reflection. We denote xy the conjugation
y~xxy.
1. The group F\\ (symbol p l l , 11 or pi) is generated by one translation denoted by t
Fn =< t >
It represents the repetition of the same pattern
P P P P P P
For example, the same rhythm
n n. n. n n n
2. The group F\m (symbol plm or lm, or pm) is generated by a translation t and a
reflection
Flm=<t,r \r2 = l,tr = t>
It represents the repetition of the same pattern with reflection through the hori-
zontal axis
P P P P P P
b b b b b b
For example,
nnnnnnnn.
wwu wwv wu
3. The group Fmi (symbol ml or p/m) is generated by a translation t and a mirror
reflection perpendicular to the horizontal axis
Flm=<t,r \r2 = l,tr = t~1 >
It represents the repetition of the same pattern vertically mirroring at each step
p q p q p q p q p q p
For example,
nnnn.Dn.nn.
175
10. Groups and Symmetries
4. The group F\g (symbol pg) is generated by a translation t of two units and a glide
reflection
Flg=<t,r \r2 = t,tr = t>
It represents the repetition of the same pattern with reflection through the hori-
zontal axis followed by a translation of one step
P P P P P
b b b b b
For example,
nj ni ni n\
r w r cj" r u r u
5. The group F\i (symbol 12) is generated by a translation t and a rotation about a
2-fold axis
F12=<t,s | s2 = t,t8 =t~1 >
It represents the repetition of the same pattern with a rotation about 2TT.
P P P P P P
d d d d d d
For example,
nn.nn.nnn.nn
uuuuuuuuu
6. The group Fmm (symbol pmm) is generated by a translation t, a reflection and a
glide reflection
p q p q p q p q p q
b d b d b d b d b d
For example,
n nn nn nn n
u uu uu wu w
176
Polyhedral Groups
7. The group Frag (symbol pmg) is generated by a translation t of two units, a rotation
about a 2-fold axis and a glide reflection
It represents the repetition of the same pattern with a rotation about 2ir.
p q p q p q p q p q
d b d b d b d b d b
For example,
nnn.nn.nnn
u wcr wu uu w
10.3 Polyhedral Groups
Definition 175 A polytope in W1 is an intersection of finitely many closed half-spaces
that is bounded and has non-empty interior. Polytopes are named polygons {when n = 2)
and polyhedra (when n = 3).
Definition 176 A flag is the n-tuple (F^F\^.^Fn-i) where Fk is a closed face of di-
mension k such that Fk-i C Fk for k = 0,1, ...,n — 1.
Proposition 177 A polytope with no holes (of genus 0) f faces, v vertices and e edges
verifies the Euler's formula
v-e+f=2
Definition 178 A polytope P is regular if the symmetry group of P acts transitively on
the set of its flags.
It can be shown that P is regular if all its faces are isometric regular polytopes in
dimension n — 1 and all its dihedral angles are equal. In the space M3, convex regular
polyhedra are called platonic solids.
Definition 179 The polyhedral group of the polytope P o / R n is the symmetry group
of P.
Definition 180 Let P be a polytope ofRn. The sequence {pi,p2j--->Pn-i} where p^ is
the number of k-dimensional (or (k — l)-dimensional) faces of the polytope F^+i con-
taining in the face Fk-2 (with F-i = 0, and Fn = P) is called the Schlafli symbol.
It can be shown that the presentation of the polyhedral group as the following form.
Let P be a polytope of W1. li Pk denotes the fc-th Schlafli symbol, the presentation of
the polyhedral group is given by
177
10. Groups and Symmetries
Polyhedral groups are Coexeter groups. In the space M3, the polyhedra are designed by
the Schl&fli symbol {p, g}, in which q > 3 is the number of regular p-gons (p > 3) that
meet at any one vertex of the polyhedron. The polyhedral group is presented by
It has two generators: a rotation r about axes passing through a q-iold vertex and a
rotation s about axes passing through the mid-point of a p-fold face from the center of
the polyhedron.
Theorem 181 In the euclidean space R3, there are exactly five convex regular polyhedra,
named platonic solids. / / G is the polyhedral group of the polyhedron P given by the
Schlafli symbol {p, q} with \G\ /p faces, \G\ jq vertices and \G\ /2 edges, then we have the
Poincare- Wyle formula
1 1 ^ 2
p q \G\
The characteristics of the platonic solids are as follows
G=(ru...rn | (rirj)m* = l)
where mij is a matrix called the Coxeter matrix of integers verifying ma = 1 (so that
r\ = 1) and mij = rriji (for j ^ i) is an integer > 2 or equal to oo (in this case, there
is no relation between r^ and rj). The integer n is the rank of the group and mij are its
exponents.
The polyhedral groups are Coxeter groups. The symmetric group Sn is a Coxeter group
with ri = (i,i + 1) for i = 1,2,..., (n — 1). Every group generated by two elements is a
Coxeter group.
178
Crystallographic Groups
Definition 183 A Dynkin diagram is a graph with n vertices, one for each generator
and edges defined by the following rules
(1) a thick edge (or an edge labelled with oo) if mij = 0,
(2) no edge between i and j if m^ = 2
(3) an edge between i and j if m^ = 3
(4) a (mtj — 2)-fold edge between i and j ifrnij > 4.
The classification of the finite Coxeter groups is presented in the following table
Groups Exponents | | Groups Exponents
An,n > 1 1 , 2 , . ..,71 #8 1,7,11,13,17,19,23,29
Bn,n>2 l,3,..,2n-l F4 1,5,7,11
Dn,n>4 l,3,..,2n- l , n - l Hz 1,5,9
-^6 1,4,5,7,8,11 HA 1,11,19,28
E7 1,5,7,9,11,13,17 | 1 GJ\m>5 1,771 — 1
179
10. Groups and Symmetries
T h e o r e m 190 Let n be a positive integer n € N*. There are finitely many isomorphism
classes of finite subgroups in GL(n,Z).
Definition 191 A T-subgroup of a group G is a normal abelian subgroup TofG without
torsion and such that G/T is finite and T is equal to its own commutant
T={geG,yt€T,gt = tg}
Definition 192 A group T is a crystallographic group ifT has a free abelian T-subgroup
of finite type. The rank ofT is called the dimension ofT.
The group T is the set of all parallel translations in the crystallographic group T. It
is a normal subgroup of finite index isomorphic to Z n . T is the lattice. The symmetry
group G = T/T preserves the lattice T. Relative to a basis of T, the elements of G
are represented by matrices with integers entries. The description of the n-dimensional
crystallographic groups is then reduced to the description of all finite groups of squares
matrices with integer entries of M{n, Z). The problem is then to compute the cohomology
group H\G,RnIZn).
P r o p o s i t i o n 193 Two cohomology classes define equivalent crystallographic groups if
and only if they are transformed into one another by the normalizer of G in GL(n, Z).
T h e o r e m 194 Let G be a symmetry group of a crystallographic group T. The natural
homomorphism
H\G,Rn/Zn) - • iJ 2 (G,Z n )
is an isomorphism.
Definition 195 A Bravais group is a subgroup of GL(n, Z) that are lattice symmetry
groups.
The Bravais groups are the stabilizer subgroups for the natural action of GL(n, Z) on
the set of positive definite quadratic forms in n variables. The number of Bravais groups
are 1 (n = 1), 5 (n = 2), 14 (n = 3), 64 (n = 4), etc.
Planar Groups
Definition 196 A discrete subgroup G of the Euclidean plane with its translation group
isomorphic to Z 2 is called a plane crystallographic group or a planar group.
Definition 197 A tiling of the plane is a covering ofR2 with non-overlapping polygons.
Definition 198 A crystal system is a group of transformations that leave a lattice in-
variant. It is characterized by the repetition of a unit cell.
T h e o r e m 199 In the euclidean plane, there are 4 crystal systems and 5 Bravais groups.
The unit cell is defined by two axes of vectors (a, b) and an angle a between the axis
fa).
180
Crystallographic Groups
By adding to the primitive unit cell of the four crystal systems an atom centered, we see
that only rectangular lattices can be centered. Thus there are five different symmetry
types of plane lattices, the five Bravais groups. This new class is called rhombic or centered
lattice. Another way to show that there are only five Bravais groups is to show that in the
euclidean plane, repeating patterns could not have 5-fold symmetry. Let 0 be a point of
the lattice. This point is an axis of n-fold symmetry. Any non-trivial rotation is a power
of the rotation r = r(0,27r/n) where O is the center of r. Let A, A' be two points in
the neighborhood of O. As the lattice is isomorphic to Z 2 , we could choose OA = 1,
OA! = — 1. Let a be the angle between A and B = r(A). Denote B' the point r~1(A).
B and B' belong to the same row, so BB' = 2cos(a) is an integer and can only take
on values -2, - 1 , 0, 1, 2. Thus cos{a) can only be equal to - 1 , -1/2, 0, 1/2, 1, that is
a = 27r/n with n = 1,2,3,4,6. So there are only five types of symmetry. Materials with
5-fold symmetry (which are not repeating patterns) are called quasicrystals.
Theorem 200 In the Euclidean plane, there are exactly 17 planar crystallographic groups.
Five of them are orientation preserving subgroups and tillings of the plane. The twelve
remaining subgroups are orientation reversing.
We now describe the 17 planar crystallographic groups. The first five groups are the
symmetry groups of tesselations of the plane. The others are the twelve orientation
reversing groups of the plane. We denote the conjugation ax = x~lax.
1. The first group (symbol pi) is generated by the translations a and b (of vectors a
and b) in the oblique crystal system
Pi = (a, b | ab = 6a)
ru r n
r7Ji
inWi J"i J\ ' " /
R
2. The second group (symbol p2) has a 2-fold symmetry in the oblique crystal system
181
10. Groups and Symmetries
3. The third group (symbol p3) is a trigonal lattice with a 3-fold symmetry
p 3 = ( a , 6, s | ab = 6a, ss = 1, a s = a _ 1 6, 6s = a" 1 )
4. The fourth group (symbol p4) is a square lattice with a 4-fold symmetry
p 4 = ( a , 6, s | a6 = 6a, s 4 = 1, as = 6, 6s = a" 1 )
rT
i-L . "U
nn
•u
a 6 D
h A
0
f'
3
w 0
r0
a 0
J) r P
h1- °i—Di
, , — • «
*
^
4
f
r*
K 3
^0
h
MV*
p J!i p
r # of
ri ^L 1 Lh/ h V fp >J rp
5. The fifth group (symbol p6) is an hexagonal lattice with a 6-fold symmetry
P 6 = (a, b, s | ab = 6a, s 6 = 1, as = b, bs = a _ 1 6)
W
•0..
k\M N\/&\ /A "\/A ; /K\ /I \ n
••<>• 6 - ^^ • A H-.L5&
182
Crystallographic Groups
6. The first group (symbol pm) of the twelve orientation reversing groups is a rectan-
gular lattice with a mirror symmetry
n7 J> w J71
rp rp rp rp
7. The group G 2 (symbol cm) is a centered lattice with parallel mirror and glide planes
G\ = (ayb,r \ ab = 6a, r2 = 1, ar = 6, br = a)
LiJiLiJjLil
JiLUILil
__.
JlLlJiL!
JJ> J>
rp rp rp
JJ
rt rp rp
—J —J JJ> J> J>
~""> "~" "-il rp rP rp
G\ = (a,b,r \ ab = ba, r2 = a, a r = a, 6 r = 6 _ 1 )
u7 JJ>
rp rp
j> JJ>
r19 — (/ a
A r > 52 I a& = &r a> fi2 =
1) a* = a_1
> &5 = fr-1>
° ~ r = 1, a = a, 6' = b~\ (sr)2 = 1
183
10. Groups and Symmetries
rp j ?r j rp jpr irp 1
'jjr>jT"'jJilJT"jJ!'
'"""T''pr'l'"rp"]lr''['rp''r
iS'EMffliSJ
J.. D r i rp i pr ! rp i
10. The group G\ (symbol pmg) is a rectangular lattice with mirror, glide reflection
and a 2-fold symmetry
n2 _ I a, 6, r, s | a6 = 6a, s2 = 1, a3 = a \ b3 = 6 *,
I LrmJ Lirxl iJ I Al I AM
I r^"] j —-y j~| ipr ! I
7JT lW UJH
I
L j. H .—j
i i
j.
\ I
-.—I
rp I i rp ! i rp I
L L ..........J..... . . . . . . ( . . . — . . . . : . . . . . . — i - J
fiji i'xii r
' ! PM i Pr i I
11. The group G | (symbol cmg) is a centered lattice with alternating mirrors and glide
planes in both directions and a 2-fold symmetry
i i —i Al !JJ> i
r — ! r \"\
. ".j..,. \ j ..,.
i i i I
12. The group G\ (symbol pgg) is a rectangular lattice with perpendicular glide planes
and a 2-fold symmetry
_ I a,6,r,s | ab = 6a, s 2 = 1, a3 = a *, 6s = 6 *,
Gi =
r 2 = a, ar = a, 6 r = 6" 1 , (sr) 2 = 6
184
Crystallographic Groups
1 1 4 1 i -1
00
Ill ^ j [L \ fp \
r-.j.
;"ZJ'1
l.cr„...j
p^j-j |
....{...CTJ K1 i*lj K... ^1 VI
=i.j |.,.m] L"n..l r_pj - . j :P...•^ 1 rp I
J
ir-.Lt^...i..nr...i |....r-".j P f l J > l Pr ^ i p r l
LciL^J IL •*J 1 rp 1
1 i J>l 11] !|
13. The group G\ (symbol p3ml) is a triangular lattice with an horizontal mirror plane
and a 3-fold symmetry
1
/•>A ! / ^ « \ , {#?
9 g
, ! , .£A? , '
^ K-f i rWr
14. The group G 2 (symbol p31m) is a triangular lattice with bisector mirror planes
and a 3-fold symmetry
r\/£Kr/fl \/Ami/il
V^wA^Aw/^ ^ \ j / \ A ; \ I /\y
( ^-.=^f ^~~ L ^-M- m*
L\///\\7&\//A\/4
)f'-A..J ^•^•••1 j^....V.j -V--;
A^A^I A 4 A
'-=& *-=&
15. The group G\ (symbol p4m) is a square lattice with horizontal mirror planes and
a 4-fold symmetry
_ I a,b,r,s | ab = ba, s4 = 1, as = b, b3 = a 1
,
G\ =
r2 = 1, ar = a, V = b'1, (sr)2 = 1
185
10. Groups and Symmetries
^4?.r^^f:4,iP.i?.f\i
» v #1 * V * I * \ * i * •' * i P \ * i * s * I
16. The group G\ (symbol p4g) is a square lattice with glide planes and a 4-fold
symmetry
_ / a, 6, r, 5 I aft = 6a, s 4 = 1, as = b, bs = a ,
ci = r 2 = a, a r = a, 6 r = 6 _ 1 , (sr) 2 = 1
17. The group Gj (symbol p6m) is an hexagonal lattice with a 6-fold symmetry
1 _ / o, 6, r, a I 06 = 6a, s 6 = 1, a* = 6, 6s = i
Gl r 2 = 1, ar = b, br = a, (sr)2 = 1
K- K'
^
^ ^ j ^ ^
•nr-' •&•• '-' A....' hr.:...' r> ?.....*
Space Groups
186
Crystallographic Groups
The axis order {n, m, k} is characteristic of the group G: Cn = {n, 0,0}, Dn = {n, 2,2},
T = {3,3,2}, O = {4,3,2} and Y = {5,3,2}. Combining with the inversion, we get nine
types of symmetry groups of finite order denoted by Klim notation Cni, Cnp = {2n, 0,0},
Cnv = {n,2,2}, Dni, Dnp = {2^,2,2}, Tu Tp = {4,3,2}, Ou Yi and three groups of
infinite order Cooi, Ax>v> A»i- The fullerene discovered in 1985 by Kroto and Smalley is a
molecule of 60 carbon atoms which belongs to the icosahedral inversion group Yi = YxCi.
The seven crystal systems are characterized by the coordinates of the unit cell of axes
{a, b, c) and of interaxial angles {a, /?, 7).
1. Triclinic. This crystal has no rational symmetry other than 1 or 1. Its symmetry
group is the group C%.
2. Monoclinic. The symmetry elements are the 2-fold rotation (2) and/or the re-
flection (2 or m). In this case, the interaxial angles verify a = j3 = n/2 and
a = 7 = 7r/2. The symmetry group is C<n.
5. Tetragonal. Tetragonal systems have one 4-fold or one 4 symmetry element. The
axes \a\ = \b\ and the interaxial angles verify a = fi = 7 = 7r/2. The symmetry
group is D4i.
7. Cubic. Cubic (or isometric) systems have four 3-fold or 3 axes. In this case, \a\ = \b\,
and a = ft = 7 = 7r/2. The symmetry group is 0{.
By adding to the primitive cell of the seven crystal systems (P) an atom body centered
(I), one face centered (A, B or C) or to each faces, all face centered (F), special centered
(R) we define the Bravais lattices. There are exactly 14 Bravais lattices. Associated to
each Bravais lattice, there are 32 crystal classes or symmetry groups presented in the
187
10. Groups and Symmetries
following table.
where the number p is chosen such that the ratio r belongs to the octave [1,2[. The
temperament thus obtained has the symmetry group of the Bravais lattice. For instance,
we consider the orthorhombic lattices. The orthorhombic P cell is a rectangular box with
atoms at each vertex. Suppose that the unit vectors have lengths a = 1, b = 2 and c = 3,
and are associated with the three prime numbers p\ = 3, P2 = 5, ps = 7. Each point
of the eight points of the unit cell is associated to a ratio 3X • 5y • 7*, as in the Euler or
Johnson lattices. The computation of all ratios leads to the following table
In the orthorombic C, we have to add two ratios placed at the center of the horizontal
faces. In the (a,b,c) basis, these points have coordinates (1/2, 3/2, 0) and (1/2, 3/2, 2).
188
Sporadic Groups
They give two ratios 3 1 / 2 • 5 3 / 2 /16 of 330 cents and 3 1 / 2 • 5 3 / 2 • 7 2 /512 of 1068 cents. In
the orthorhombic I lattice, we have to add a point in the center of the box of coordinates
(1/2, 3/2, 1) of 99 cents. In the all face centered orthorhombic F, the points (1/2, 3/2,
0) of 330 cents, (1/2, 3/2, 2) of 1068 cents, (1/2, 0, 1) of 720 cents, (1/2, 3, 1) of 679
cents, (0, 3/2, 1) of 348 cents and (1, 3/2, 1) of 1050 cents are added.
and
{5, 6, 4, 7, 3, 8, 2, 9, 1, 10, 0, 11}
These two rows are described by the knot -D358 which is also used by Anton Webern in
his Symphony} opus 21.
189
10. Groups and Symmetries
planar map, there is one associated rational function. The action of the Galois group on
maps has various invariants. One of them plays an important role: it is the cartographic
group. This group is directly readable on maps (that is graphs drawn on two-dimensional
manifolds). A map is a very concise presentation of permutations groups. For more de-
tails, see [Zvonkin, 2004]. Some musical works are based on the user of permutations:
Olivier Messiaen's Chronochromie, He de feu #, etc. Iannis Xenakis' Metastasis, etc.
To understand the power of permutations and to give a classification of the cartographic
groups, we review the elementary combinatoric theory of maps. This leads to the map of
the Mathieu groups used by Olivier Messiaen.
Definition 202 Let Sn be the symmetric group that acts on n objects. A constellation
of length k and degree n or a k-constellation is a sequence [<JI,<J2, •••0'fc] where Gj G Sn
verifying
(1) the group G = < cri,<72, ...,<7fc > acts transitively on the set of n points
(2) the product G\G<i—<Jk is the identity
GiG2~-crk = id
The permutation <72 associated with the four edges is made of the cycles of each labeled
edge.
a 2 = (l,2)(3,4)(5,6)(7,8)
The permutation G\ is given by the four vertices. It is made of the cycles composed by
the number around each vertex
a 1 = (2,5,3)(l,7,6)(4)(8)
190
Constellations and Cacti
The permutation a^ is associated with faces. Faces are the resulting areas after a cutout
along the graph.
a 3 = a r 1 ^ 1 = (l,3,4,5,7,8)(2,6)
Definition 206 Let V be the number of vertices, E the number of edges and F be the
number of faces. The Euler formula of a map verifies
V-E +F = 2-2g
The number g is called the genus of the map. A map of genus g = 0 is called plane.
The Mathieu group M\2 is associated with the icosaedron. Let ri be the permutation
of the neighbor of i in the rotation about the vertex i. The Mathieu group M\i is the
group generated by urj1 for z,j = 1,2,..., 12. Many maps are associated with M12. For
example, the map of the following figure has 4 vertices, 6 edges and 4 faces.
a3 = o^o? = (1,3)(2,5,12,9,10,8,6,4)(7)(11)
generate the group M\i- The passport of this map is [732,2 6 ,821 2 ]. Another representa-
tion is given by the following figure
191
10. Groups and Symmetries
a2 = (1,2)(3,4)(5,6)(7,8)(9,10)(11,12)
a3 = a^a^1 = (1)(3)(2,7,9,11,12,10,6,4)(5,8)
generate the same group M12. The passport of this map is [6321,26,8212]. One can show
that the group M\i has 50 maps and 1437 hypermaps. The first 50 hypermaps are given
in the table.
Passport Maps Hypermaps
3 e
\ 3 ,2 ,(ll,l) 4 12
4 ^ , 2 ° , (11,1) 4 12
5 ^ , 2 ° , 6321 12 36
5*1^ 28,821 s 6 18
6321,2°, 6321 8 24
6321,2 e , 821* 12 36
82P,2 B ,821 : J 4 12
Total 50 150
Some passports have no maps. They are associated with 1287 hypermaps. For example,
the passport [3 4 ,2 4 1 4 , (10,2)] has no map and 12 hypermaps.
Definition 208 A cactus is a plane constellation [<7i,<72, ...0"/c,c] where c is the cyclic
permutation c= (1,2,.., n). Thus
CTi<72...<Tfc = C
Definition 209 Two cacti are isomorphic if they are isomorphic as constellations.
Cacti are represented by polygons. A rooted cactus is a cactus in which one polygon is
considered as the root of the cactus and marked 1.
Example. Major triads having one common pitch-class are represented by a cactus. Three
permutations are associated with these chords and verify
<7l<72<73 = C
192
Constellations and Cacti
The permutation 0-3 is obtained from black vertices around the white vertices labelled 1.
a3 = (1,2,3X4,5,7X6)
The second permutation represents the cycles made from the white vertices labelled 2.
a 2 = (l,4)(2)(3)(5)(6)(7)
<n = (l)(2)(3)(4)(5)(6,7)
<7i<72<73 = (1,2,3,4,5,6,7)
Harmonic progressions of triads with one common pitch-class can be classified by their
passports. In our example, the passport of the cactus is
One can show that the number of cactus having this passport is 49 (see [Zvonkin, 2004]).
193
11
Tessellations and Canons
Tilings have many applications in music theory. They have been used for building tem-
peraments on a lattice and in canons theory, and could also be very useful for covering
the plane with twelve tone rows, chords and scales. The first mathematical introduction
on Tilings and Patterns has been written by Branko Griinbaum and Geoffrey Shepard
in 1989. Canons are special case of tilings. Moreno Andreatta has shown that Regular
Canons are related to the classification of abelian finite groups.
Z12 = 4Z 3 0 3Z 4
A triangulation of each square leads to major and minor triads. This triangulation cor-
responds to the triangulation of a manifold M, and verifies the Euler-Poincare formula.
Let F be the number of faces, V the number of the vertices and E the number of edges.
11. Tessellations and Canons
F + V - E = x(M)
C Eb F# A C
But the torus is not the only possibility of chords tesselations. If we consider the chain
of fifths represented by the translation Tj and the chains of tritone, represented by the
translations TQ, we get a tiling in which the gluing of the borders in the direction of the
arrows leads to the Klein bottle. A triangulation of this tilings verifies the Euler-Poincare
formula F + V - E = 0.
G 0 ~" A E B F*|
Db Ab Eb Bb F
6 1 8 3 10 5
7 * _ ^» 9 i Jj I
C G D A^ E B F#
Another instructive example is given by the tiling of the projective plane PR 2 with
Steiner systems. Let X be the set {1, 2, ..., n}.
Example. The Steiner triplet S(2,3,7) is the set of 7 triplets: {1,2,4}, {2,3,5}, {3,4,6},
{4,5,7}, {1,5,6}, {2,6,7} and {1,3,7}. The unique Steiner triple system of order 7 is
known as the Fano plane.
The Steiner triple system 5(2, s + 1, s2 + s + 1) is, for each integer 5, a finite project-
ive plane. Suppose now that we have a set of six musical objects, denoted by the first
196
Musical Tilings
positive numbers X = {1,2,..., 6}. The set of triplets such that all pair of numbers in X
occurs exactly in two triplets is not a Steiner triple system, but is related to 5(2,3,7).
It is a set of 10 triplets: Tx = {1,2,3}, T2 = {1,3,5}, T 3 = {1,5,6}, T4 = {1,2,4},
T 5 = {1,4,6}, T6 = {2,3,6}, T7 = {2,4,5}, T8 = {2,5,6}, T9 = {3,4,5}, Tl0 = {3,4,6}.
The triangulation of the projective plane is shown in the following figure.
The Euler characteristic of the projective plane (x = 1) verifies the relation F+V—E = x,
with F = 10, V = 6 and E = 15.
Archimedean tilings are uniform tilings of the plane by regular polygons, n-gons,
as squares (denoted 4), triangles (3) and hexagons (6), octogons (8) and dodecagons (12
or C). Each vertex has the same adjacent tiles and is named a\...an by these tiles. For
example 3463 means, that at each vertex we find a triangle (3), a square (4), an hexagon
(6) and another triangle (3). Each number of the tiling represents a sector of a 2n angle:
4 means a n/2 angle, 3 means a 7r/3 angle, 6 corresponds to a 27r/3, 8 is a 37r/4 angle
and 12 means a 57r/6 angle. This determines a function </> verifying
The sum of each components of the tiling name is always 2ir. For example for the tiling
x given by C3C (where C is the hexadecimal notation of 12) the function
It has been shown that there are only eleven archimedean tilings: 333333, 4444, 666,
33336, 33344, 33434, 3463, 3636, 3CC, 46C and 488. Each tiling has a group of sym-
metry. The tiling 43433 has a right and left handed version. Its symmetry group is p4girn.
Archimedean tilings have been generalized to k-uniform tilings, that is tilings with k
kinds of vertices. Archimedean tilings have only one type of vertices with regular poly-
gons, fc-uniform tilings have been enumerated by Otto Krotenheerdt. They are 20 tilings
with two kinds of vertices, 39 tilings with 3 kinds of vertices, 33 tilings with 4 kinds of
vertices, 15 tilings with k = 5, 10 tilings with k = 6, 7 tilings with k = 7 and no tilings
for k greater or equal 8.
197
11. Tessellations and Canons
The following picture shows a tiling 333333 filled by twelve tone rows. In each triangle,
the twelve notes are shared with the neighboring triangles. The period of this tilling is
equal to 12.
A
1
•/ A
A J
0 3 7 11 8 3 7 11 4 3 7 11 0
The next picture shows a 6434 tiling filled by twelve tone rows. The wheels share the
notes of each row. The period is equal to 12.
Twelve \
Tono
G 7 0 Row I S l ^ ^ H ^
Tiling /
3
11
11
v
\<{o] 5 ]8>^ > 10
1 9 \ 1 / 3
4 5 4
' iQ ) 'A
0 7 6
Polygons tilings are figures assembled from other polygons as in the Tetris game.
For figure made from squares Solomon W. Golomb coined the term polyominoes in 1954.
Tilings made with figures from triangles are called polyamonds and tilings made with fig-
ures from hexagons are called polyhexes. There are only one monomino (a single square),
one domino (a double square), two triominoes (assemblies of three squares), five tet-
rominoes, twelve pentominoes, 35 hexominoes, 108 heptominoes, etc. Only four hep-
tominoes do not tile the plane.
198
Musical Tilings
For a tiling of polyominoes, one can associate a rhythm with each square. The polyominoe
and its rhythms are repeated through the whole tiling. In the following figure, the tile is
represented by five elementary squares (see for example the black pentomino). The tiling
of the whole rectangle is made of ten tiles. The distribution of the rhythmic figues is the
same in each tile.
Pentilings are tesselations made with plane pentagons. Figures with five-fold sym-
metry have been associated with magic and mysticism since ancient times, as it appears
to be a non natural symmetry. They are sometimes considered as an anticipation of
Penrose's aperiodic tilings. From the musical point of view, the vertices of a pentagon
are associated with a musical object such as a rhythm figure or a pentachord. The first
arrangements of pentagons were described by Diirer in a book published in 1525. In one
of them, the tile consists of two pentagons and one lozenge gap. Six pentagons are linked
by an hexagon as shown in the following left figure.
10 6
The right figure is built by applying a five-fold rotation symmetry about a central
pentagon, associated with the chord [0, 2, 5, 7, 10]. Five hexagons are distributed around
the central point of the figure. Six pentagons are included in a bigger pentagon. Two
adjacent pentachords have two common pitch classes.
199
11. Tessellations and Canons
Aperiodic tilings are tilings in which no subset of polygons tiles the plane period-
ically. Robert Berger has found an aperiodic tiling in 1964. Ten years later, Raphael
Robinson simplified the number of the tiles to six. Robert Ammann discovered a number
of aperiodic tilings and in 1974, Robert Penrose found an aperiodic tiling that uses only
two tiles. The first tile is a rhombus containing an angle of 36° and the second tile is a
rhombus containing an angle of 72°. If we locate the origin at one of the vertices of the
tiling and define the x and y axes such that the angles between edges of tiles and the x
axis are multiple of 7r/5, then we could represent the Penrose tilings by integers points
of R 4 . Using the rational expressions of angles
C O S(*V5)
v../v/ = —- A — sin(7r/5) = V2
4
and
to /K\ -! + ^ • /« ,K\ /zVs + VS
v
cos(27r/5) = 4 sin(27r/5)
' ' = v2- 4
we can represent the coordinates (x,y) of each vertex by four integers (#1,0:2,2/1,2/2)
defined by
J x = x\ cos(7r/5) + #2 cos(27r/5)
\ V = 2/1 sin(7r/5) + 2/2 sin(27r/5)
Wang's conjecture (1961) states that all tilings could always be arranged to tile the plane
periodically.
Reptiles. A Reptile (Replicating tile) is a shape that can be tiled with smaller congru-
ent copies of itself. The rep number, denoted by rep-n, is the number n of tiles required
to build a larger version. All triangles and parallelograms are rep-n2 (for some integer
n > 1). The 45-degree right triangle is rep-2. It is easy to show that if a figure is a rep-n
tile then it is also a rep-(2n-l) tile.
The reptile is said to be regular if the copies used to build the larger version have the
same size (self-similarity). If a figure is a regular rep-n tile then it is also a regular rep-n2
tile. It has been shown by Sze-Man Ngai, Victor Sirvent, J.J.P. Veerman and Yang Mang
that there are only six regular rep-2: the 45-degree right isocele triangle, the rectangle
of sides 1 and y/2, and four fractal reptiles: the Levy dragon, the Heighway dragon, the
twindragon and the tame twindragon. These tiles are used in fractal music. Fractal music
is the use of fractals computation to create music by a mapping of numerical values issued
by the computation to musical parameters.
200
Perfect Tilings
Definition 211 Let m, k be two non negative integers. A linear perfect tiling of length
n = mk and of order k is a partition of Zn in m sets of k elements Aij = Mij(Zk)
(hj)es
such that all coefficients i are distinct. The mutiplicative coefficients i are called the
augmentations and S the set of affine coefficients of the tiling.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14]
A+2 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0
2A + 4 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0
4A + 5 0 0 0 0 0 1 0 0 0 1 0 0 0 1 0
5A + 1 0 1 0 0 0 0 1 0 0 0 0 1 0 0 0
7A 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1
It can be written in a compact form, keeping only the multiplicative coefficient of the
affine transformation,
(7,5,1,1,1,4,5,7,2,4,2,5,2,4,7)
For n = 21, there are only 9 perfect linear tilings (if we do not count the mirror tilings).
One of them is given by the set
(9,5,6,1,1,1,5,3,6,9,3,5,4,3,6,2,4,2,9,2,4)
Proposition 212 The number of the linear perfect tilings of order k = 3 are given in
the table
m n mn m n m
5 15 2 11 33 4 138
6 18 0 12 36 15 324
7 21 18 13 39 61 644
8 24 66 14 42 328 800
9 27 382 15 45 2 151878
10 30 1 104 L*6 48 17 660 110
201
11. Tessellations and Canons
Perfect tilings have been defined for any shape. Perfect square tilings are the tiling of a
square with smaller squares.
Definition 213 A perfect square tiling is a tiling of a square with finitely many similar
but non-congruent copies of the same square. The order of the tiling is the smallest
number of copies needed in a perfect tiling.
Example. Duijvestijn [Duijvestijn, 1978] has proved that the smallest order for a perfect
square tiling is 21.
WXm
50
1
15 17
9
I 18 24
29 : 25 16
1
33
37 g m
Proposition 214 The number of perfect square tilings of order k is given in the following
table
\ k \P\\ k W\
21 1 24 26
22 8 25 160
23 12 26 441 \
202
Messiaen Chords
1 5 9 1
2
o
In the molecular presentation, the chord shows two limited transpositions sets, the
double fifth {0, 5, 6, 11} and the augmented fifth {2, 6, 10}.
The second type of chords used by Olivier Messiaen is called the chord of contracted
resonance (accord a resonance contract&e, Tome 7, p. 150). These chords have been used
for the first time in the Quatuor pour la fin du temps. The molecular presentation of
CCR1A {1, 2, 3, 4, 6, 7, 9}
203
11. Tessellations and Canons
shows the limited transpositions subsets, the altered seventh {1, 3, 7, 9} and the two
tritones.
The molecular presentation of CCR1B {1, 2, 3, 4, 5, 8, 10} shows the same subsets,
the altered seventh {2, 4, 8, 10} and the two tritones.
These two chords have the same number of fifths (notes located on the diagonal axis)
and the same limited transpositions subsets.
The second chords of contracted resonance (CCR2) A and B have the following present-
ation. The molecular presentation of CCR2A {0, 1, 5, 6, 9, 10}
shows the limited transpositions subsets, the chord of augmented fifth {1, 5, 9} and the
tritones {0, 6}.
The molecular presentation of CCR2B {0, 3, 4, 6, 10, 11} shows two different subsets,
the altered seventh {0, 4, 6, 10} and two tritones. We can easily see that this chord is
included in CCR1A.
6 10
204
Messiaen Chords
The third type of chords studied by Messiaen is the revolving chord (accord tournant,
Tome 7, p. 165). The molecular presentation of the revolving chord A {0, 1, 2, 3, 4, 6,
7, 8} looks like a long troncated triangle. The chord includes the mode 5 {2, 3, 4, 8, 9,
10}, the chord of altered seventh {2, 4, 8, 10}, the double fifth {2, 3, 8, 9}, the chord of
augmented fifth {0, 4, 8} and three tritones.
The molecular presentation of the revolving chord B {0, 1, 2, 3, 7, 8, 10, 11} has also a
triangular shape. The chord includes the double fifth {1, 2, 7, 8}, the chord of augmented
fifth {3, 7, 11} and two tritones.
The molecular presentation of the revolving chord C {0, 1, 3, 4, 5, 8, 10, 11} is built
on the axis of fifths. The chord includes the limited transpositions sets, the double fifth
{4, 5, 10, 11}, the chord of augmented fifth {0, 4, 8} and two tritones.
In the harmonic progression, the molecular representation helps us to follow the evol-
ution of the kernel of included limited transpositions sets. In the final movement of Et
exspecto resurrectionem mortuorum, the melodies uses the previous chords. The kernel
is a chord of altered seventh that we can follow as a rectangle shape (CCR1 mes. 9: 1A,
mes 10: IB, mes 11: 10A, mes. 12: 10B). Messiaen uses the revolving chord 12A which
includes the altered seventh and the double fifth. The double fifth becomes the new ker-
nel. This produces a kind of modulation. The double fifth is the only subset of limited
transpositions in the following chords (revolving chords 12B et 12C, mes 12-13).
205
11. Tessellations and Canons
Definition 215 The volume of a reflexive chord associated with a polytope P is the
volume of the polytope P. Let dP be the border of P and \P\ be the cardinal of P. The
volume of P is given by the Pick's formula.
vol{P) = \P\-\\dP\-l
Example. In the previous example, the chord [0, 3, 4, 7, 8, 10, 11] has the volume equal
to vol(P) = 7 - 6 / 2 - 1 = 3.
In the plane, there are only 16 reflexive chords. They are given in the following figure.
206
Young Diagrams
J. (6) I I I I I I I
*• J- (3,3)
i J.J> F"
(3,2,"lT
(3,1,1,1)
JJi)i)(2,2,1,1)P
J>J>J>J>J>J> (1.1.1.1.1.1)0
207
11. Tessellations and Canons
From a mathematical point of view, canons are linked to the decomposition of finite
abelian groups.
Definition 216 Let G be an abelian group and R, S, X be three subsets of G. A set R
tiles the set X with S if X is a direct sum of R and S
X = R®S
i.e. if every element of x G X has a unique decomposition x = r + s with r G R and
s G S. R is called the tile and S is the translation set. When X = G, the couple (iJ, S)
is called a factorization of the group G.
Examples. The tile R = {0,2} tiles the set, Z4 = {0,1,2,3} with two translations S =
{0,1}, i.e. the sets R and R + 1.
0 1 2 ~3~1
R 1 0 1 0
R+ l 0 1 0 1
208
Prime Canons
Remark that the number 1 appears only one time by column. The motive R is also
written 101. The tiling corresponds to the decomposition of Z4 in the direct sum
Z4 = # 0 S ={0,2} 0(0,1}
= 2Z 2 0 Z 2
In the musical field, a factorization (i?, S) is called a canon or a tiling canon. The set R
represents the set of events of the ground voice (dux) and the set S represents the time
attacks: R + s with s G S are the different voices (comes) of the canon.
A(x) = £>»
aeA
When working with polynomials, the direct sum of sets is interpreted as a product
of polynomials and the union of sets as a sum of the corresponding polynomials. In the
previous example, R = (0,2} is associated with R(x) = x2 + x° = x2 + 1 and the set
S = {0,1} corresponds to the polynomial S(x) = x + 1. The product
R(x)S(x) = (z 2 + l)(a; + l)
= x3 + x2 + x + 1
\A\ = A{\)
(2) Let Pm(x) be the characteristic polynomial ofZm. Then the characteristic polynomial
of the set kZm is Pm(xk).
(3) If E = i? 0 S, then the characteristic polynomial of E is
E(x) = R(x)S(x)
Remark that the canons are related to the underlying group (which defines the sum).
For example, the tile 10110100 associated with R = {0,2,3,5} tiles the integers Z, but
not the nonnegative integers N (there is always a hole in position 1). The tile R tiles
the group Zs = {0,1,2, ..,7} equipped with the sum modulo 8, but not the same set
Ns = {0,1,2, ...,7} equipped with the usual sum. The first example is a tiling of the
circle, whereas the second is a tiling of the line.
209
11. Tessellations and Canons
Tj{R+si)=R' + 8aW
for i = 1,.., | 5 | . In other words, E and E' are equivalent, if the voices of the canon E
taken at any time is a permutation of the voices of the canon E\
Example. The canon E = {0,1} 0 {0,2} is equivalent to the canon E' = {0,3} 0 {0,2},
because T^(R) = R' and T^(R + 2) = R' + 2. The table of the second canon is obtained
from the table of the first by a translation of each element by T 3 (or by a circular
permutation of the columns)
h
0 1 2 ~3~] 0 2 3
R 1 1 0 0 R' 1 0 0 1
R+2 0 0 1 1 R' + 2 0 1 1 0
Definition 220 Two canons E and E* of Zn are called dual if E = i? 0 5 and E* =
S®R.
Example. The canon E = {0,1} 0 {0,2} is dual of the canon E* = {0,2} 0 {0,1}.
Remark that the tables can be very different, by exchanging the role of R and S.
0 1 2 31 0 1 2 3
-
R 1 1 0 0 s 0 0
R+2 0 0 1 1 5+1 0 1 0 1
Definition 221 Two canons E and E' ofZn are isomorphic if E' is equivalent to E or
to its dual E*. Non-isomorphic canons are called prime canons.
An important problem is to build and to determine the number of prime canons. The
problem seems to be easier on the line than on the circle. The first way of building
prime canons on the line (i.e. we work on Z n with the usual sum) is to consider the
decomposition of n in powers of primes. For n = qi...qr, we have
Example. Writing all the decomposition products for n = 12, we get eight canons
n R
1x12 Zi 0 Z12
2x6 Z2 0 2Z6
3x4 Z 3 0 3Z4
4x3 Z4 0 4Z3
6x2 Z6 0 6Z2
2x2x3 Z 2 0 2Z 2 0 4Z3
2x3x2 Z 2 0 2Z 3 0 6Z2
3x2x2 Z 3 0 3Z 2 0 6Z2
210
Prime Canons
Another way to do the same thing is to use tensorial products. Define the horizontal
elements
Hrj i i •• i
1 0 "ol 1 0
T = H2 ® V3 ® V2 1 1 ® 0 1 0 ®
0 1
0 0 1
pTfl | 0 0
1 0
T = 0 111 1 1 0
0 1
0 0 111 1 1
which leads to the following canon
\1 1 0 0 0 0 0 0 0 0 0 0
0 0 1 1 0 0 0 0 0 0 0 0
0 0 0 0 1 1 0 0 0 0 0 0
T = 0 0 0 0 0 0 1 1 0 0 0 0
0 0 0 0 0 0 0 0 1 1 0 0
0 0 0 0 0 0 0 0 0 0 1 1I
which can be written
T = {0,1} 0 {0,2,4,6,8,10} = Z 2 0 2Z 6
But these generators could not be used for tiling the circle (when we use the sum
modulo n). The first prime canons are given in the following table. The column N n
gives the number of canons for tiling the line, and the column Z n gives the number of
211
11. Tessellations and Canons
n Nn Zn Total nr 2 2 4
2 1 0 1 10 3 3 6
3 1 0 1 11 1 0 1
4 2 0 2 12 8 15 23
5 1 0 1 13 1 0 1
6 3 0 3 14 3 10 13
7 1 0 1 15 3 22 25
8 4 2 6 16 8 41 49
We see in the table that for n = 8, two new prime canons are given. These canons are
E1 = 3Z 4 0 4Z 2 = {0,3,6,9} 0 {0,4}
and
E2 = 3Z 2 0 2Z 4 = {0,3} 0 {0,2,4,6}
which are not obtained from tensorial product of a combination of the elements
Hp and Vq.
0 1 2 3 4 ~5l
0 1 2 3 4 5
A 1 1 0 0 0 0
B 1 0 1 0 1 0
A+2 0 0 1 1 0 0
B+ l 0 1 0 1 0 1
A+4 0 0 0 0 1 1
These two tiles are embedded in Zi 2 , and then become A' = {0,2} and B' = {0,4,8}.
The new canon is A! 0 {0,4,8} U 5 ' 0 {1,3}. In the space Zi 2 , the tile A is written
according to the space {0, 2, 4, 8, 10} and the tile B is written according to the space
{1, 3, 5, 7, 9, 11}.
0 1 2 3 4 5 6 7 8 9 10
A' 1 0 1 0 0 0 0 0 0 0 0
in0
A' + 4 0 0 0 0 1 0 1 0 0 0 0 0
\ A' + 8 0 0 0 0 0 0 0 0 1 0 1 0
\ B' + l 0 1 0 0 0 1 0 0 0 1 0 0
\ B' + 3 0 0 0 1 0 0 0 1 0 0 0 1
Prime canons are used to tile the line with chords. If we call the notes and the voices 1,
2, 3, etc., the problem is to write a tiling in which each column is a complete chord with
no repeated note and is a different circular permutation compared to the other columns.
212
Multiple Canons
For a chord of three notes, the set {1, 2, 3} and its permutation {2, 3, 1} and {3, 1, 2}
form three columns of different transpositions of the same chord. The motive 100 tiles a
chord of 3 notes on 3 voices. Each prime canon is associated with a word. The tiling V3
in Z3 is associated with the word 123 composed of the number of the line in which the
1 appears, when we read the canon from the column 0 to the column 2.
0 1 ~2~|
Voice 1 1 0 0
Voice 2 0 1 0
Voice 3 0 0 1
Prom the word w, one constructs the different permutations an(w) with no equal number
in the same column
Hi 2 3 w
2 3 1 oi(w)
3 1 2 a2(w)
Thus the tiling of a three notes chord is derived from the prime canon V3.
Tiling chords could also be constructed starting from canons with different tiles. For
example, from the previous two tiles canon
A' 0 { 0 , 4 , 8 1 0 5 ' 0 ( 1 , 3 }
we get a word
w = 114522453345
This word and the permutation <72 (w) lead to a tiling chord with two notes on five voices
pr 1 4 5 2 2 4 5 3 3 4 5 w
4 5 2 2 4 5 3 3 4 5 1 1 a2{w)
f °"1 1
1
2
2
3
2
4
0
5
0
6
0
7
0
8
0
9
0
10 " 111
0 0
0 0 0 0 1 1 2 2 0 0 0 0
0 0 0 0 0 0 0 0 1 1 2 2
2 0 1 0 2 0 1 0 2 0 1 0
0 2 0 1 0 2 0 1 0 2 0 1
Remark that the permutation GQ(W) and aio(w) are other solutions of tiling on two
voices. But there are also five solutions on five voices. For example, one solution is given
in the following table
FT 1 4 5 2 2 4 5 3 3 4 5 w
3 4 5 1 1 4 5 2 2 4 5 3 a3(w)
5 3 3 4 5 1 1 4 5 2 2 4 a5(w)
4 5 2 2 4 5 3 3 4 5 1 1 o-ioM
213
11. Tessellations and Canons
The motive of the canon tiles a chord of four notes (1, 2, 3, 4) on five voices.
n 1 ?, 3 4 5 6 7 8 9 10 11
l 1 0 2 2 3 3 0 0 0 4 4
0 0 4 4 1 1 0 2 2 3 3 0
2 3 3 0 0 0 4 4 1 1 0 2
4 2 1 3 4 2 1 3 4 2 1 3
3 4 2 1 3 4 2 1 3 4 2 1
Definition 223 Let G be an abelian group. G is called an Hajds group if for each fac-
torization of G = i? © S at least one factor is periodic.
Remark that if G is a finite abelian group, the factors are not necessary subgroups of G.
For example, if R = {0, 1, 4, 5} and S = {0, 2}, R®S = Z 8 The set R is 4-periodic, but
none of R and S are subgroups of Zg.
In 1962, A.D. Sands gave the classification of all finite abelian groups which are Haj6s
groups.
Theorem 224 (Sands) The group Zn is an Hajds group if and only ifn is of the form:
pk for k > 0, pkq for k>l, p2q2, pqr, p2qr or pqrs for distinct primes p, q, r, s.
Theorem 225 (De Bruijn) The group Zn is a non-Hajds group if and only if n can
be expressed in the form piP2^i^2^3 where p\, P2 are primes, piUi > 2 for i = 1, 2, 3
and gcd(n\p\,nvp2) = 1.
Proof We will show that the De Bruijn's theorem is equivalent to the Sands's theorem.
Define the following sets
214
Regular Canons
•
The smallest values of n for which Z n is a non-Haj6s group are: 72, 108, 120, 144, 168,
180, 200, 216, 240, 252, 264, 270, 280, 288, 300, 312, 324, 336, 360, 378, 392, 396, 400,
408, 432, 440, 450, 456, 468, 480, 500, 504, 520, 528, 540, 552, 560, 576, 588, 594, 600,
612, 616, 624, 648, 672, 675, 680, 684, 696, 700, 702, 720, 728, 744, 750, 756, 760, 784,
792, 800, 810, 816, 828, 864, 880, 882, 888, 900, 912, 918, 920, 936, 945, 952, 960, 968,
972, 980, 984, 1000.
Definition 226 A regular canon (R, S) is a factorization of a non-Hajds group Z n into
two aperiodic subsets
Zn = R®S
Theorem 227 Let Z n be a non-Hajds group with n = p\P2n\n2n^. Let A and B be the
sets
A = n 2 n 3 (Zp 2 0 p 2 ^ i Z P l )
B = nin 3 (Z P l 0 p i n 2 Z P 2 )
and let Tj(B) = {j} 0 B be the translation of length j . Then the canon (R, S) defined by
S = n3(p2n2Zni 0piniZn2)
R = AUTi(JB)U...UV1(5)
is a regular canon.
Proof Using the following properties
Za0aZ6 = Za6
cLa 0 aZbc = %abc mod abc gcd(a, c) = 1
215
11. Tessellations and Canons
S®A = p2n2n3Zni 0 p \ n i n 3 Z n 2 0 n 2 n 3 Z P 2 0p 2 n x n 2 n 3 Z P x
= 72 2 72 3 (p 2 £i n i ® Zp2) 0 77-172.3 ( p i Z n 2 0 p 2 n 2 Z P l )
= ra2n3(ZniP2) 0 n i n 3 ( p i Z n 2 0 p 2 n 2 Z P l )
= 7 i 2 n 3 ( ^ » n i P 2 0 p2n{ZPx) 0 nin 3 piZ n 2
= n2n3ZPlP2ni 0nin3piZn2
= ^3(^iPi^n 2 0 n2ZPlP2ni)
PlP2^1^2
Thus, we get
,4 = 6 Z 3 0 36Z 2 = {O,6,12,36,42,48}
B = 4 Z 2 0 24Z 3 = {O,4,24,28,48,52}
S = 1 8 Z 2 0 8Z 3 = {O,8,16,18,26,34}
R = A UTi (B) = {0,1,5,6,12,25,29,36,42,48,49,53}
is a regular canon.
Conjecture 228 All regular canons are generated by transformations of the sets A, B
and S. In other words, it exists some applications Uj, Vj (non necessary linear) such that
216
Magic Squares
A = 9 Z 3 0 54Z2 = {O,9,18,54,63,72}
B = 6 Z 2 0 36Z 3 = {O,6,36,42,72,78}
Rij = A\JTi(B)\JTj(B)
for different values (i,j) are regular canons.
A = 6 Z 5 e 6 0 Z 2 = {0,6,12,18,24,60,66,72,78,84}
B = 4Z 2 0 24Z5 = {0,4,24,28,48,52,72,76,96,100}
[4 9 2
3 5 7
8 1 6
217
11. Tessellations and Canons
In the representation of the i" King, numbers are represented by points. According to
the legend, the emperor (ca. —2200) saw it on the shell of a tortoise near the Yellow
river.
It is easy to show that the set of magic squares (a^) formed by the n2 non-negative
integers is a vector space Mn of dimension
dimM n = n{n- 2)
The group A4 of the isometries of the square acts on the space Mn. This action is faithful.
A4 is composed of the eight transformations
A4 = {l,si,s 2 ,s 3 ,s 4 ,7*1,^2,7*3}
si(aid) = a
i,n—j+\
s2(aij) = a
n—i+lJ
s a
s( hj) = Q"n—j+l,n- - i + 1
S4(aii:j) = a
3,i
r
i(aij) = a
n—j+l,i
r a
2( ij) = Q"n—i+l,n- -3 + 1
r a
s( ij) = Q>j,n—i+1
The Roman knew an alphabetical magic square that can be read just as well horizont-
ally as vertically:
rsA A T 0 R\
R E P 0
T E N E T
0 P E R A
R 0 T A S
In the 13 th century, Manual Moschopoulos, raises of Maxime Planude stated a series
of rules to build a magic square. In the 16 th century, Albrecht Diirer (1471-1528) quoted
the magic square below in its famous engraving the Melancholy.
ri6~ 3 2 "131
5 10 11 8
9 6 7 12
4 15 14 1 I
The date of the engraving (1514) is formed by the two numbers in the center of the
last line. Each line, each column or diagonal gives a linear sum of 34. If this square is
divided in four smaller squares and a central square, each subsquare being made of four
elements, one notes that the sum of these four elements is always equal to the linear sum
(34). This property is not true for all the subsquares. For example, the square (3, 2, 10,
11) is a subsquare of sum 3 + 2 + 10 + 11 = 26, and not 34. On the other hand, the
Jupiter square admits a perfect subdivision in subsquare: the sum of the four elements
of an unspecified subsquare is equal to the constant of the Jupiter square.
218
Magic Squares
15 10 3 ~wi
4 5 16 9
14 11 2 7
1 8 13 12j
Dtirer's square and the Jupiter square are constructed by using the first 16 positive
integers (from 1 to 16) but are laid out in a particular order, to preserve the magic
constant. The Mars square uses the first 25 integers.
2 23 25 7 8
4 16 9 14 22
21 11 13 15 5
20 12 17 10 6
18 3 1 19 24
In the center, the subsquare of nine elements is also a magic square. The Venus square
uses the first 49 integers.
[4 29 12 37 20 45 281
35 11 36 19 44 27 3
10 42 18 43 26 2 34
41 17 49 25 1 33 9
16 48 24 7 32 8 40
47 23 6 31 14 39 15
22 5 30 13 38 21 46
The Mercury square is constructed by using the first 64 positive integers.
H*
49
58
15
59
14
5
52
4
53
62
11
63
10
1
56
41 23 22 44 45 19 18 48
32 34 35 29 28 38 39 25
40 26 27 37 36 30 31 33
17 47 49 20 21 43 42 24
9 55 54 12 13 51 50 16
64 2 3 61 60 6 7 57
In music, magic squares have been used in the 20 th century. Peter Maxwell Davies' piece
Ave Maris Stella composed in 1975 is based on the following square.
ri 6 2 7 3 8 4 9 5
6 2 7 3 8 4 9 5 1
2 7 3 8 4 9 5 1 6
7 3 8 4 9 5 1 6 2
3 8 4 9 5 1 6 2 7
8 4 9 5 1 6 2 7 3
4 9 5 1 6 2 7 3 8
8 5 1 6 2 7 3 8 4
5 1 6 2 7 3 8 4 9|
219
11. Tessellations and Canons
nr 1 ~8~|
7 5 3
2 9 4
In Tabula Smaragdina, he uses the Jupiter square in which the integers are transformed
into rhythmic figures according to pre-established rules of transposition. In his Etude sur
le carre* magique sonore for piano, Ivan Wyschnegradsky uses a cyclic melody structure of
six measures, founded on non-octaviant spaces. The period is a seventh major of volume
11, i.e. consisted of 11 semitones. It is divided into four parts (E\>, F|J, C, D). This period
generates a complete cycle by repetition of the initial structure (3 semitones, 3 semitones,
3 semitones and 2 semitones) extended to the whole audible space.
220
Appendix A
Dodecaphonic Knots
The diagrams give the 554 chord diagrams of the dodecaphonic knots and for each,
the following properties: the Gauss word, the structural vector and the associated per-
mutations. The knots are classified by increasing Gauss words.
(0 1) (2 3) (4 5) (6 11) (7 8) (9 10)
Gauss word 112233456654
Structural vector 401010
(0 1) (2 3) (4 5) (6 11) (7 10) (8 9)
D257 x
= adef2b D262 X = ade-'dfb
Gauss word 112345263645 Gauss word 112345264635
276
Appendix B
All-Interval Series
The tables give a description of all-interval series classified by their chord diagram. The
series are denoted by their structural vector, whose components are the numbers of
semitones between two adjacent pitches.
Knot 14 1 n. 20
n. 21
3 6 7 10 81 9 5 4 2 11
3 6 8 11 7 2 9 4 5 1 10
1 n. 1 1 4 2 7 9 11 10 8 5 6 3 n. 22 3 6 10 1 5 4 9 2 7 11 8
n. 2 1 6 3 4 2 5 11 9 10 8 7 n. 23 3 6 11 2 4 5 9 1 8 10 7
n. 3 1 6 7 8 10 3 5 11 2 4 9 n. 24 3 8 10 1 7 9 2 4 5 6 11
n. 4 1 6 7 8 10 9 11 5 2 4 3 n. 25 38 10 1 7 9 2 4 11 65
n. 5 1 6 9 4 2 11 53 10 87 n. 26 415 10 3 8 7 11 269
n. 6 18 10 352 11 7964 n. 27 4 6 1 8 10 352 11 79
n. 7 18 10 35 11 2 4 9 6 7 n. 28 4 6 3 1 5 2 11 9 10 87
n. 8 18 10 9 11 5 2 4 3 6 7 n. 29 4 6 7 8 10 9 11 2 5 1 3
n. 9 1 10 873 11 4 2 5 6 9 n. 30 4 6 9 7 11 253 10 81
n. 10 2 5 1 8 9 10 11 7463 n. 31 47 11 10 9 8 1 5 2 6 3
n. 11 2 11 783 10 5 1 4 6 9 n. 32 5 24 11378 10 169
n. 12 3 1 5 2 11 9 10 8 7 6 4 n. 33 5 4 2 3 1 7 10 896 11
n. 13 3 4 2 5 11 9 10 8 1 6 7 n. 34 5 4 2 3 1 10 7 11 968
n. 14 3 4 2 5 11 9 10 8761 n. 35 5 4 2 9 7 1 10 836 11
n. 15 351 10 7 9 2 4 11 68 n. 36 5 6 3 8 10 1 7 9 2 4 11
n. 16 3 6 1 4 2 7 9 11 10 85 n. 37 5 6 9 8 10 7 1 3 2 4 11
n. 17 3 6 2 5 1 8 9 10 11 74 n. 38 56 11 4 2 3 1 7 10 89
n. 18 3 6 4 7 11 10 9 8 1 5 2 n. 39 56 11 4 2 9 7 1 10 83
n. 19 3 6 5 8 10 11 9 7 2 4 1 | n. 40 58 10 11 9 7 2 4 1 6 3 |
B. All-Interval Series
Knot 19
1 n. 1 1 2 4 10 9 7 11 3 5 6 8 1 n. 33 3 6 2 11 7 8 1 5 9 4 10
n. 2 1 3 7 11 9 2 8 10 5 6 4 1 n. 34 3 6 4 1 5 10 11 7 9 2 8
n. 3 1 3 7 11 9 8 2 4 5 6 10 n. 35 3 6 4 10 9 11 7 2 1 5 8
n. 4 1 8 10 4 9 7 11 3 5 6 2 n. 36 3 6 5 2 4 11 10 8 9 1 7
n. 5 2 6 1 8 10 4 9 7 11 3 5 n. 37 3 6 5 11 9 10 8 1 2 4 7
n. 6 2 6 5 3 11 7 9 4 10 8 1 n. 38 3 6 7 1 9 8 10 11 4 2 5
n. 7 2 6 7 8 10 4 3 1 5 9 11 n. 39 3 6 7 4 2 1 8 10 9 11 5
n. 8 2 6 11 9 5 1 3 4 10 8 7 n. 40 3 6 8 2 9 7 11 10 5 1 4
n. 9 4 6 1 3 7 11 9 2 8 10 5 n. 41 3 6 8 5 1 2 7 11 9 10 4
n. 10 4 6 5 10 8 2 9 11 7 3 1 n. 42 3 6 10 4 9 5 1 8 7 11 2
n. 11 4 6 7 9 1 5 3 2 8 10 11 n. 43 3 6 10 7 11 4 5 1 9 8 2
n. 12 4 6 11 10 8 2 3 5 1 9 7 n. 44 3 6 11 5 9 4 2 7 8 10 1
n. 13 5 3 11 7 9 4 10 8 1 6 2 n. 45 3 6 1 1 8 10 5 4 2 9 7 1
n. 14 5 3 11 7 9 10 4 2 1 6 8 n. 46 3 7 1 10 8 9 2 4 5 6 11
n. 15 5 4 2 8 9 11 7 3 1 6 10 n. 47 3 7 1 10 8 9 2 4 11 6 5
n. 16 5 10 8 2 9 11 7 3 1 6 4 | n. 48 3 7 11 8 2 4 5 9 1 6 10
n. 49 3 10 1 7 11 9 2 4 5 6 8
n. 50 3 10 8 11 2 4 5 9 1 6 7
n. 51 3 10 8 11 2 4 5 9 7 6 1
n. 52 3 11 5 2 4 9 10 8 1 6 7
n. 53 3 11 5 2 4 9 10 8 7 6 1
n. 54 3 11 7 4 10 8 1 9 5 6 2
1 n. 1 1 3 11 4 2 5 8 10 9 6 7
n. 55 4 1 5 10 11 7 9 2 8 6 3
n. 2 1 4 2 7 8 10 3 5 11 6 9
n. 56 4 6 1 8 10 9 7 11 5 2 3
n. 3 1 6 3 10 8 11 2 4 5 9 7
n. 57 4 6 3 2 5 11 7 9 10 8 1
n. 4 1 6 3 11 5 2 4 9 10 8 7
n. 58 4 6 7 8 10 3 1 5 11 2 9
n. 5 1 6 7 3 11 4 2 5 8 10 9
n. 59 4 6 9 2 11 5 1 3 10 8 7
n. 6 1 6 7 8 10 3 4 2 11 5 9
n. 60 4 7 11 10 5 1 3 2 8 6 9
n. 7 1 6 7 8 10 9 4 2 5 11 3
n. 61 4 10 3 5 1 2 7 11 8 6 9
n. 8 1 6 7 9 5 4 2 11 8 10 3
n. 62 4 10 9 11 7 2 1 5 8 6 3
n. 9 1 6 9 5 11 2 4 3 10 8 7
n. 63 5 2 4 11 10 8 9 1 7 6 3
n. 10 1 6 9 10 8 5 2 4 11 3 7
n. 64 5 3 7 8 10 1 4 2 9 6 11
n. 11 1 7 3 8 10 5 4 2 11 6 9
n. 65 5 4 2 3 8 10 7 1 9 6 11
n. 12 1 7 9 2 4 5 10 8 11 6 3
n. 66 5 4 2 9 8 10 1 7 3 6 11
n. 13 1 8 10 3 4 2 1 1 5 9 6 7
n. 67 5 4 2 9 11 7 1 10 3 6 8
n. 14 1 8 10 9 4 2 5 1 1 3 6 7
n. 68 5 6 3 2 4 7 10 8 1 9 11
n. 15 1 8 10 9 7 11 5 2 3 6 4
n. 69 5 6 3 7 1 10 8 9 2 4 11
n. 16 1 9 5 4 2 8 11 7 3 6 10
n. 70 5 6 9 1 7 10 8 3 2 4 11
n. 17 1 9 5 4 2 11 8 10 3 6 7
n. 71 5 6 9 2 4 1 10 8 7 3 11
n. 18 1 10 8 7 2 4 9 5 11 6 3
n. 72 5 6 11 3 7 8 10 1 4 2 9
n. 19 2 5 1 8 7 11-3 4 10 6 9
n. 73 5 6 11 4 2 3 8 10 7 1 9
n. 20 2 6 3 11 7 4 10 8 1 9 5
n. 74 5 6 11 4 2 9 8 10 1 7 3
n. 21 2 6 5 9 1 8 10 4 7 11 3
n. 75 5 6 11 9 1 8 10 7 4 2 3
n. 22 2 6 9 5 1 4 10 8 7 3 11
n. 76 5 8 10 11 4 2 3 1 7 6 9
n. 23 2 6 11 3 7 8 10 4 1 5 9
n. 77 5 9 1 8 10 4 7 11 3 6 2
n. 24 2 8 3 7 11 4 5 1 10 6 9
n. 78 5 9 1 8 10 7 4 2 3 6 11
n. 25 2 8 9 1 5 4 11 7 10 6 3
n. 79 5 11 3 4 2 1 8 10 7 6 9
n. 26 2 11 7 8 1 5 9 4 10 6 3
n. 80 5 11 9 10 8 1 2 4 7 6 3 |
n. 27 3 2 4 7 10 8 1 9 5 6 11
n. 28 3 2 4 7 10 8 1 9 11 6 5
n. 29 3 2 5 11 7 9 10 8 1 6 4
n. 30 3 6 1 7 9 2 4 5 10 8 11
n. 31 3 6 1 10 8 7 2 4 9 5 11
n. 32 3 6 2 8 9 1 5 4 11 7 10 |
278
B. All-Interval Series
Knot 33 Knot 35
1 n. 1 1 6 4 3 8 10 7 2 11 5 9 1 n. 1 1 4 5 9 2 7 11 8 3 6 10
n. 2 1 6 4 11 5 2 3 8 10 79 n. 2 1 8 7 11 2 5 9 4 3 6 10
n. 3 1 6 9 5 2 4 11 8 7 3 10 n. 3 2 6 3 4 7 11 10 9 1 8 5
n. 4 1 6 9 5 11 2 7 10 8 3 4 n. 4 2 6 3 8 9 1 10 7 11 4 5
n. 5 1 6 9 7 3 8 5 2 4 11 10 n. 5 2 6 5 4 11 7 10 1 9 8 3
n. 6 1 6 9 7 10 8 3 2 5 11 4 n. 6 2 6 5 8 1 9 10 1 1 7 4 3
n. 7 16 10 3 7 8 11 4 2 5 9 n. 7 2 6 9 4 1 5 10 3 7 8 11
n. 8 16 10 11 4 2 5 8 3 7 9 n. 8 2 6 9 8 3 7 10 1 5 4 11
n. 9 2 1 8 10 7 4 9 5 3 6 11 n. 9 26 11 4 5 1 10 7 3 8 9
n. 10 23 11 4 7 8 10 1 9 6 5 n. 10 26 11 8 7 3 10 5 1 4 9
n. 11 2 7 8 10 1 4 3 11 9 6 5 n. 11 3 4 7 11 10 9 1 8 5 6 2
n. 12 2 9 5 4 1 8 10 7 3 6 11 n. 12 3 4 9 5 2 11 7 8 1 6 10
n. 13 3 1 9 8 11 2 4 5 10 6 7 n. 13 3 8 9 1 10 7 11 4 5 6 2
n. 14 31 10 8 9 2 11 5 4 6 7 n. 14 3 8 11 7 2 9 5 4 1 6 10
n. 15 3 5 2 4 9 10 7 1 8 6 11 n. 15 5 4 11 7 10 1 9 8 3 6 2
n. 16 3 5 9 4 7 10 8 1 2 6 11 n. 16 5 8 1 9 10 11 7 4 3 6 2 |
n. 17 3 7 1 10 5 2 4 9 8 6 11
n. 18 3 7 10 8 1 4 5 9 2 6 11
n. 19 3 11 2 4 5 8 1 9 10 6 7
n. 20 3 11 5 2 1 10 8 9 4 6 7
n. 21 4 3 8 10 72 11 5 9 6 1
n. 22 4 5 11 2 9 8 10 1 3 6 7
n. 23 4 9 8 10 1 2 5 11 3 6 7
n. 24 4 11 5 2 3 8 10 7 9 6 1 Knot 38
n. 25 5 6 2 3 11 4 7 8 10 19
1 n. 1 1 3 7 11 5 4 2 8 9 6 10
n. 26 5 6 2 7 8 10 1 4 3 11 9
n. 2 1 3 10 8 5 4 2 11 9 6 7
n. 27 5 6 8 3 4 2 11 10 7 1 9
n. 3 1 6 3 5 2 4 11 8 10 9 7
n. 28 5 6 8 7 1 1 0 3 4 2 119
n. 4 1 6 7 3 10 8 5 4 2 11 9
n. 29 5 6 9 1 7 10 11 2 4 3 8
n. 5 1 6 7 9 10 8 11 4 2 5 3
n. 30 5 6 9 1 10 8 7 4 11 3 2
n. 6 1 6 9 11 2 4 5 8 10 3 7
n. 31 5 6 9 11 2 4 3 10 1 7 8
n. 7 1 7 2 4 9 8 10 11 5 6 3
n. 32 5 6 9 11 3 4 1 10 8 7 2 |
n. 8 1 7 8 10 3 2 4 1 1 5 6 9
n. 9 19 10 8 11 4 2 5 3 6 7
n. 10 2 6 3 4 10 8 7 1 5 9 11
n. 11 2 6 5 3 11 7 1 8 10 4 9
n. 12 2 6 9 4 10 8 1 7 11 35
n. 13 26 11 9 5 1 7 8 10 4 3
Knot 34 n. 14 2 8 1 5 9 7 11 10 4 6 3
n. 15 2 8 7 11 3 1 5 10 4 6 9
n. 1 1 8 7 3 4 11 5 2 9 6 10 n. 16 31 10 8 7 4 2 9 5 6 11
n. 2 2 6 3 10 7 1 8 9 5 4 11 n. 17 31 10 8 7 4 2 9 11 6 5
n. 3 2 6 5 4 11 3 8 7 1 10 9 n. 18 3 4 10 8 7 1 5 9 11 6 2
n. 4 2 6 9 10 1 7 8 3 11 4 5 n. 19 3 5 2 4 11 8 10 9 1 6 7
n. 5 2 6 11 4 5 9 8 1 7 10 3 n. 20 3 5 2 4 11 8 10 9 7 6 1
n. 6 3 2 11 5 4 9 1 8 7 6 10 n. 21 3 6 1 7 2 4 9 8 10 11 5
n. 7 3 10 7 1 8 9 5 4 11 6 2 n. 22 3 6 2 8 1 5 9 7 11 10 4
n. 8 54 11387110 9 6 2 n. 23 3 6 4 10 11 7 9 5 1 8 2
n. 24 3 6 5 11 10 8 9 4 2 7 1
n. 25 3 6 7 1 8 10 9 2 4 5 11
n. 26 3 6 8 2 7 11 9 1 5 4 10
n. 27 36 10 4 5 1 9 11 7 2 8
n. 28 36 11 5 4 2 9 10 8 1 7
n. 29 3 8 2 4 11 5 1 9 7 6 10 |
279
B. All-Interval Series
1 n. 30
n. 31
4 10 5 1 3 11 7 8 2 6 9
4 10 11 7 9 5 1 8 2 6 3
Knot 49
n. 32 5 3 2 4 1 8 10 7 9 6 11 1 n. 1 1 4 9 7 10 8 5 3 11 6 2
n. 33 5 3 11 7 1 8 10 4 9 6 2 n. 2 1 9 7 4 2 5 3 8 11 6 10
n. 34 5 6 3 1 10 8 7 4 2 9 11 n. 3 1 9 7 10 8 5 3 2 11 6 4
n. 35 5 6 9 7 10 8 1 4 2 3 11 n. 4 1 10 9 7 4 2 5 3 11 6 8
n. 36 5 6 11 3 2 4 1 8 10 7 9 n. 5 2 6 1 4 9 7 10 8 5 3 11
n.37 5 6 11 9 2 4 7 8 10 1 3 n. 6 2 6 5 9 11 8 10 1 3 4 7
n. 38 5 9 2 4 7 8 10 1 3 6 11 n. 7 2 6 7 4 3 1 10 8 11 9 5
n. 39 5 11 4 2 3 10 8 7 1 6 9 n. 8 26 11 3 5 8 10 7 9 4 1
n. 40 5 11 10 8 9 4 2 7 1 6 3 | n. 9 4 6 1 9 7 10 8 5 3 2 11
n. 10 4 6 5 2 9 11 8 10 1 3 7
n. 11 4 6 7 3 1 10 8 11 9 2 5
n. 12 4 6 11 2 3 5 8 10 7 9 1
n. 13 5 2 9 11 8 10 1 3 7 6 4
Knot 39 n. 14 5 8 9 11 2 4 1 3 7 6 10
n. 15 59 11 2 4 1 3 10 7 6 8
1 n. 1 1 6 2 5 8 10 9 11 7 4 3 n. 16 59 11 8 10 1 3 4 7 6 2 |
n. 2 1 6 3 4 7 11 9 10 8 5 2
n. 3 1 6 8 5 2 4 3 11 7 10 9
n. 4 1 6 9 10 7 11 3 4 2 5 8
n. 5 2 5 8 10 9 11 7 4 3 6 1
n. 6 2 11 8 10 3 5 1 4 9 6 7
Knot 50
n. 7 3 2 5 1 9 8 10 7 4 6 11 1 n. 1 1 4 5 10 3 8 7 2 11 6 9
n. 8 3 4 7 11 9 10 8 5 2 6 1 n. 2 1 6 3 10 9 4 5 2 11 8 7
n. 9 3 8 11 7 9 2 4 1 10 6 5
n. 3 1 6 4 3 2 5 11 9 10 7 8
n. 10 3 10 1 5 9 4 2 11 8 6 7 n. 4 1 6 78 5 2 11 4 3 10 9
n. 11 41 10 8 3 7 11 2 9 6 5
n. 5 1 6 7 8 11 2 5 4 9 103
n. 12 4 7 10 8 9 1 5 2 3 6 11
n. 6 1 6 8 7 4 3 11 5 2 9 10
n. 13 5 6 3 8 11 7 9 2 4 1 10
n. 7 1 6 8 7 10 9 11 5 2 3 4
n. 14 5 6 4 1 10 8 3 7 11 2 9 n. 8 1 6 9 10 3 4 11 2 5 8 7
n. 15 5 6 9 2 11 7 3 8 10 14
n. 9 16 10 9 2 5 11 3 4 7 8
n. 16 56 10 1 4 2 9 7 11 8 3 |
n. 10 1 8 2 5 11 4 3 7 9 6 10
n. 11 1 8 5 2 11 4 3 10 9 6 7
n. 12 18 11 2 5 4 9 10 3 6 7
n. 13 1 10 5 4 9 2 7 8 11 6 3
n. 14 23 10 7 1 9 8 5 4 6 11
Knot 46 n. 15 2 5 4 11 3 7 8 1 10 6 9
n. 16 2 6 3 5 9 8 1 7 10 4 11
1 n. 1 1 6 3 7 4 5 2 11 8 9 10 n. 17 2 6 5 4 10 1 7 8 3 11 9
n. 2 1 6 3 11 8 9 2 7 4 5 10 n. 18 2 6 9 11 3 8 7 1 10 4 5
n. 3 1 6 10 5 4 7 2 9 8 11 3 n. 19 26 11 4 10 7 1 8 9 5 3
n. 4 1 6 10 9 8 1 1 2 5 4 7 3 n. 20 29 10 1 7 3 8 11 4 6 5
n. 5 2 1 8 11 10 9 4 7 3 6 5 n. 21 2 11 4 5 9 1 8 7 10 6 3
n. 6 2 3 4 1 10 7 8 5 9 6 11 n. 22 3 1 9 4 5 11 2 8 7 6 10
n. 7 2 7 8 5 10 3 4 1 9 6 11 n. 23 3 2 9 8 1 10 7 4 5 6 11
n. 8 2 9 4 7 10 1 8 11 3 6 5 n. 24 3 2 9 8 1 10 7 4 11 65
n. 9 3 7 4 5 2 11 8 9 10 6 1 n. 25 3 5 9 8 1 7 10 4 11 6 2
n. 10 3 7 4 9 10 11 8 1 2 6 5 n. 26 3 6 1 10 5 4 9 2 7 8 11
n. 11 3 11 8 1 10 7 4 9 2 6 5 n. 27 3 6 2 11 4 5 9 1 8 7 10
n. 12 3 11 8 9 2 7 4 5 10 6 1 n. 28 3 6 4 1 2 7 9 11 10 58
n. 13 5 6 2 1 8 11 10 9 4 7 3 n. 29 3 6 5 2 1 8 9 10 11 4 7
n. 14 5 6 2 9 4 7 10 1 8 11 3 n. 30 3 6 7 4 11 10 9 8 1 2 5
n. 15 5 6 3 7 4 9 10 11 8 1 2 n. 31 3 6 8 5 10 11 9 7 2 1 4
n. 16 5 6 3 11 8 1 10 7 4 9 2 | n. 32 36 10 7 8 1 9 5 4 11 2 |
280
B. All-Interval Series
1 n. 33 3 6 11 8 7 2 9 4 5 10 1 1 n. 8 1 1 4 2 9 7 3 8 11 6 5 10
n. 34 3 10 9 4 5 2 11 8 1 6 7 n. 9 1 4 3 11 9 10 8 5 6 2 7
n. 35 3 11) 9 4 5 2 11 8 7 6 1 n. 10 1 4 6 9 7 11 2 5 10 3 8
n. 36 4 1 2 7 9 11 10 5 8 6 3 n. 11 1 4 9 5 3 10 8 11 6 2 7
n. 37 4 3 2 5 11 9 10 7 8 6 1 n. 12 1 7 6 8 9 10 11 5 2 4 3
n. 38 4 5 2 3 1 7 10 9 8 6 11 n. 13 1 7 6 9 4 2 11 5 10 3 8
n. 39 4 5 8 9 1 7 10 3 2 6 11 n. 14 1 8 10 3 5 2 11 4 6 9 7
n. 40 4 7 2 1 3 5 10 11 8 6 9 n. 15 1 8 10 9 11 2 5 4 6 3 7
n. 41 4 9 2 11 5 3 10 1 8 6 7 n. 16 1 8 10 9 11 5 2 7 6 3 4
n. 42 4 i:L 2 9 7 1 10 3 8 6 5 n. 17 1 9 6 4 11 2 5 3 10 8 7
n. 43 4 11 8 3 7 1 10 9 2 6 5 n. 18 2 1 6 7 4 3 11 9 10 8 5
n. 44 5 2 1 8 9 10 11 4 7 6 3 n. 19 2 1 6 7 4 9 5 3 10 8 11
n. 45 5 4 1 10 7 8 3 2 9 6 11 n. 20 2 7 6 1 4 3 11 9 10 8 5
n. 46 5 4 7 10 1 8 9 2 3 6 11 n. 21 2 7 6 1 4 9 5 3 10 8 11
n. 47 5 4 10 1 7 8 3 11 9 6 2 n. 22 3 1 5 2 11 10 9 8 6 7 4
n. 48 5 6 2 9 10 1 7 3 8 11 4 n. 23 3 1 6 4 7 2 9 11 10 8 5
n. 49 5 6 3 2 9 8 1 10 7 4 11 n. 24 3 1 6 7 8 10 9 11 2 5 4
n. 50 5 6 4 11 2 9 7 1 10 3 8 n. 25 3 4 2 5 11 10 9 8 6 7 1
n. 51 5 6 4 11 8 3 7 1 10 9 2 n. 26 3 4 6 1 8 10 9 11 5 2 7
n. 52 5 6 8 3 10 1 7 9 2 1 1 4 n. 27 3 5 1 10 7 2 9 4 6 11 8
n. 53 5 6 9 2 3 8 7 10 1 4 11 n. 28 3 5 6 8 11 10 9 7 2 4 1
n. 54 5 6 11 4 1 10 7 8 3 2 9 n. 29 3 5 6 11 4 2 9 7 10 1 8
n. 55 5 6 11 4 7 10 1 8 9 2 3 n. 30 3 7 6 1 8 10 9 11 2 5 4
n. 56 5 8 1 2 3 4 11 10 7 6 9 | n. 31 3 8 6 5 4 2 9 7 1 10 11
n. 32 3 8 10 1 7 2 9 4 6 11 5
n. 33 3 11 6 5 4 2 9 7 10 1 8
n. 34 4 1 2 3 5 10 8 11 6 9 7
Knot 52 n. 35 4 1 6 8 3 10 5 2 11 7 9
n. 1 1 8 2 5 11 7 9 10 3 6 4 n. 36 4 3 2 1 7 10 8 9 6 11 5
n. 2 1 10 5 11 7 9 2 8 3 6 4 n. 37 4 3 2 1 10 7 11 9 6 8 5
n. 3 3 2 9 11 7 1 10 4 5 6 8 n. 38 4 3 6 7 2 5 11 9 10 8 1
n. 4 3 4 10 9 11 7 1 2 5 6 8 n. 39 4 5 2 11 9 10 8 1 6 7 3
n. 5 3 8 2 9 7 11 5 10 1 6 4 n. 40 4 5 2 11 9 10 8 7 6 1 3
n. 6 3 10 9 7 11 5 2 8 1 6 4 n. 41 4 7 2 9 11 10 8 5 6 3 1
n. 7 n. 42 4 7 6 8 9 10 11 2 5 1 3
4 6 1 8 2 5 11 7 9 10 3
n. 8 4 6 1 10 5 11 7 9 2 8 3 n. 43 4 9 2 7 1 10 8 3 6 5 11
n. 9 n. 44 4 9 2 7 10 1 5 3 6 8 11
4 6 3 8 2 9 7 11 5 10 1
n. 10 4 6 3 10 9 7 11 5 2 8 1 n. 45 4 9 6 1 2 11 5 3 10 8 7
n. 11 n. 46 4 11 2 5 3 10 8 1 6 7 9
4 6 7 8 2 11 5 1 3 10 9
n. 12 4 6 7 10 11 5 1 3 2 8 9 n. 47 4 11 2 5 3 10 8 7 6 1 9
n. 13 4 6 9 8 2 3 1 5 11 10 7 n. 48 5 3 6 1 4 2 7 9 10 11 8
n. 14 n. 49 5 3 6 8 1 10 7 9 2 4 11
4 6 9 10 3 1 5 11 2 8 7
n. 15 5 2 1 7 11 9 10 4 3 6 8 n. 50 5 4 2 3 1 10 7 8 6 9 11
|_n. 16 n. 51 5 4 2 9 7 1 10 11 6 3 8
5 4 10 1 7 11 9 2 3 6 8 |
n. 52 5 4 2 9 7 10 1 8 6 3 11
n. 53 5 8 3 7 9 2 4 1 6 10 11
n. 54 5 8 6 9 11 7 10 1 2 3 4
Knot 74 n. 55 5 8 9 1 3 2 4 7 6 10 11
n. 56 5 8 10 9 11 3 4 1 6 7 2
n.~T~ 1 2 6 5 8 10 9 11 3 4 7 n. 57 5 8 10 9 11 3 4 7 6 1 2
n. 2 1 2 6 11 8 10 3 5 9 4 7 n. 58 5 8 10 11 9 2 7 4 6 1 3
n. 3 1 2 11 5 3 10 8 7 6 4 9 n. 59 5 9 6 8 7 10 1 3 2 4 11
n. 4 1 3 6 4 5 2 11 9 10 8 7 n. 60 5 10 6 1 4 2 9 7 3 8 11
[ n. 5 1 3 6 5 8 10 11 9 2 7 4 n. 61 5 10 6 7 4 2 3 1 9 8 11
n. 6 1 4 2 7 9 10 11 8 6 5 3 n. 62 5 10 7 1 3 2 4 11 6 8 9
n. 7 1 4 2 9 7 3 8 5 6 11 10 n. 63 5 11 6 4 9 2 7 1 10 8 3
n. 64 5 11 6 9 8 10 7 1 2 3 4 |
281
B. All-Interval Series
Knot 75 Knot 80
1 n. 1 1 1 4 2 9 6 5 8 3 7 11 10 1 n. 1 1 2 5 3 10 8 11 6 9 7 4
n. 2 1 4 3 11 7 2 6 5 8 10 9 n. 2 1 2 5 6 8 7 11 3 4 10 9
n. 3 1 7 11 10 9 8 6 5 2 4 3 n. 3 1 4 2 9 7 10 11 6 8 5 3
n. 4 1 1 0 8 3 6 5 11 7 2 9 4 n. 4 1 10 5 6 9 2 8 3 7 11 4
n. 5 2 1 5 9 4 7 6 3 10 8 11 n. 5 2 1 4 6 10 11 7 3 8 5 9
n. 6 2 1 5 9 4 10 6 3 7 11 8 n. 6 2 1 5 9 4 7 3 6 10 11 8
n. 7 2 3 10 5 1 4 6 9 7 11 8 n. 7 2 1 10 6 4 11 7 3 8 5 9
n. 8 2 5 1 3 6 4 7 11 10 9 8 n. 8 2 5 10 6 9 1 4 3 11 7 8
n. 9 2 5 1 9 6 10 4 3 11 7 8 n. 9 2 7 4 6 10 5 1 9 8 11 3
n. 10 2 7 11 3 4 1 6 9 10 8 5 n. 10 2 7 10 6 4 5 1 9 8 11 3
n. 11 2 7 11 3 4 10 6 9 1 5 8 n. 11 2 7 11 3 4 1 9 6 10 5 8
n. 12 2 8 3 7 11 4 6 10 1 5 9 n. 12 2 11 10 6 3 7 4 9 5 1 8
n. 13 2 8 3 7 11 10 6 4 1 5 9 n. 13 3 1 4 6 7 2 11 9 10 8 5
n. 14 2 8 9 1 5 4 6 10 7 11 3 n. 14 3 2 8 9 1 5 4 6 7 10 11
n. 15 2 8 9 1 5 10 6 4 7 11 3 n. 15 3 5 8 6 11 10 7 9 2 4 1
n. 16 2 9 10 11 7 4 6 3 1 5 8 n. 16 3 7 4 9 5 1 2 6 8 11 10
n. 17 2 11 7 3 6 10 4 9 5 1 8 n. 17 3 7 4 9 5 1 8 6 2 11 10
n. 18 2 11 7 9 6 4 1 5 10 3 8 n. 18 3 10 4 9 5 1 8 6 11 2 7
n. 19 3 1 5 2 6 8 9 10 11 7 4 n. 19 3 11 8 9 1 5 4 6 10 7 2
n. 20 3 1 5 8 6 2 9 10 11 7 4 n. 20 3 11 8 9 1 5 10 6 4 7 2
n. 21 3 2 4 7 6 10 5 1 9 8 11 n. 21 4 1 2 6 9 5 8 3 7 11 10
n. 22 3 4 2 5 6 8 9 10 11 7 1 n. 22 4 1 3 6 5 8 10 9 11 2 7
n. 23 3 5 1 4 6 10 9 2 7 11 8 n. 23 4 5 1 9 8 2 3 6 11 10 7
n. 24 3 5 1 10 6 4 9 2 7 11 8 n. 24 4 5 1 9 8 11 3 6 2 7 10
n. 25 3 7 11 2 6 8 1 5 9 4 10 n. 25 4 7 2 6 3 11 8 9 1 5 10
n. 26 3 7 11 8 6 2 1 5 9 4 10 n. 26 4 7 9 6 11 8 10 3 5 2 1
n. 27 3 8 10 1 6 4 9 2 7 11 5 n. 27 4 11 7 3 8 2 9 6 5 10 1
n. 28 3 10 8 11 6 2 1 5 9 4 7 n. 28 4 11 7 3 8 5 9 6 2 1 10
n. 29 3 11 7 4 6 10 5 1 9 8 2 n. 29 5 2 1 6 9 10 4 3 11 7 8
n. 30 3 11 7 10 6 4 5 1 9 8 2 n. 30 5 8 10 9 11 2 7 6 4 1 3
n. 31 4 1 5 3 6 8 11 7 2 9 10 n. 31 5 10 1 3 2 4 7 6 9 11 8
n. 32 4 1 5 9 6 2 8 3 7 11 10 n. 32 5 10 1 6 4 11 7 3 8 2 9 |
n. 33 4 1 5 10 3 2 6 8 11 7 9
n. 34 4 1 5 10 3 8 6 2 11 7 9
n. 35 4 3 2 1 5 8 6 9 11 7 10
n. 36 4 3 2 1 5 11 6 9 8 10 7
n. 37 4 5 1 9 8 2 6 3 11 7 10 Knot 81
n. 38 4 7 11 3 6 2 8 9 1 5 10
1 n. 1 2 3 8 6 4 11 10 7 1 5 9
n. 39 4 7 11 9 6 8 5 1 2 3 10
n. 2 2 5 4 6 8 9 10 1 7 11 3
n. 40 4 7 11 10 9 2 6 8 5 1 3
n. 3 2 9 8 6 4 5 10 1 7 11 3
n. 41 4 7 11 10 9 8 6 2 5 1 3
n. 4 2 11 4 6 8 3 10 7 1 5 9
n. 42 4 9 2 7 11 5 6 3 8 10 1
n. 5 3 7 11 5 2 1 8 6 4 9 10
n. 43 4 9 2 7 11 8 6 3 5 1 10
n. 6 3 7 11 5 2 9 4 6 8 1 10
n. 44 4 11 7 3 8 2 6 9 5 1 10
n. 7 3 11 7 1 10 5 4 6 8 9 2
n. 45 5 2 4 3 6 1 7 11 10 9 8
n. 8 3 11 7 1 10 9 8 6 4 5 2
n. 46 5 8 3 7 11 10 6 1 4 2 9
n. 9 4 3 2 11 5 1 9 6 10 7 8
n. 47 5 8 10 9 6 1 4 3 11 7 2
n. 10 4 3 10 6 9 1 5 1 1 2 7 8
n. 48 5 11 7 2 9 4 6 1 10 8 3 |
n. 11 4 5 2 6 3 11 7 1 10 9 8
n. 12 4 5 10 1 7 11 3 6 2 9 8
n. 13 4 9 2 5 11 7 3 6 10 1 8
n. 14 4 9 10 6 3 7 11 5 2 1 8
n. 15 4 11 2 6 9 5 1 7 10 3 8
n. 16 4 11 10 7 1 5 9 6 2 3 8
282
B. All-Interval Series
Knot 87 Knot 89
1 n. 1 1 3 6 7 10 89 11 254 n. 1 1 2 4 9 10 7 11 3 6 5 8
n. 2 1 9 4 5 2 7 11 863 10 n. 2 1 4 6 5 10 892 1173
n. 3 110 8 3 5 2 11 4 6 7 9 n. 3 37 11 298 10 5 6 4 1
n. 4 2 3 6 4 7 11 10 1895 n. 4 3 11 7 10 9 4 2 1 6 8 5
n. 5 2 9 6 4 1 5 10 783 11 n. 5 4 1 6 3 7 11 298 10 5
n. 6 3 1 6 4 5 2 11 98 10 7 n. 6 4 7 6 9 1 5 2 3 8 10 11
n. 7 3 2 6 5 9 8 1 10 11 74 n. 7 5 8 6 1 2 4 9 10 7 11 3
n. 8 3 5 6 8 1 10 7 9 4 2 11 n. 8 5 10 8 9 2 11 7 3 6 1 4
n. 9 3 10 6 1 9 4 5 2 7 11 8
n. 10 415 10 783 11 629
n. 11 452 11 98 10 7 6 3 1
n. 12 47 11 10 1 8 9 5 6 2 3
n. 13 4 11 2 5 3 8 10 1 6 9 7
n. 14 5 2 4 3 1 10 786 11 9
n. 15 536 11 2 4 9 7 10 18 Knot 96
n. 16 5 9 8 1 10 11 7 4 6 3 2 |
1 n. 1 1 3 6 5 11 7 8 2 4 9 10
n. 2 1 7 11 4 10 8 9 2 6 5 3
n. 3 1 9 6 5 2 8 10 3 7 11 4
n. 4 1 10 4 2 3 11 7 8 6 5 9
Knot 95 n. 5 2 1 5 9 11 7 10 4 6 3 8
n. 6 2 3 6 4 10 7 11 9 5 1 8
1 n. 1 1 2 4 9 7 10 11 6 3 5 8 n. 7 236 10 7 11 4 1 5 9 8
n. 2 1 2 5 3 8 10 11 6 4 7 9 n. 8 238 10 4 5 1 7 6 9 11
n. 3 1 2 5 6 3 4 10 97 11 8 n. 9 23 11 7 4 5 1 10 698
n. 4 1 4 2 8 5 3 11 6 10 79 n. 10 2 5 1 8 11 7 3 4 6 10 9
n. 5 1456 10 982 11 73 n. 11 2 5 1 8 11 73 10 649
n. 6 1 8 5 6 3 11 7 10 492 n. 12 27 11 351 10 4 6 9 8
n. 7 1 9 7 4 10 8 11 6 3 5 2 n. 13 2 9 5 1 4 11 7 10 638
n. 8 1 10 5 6 4 7 11 9 2 8 3 n. 14 2 9 6 4 10 153 11 78
n. 9 234 10 1596 11 87 n. 15 296 10 1 5 4 7 11 38
n. 10 2 5 3 6 11 8 10 4 7 9 1 n. 16 298 10 4 11 7 1 6 3 5
n. 11 294 10 7 11 3 6 5 8 1 n. 17 2 11 7 8 5 1 9 4 6 10 3
n. 12 2 11 9 6 5 8 10 4 1 3 7 n. 18 2 11 7 8 5 1 9 10 643
n. 13 3 1 4 6 5 10 89 11 27 n. 19 316 10 9 4 2 8 7 11 5
n. 14 31 10 659 11 8 2 4 7 n. 20 3 2 6 8 1 5 9 11 7 10 4
n. 15 34 10 97 11 8 6 1 2 5 n. 21 3 2 6 8 9 5 1 4 11 7 10
n. 16 3 5 2 6 1 9 7 4 10 8 11 n. 22 346 10 5 1 9 7 11 28
n. 17 3 5 8 6 1 2 4 9 7 10 11 n. 23 346 10 9 1 5 8 7 11 2
n. 18 37 11 289 10 6 5 4 1 n. 24 3 5 6 2 9 8 10 4 11 71
n. 19 3 8 2 9 11 7 4 6 5 10 1 n. 25 3 7 6 4 5 1 9 10 82 11
n. 20 3 11 7 10 4 9 2 6 1 8 5 n. 26 3 8 6 2 1 5 9 11 7 10 4
n. 21 4 1 3 6 7 2 11 98 10 5 n. 27 3 8 6 2 9 5 1 4 11 7 10
n. 22 4 1 5 3 2 8 9 6 7 10 11 n. 28 3 10 6 4 5 1 9 7 11 28
n. 23 4 7 9 6 1 2 5 3 8 10 11 n. 29 3 10 6 4 9 1 5 8 7 11 2
n. 24 47 11 9 2 8 3 6 1 10 5 n. 30 3 11 6 8 1 5 9 2 4 10 7
n. 25 5 2 1 6 8 11 79 10 43 n. 31 4 3 6 2 11 7 8 5 1 9 10
n. 26 5 4 1 6 3 7 11 289 10 n. 32 4 3 6 8 2 11 7 9 1 5 10
n. 27 5 8 1 6 2 9 4 10 7 11 3 n. 33 4 3 7 11 8 1 5 2 6 9 10
n. 28 58 10 4 1 3 7 6 2 11 9 n. 34 4 5 1 9 7 11 2 8 6 3 10
n. 29 59 11 8 2 4 7 6 3 1 10 n. 35 4 5 1 9 10 82 11 637
n. 30 5 10 1 3 4 2 7 6 8 11 9 n. 36 4 9 1 5 8 7 11 263 10
n. 31 5 10 1 6 3 8 2 9 11 74 n. 37 4 9 6 2 5 1 8 11 73 10
L_n. 32 5 10 89 11 2 7 6 3 1 4 | n. 38 4 9 6 8 2 5 1 3 7 11 10 |
283
B. All-Interval Series
1 n. 39 4 11) 1 5 3 11 7 2 6 8 9 n. 41 5 8 10 3 2 4 7 6 1 9 11
n. 40 4 11) 1 5 3 11 7 8 6 2 9 n. 42 5 8 10 9 2 4 1 6 7 3 11
n. 41 4 10 7 11 9 5 1 2 6 8 3 n. 43 5 8 10 9 2 4 7 6 1 3 11
n. 42 4 10 7 11 9 5 1 8 6 2 3 n. 44 5 9 1 6 7 4 2 3 10 8 11
n. 43 4 11 7 3 1 5 2 8 6 9 10 n. 45 5 9 7 6 1 4 2 3 10 8 11
n. 44 4 11 7 3 10 8 2 5 6 9 1 n. 46 5 10 8 11 4 2 3 6 7 1 9
n. 45 5 2 8 10 3 7 11 4 6 1 9 n. 47 5 11 3 6 7 2 4 1 8 10 9
n. 46 5 3 6 1 7 11 4 10 8 9 2 n. 48 5 11 9 6 7 8 10 1 2 4 3 |
n. 47 5 9 6 1 10 4 2 3 11 7 8
n. 48 5 11 7 8 2 4 9 10 6 1 3 |
284
B. All-Interval Series
285
B. All-Interval Series
Knot 131
1 n. 1 1 1 2 4 7 9 5 6 11 8 3 10 1 n. 55 3 8 2 1 5 9 6 7 4 11 10
n. 2 1 2 4 7 9 11 6 5 8 3 10 n. 56 3 8 5 10 6 11 2 4 9 7 1
n. 3 1 2 4 10 3 8 6 11 5 9 7 n. 57 3 8 9 2 6 1 4 10 7 11 5
n. 4 1 2 4 10 9 8 6 5 11 3 7 n. 58 3 8 9 5 1 2 6 7 4 10 11
n. 5 1 2 7 4 6 3 8 10 11 5 9 n. 59 3 8 10 4 7 2 6 1 9 5 11
n. 6 1 2 8 5 6 10 11 7 3 4 9 n. 60 3 8 10 11 5 9 6 1 2 7 4
n. 7 1 3 5 8 10 11 6 2 9 4 7 n. 61 3 10 4 11 7 9 6 5 2 1 8
n. 8 1 3 11 5 6 8 9 10 4 2 7 n. 62 3 10 8 1 7 9 6 11 4 5 2
n. 9 1 4 3 2 6 5 10 8 11 9 7 n. 63 3 11 5 10 8 9 6 4 1 2 7
n. 10 1 4 5 10 6 9 8 2 7 1 1 3 n. 64 3 11 7 2 8 9 6 10 5 4 1
n. 11 1 4 9 2 6 11 10 8 5 3 7 n. 65 4 1 2 7 6 3 11 5 10 8 9
n. 12 1 4 10 5 6 9 8 3 11 7 2 n. 66 4 3 2 8 10 5 6 11 9 1 7
n. 13 1 4 10 7 11 5 6 3 8 9 2 n. 67 4 3 2 8 10 11 6 5 9 1 7
n. 14 1 7 3 5 6 11 10 8 2 9 4 n. 68 4 5 10 1 6 9 11 7 8 2 3
n. 15 1 7 3 11 6 5 10 8 2 9 4 n. 69 4 7 2 1 6 9 5 11 10 8 3
n. 16 1 7 3 11 6 10 5 8 2 4 9 n. 70 4 9 2 8 10 5 6 11 3 7 1
n. 17 1 7 9 4 2 11 6 10 5 8 3 n. 71 4 9 2 8 10 11 6 5 3 7 1
n. 18 1 7 11 2 8 5 6 10 9 4 3 n. 72 4 11 10 7 6 3 5 1 8 2 9
n. 19 1 8 7 10 6 9 2 4 11 5 3 n. 73 5 2 1 8 6 3 10 4 11 7 9
n. 20 1 9 5 11 6 3 8 10 4 7 2 n. 74 5 2 4 3 1 7 6 9 8 11 10
n. 21 1 9 5 11 6 8 3 10 4 2 7 n. 75 5 3 1 4 2 7 6 10 9 8 11
n. 22 1 9 11 8 10 5 6 2 3 4 7 n. 76 5 3 7 1 6 4 9 2 8 10 11
n. 23 1 10 5 4 6 3 2 8 7 11 9 n. 77 5 4 11 2 6 9 10 8 7 1 3
n. 24 1 10 8 3 5 11 6 9 4 7 2 n. 78 5 8 1 2 6 9 4 10 11 7 3
n. 25 2 1 4 3 6 5 11 9 8 10 7 n. 79 5 8 2 1 6 9 4 3 7 11 10
n. 26 2 1 4 10 8 9 6 5 11 3 7 n. 80 5 8 2 11 7 1 6 3 4 9 10
n. 27 2 1 5 9 8 3 6 11 10 4 7 n. 81 5 8 3 10 6 1 2 4 7 9 11
n. 28 2 1 8 5 6 3 7 11 10 4 9 n. 82 5 8 9 10 6 7 2 4 1 3 11
n. 29 2 3 4 1 6 7 9 11 8 10 5 n. 83 5 9 1 7 6 3 4 2 8 11 10
n. 30 2 3 4 7 6 1 9 11 8 10 5 n. 84 5 9 1 7 6 4 3 2 8 10 11
n. 31 2 3 8 9 6 11 5 1 10 4 7 n. 85 5 9 7 4 2 1 6 10 3 8 11
n. 32 2 5 4 11 6 9 7 1 8 10 3 n. 86 5 10 4 1 6 2 7 11 3 8 9
n. 33 2 7 4 9 6 11 5 3 8 10 1 n. 87 5 10 8 2 3 4 6 7 1 9 11
n. 34 2 7 4 10 8 3 6 11 5 9 1 n. 88 5 10 8 2 9 4 6 1 7 3 11
n. 35 2 7 8 11 6 9 1 5 10 4 3 n. 89 5 10 8 11 9 1 6 7 4 3 2
n. 36 2 7 11 3 8 9 6 5 10 4 1 n. 90 5 10 8 11 9 7 6 1 4 3 2
n. 37 2 9 4 1 6 7 3 5 8 10 11 n. 91 5 10 11 8 6 3 4 2 7 1 9
n. 38 2 9 4 7 6 1 3 5 8 10 11 n. 92 5 11 3 1 6 7 2 4 10 9 8
n. 39 2 9 8 3 6 5 11 7 10 4 1 n. 93 5 11 3 7 6 1 2 4 10 9 8
n. 40 2 11 4 5 6 3 1 7 8 10 9 n. 94 5 11 3 7 6 2 1 4 10 8 9
n. 41 3 1 5 4 10 9 6 8 7 2 11 n. 95 5 11 7 10 4 1 6 2 9 8 3
n. 42 3 1 7 8 10 9 6 2 11 4 5 n. 96 5 11 9 8 10 7 6 2 1 4 3 |
n. 43 3 2 4 5 11 9 6 7 8 1 10
n. 44 3 2 8 7 11 9 6 1 10 5 4
n. 45 3 4 1 2 6 7 10 8 9 11 5
n. 46 3 4 2 7 1 9 6 5 10 11 8
n. 47 3 4 2 8 11 10 6 5 9 1 7
n. 48 3 4 9 1 5 10 6 11 8 2 7
n. 49 3 4 9 10 6 5 8 2 1 1 7 1
n. 50 3 4 10 5 1 9 6 11 8 7 2
n. 51 3 5 1 8 2 9 6 4 11 10 7
n. 52 3 5 11 4 2 9 6 10 7 8 1
n. 53 3 7 1 2 4 9 6 8 5 10 11
n. 54 | 3 7 11 10 4 9 6 2 1 8 5 |
B. All-Interval Series
287
B. All-Interval Series
288
B. All-Interval Series
Knot 152
1 n. 1 1 3 11 7 4 5 2 8 6 9 10 1 n. 33 4 7 10 1 9 8 2 3 6 1 1 5
n. 2 1 9 6 7 3 5 8 2 4 11 10 n. 34 4 10 1 7 9 2 11 3 6 8 5
n. 3 1 9 11 8 2 4 5 10 6 7 3 n. 35 4 10 7 1 3 2 5 9 6 8 11
n. 4 2 1 8 10 4 7 9 5 6 3 11 n. 36 5 3 6 7 4 10 11 9 8 1 2
n. 5 2 3 6 4 10 7 8 5 1 9 11 n. 37 5 3 8 7 2 9 4 1 6 10 11
n. 6 2 7 8 10 4 1 3 1 1 6 9 5 n. 38 541 10 783 11 692
n. 7 2 9 6 4 10 1 8 11 7 3 5 n. 39 5 8 2 1 3 4 11 10 679
n. 8 3 2 6 11 9 1 5 8 7 10 4 n. 40 5 8 6 3 11 2971 10 4
n. 9 3 7 6 10 5 4 2 8 11 9 1 n. 41 589 10 11 4 3 1 6 7 2
n. 10 3 10 6 7 9 5 1 4 11 2 8 n. 42 589 10 11 4 3 7 6 1 2
n. 11 3 11 6 2 1 8 10 4 7 9 5 n. 43 5 9 8 1 2 3 4 7 6 10 11
n. 12 4 10 1 8 11 7 3 5 6 2 9 n. 44 5981 10 74 11 623
n. 13 4 10 7 8 5 1 9 11 6 2 3 n. 45 5 10 6 1 4 9 2 7 8 3 11
n. 14 5 3 7 11 8 1 10 4 6 9 2 n. 46 5 10 6 7 4 3 2 1 8 9 11
n. 15 5 9 6 11 3 1 4 10 8 7 2 n. 47 5 11 6 3 2 8 9 1 10 74
n. 16 5 9 7 4 10 8 1 2 6 11 3 | n. 48 5 11 641 10 7 3 8 2 9 |
289
B. All-Interval Series
1 n. 33 3 8 6 5 4 11 7 1 10 9 2 1 n. 20 4 3 7 9 11 10 5 8 6 2 1
n. 34 3 8 9 2 7 11 5 10 6 1 4 n. 21 4 5 6 8 3 11 9 7 2 1 10
n. 35 3 8 11 7 9 2 5 4 6 1 10 n. 22 4 7 2 1 3 5 9 8 6 11 10
n. 36 3 10 6 1 8 5 2 11 7 9 4 n. 23 4 7 3 1 2 5 10 11 6 8 9
n. 37 4 1 6 10 5 11 7 2 9 8 3 n. 24 4 9 1 3 5 10 11 2 6 8 7
n. 38 4 3 1 5 2 11 8 7 6 10 9 n. 25 4 9 1 3 5 10 11 8 6 2 7
n. 39 4 3 6 10 9 2 5 11 7 8 1 n. 26 4 9 6 8 11 3 5 10 1 2 7
n. 40 4 3 7 11 10 9 2 5 6 8 1 n. 27 4 11 6 8 9 5 3 1 2 7 10
n. 41 4 5 2 9 7 11 8 3 6 10 1 n. 28 5 4 6 10 1 2 7 9 11 3 8
n. 42 4 5 6 8 2 9 10 11 7 3 1 n. 29 5 8 1 7 10 4 9 2 6 11 3
n. 43 4 5 6 8 3 11 7 2 9 10 1 n. 30 5 9 6 11 8 7 1 10 4 3 2
n. 44 4 7 6 10 11 5 1 2 3 8 9 n. 31 5 10 6 4 1 2 7 9 11 3 8
n. 45 4 9 1 5 10 3 2 11 6 8 7 n. 32 5 10 11 2 7 9 1 4 6 3 8 |
n. 46 4 9 6 10 3 2 11 5 1 8 7
n. 47 4 9 7 11 2 5 8 1 6 10 3
n. 48 4 10 3 2 1 5 9 11 6 8 7
n. 49 4 10 9 2 7 11 3 5 6 8 1
n. 50
n. 51
4 11 2 3 1 5 8 9 6 10 7
4 11 6 8 2 3 10 5 1 9 7
Knot 172
n. 52 4 11 6 8 9 5 1 2 3 10 7 1 n. 1 1 2 8 10 11 9 5 6 3 7 4
n. 53 5 2 6 3 4 7 11 9 10 1 8 n. 2 1 4 5 6 3 8 11 7 2 9 10
n. 54 5 2 9 10 11 7 3 4 6 1 8 n. 3 1 8 5 6 2 9 10 11 7 4 3
n. 55 5 3 11 7 2 9 10 4 6 1 8 n. 4 1 9 7 2 4 10 5 6 8 11 3
n. 56 5 4 1 10 7 11 9 8 6 3 2 n. 5 2 3 10 5 1 4 9 6 7 8 11
n. 57 5 4 6 1 3 7 11 10 9 2 8 n. 6 2 3 10 5 1 7 9 6 4 11 8
n. 58 5 4 6 1 10 9 2 7 11 3 8 n. 7 2 5 4 6 3 1 7 11 10 9 8
n. 59 5 4 11 7 1 10 9 2 6 3 8 n. 8 2 9 10 11 7 1 3 6 4 5 8
n. 60 5 8 6 3 4 9 10 11 7 1 2 n. 9 2 9 10 11 7 4 3 6 1 8 5
n. 61 5 8 7 11 3 4 9 10 6 1 2 n. 10 2 11 4 6 9 7 1 5 10 3 8
n. 62 5 9 8 3 7 11 4 10 6 1 2 n. 11 3 1 7 11 10 9 2 6 8 5 4
n. 63 5 10 6 1 4 11 7 3 8 9 2 n. 12 3 1 7 11 10 9 8 6 2 5 4
n. 64 5 10 6 1 9 4 3 11 7 8 2 | n. 13 3 4 7 11 10 9 2 6 5 8 1
n. 14 3 5 11 7 2 9 4 6 10 1 8
n. 15 3 5 11 7 2 9 10 6 4 1 8
n. 16 3 7 4 6 1 2 8 10 11 9 5
Knot 159 n. 17 3 8 11 7 2 9 10 6 1 4 5
n. 18 3 11 8 6 5 10 4 2 7 9 1
1 n. 1 1 2 6 8 5 10 11 9 7 3 4
n. 19 4 1 8 6 3 5 11 7 2 9 10
n. 2 1 2 7 10 11 9 5 8 6 3 4
n. 20 4 1 9 6 11 3 5 10 8 2 7
n. 3 1 4 5 11 2 8 9 10 6 7 3
n. 21 4 3 2 1 5 11 9 6 8 7 10
n. 4 1 8 6 2 5 10 11 9 7 3 4
n. 22 4 5 2 6 8 9 10 11 7 1 3
n. 5 1 9 6 7 4 11 5 2 8 3 10
n. 23 4 5 8 6 2 9 10 11 7 1 3
n. 6 2 1 6 4 3 7 9 11 10 5 8
n. 24 4 7 3 6 5 9 11 10 8 2 1
n. 7 2 3 4 10 1 7 8 11 6 9 5
n. 25 4 7 8 6 9 11 5 1 2 3 10
n. 8 2 5 10 11 9 7 3 4 6 1 8
n. 26 4 9 2 7 11 5 3 6 8 1 10
n. 9 2 7 6 4 9 1 3 5 10 11 8
n. 27 4 11 2 6 8 3 10 5 1 7 9
n. 10 2 9 4 10 7 1 8 5 6 3 11
n. 28 4 11 8 6 2 3 10 5 1 7 9
n. 11 2 11 10 5 3 1 9 4 6 7 8
n. 29 5 4 1 6 10 9 2 7 11 8 3
n. 12 3 4 6 1 2 7 10 11 9 5 8
n. 30 5 8 1 6 3 4 7 11 10 9 2
n. 13 3 7 6 10 9 8 2 11 5 4 1
n. 31 5 9 11 10 8 2 1 6 4 7 3
n. 14 3 8 6 5 10 11 2 7 9 1 4
n. 32 5 10 4 2 7 9 1 6 3 11 8 |
n. 15 3 11 6 2 9 4 10 7 1 8 5
n. 16 4 1 2 7 9 11 3 8 6 5 10
n. 17 4 1 9 7 2 11 10 5 6 8 3
n. 18 4 3 6 8 5 9 11 10 7 2 1
n. 19 4 3 7 9 11 10 5 2 6 8 1 |
290
B. All-Interval Series
Knot 173 1 n. 15 2 11 4 3 1 10 9 6 5 8 7
n. 16 2 11 10 9 7 4 3 6 5 8 1
1 n. 1 1 3 4 11 7 9 10 6 2 5 8 n. 17 32 11 981 10 6 5 4 7
n. 2 1 9 10 11 7 3 4 6 8 5 2 n. 18 3 4 7 9 10 11 2 6 1 8 5
n. 3 2 3 5 1 8 9 11 6 4 7 10 n. 19 38 11 927 10 6541
n. 4 2 3 8 6 4 5 11 7 10 1 9 n. 20 3 10 7 9 4 5 2 6 1 8 11
n. 5 2 5 4 6 8 3 11 7 10 1 9 n. 21 4 3 7 9 11 10 1 6 2 8 5
n. 6 2 5 8 6 1 3 4 11 7 9 10 n. 22 4 9 1 3 5 10 7 6 2 8 11
n. 7 2 5 8 6 4 3 7 11 10 9 1 n. 23 4 10 1629 11 3785
n. 8 2 9 8 6 4 11 5 1 10 7 3 n. 24 4 10 7 6 2 3 5 9 1 8 11
n. 9 2 9 11 7 8 3 5 6 4 1 10 n. 25 5 2 7 9 11 3 8 6 1 4 10
n. 10 2 11 4 6 8 9 5 1 10 7 3 n. 26 5 2 8 6 1 4 11 379 10
n. 11 2 11 8 6 4 9 1 5 10 3 7 n. 27 5 4 1 6 3 8 11 927 10
n. 12 2 11 8 6 7 9 4 5 1 3 10 n. 28 5 4 7 6 3 2 11 981 10
n. 13 3 7 10 1 5 9 8 6 4 11 2 n. 29 5 8 1 6 2 11 10 9 7 4 3
n. 14 3 7 10 1 5 11 4 6 8 9 2 n. 30 5 8 2 6 1 10 11 9 7 3 4
n. 15 3 11 2 5 1 7 8 6 4 9 10 n. 31 5 8 7 3 11 9261 10 4
n. 16 3 11 2 5 1 9 4 6 8 7 10 n. 32 5 8 7 6 2 11 431 10 9 |
n. 17 4 1 10 6 2 9 11 7 8 3 5
n. 18 4 1 10 6 5 9 2 7 11 3 8
n. 19 4 3 7 11 2 5 9 6 10 1 8
n. 20
n. 21
4 3 7 11 10 9 1 6 2 5 8
4 3 10 6 9 5 2 1 1 7 1 8
Knot 183
n. 22 4 5 2 6 9 1 10 7 11 3 8 n. 1 1 4 11 3 7 8 5 6 2 9 10
n. 23 4 5 11 7 10 1 9 6 2 3 8 n. 2 1 9 5 6 2 8 7 3 11 4 10
n. 24 4 7 10 6 2 3 5 1 8 9 11 n. 3 2 3 10 6 7 4 5 9 1 8 11
n. 25 4 7 10 6 11 3 2 1 5 9 8 n. 4 2 8 1 9 5 4 10 6 7 3 11
n. 26 4 9 1 5 2 11 3 6 10 7 8 n. 5 2 8 7 3 11 4 10 6 1 9 5
n. 27 4 9 1 5 10 3 7 6 2 11 8 n. 6 2 9 10 6 1 4 11 3 7 8 5
n. 28 4 9 10 6 3 11 2 5 1 7 8 n. 7 5 8 7 3 11 4 1 6 10 9 2
n. 29 4 11 2 6 3 7 10 1 5 9 8 n. 8 5 9 1 6 10 4 11 3 7 8 2
n. 30 4 11 5 1 10 7 3 6 2 9 8
n. 31 5 3 8 7 11 9 2 6 10 1 4
n. 32 5 9 2 7 11 3 8 6 4 1 10 |
Knot 184
1 n. 1 134 11 2 5 9 6 10 78
n. 2 1 3 5 6 8 11 10 9 2 7 4
n. 3 13 10 52 11 9 6 4 7 8
Knot 179 n. 4 1729 10 11 5 6 8 3 4
n. 5 1 7 9 6 4 11 25 10 38
1 n. 1456 10 729 11 83
1 n. 6 1796 10 52 11 438
n. 14 10 6 5 2 7 9 1138
2 n. 7 21 10 7 8 3 4 6 5 11 9
n. 14 11 379 10 6 5 2 8
3 n. 8 2 5 4 6 11 981 10 73
n. 14 11 6 10 7 8 3 5 2 9
4 n. 9 27 10 1 8 9 4 6 11 53
n. 1 8 5 6 3 4 7 9 10 11 2
5 n. 10 2 11 4 6 5 3 8 7 10 19
n. 18 11 63 10 7 9 4 5 2
6 n. 11 3 1 7 6 8 9 4 5 2 11 10
n. 1 10 4 6 5 8 7 3 11 92
7 n. 12 3 1 7 6 8 9 10 11 254
n. 1 10 11 9 7 3 4 6 5 8 2
8 n. 13 31 10 729 11 6 4 5 8
n. 2 3 5 9 1 8 11 64 10 7
9 n. 14 352 11 10 9 7 6 8 1 4
n. 2 5 4 9 7 10 36 11 81
10 n. 15 35 11 6 4 9 2 7 10 18
n. 25 10 3 1 4 9 6 11 87
11 n. 16 35 11 6 4 9 8 1 10 72
n. 2 8 5 6 4 3 7 9 11 10 1
12 n. 17 37 10 189 11 6 4 5 2
n. 28 11 6 4 9 1 3 5 10 7
13 n. 18 3 11 2 5 4 9 7 6 8 1 10
[n. 14 29 11 3 7 8 5 6 4 10 1 n. 19 4 1 2 3 10 586 11 97 |
291
B. All-Interval Series
1 n. 20 4 1 8 6 7 9 10 11 2 5 3 1 n. 11 2 3 5 11 6 4 1 8 9 10 7
n. 21 4 3 2 1 10 7 8 6 9 11 5 n. 12 2 5 103 8 11 6 4 9 1 7
n. 22 4 3 8 6 5 11 10 9 2 7 1 n. 13 2 7 4 3 6 5 8 9 11 10 1
n. 23 4 3 8 7 10 1 2 6 9 11 5 n. 14 2 7 10 3 6 11 8 9 5 4 1
n. 24 4 5 2 6 3 7 10 1 8 9 11 n. 15 2 9 11 5 6 4 7 8 3 10 1
n. 25 4 5 2 11 10 9 8 6 7 1 3 n. 16 2 11 10 9 8 5 6 4 3 7 1
n. 26 4 5 8 6 3 1 10 7 2 9 11 n. 17 3 2 11 10 6 5 8 1 9 4 7
n. 27 4 7 2 9 10 11 8 6 5 3 1 n. 18 3 4 7 2 6 1 10 11 9 8 5
n. 28 4 7 8 6 1 3 10 5 2 11 9 n. 19 3 8 11 10 6 5 2 7 9 4 1
n. 29 4 9 2 7 10 1 8 6 3 5 11 n. 20 3 10 7 2 6 1 4 5 9 8 11
n. 30 4 9 8 1 10 7 2 6 3 5 11 n. 21 4 1 8 9 10 7 6 2 3 5 11
n. 31 4 9 8 6 11 5 10 3 2 1 7 n. 22 4 3 7 1 6 2 11 10 9 8 5
n. 32 4 11 2 5 10 3 8 6 1 7 9 n. 23 4 7 8 3 10 1 6 2 9 11 5
n. 33 4 11 2 6 9 1 10 7 8 3 5 n. 24 4 9 1 7 6 2 5 10 3 8 11
n. 34 4 11 8 6 9 7 10 1 2 3 5 n. 25 5 2 3 4 11 8 6 1 7 9 10
n. 35 5 3 1 6 4 7 2 9 10 11 8 n. 26 5 2 7 9 4 1 6 3 8 11 10
n. 36 5 3 2 1 10 7 9 6 8 11 4 n. 27 5 8 1 9 4 7 6 3 2 11 10
n. 37 5 3 8 7 10 1 9 6 2 11 4 n. 28 5 8 3 11 4 7 6 2 1 10 9
n. 38 5 1 1 9 6 2 1 10 7 8 3 4 n. 29 5 8 9 10 11 2 6 1 7 3 4
n. 39 5 11 9 6 8 7 10 1 2 3 4 n. 30 5 8 9 11 10 1 6 2 7 4 3
n. 40 5 11 10 9 2 7 1 6 4 3 8 | n. 31 5 11 3 8 6 1 4 9 2 7 10
n. 32 5 11 9 2 6 1 10 3 8 7 4 |
Knot 186
1 n. 1 1 8 7 3 10 6 9 2 5 11 4 Knot 207
n. 2 1 9 7 2 8 6 11 3 5 10 4
1 n. 1 1 2 8 5 6 9 7 3 4 11 10
n. 3 1 1 0 5 3 8 6 11 2 7 9 4
n. 2 1 4 9 7 2 11 6 3 8 5 10
n. 4 2 3 11 4 5 6 8 7 1 10 9
n. 3 1 4 10 5 6 2 7 8 3 11 9
n. 5 2 9 5 4 11 6 8 1 7 10 3
n. 4 1 10 5 3 8 11 6 2 7 4 9
n. 6 3 2 11 5 4 6 7 8 1 9 10
n. 5 2 1 4 3 6 7 10 11 9 8 5
n. 7 3 10 7 1 8 6 11 4 5 9 2
n. 6 2 1 8 9 5 3 6 7 4 10 11
n. 8 4 3 1 2 5 6 8 9 11 10 7
n. 7 2 7 4 9 6 1 10 5 3 8 11
n. 9 4 5 11 2 3 6 10 9 1 8 7
n. 8 2 7 8 3 11 9 6 1 4 10 5
n. 10 4 9 7 2 11 6 8 3 5 10 1
n. 9 3 1 9 4 5 10 6 7 2 8 11
n. 11 4 10 5 3 11 6 8 2 7 9 1
n. 10 3 4 1 2 6 5 8 9 11 10 7
n. 12 4 10 11 9 5 6 8 2 1 3 7
n. 11 3 5 9 8 1 2 6 11 10 4 7
n. 13 4 11 5 2 9 6 10 3 7 8 1
n. 12 3 8 5 10 6 1 4 9 7 2 11
n. 14 5 2 1 3 4 6 7 10 11 9 8
n. 13 5 2 1 3 4 7 6 10 11 8 9
n. 15 54 11 3 2 6 9 10 178
n. 14 5 8 2 1 6 10 11 4 3 7 9
n. 16 59 11 10 4 6 7 3 1 2 8 |
n. 15 5 8 9 11 10 7 6 3 4 1 2
n. 16 5 10 4 1 6 9 11 3 8 7 2 |
Knot 206
n. 1 1 4 3 7 8 11 6 10 5 2 9 Knot 208
n. 2 1 4 5 9 8 11 63 10 72
n. 3 1 4 9 2 7 10 65 11 38 n. 1 172 11 6 8 5 3 10 94
n. 4 1 4 9 7 2 5 6 10 11 83 n. 2 1 7 9 2 3 11 685 10 4
n. 5 1 7 3 4 6 5 8 9 10 11 2 n. 3 1 8 5 3 6 4 7 10 11 92
n. 6 179 10 6 5 2 3 4 11 8 n. 4 18 11 36 10 7 4 5 9 2
n. 7 1 10 3 8 7 4 6 5 11 92 n. 5 1 9 5 4 7 3 6 10 11 82
n. 8 1 10 11 9 8 5 6 3 4 7 2 n. 6 19 10 35 1168274
n. 9 2 1 4 9 6 11 835 10 7 n. 7 19 11 10 7 3 6 4 5 8 2
n. 10 21 10 9 6 5 8 3 11 47 n. 8 110 5 11 6 8 3 2 9 7 4
292
B. All-Interval Series
1 n. 9 2 1 4 10 6 11 3 7 8 5 9 1 n. 11 2 7 4 3 11 8 6 59 1 10
n. 10 2 3 5 10 1 4 6 9 11 8 7 n. 12 2 7 4 9 5 8 6 11 3 1 10
n. 11 2 3 11 41 10 6 9 5 8 7 n. 13 2 11 3 7 6 4 1 9 8 5 10
n. 12 2 5 8 7 3 10 6 11 4 1 9 n. 14 2 11 9 1 6 4 7 3 8 5 10
n. 13 2 7 4 1 0 6 5 9 1 8 11 3 n. 15 3 7 1 8 6 4 9 2 5 11 10
n. 14 2 8 5 4 6 3 7 10 11 9 1 n. 16 3 7 9 4 6 8 1 2 5 11 10
n. 15 2 8 5 1069 14 11 3 7 n. 17 3 11 5 4 6 8 9 10 1 7 2
n. 16 28 11 4 6 9 1 10 5 3 7 n. 18 3 11 9 8 6 4 5 10 1 7 2
n. 17 2 8 11 10 6 3 7 4 5 9 1 n. 19 4 1 9 8 5 10 6 2 11 3 7
n. 18 2 9 5 4 7 10 6 3 11 81 n. 20 4 1 9 8 11 10 6 2 5 3 7
n. 19 29 11 10 7 4 6 3 5 8 1 n. 21 4 3 1 9 6 10 5 11 2 7 8
n. 20 2 11 8 1 9 10 6 5 4 7 3 n. 22 4 3 2 11 5 10 6 9 1 7 8
n. 21 3 1 4 5 6 10 9 7 2 11 8 n. 23 4 5 10 1 7 2 6 3 11 9 8
n. 22 3 5 8 1 6 2 9 11 10 7 4 n. 24 4 5 11 3 6 2 7 1 10 9 8
n. 23 3 7 4 5 6 10 9 1 8 11 2 n. 25 4 7 3 8 5 10 6 2 11 91
n. 24 3 7 4 5 9 1 6 2 8 11 10 n. 26 4 7 3 8 11 10 6 2 5 9 1
n. 25 3 7 10 11 9 1 6 2 8 5 4 n. 27 4 9 2 5 11 10 6 3 7 1 8
n. 26 3 11 2 7 9 5 6 10 4 1 8 n. 28 4 9 7 3 6 10 11 5 2 1 8
n. 27 3 11 8 1 6 2 9 5 4 7 10 n. 29 4 11 5 9 6 2 1 7 10 3 8
n. 28 3 11 8 1 9 5 6 10 4 7 2 n. 30 4 11 10 7 1 2 6 9 5 3 8
n. 29 4 1 2 8 6 5 11 9 10 3 7 n. 31 5 9 1 10 6 2 7 4 3 11 8
n. 30 4 1 3 2 9 8 6 5 11 10 7 n. 32 5 9 7 10 6 2 1 4 3 11 8 |
n. 31 41 10 5 3 2 6 7 8 11 9
n. 32 4 3 10 9 11 8 6 5 2 1 7
n. 33 4 5 8 2 6 1 9 11 10 73
n. 34 4 7 2 8 6 11 53 10 91 Knot 213
n. 35 4 7 9 2 3 8 6 11 5 10 1
n. 36 4 7 10 11 9 2 6 1 8 5 3 1 n. 1 1 4 9 6 11 3 5 8 10 7 2
n. 37 4 9 10 3 5 8 6 11 2 7 1 n. 2 1 9 7 4 2 5 10 6 11 8 3
n. 38 4 10 5 8 6 1 1 3 2 9 7 1 n. 3 1 9 10 6 5 2 8 1 1 7 4 3
n. 39 4 10 11 8 6 5 9 2 3 1 7 n. 4 1 10 4 7 11 8 3 6 5 9 2
n. 40 4 11 8 2 6 7 3 5 10 19 n. 5 2 1 10 8 11 9 5 6 3 4 7
n. 41 5 2 1 7 6 4 3 10 9 11 8 n. 6 2 3 11 6 9 8 5 1 4 10 7
n. 42 5 4 1 3 6 8 11 2 7 9 10 n. 7 2 7 10 8 5 3 11 6 9 4 1
n. 43 5 4 7 3 6 2 11 8 1 9 10 n. 8 2 9 5 6 3 8 11 7 4 10 1
n. 44 5 9 1 8 11 3 6 2 7 4 10 n. 9 3 4 7 6 2 1 10 8 11 9 5
n. 45 5 9 2 3 1 7 6 4 10 11 8 n. 10 3 4 7 11 8 2 5 6 10 9 1
n. 46 5 9 7 2 11 3 6 8 1 4 10 n. 11 3 8 11 6 10 5 2 4 7 9 1
n. 47 5 11 9 10 3 7 6 4 1 2 8 n. 12 3 8 11 7 4 10 1 6 2 9 5
n. 48 5 11 10 7 6 4 1 3 2 9 8 | n. 13 5 2 8 11 7 4 3 6 1 9 10
n. 14 5 8 9 6 7 3 1 4 2 11 10
n. 15 5 9 2 6 1 10 4 7 11 8 3
n. 16 59 11 8 10 1 2 6 7 4 3 |
Knot 210
n. 1 1 9 5 2 6 10 11 8 3 7 4 Knot 214
n. 2 1 9 11 2 6 10 5 8 3 7 4
n. 3 2 1 4 3 11 8 6 5 9 7 10 n. 1 1 2 7 6 4 3 10 8 9 5 11
n. 4 2 1 4 9 5 8 6 11 3 7 10 n. 2 1 2 7 6 5 11 3 8 10 9 4
n. 5 2 1 7 10 3 8 6 4 11 5 9 n. 3 1 4 9 6 2 5 8 10 7 3 11
n. 6 2 1 7 10 11 4 6 8 3 5 9 n. 4 1 7 3 4 2 9 8 6 5 10 11
n. 7 2 5 3 7 6 4 1 9 8 11 10 n. 5 1 7 3 10 8 9 2 6 5 4 11
n. 8 2 5 9 1 6 4 7 3 8 11 10 n. 6 1 8 7 6 5 11 3 2 4 9 10
n. 9 2 7 1 10 5 4 6 8 9 11 3 n. 7 1 8 7 6 10 3 4 2 9 5 11
n. 10 2 7 1 10 9 8 6 4 5 1 1 3 n. 8 1 9 5 2 4 7 10 6 3 8 11
293
B. All-Interval Series
1 n. 9 1 9 5 8 10 7 4 6 3 2 11 1 n. 8 173 10 8 2 9 6 5 11 4
n. 10 1 109 68 5 24 73 11 n. 9 1 7 8 6 5 11 324 10 9
n. 11 238 10 9 1 7 6 5 4 11 n. 10 195 11 7 10 4 6 3 8 2
n. 12 258 10 73 11 6 1 4 9 n. 11 1976 10 524 11 83
n. 13 298 10 3 7 1 6 1145 n. 12 1 9 11 6 2 3 8 10 5 4 7
n. 14 2 1L 8 10 1 9 5 6 7 4 3 n. 13 1 9 11 8 10 4 7 6 3 5 2
n. 15 32 11 6 1 9 5 8 10 74 n. 14 21 10 8 7 4 3 6 5 9 11
n. 16 3 4 7 6 5 9 1 10 8 11 2 n. 15 238 10 5 4 1 6 7 9 11
n. 17 38 11 6 1 9 5 2 4 7 10 n. 16 238 10 5 4 7 6 1 9 11
n. 18 3 107659 142 11 8 n. 17 2 5 3 6 7 4 10 8 11 91
n. 19 41 10 8 11 3 7 6 5 2 9 n. 18 27 10 8 1 4 9 6 11 35
n. 20 43 10 895 11 6 1 2 7 n. 19 2 8 3 6 4 10 7 11 591
n. 21 47 10 8 5 9 1 6 11 23 n. 20 2 8 9 6 4 10 15 1137
n. 22 49 10 83 11 5 6 7 2 1 n. 21 298 10 11 4 1 6 7 3 5
n. 23 5 2 9 6 4 1 10 8 11 37 n. 22 298 10 11 4 7 6 1 3 5
n. 24 54 11 6 1 7 3 10 892 n. 23 2 11 9 6 1 4 10 8 5 3 7
n. 25 54 11 6 2 3 8 10 9 1 7 n. 24 31 10 6 5 9 7 4 2 8 11
n. 26 5 8 9 6 10 142 11 37 n. 25 328 10 9 1 7 6 4 5 11
n. 27 5 9 1 4 2 11 863 10 7 n. 26 3 4 7 8 10 126 11 95
n. 28 591 10 8 11 2 6 3 4 7 n. 27 34 10 6 5 9 1 7 11 28
n. 29 5 10 11 6 1 7 3 4 2 9 8 n. 28 3 5 2 6 1 9 11 8 10 47
n. 30 5 10 11 6 8 3 2 4 9 1 7 n. 29 3 8 2 6 1 9 5 11 7 10 4
n. 31 5 11 3 2 4 9 10 6 1 8 7 n. 30 38 11 425 10 6 7 9 1
n. 32 5 11 38 10 9 4 6 1 2 7 | n. 31 3 10 4 2 9 5 11 6 8 1 7
n. 32 45 11 6 3 2 8 10 9 1 7
n. 33 4 10 15 11 3 7 6 2 8 9
Knot 220 n. 34
n. 35
4 10 7 11 5 9 1 6 2 8 3
4 11 5 6 9 2 8 10 371
1 n. 1 1 2 8 6 9 5 3 11 7 4 10 n. 36 5 3 1 6 7 4 11 10 8 9 2
n. 2 1 8 2 4 11 5 9 6 7 3 10 n. 37 5 3 1 6 10 9 4 2 7 8 11
n. 3 19 10 6 7 8 2 4 5 11 3 n. 38 5 3 7 6 1 4 11 10 892
n. 4 23 11 6 9 1 7 8 10 45 n. 39 53 11 6 9 4 1 8 10 72
n. 5 2 8 5 1 9 7 3 6 4 10 11 n. 40 5 8 1 2 4 3 10 679 11
n. 6 28 11 7 3 1 9 6 4 10 5 n. 41 5 8 2 4 1 3 11 6 10 79
n. 7 2 9 5 6 3 7 1 8 10 4 11 n. 42 5 8 7 2 4 9 10 613 11
n.8 3 7 1 8 10 4 11 6 2 9 5 n. 43 5 9 1 7 11 2 8 6 3 4 10
n. 9 3 7 9 1 5 8 2 6 11 10 4 n. 44 5 9 7 4 2 8 11631 10
n. 10 3 11 5 4 2 8 7 6 10 91 n. 45 5976 10 3 4 2 1 8 11
n. 11 3 11 9 5 1 4 10 6 7 2 8 n. 46 59 11621 10 8 7 4 3
n. 12 4 10 5 6 2 8 11 7 3 1 9 n. 47 5 11 324 10 9 6 1 7 8
n. 13 4 10 11 6 2 8 5 1 9 7 3 n. 48 5 11 4 6 1 7 3 10 829 |
n. 14 54 10 8 7 1 9 6 1132
n. 15 5 9 2 6 11 4 10 8 1 7 3
n. 16 5 10 4 6 9 1 3 7 11 82 |
Knot 241
Knot 221 n. 1 1 2 5 9 6 8 3 11 7 10 4
n. 2 127 11 5 3 6 8 9 4 10
n. 1 1 3 5 6 2 9 8 10 11 47 n. 3 13 11 7 10 9 6 8 2 5 4
n. 2 1 3 5 6 11 8 7 2 4 9 10 n. 4 1 9 4 3 6 8 2 7 11 5 10
n. 3 1 3 7 6 9 8 5 4 2 11 10 n. 5 217 11 10 4 6 3 8 9 5
n. 4 13 11 6 5 8 7 2 4 9 10 n. 6 2 7 1 5 10 4 6 9 8 3 11
n. 5 145 10 8 3 2 6 1197 n. 7 2 8 3 4 6 9 7 1 5 10 11
n. 6 14 10 8 5 3 7 6 2 11 9 n. 8 2 8 9 4 6 3 1 7 11 10 5
n. 7 14 11 10 8 9 2 6 5 3 7 n. 9 317 11 10 5 6 2 8 9 4
294
B. All-Interval Series
1 n. 10 3 2 5 1 9 11 6 8 7 10 4 1 n. 9 1 9 10 8 7 4 6 5 3 2 11
n. 11 3 4 9 1 6 10 5 11 7 2 8 n. 10 1 10 9 7 6 8 5 4 2 3 11
n. 12 3 5 11 7 2 1 6 10 4 9 8 n. 11 2 5 10 8 9 11 6 1 4 3 7
n. 13 3 7 10 11 6 8 2 5 1 9 4 n. 12 2 11 10 8 3 5 6 7 4 9 1
n. 14 3 8 9 5 6 2 1 7 11 10 4 n. 13 3 1 9 7 6 5 10 4 2 11 8
n. 15 3 10 1 5 9 7 6 4 11 2 8 n. 14 3 1 10 5 6 7 9 4 2 11 8
n. 16 3 H 2 7 6 4 10 1 5 9 8 n. 15 3 5 2 1 6 11 9 8 10 7 4
n. 17 4 3 7 11 2 8 6 5 10 1 9 n. 16 3 5 9 11 6 1 2 8 10 7 4
n. 18 4 3 8 2 6 11 10 5 1 7 9 n. 17 4 1 8 10 3 7 6 5 2 9 11
n. 19 4 5 2 8 6 9 10 7 11 3 1 n. 18 4 1 10 8 2 7 6 5 3 11 9
n. 20 4 9 1 5 2 8 6 11 10 7 3 n. 19 4 1 10 8 3 5 6 7 2 11 9
n. 21 4 9 8 2 6 5 10 11 7 1 3 n. 20 4 7 8 10 9 1 6 11 2 3 5
n. 22 4 10 1 5 9 8 6 3 11 2 7 n. 21 4 7 10 8 2 1 6 11 9 5 3
n. 23 4 10 1 8 6 5 3 7 11 2 9 n. 22 4 7 10 8 9 11 6 1 2 5 3
n. 24 4 10 5 1 7 2 6 11 3 8 9 n. 23 5 2 9 11 6 4 1 8 10 3 7
n. 25 4 10 7 8 6 11 9 1 5 2 3 n. 24 5 3 2 11 6 1 9 10 8 7 4
n. 26 4 10 7 11 3 8 6 9 5 2 1 n. 25 5 3 8 10 1 4 6 9 11 2 7
n. 27 4 10 11 7 1 2 6 5 9 8 3 n. 26 5 3 8 10 11 2 6 1 9 4 7
n. 28 4 11 2 8 6 3 10 1 5 9 7 n. 27 5 3 11 9 6 4 1 10 8 2 7
n. 29 5 3 7 11 2 9 6 4 10 1 8 n. 28 5 8 3 11 6 10 1 2 4 9 7
n. 30 5 9 8 3 6 4 10 11 7 1 2 n. 29 5 9 2 4 11 8 6 1 3 10 7
n. 31 5 10 1 9 6 4 3 7 11 2 8 n. 30 5 9 8 11 6 1 3 4 2 7 10
n. 32 5 10 11 7 1 3 6 4 9 8 2 | n. 31 5 10 1 3 6 8 11 2 4 9 7
n. 32 5 10 4 2 11 8 6 3 1 9 7 |
Knot 242
1 n. 1 1 2 11 7 8 5 6 3 4 9 10
n. 2 1 3 11 5 6 8 7 2 4 10 9
n. 3 1 4 11 7 10 5 6 2 9 8 3
n. 4 1 7 3 5 6 9 2 8 10 11 4
Knot 252
n. 5 2 3 8 9 6 7 4 5 1 10 11 1 n. 1 1 4 9 7 2 8 3 6 11 5 10
n. 6 2 9 8 3 6 1 4 11 7 10 5 n. 2 1 7 2 6 11 8 3 5 10 4 9
n. 7 3 2 8 10 5 4 6 7 1 9 11 n. 3 2 1 4 6 3 7 10 11 9 5 8
n. 8 3 4 9 10 6 1 2 11 7 8 5 n. 4 2 1 7 6 3 4 10 11 9 8 5
n. 9 3 8 9 2 6 5 10 7 11 4 1 n. 5 2 5 9 11 10 7 3 6 4 1 8
n. 10 3 10 4 2 1 8 6 11 5 9 7 n. 6 2 7 1 6 9 4 10 5 3 8 11
n. 11 4 5 10 8 2 3 6 11 9 1 7 n. 7 2 7 4 6 9 1 10 5 3 11 8
n. 12 4 11 10 8 2 9 6 5 3 7 1 n. 8 2 11 3 5 10 1 9 6 4 7 8
n. 13 5 3 7 1 6 4 11 10 8 2 9 n. 9 3 4 10 11 9 8 5 6 2 1 7
n. 14 5 8 7 1 1 2 1 6 10 9 4 3 3 7 10 11 9 5 2 6 8 1 4
n. 10
n. 15 5 10 7 11 4 1 6 3 8 9 2 n. 11 3 7 10 11 9 5 8 6 2 1 4
n. 16 5 11 3 1 6 9 10 4 2 7 8 | n. 12 3 8 2 7 9 4 1 6 10 5 11
n. 13 3 11 2 7 9 1 4 6 10 5 8
n. 14 3 11 2 7 9 1 10 6 4 5 8
Knot 243 n. 15 4 1 2 6 8 5 9 11 10 7 3
n. 16 4 1 8 6 2 5 9 11 10 7 3
n. 1 1 2 5 3 6 4 7 10 8 9 11 n. 17 4 1 9 7 2 11 3 6 8 5 10
n. 2 1 2 8 10 7 4 6 3 5 9 11 n. 18 4 5 8 6 3 11 2 7 9 1 10
n. 3 1 3 4 2 5 8 6 9 7 10 11 n. 19 4 7 2 6 8 11 3 5 10 1 9
n. 4 1 3 4 2 7 10 6 5 9 8 11 n. 20 4 7 3 1 2 5 9 6 8 11 10
n. 5 1 3 7 9 6 8 5 2 4 10 11 n. 21 4 7 8 6 2 11 3 5 10 1 9
n. 6 1 3 10 7 6 5 9 2 4 11 8 n. 22 4 11 8 6 9 5 2 1 3 7 10
n. 7 1 4 3 7 6 2 5 10 8 9 11 n. 23 5 8 9 11 10 4 3 6 7 1 2
n. 8 1 9 4 7 6 5 3 8 10 11 2 n. 24 5 11 10 6 7 4 3 1 2 8 9 |
295
B. All-Interval Series
296
B. All-Interval Series
Knot 354
1 n. 1 1 3 5 2 9 6 8 7 10 4 11 1 n. 35 1 2 11 9 5 8 6 4 7 3 1 10
n. 2 1 3 10 7 8 6 9 2 5 4 11 n. 36 2 11 9 7 4 6 8 5 3 1 10
n. 3 1 8 2 5 4 6 3 10 7 9 11 n. 37 3 1 4 5 10 6 2 7 8 11 9
n. 4 1 8 7 10 3 6 4 5 2 9 11 n. 38 3 1 10 5 4 6 8 7 2 11 9
n. 5 3 2 11 4 5 6 7 9 10 1 8 n. 39 3 2 5 4 11 6 1 8 7 10 9
n. 6 3 2 11 9 7 6 5 4 10 1 8 n. 40 3 2 11 4 5 6 7 8 1 10 9
n. 7 3 10 7 8 1 6 11 9 2 5 4 n. 41 3 4 1 2 7 6 5 10 11 8 9
n. 8 3 10 7 9 11 6 1 8 2 5 4 n. 42 3 4 7 2 1 6 11 10 5 8 9
n. 9 4 5 2 8 1 6 11 9 7 10 3 n. 43 3 5 2 1 8 6 4 11 10 7 9
n. 10 4 5 2 9 11 6 1 8 7 10 3 n. 44 3 5 8 1 2 6 10 11 4 7 9
n. 11 4 11 2 3 5 6 7 8 1 10 9 n. 45 3 7 4 11 10 6 2 1 8 5 9
n. 12 4 11 2 8 7 6 5 3 1 10 9 n. 46 3 7 10 11 4 6 8 1 2 5 9
n. 13 5 3 1 10 9 6 4 11 2 8 7 n. 47 3 8 5 10 11 6 1 2 7 4 9
n. 14 5 3 2 11 4 6 9 10 1 8 7 n. 48 3 8 11 10 5 6 7 2 1 4 9
n. 15 5 4 10 1 8 6 3 2 11 9 7 n. 49 3 10 1 8 7 6 5 4 11 2 9
n. 16 5 4 11 2 3 6 8 1 10 9 7 | n. 50 3 10 7 8 1 6 11 4 5 2 9
n. 51 3 11 2 7 8 6 4 5 10 1 9
n. 52 3 11 8 7 2 6 10 5 4 1 9
n. 53 4 1 3 5 2 6 10 7 9 11 8
n. 54 4 1 3 7 10 6 2 5 9 11 8
1 n. 1 1 2 7 4 3 6 9 8 5 10 11 n. 55 4 1 9 5 2 6 10 7 3 11 8
n. 2 1 2 7 4 9 6 3 8 5 10 11 n. 56 4 1 9 7 10 6 2 5 3 11 8
n. 3 1 3 4 7 2 6 10 5 8 9 11 n. 57 4 3 1 7 2 6 10 5 11 9 8
n. 4 1 3 10 7 8 6 4 5 2 9 11 n. 58 4 3 7 1 2 6 10 11 5 9 8
n. 5 1 4 3 2 5 6 7 10 9 8 11 n. 59 4 5 2 3 11 6 1 9 10 7 8
n. 6 1 4 9 2 5 6 7 10 3 8 11 n. 60 4 5 2 9 11 6 1 3 10 7 8
n. 7 1 8 7 10 3 6 9 2 5 4 11 n. 61 4 5 10 1 3 6 9 11 2 7 8
n. 8 1 8 7 10 9 6 3 2 5 4 11 n. 62 4 5 10 1 9 6 3 11 2 7 8
n. 9 1 9 4 7 2 6 10 5 8 3 11 n. 63 4 5 11 3 2 6 10 9 1 7 8
n. 10 1 9 10 7 8 6 4 5 2 3 11 n. 64 4 5 11 9 2 6 10 3 1 7 8
n. 11 1 10 3 8 5 6 7 4 9 2 11 n. 65 4 7 3 1 10 6 2 11 9 5 8
n. 12 1 10 9 8 5 6 7 4 3 2 11 n. 66 4 7 3 11 2 6 10 1 9 5 8
n. 13 2 1 4 3 7 6 5 9 8 11 10 n. 67 4 7 9 1 10 6 2 11 3 5 8
n. 14 2 1 4 9 7 6 5 3 8 11 10 n. 68 4 7 9 11 2 6 10 1 3 5 8
n. 15 2 1 7 3 4 6 8 9 5 11 10 n. 69 4 9 1 7 2 6 10 5 11 3 8
n. 16 2 1 7 9 4 6 8 3 5 11 10 n. 70 4 9 7 1 2 6 10 11 5 3 8
n. 17 2 1 8 5 3 6 9 7 4 11 10 n. 71 4 11 2 3 5 6 7 9 10 1 8
n. 18 2 1 8 5 9 6 3 7 4 11 10 n. 72 4 11 2 9 5 6 7 3 10 1 8
n. 19 2 3 5 11 4 6 8 1 7 9 10 n. 73 4 11 5 3 2 6 10 9 7 1 8
n. 20 2 3 11 5 4 6 8 7 1 9 10 n. 74 4 11 5 9 2 6 10 3 7 1 8
n. 21 2 5 3 1 4 6 8 11 9 7 10 n. 75 4 11 10 7 3 6 9 5 2 1 8
n. 22 2 5 3 11 8 6 4 1 9 7 10 n. 76 4 11 10 7 9 6 3 5 2 1 8
n. 23 2 5 9 1 4 6 8 11 3 7 10 n. 77 5 2 3 4 1 6 11 8 9 10 7
n. 24 2 5 9 11 8 6 4 1 3 7 10 n. 78 5 2 9 4 1 6 11 8 3 10 7
n. 25 2 7 1 3 4 6 8 9 11 5 10 n. 79 5 3 2 11 4 6 8 1 10 9 7
n. 26 2 7 1 9 4 6 8 3 11 5 10 n. 80 5 3 8 11 10 6 2 1 4 9 7
n. 27 2 7 4 3 1 6 11 9 8 5 10 n. 81 5 4 11 2 3 6 9 10 1 8 7
n. 28 2 7 4 9 1 6 11 3 8 5 10 n. 82 5 4 11 2 9 6 3 10 1 8 7
n. 29 2 7 8 11 3 6 9 1 4 5 10 n.83 5 8 3 10 1 6 11 2 9 4 7
n. 30 2 7 8 11 9 6 3 1 4 5 10 n. 84 5 8 9 10 1 6 11 2 3 4 7
n. 31 2 9 5 11 4 6 8 1 7 3 10 n. 85 5 9 2 11 4 6 8 1 10 3 7
n. 32 2 9 11 5 4 6 8 7 1 3 10 n. 86 5 9 8 11 10 6 2 1 4 3 7
n. 33 2 11 3 5 8 6 4 7 9 1 10 n. 87 5 10 11 8 3 6 9 4 1 2 7
n. 34 2 11 3 7 4 6 8 5 9 1 10 | n. 88 5 10 11 8 9 6 3 4 1 2 7 |
297
Bibliography
[1] AGMON (Etyan), "A Mathematical Model of the Diatonic System", Journal of
Music Theory, 33, 1989, pp. 1-25.
[2] AGMON (Etyan), "Coherent Tone-Systems: a study in the theory of diatonicism",
Journal of Music Theory, 40(1), 1996, pp. 39-59.
[3] AGMON (Etyan), "The Multiplicative Norm and its Implications for Set-Class
Theory", Perspectives of New Music, 40(1), 2002, pp. 216-234.
[4] AGON (Carlos), OpenMusic: Un langage visuel pour la composition musicale assis-
tte par ordinateur, Th&se, Paris VIII, 1998.
[5] AGON (Carlos), "Mixing Visual Programs and Music Notation in OpenMusic" in
Mazzola, Noll and Lluis-Puebla, Perspectives in Mathematical and Computational
Music Theory, Epos, 2004, pp. 233-244.
[6] ALEMBERT (Jean Henri d'), Elements de Musique tMorique et pratique selon les
principes de M. Rameau, Paris, 1752. Reprint Slatkine, Geneve.
[7] ALESSI (Rosario), Sonometro. Misura matematica del suono musicale. Metodo teo-
rico per la giusta divisione della scala cromatica (...) per accordare i pianoforti.
A. Morano, Naples 1881.
[8] A M I O T (Emmanuel), "Pour en finir avec le desir: la notion de symetrie en analyse
musicale", Analyse musicale, 22, Paris, 1991.
[9] AMIOT (Emmanuel), "Mathematique et analyse musicale: une fecondation reci-
proque", Analyse musicale, 28, Paris, 1992.
[10] AMIOT (Emmanuel), "La serie dodecaphonique et ses symetries", Quadrature, 19,
Paris, editions du choix, 1994, pp. 15-32.
[11] AMIOT (Emmanuel), "Des Vuza canons aux canons rythmiques 6conomiques", Ma-
MuX Seminar (Mathematiques, Musique et relations avec d'autres disciplines),
Ircam, 2002.
Bibliography
300
Bibliography
301
Bibliography
302
Bibliography
303
Bibliography
[93] BUTEAU (Chantal), Motivic Topologies and their Signification in Musical Motivic
Analysis, Masters Thesis, Universite de Laval, Quebec, 1998.
[94] BUTEAU (Chantal), A topological model of motivic structure and analysis of music-
Theory and Operationalization, PhD, University of Zurich, 2003.
[95] BUTEAU (Chantal), "Automatic Motivic Analysis Including Melodic Similarity For
Different Contour Cardinalities: Applications to Schumann's Of Foreign Lands and
People", Barcelona, Proceedings of the International Computer Music Conference,
2005, pp. 239-242.
[96] CAREY (Norman), "Well-Formed Scales and the Stern Brocot Tree", American
Mathematical Society Conference, Phoenix, 2004.
[97] CAREY (Norman), CLAMPITT (David), "Aspects of well-formed scales", Music
Theory Spectrum, 11, 1989, pp. 187-206.
[98] CAREY (Norman), CLAMPITT (David), "Regions: A Theory of Tonal Spaces in
Early Medieval Treatises", Journal of Music Theory, 40(1), 1996, pp. 113-147.
[99] CAREY (Norman), CLAMPITT (David), "Self-similar Pitch Structures, Their Duals,
and Rhythmic Analogues", Perspectives of New Music, 34(2), 1996, pp. 62-87.
[100] CARLOS (Wendy), "Tuning: At the Crossroads", Computer Music Journal, 11(1),
1987, pp. 29-43.
[101] CARLOS (Wendy), "Three Asymmetrical Divisions of the Octave", Pitch, 4(1),
1990, pp. 2-3.
[102] CASTELLENGO (Michele), FABRE (Benolt), G U Y O T (Frederique), "Etude de la ca-
tegorisation d'un corpus de bruits domestiques", in Daniele Dubois, Categorisation
et cognition, Paris, Kime, 1997, pp. 41-58.
[103] CAZDEN (Norman), "Musical consonance and dissonance: a cultural criterion",
Journal of Aesthetics, 4, 1945, pp. 3-11.
[104] CAZDEN (Norman), "Sensory theories of musical consonance", Journal of Aesthe-
tics, 20, 1962, pp. 310-319.
[105] CAZDEN (Norman), "The definition of consonance and dissonance", International
Review of Aesthetics and Sociology of Music, 11(2), 1980, pp. 123-168.
[106] CHALMERS (John H, Jr.), "The Application of Rothenberg's Pattern Recognition
Model to the Structure of Tetrachords and Tetrachordal Scales", Xenharmonikdn,
3, 1975.
[107] CHALMERS (John H, Jr.), "Microtonal Mathematics", Interval, 1(2), 1978.
[108] CHALMERS (John H, Jr.), "Construction and Harmonization of Microtonal Scales in
Non-12-Tone Equal Temperaments", Proceedings of the 1982, International Com-
puter Music Conference, ICMA, 1982, pp. 534-555.
[109] CHALMERS (John H, Jr.), "Polychordal Arrays of MOS Scales", Xenharmonikdn,
7-8, 1979.
[110] CHALMERS (John H, Jr.), "Some Additional With Nineteen-tone Scales", Xenhar-
monikdn, 7-8, 1979.
[Ill] CHALMERS (John H, Jr.), "Tritriadic Scales With Seven Tones", Xenharmonikdn,
9, 1986.
[112] CHALMERS (John H, Jr.), "Tritriadic Transformations", One/one, 3(1), 1987.
304
Bibliography
[113] CHALMERS (John H, Jr.), "Tritriadic Scales with Seven Tones, Part Two: Derived
Forms and Structural Properties", Xenharmonikdn, 10, 1988.
[114] CHALMERS (John H, Jr.), "Third Coast New Music Project", One/one, 4(3), 1988,
pp. 15.
[115] CHALMERS (John H, Jr.), "Tritriadic Scales with Seven Tones. Part Three: The
M -> T and D - • M Matrices", Xenharmonikdn, 11, 1989, pp. 40-68.
[116] CHALMERS (John H, Jr.), "Tetrachordal Scales and Complexes", Xenharmonikdn,
11, 1989, pp. 1-20.
[117] CHALMERS (John H, Jr.), "Three Approaches to Harmony in 13-TET", Xenhar-
monikdn, 13, 1991, pp. 53-64.
[118] CHALMERS (John H, Jr.), Divisions of the Tetrachord, Hanover NH, Prog Peak
Music, 1993.
[119] CHEMILLIER (Marc), "MonoYde libre et musique", RAIRO, Informatique tMo-
rique et applications, Paris, Gauthier Villars, 21(3), 1987, pp. 341-371, and 21(4),
pp. 379-418.
[120] CHEMILLIER (Marc), Structure et m&thodes algGbriques en Informatique musicale,
These Paris, L.I.T.P., Institut Blaise Pascal, 1990.
[121] CHEMILLIER (Marc), "Mathematiques et musiques de tradition orale", Musique et
mathematiques, Paris, Alea, 1997, pp. 133-143.
[122] CHEMILLIER (Marc), "Generateurs musicaux et singularites", Actes des Sixidmes
Journe'es d'Informatique Musicale, CNET-CEMAMu, 1999, pp. 167-177.
[123] CHEMILLIER (Marc), "Ethnomusicology, Ethnomathematics. The Logic Underlying
Orally Transmitted Artistic Practices", in ASSAYAG et al. 2002, pp. 161-183.
[124] CHEMILLIER (Marc), PACHET (F.), Recherches et applications en informatique mu-
sicale, Paris, Hermes, 1998.
[125] CHEMILLIER (Marc), DUCHAMP (Gerard), "Algorithme de reduction des degres
dans une gamme musicale (d'apres la theorie d'Yves Hellegouarch)", Gazette des
Mathematiques, Societe Mathematique de Prance, 82, 1999, pp. 26-30.
[126] CHEMILLIER (Marc), TRUCHET (Charlotte), "Computation of words satisfying the
rhythmic oddity property (after Simha Arom's works)", Information Processing
Letters, 86, 2003, pp. 255-261.
[127] CHILDS (Adrian), "Moving Beyond Neo-Riemannian Triads: Exploring a Trans-
formational Model for Seventh Chords", Journal of Music Theory, 42(2), 1998,
pp. 181-193.
[128] CHOUVEL (Jean-Marc), "La physique et Pesthetique. Une analyse epistemologique
des modalites de connaissance du phenomene harmonique", Paris, Editions Eska,
Musurgia, 2(4), 1995, pp. 88-102.
[129] CHOUVEL (Jean-Marc), Esquisses pour une pensie musicale. Les metamorphoses
d'Orphee, Paris, L'Harmattan, 1998.
[130] CHOUVEL (Jean-Marc), DOVAL (Boris), "La concordance harmonique: vers une
nouvelle approche de la consonance", Communication a la troisieme conference
internationale pour la perception et la cognition musicale, Liege, 23-27 juillet 1994.
[131] CHOWNING (John), "The Synthesis of Complex audio Spectra by Means of Fre-
quency Modulation", Journal of the Audio Engineering Society, 21(7), 1985,
pp. 526-534.
305
Bibliography
[132] CLAMPITT (David), "Some Refinements on the Three Gap Theorem, with Appli-
cations to Music", Muzica, 6(2), 1995, pp. 12-22.
[133] CLAMPITT (David), Pairwise Well-formed Scales: Structural and Transformational
Properties, Ph.D. Dissertation, State University of New York at Buffalo, 1997.
[134] CLOUGH (John), "Diatonic Interval Sets and Transformational Structures", Music
Theory Spectrum, 18, 1979, pp. 461-482.
[135] CLOUGH (John), "Diatonic Interval Cycles and Hierarchical Structures", Perspec-
tives of New Music, 32(1), 1994, pp. 228-253.
[136] CLOUGH (John), "A rudimentary geometric model for contextual transposition and
inversion", Journal of Music Theory, 42(2), 1998, pp. 297-341.
[137] CLOUGH (John), MYERSON (Gerald), "Variety and Multiplicity in Diatonic Sys-
tems", Journal of Music Theory, 29, 1985, pp. 249-270.
[138] CLOUGH (John), MYERSON (Gerald), "Musical Scales and the Generalized Circle
of Fifths", American Mathematical Monthly, 93(9), 1986, pp. 695-701.
[139] CLOUGH (John) and D O U T H E T T (Jack), "Maximally Even Sets", Journal of Music
Theory, 35,1991, pp. 93-173. Errata: Journal of Music Theory, 36(2), 1992, pp. 401-
402.
[140] CLOUGH (John), CUCIUREAN (John), and D O U T H E T T (Jack), "Hyperscales and
the Generalized Tetrachord", Journal of Music Theory, 41, 1997, pp. 67-100.
[141] CLOUGH (John), ENGEBRETSEN (Nora), and KOCHAVI (Jonathan), "Scales, Sets
and Interval Cycles: A Taxonomy", Music Theory Spectrum, 21(1), 1999, pp.74-104.
[142] COXETER (Harold Scott Macdonald), Introduction to geometry, John Wiley, 1959.
[143] COXETER (Harold Scott Macdonald), Regular Polytopes, Macmillan, 1963.
[144] COHN (Richard), "Maximally Smooth Cycles, Hexatonic Systems, and the Analysis
of Late-Romantic Triadic Progressions", Music Analysis 15(1), 1996, pp. 9-40.
[145] COHN (Richard), "Neo-Riemannian Operations, Parsimonious Trichords, and their
Tonnetz Representations", Journal of Music Theory, 41(1), 1997, pp. 1-66.
[146] COHN (Richard), "Introduction to Neo-Riemannian Theory: A Survey and A His-
torical Perspective", Journal of Music Theory, 42(2), 1998, pp. 167-180.
[147] COHN (Richard), "Square Dances with Cubes", Journal of Music Theory, 42(2),
1998, pp. 283-296.
[148] COLBOURN (Charles), ROSA (Alexander), Triple Systems, Oxford University Press,
1999.
[149] COSTERE (Edmond), Lois et styles des harmonies musicales, Paris, Presses Uni-
versitaires de Prance, 1954.
[150] COSTERE (Edmond), Mort ou transfiguration de Vharmonie, Paris, Presses Uni-
versitaires de Prance, 1962.
[151] COOKE (Deryck), The Language of Music, Oxford University Press, 1959.
[152] COONS (Edgar), KRAEHENBUEHL (David), "Information as a measure of structure
in Music", Journal of Music Theory, 2(2), 1958, pp. 127-161.
[153] COVACH (John R), "The Zwolftonspiel of Josef Matthias Hauer", Journal of Music
Theory, 36(1), 1992, pp. 149-184.
306
Bibliography
[154] COVEN (Ethan), MEYEROWITZ (Aaron), "Tiling the integers with translate of one
finite set", Journal of Algebra, 212, 1999, pp. 161-174.
[155] CROCKER (Malcolm), ed. Handbook of Acoustics, Wiley Interscience, 1998.
[156] DABBY (Diana S.), "Musical Variations form a Chaotic Mapping", Chaos, 6(2),
1996, pp. 95-107.
[157] DAHLHAUS (Carl), Untersuchung uber die Entstehung der harmonischen Tonalitat,
Barenreiter, Kassel, 1967.
[158] DALL'OLIO (Giambattista), "Sull'applicazione della matematica alia musica", Me-
morie delle Societa Italiana delle Scienze, IX, 1802, pp. 609-625.
[159] DAMMAN (Ralf), "Die Musica mathematica von Bartotus", Archiv fur Musikwis-
senchaft, XXVI, 1969, pp. 140-162.
[160] DANIELOU (Alain), Simantique musicale, Essai de psycho-physiologie auditive, Pa-
ris, Hermann, 1967.
[161] DELIEGE (Celestin), Les fondements de la musique tonale, Paris, Lattes, 1984.
[162] DELIEGE (Celestin), Cinquante ans de moderniti musicale: de Darmstadt a VIrcam.
Constribution historiographique a une musicologie critique, Bruxelles, Mardaga,
2003.
[163] DELIEGE (Irene), PADDISON (M.), Musique contemporaine. Perspectives th&oriques
et philosophiques, Bruxelles, Mardaga, 2000.
[164] D E L O N E (Richard), et al. eds. Aspects of 20th Century Music, Englewood Cliffs,
Prentice-Hall, 1975.
[165] DEVIE (Dominique), Le temperament musical, Societe de musicologie du Langue-
doc, 1990.
[166] D O U T H E T T (Jack), ENTRINGER (Roger), MULLHAUPOT (Andrew), "Musical Scale
Construction: The Continued Fraction Compromise", Utilitas Mathematica, 42,
1992, pp. 97-113.
[167] DOUTHETT (Jack), STEINBACH (Peter), "Parsimonious graphs: A study in parsi-
mony, contextual transformations, and modes of limited transpositions", Journal
of Music Theory, 42(2), 1998, pp. 241-263.
[168] DRIEBERG (Priedrich von), Die mathematische Intervallenlehre der Griechen, Ber-
lin, 1818.
[169] DUCHON (M.), "Survey of some topics of harmonic analysis", Tatra Mountains
Mathematical Publications, 23, 2001, pp. 165-169.
[170] DUDON (Jacques), "La gamme doree: Un temperament inegal issu du nombre
d'or", in Jean-Bernard Condat, Nombre d'or et Musique, Verlag Peter Lang, 1988,
pp. 147-157.
[171] DUDON (Jacques), "Musiques Harmoniques", One/one, 5(3), 1989, pp. 10-14.
[172] DuiJVESTUN (Arie), "A Simple Perfect Square of Lowest Order", J. Comb. Th.
Ser. B, 25, 1978, pp. 240-243
[173] DUNSBY (Jonathan), Music Analysis in Theory and Practice, New Haven: Yale
University Press, 1988.
[174] DURUTTE (Camille), Technie, ou lois g6n£rales du systeme harmonique, Mallet-
Bachelier, Paris, 1855.
307
Bibliography
308
Bibliography
[195] FAREY (John), "On different modes of expressing the magnitudes and relations
of musical intervals, with some remarks, in commendation of Professor Fischer's
proportionally-tempered douzeave", The Silliman American Journal of Science and
Arts, II, London, 1820, pp. 65-81.
[196] FAUVEL (John), FLOOD (Raymond), WILSON (Robert), Music and Mathematics,
From Pythagoras to Fractals, Oxford University Press, 2003.
[197] FEICHTINGER (Hans G.), DORFLER (M.), eds. Diderot Forum on Mathematics
and Music: Computational and Mathematical Methods in Music, Vienna, Osterrei-
chische Computergesellschaft, 1999.
[198] FERNANDEZ (Claude), PERCHERANCIER (Jean-Michel), Les oeuvres pour piano et
orchestre, Geneve, Champion-Slatkine, 1988.
[199] FERRETTI (R.), MAZZOLA (G.), "Algebraic Varieties of Musical Performances",
Tatra Mountains Mathematical Publications, 23, 2001, pp. 59-69.
[200] FlCHET (Laurent), Les theories scientifiques de la musique aux XIXe et XXe si£cle,
Vrin, Paris, 1996.
[201] FISCHER (Myriam), "Leonhard Euler et la musique", Strasbourg, L'Ouvert, 112,
2005, pp. 19-31.
[202] FLETCHER (Neville H.), and ROSSING (Thomas D.), The Physics of Musical Ins-
truments, Springer Verlag, 1991.
[203] FOKKER (Adriaan D.), "Het muzikale toonstelsel van Christiaan Huygens, de nor-
male di&zenstemming", 1942.
[204] FOKKER (Adriaan D.), "Rekenkundige bespiegeling der muziek", (Arithmetical re-
flection of music), Noorduijn, Gorinchem, 1945, review by E.J. Dijksterhuis, in
Mens en Melodie, 1, 1946, pp. 59-62.
[205] FOKKER (Adriaan D.), "Les mathematiques et la musique", in Recherches musi-
cales. Th&oriques et pratiques, Den Haag, From Teyler Museum Archives, 10, 1947,
pp. 1-31.
[206] FOKKER (Adriaan D.), "De behoefte aan grotere nauwkeurigheid in de muzikale
notatie der toonhoogte", 1953.
[207] FOKKER (Adriaan D.), "Equal Temperament and the Thirty-one-keyed Organ",
1955.
[208] FOKKER (Adriaan D.), "Optelakkoorden", 1960.
[209] FOKKER (Adriaan D.), "Met verjongde oren", 1964.
[210] FOKKER (Adriaan D.), "Simon Stevin's views on music", Introduction to Vande
spiegheling der singconst ("On the theory of the art of singing"), in Principal Works
of Simon Stevin 5, A.D. Fokker (ed.), Amsterdam, 1955-1966, pp. 413-464.
[211] FOKKER (Adriaan D.), "On the Expansion of the Musician's Realm of Harmony",
This paper is a preparatory study for a report on the problem of the notation of
microtones, which will be presented to the general assembly of the International
Musicological Society in 1967.
[212] FOKKER (Adriaan D.), "Unison vectors and periodicity blocks in the three-
dimensional (3-5-7), harmonic lattice of notes", From Koninklijke Nederlandse Aka-
demie van Wetenschappen - Amsterdam, Proceedings, Series B 72(3), 1969 (Com-
municated at the meeting of February 22, 1969).
309
Bibliography
[213] FOKKER (Adriaan D.), New Music With 31 Notes, translated by Leigh Gerdine,
Orpheus - Series of Monographs on Basic Questions in Music, vol. 5, Verlag fur
Systematische Musikwissenschaft, Bonn, 1975.
[214] FOKKER (Adriaan D.), VAN DlJK, (Jan), "Experiences musicales avec les genres
musicaux de Leonhard Euler contenant la septieme harmonique", Recherches mu-
sicales. TMoriques et pratiques, Den Haag, Martinus Nijhoff, Teyler Museum Ar-
chives, 10, 1951.
[215] FORD (L.R.), "Fractions", American Mathematical Monthly, 45(9), pp. 586-601.
[216] FORTE (Allen), "A Theory of Set-Complexes for Music", Journal of Music Theory,
8(2), 1964, pp. 136-184.
[217] FORTE (Allen), "Sets and Nonsets in Schoenberg's Atonal Music", Perspectives of
New Music, 11(1), 1972, pp. 43-64.
[218] FORTE (Allen), The Structure of Atonal Music, New Haven and London, Yale
University Press, 1973.
[219] FORTE (Allen), Introduction to Schenkerian Analysis, New York, Norton, 1982.
[220] FRANCES (Robert), La perception de la musique, Paris, Vrin, 1984.
[221] FRIPERTINGER (Harald), "Enumeration in Musical Theory", S6minaire Lotharin-
gien de Combinatoire, 476(42), 1992, pp. 26-29.
[222] FRIPERTINGER (Harald), "Endliche Gruppenaktionen auf Funktionenmengen. Das
Lemma von Burnside - Repr&sentantenkonstruktionen - Anwendungen in der Mu-
siktheorie", Bayreuther Mathematische Schriften, 45, 1993, pp. 19-135.
[223] FRIPERTINGER (Harald), "Enumeration and construction in music theory", in H.G.
FEICHTINGER and M. DORFLER eds. Computational and Mathematical Methods in
Music Theory, Vienna, Austria, 1999, pp. 179-204.
[224] FRIPERTINGER (Harald), "Enumeration of Mosaics", Discrete Mathematics, 199,
1999, pp. 49-60.
[225] FRIPERTINGER (Harald), "Enumeration of Non-Isomorphic Canons", Tatra Mt.
Math. Publ. 23, 2001, pp. 47-57.
[226] FRIPERTINGER (Harald), "Enumeration of Mosaics, Enumeration of Canons", Ma-
MuX Seminar (Mathematiques, Musique et relations avec d'autres disciplines),
Ircam, 2002.
[227] FRIPERTINGER (Harald), "Tiling Problems in Music Theory" in Mazzola, Noll and
Lluis-Puebla, eds. Perspectives in Mathematical and Computational Music Theory,
Epos, 2004, pp. 153-168.
[228] FULLER (Ramon), "A Study of Interval and Trichord Progressions", Journal of
Music Theory, 16(1-2), 1972, pp. 102-141.
[229] GALLIMARD (Jean Edme), L'arithme'tique des musiciens, Paris, 1754.
[230] GALLIMARD (Jean Edme), La tMorie des sons ou Von demontre dans une exacte
precision les rapports de tous les intervalles de la gamme, Paris, Ballard, 1754.
[231] GAMER (Carlton), LANSKY (Paul), "Fanfare for the Common Tone", Perspectives
of New Music, 1976, pp. 229-235.
[232] GAMER (Carlton), "Some combinatorial resources of equal-tempered systems",
Journal of Music Theory, 11, 1967, pp. 32-59.
310
Bibliography
311
Bibliography
[254] HALL (Donald E.), Musical Acoustics, Wadsworth Publishing Compagny, Belmont,
California, 1980.
[255] HALL (Donald E.), "A Systematic Evaluation of Equal Temperaments Through
n = 612", Interface, 14, 1985, pp. 61-73.
[256] HALSEY (D.), H E W I T T (E.), "Eine gruppentheoretische Methode in der Musiktheo-
rie", Jahresber. der Dt. Math. Vereinigung, 80, 1978, pp. 151-207.
[257] HALUSKA (Jan), "Uncertainty and tuning in music", Tatra Mountains Mathema-
tical Publications, 12, 1997, pp. 113-129.
[258] HALUSKA (Jan), "Searching the frontier of the Pythagorean system", Tatra Moun-
tains Mathematical Publications, 16, 1999, pp. 273-281.
[259] HALUSKA (Jan), ed. Harmonic Analysis and Tone Systems, Bratislava, Tatra
Mountains Mathematical Publications, 23, 2001.
[260] HALUSKA (Jan), "Equal temperament and Pythagorean Tuning: a geometrical in-
terpretation in the plane", Fuzzy Sets and Systems, 114, 2000, pp. 261-269.
[261] HALUSKA (Jan), The Mathematical Theory of Tone Systems, Marcel Dekker, New-
York, 2003.
[262] HALUSKA (Jan), RIECAN (B.), "On a lattice of tetrachords", Tatra Mountains
Mathematical Publications, 16, 1999, pp. 283-294.
[263] HANSON (Howard), Harmonic Materials of Modern Music: Resources of the Tem-
pered Scale, New York, Appleton-Century-Crofts Inc., 1946.
[264] HARDCASTLE (David), KHANIN (Kostya), "Continued Fractions and the d-
dimensional Gauss Transformation", Communications in Mathematical Physics,
215, 2001, pp. 487-515.
[265] HARDY (Geoffrey), W R I G H T (Edward), An introduction to the theory of numbers,
Oxford University Press, 1980.
[266] HARRISON (Daniel), Harmonic function in chromatic music: A renewed dualist
theory and an account of its precedents, University of Chicago Press, 1994.
[267] HARRISON (Daniel), "Some Group Properties of Triple Counterpoint and Their
Influence on Compositions by J.S. Bach", Journal of Music Theory, 32(1), 1988,
pp. 23-49.
[268] HARTMANN (George C ) , "A numerical exercise in musical scales", American Jour-
nal of Physics, 55(3), 1987, pp. 223-226.
[269] HASCHER (Xavier), "Liszt et les sources de la notion d'agregat", Analyse Musicale,
43, 2002, pp. 48-57.
[270] HASCHER (Xavier), "Une analyse transformationnelle de l'opus 19 n° 4 de Schoen-
berg au moyen des K-reseaux", paper delivered at the International Symposium
"Around Set Theory", Ircam, 2003 (to appear).
[271] HASCHER (Xavier), "Famille cordique, relation de similarity et uniformite maxi-
male", LVuvert, 112, 2005, pp. 33-46.
[272] HAUER (Josef Matthias), Zwolftontechnik, Die Lehre von den Tropen, Vienna, Uni-
versal Edition UE 8438, 1926.
[273] HELLEGOUARCH (Yves), "Scales", Comptes Rendus de la Societe Royale du Ca-
nada, IV(5), 1982 and V(2), 1983.
312
Bibliography
313
Bibliography
[292] HURON (David), "Tone and Voice: A Derivation of the Rules of Voice-Leading from
Perceptual Principles", Music Perception, 19(1), 2001, pp. 1-64.
[293] HuSMANN (Heinrich), Einfwrung in die Musikwissenschaft, Heinrichschofen, Wil-
helmshaven, 1975.
[294] HYER (Brian), Tonal Intuitions in Tristan und Isolde, Ph.D. diss., Yale University,
1989.
[295] HYER (Brian), "Reimag(in)ing Riemann", Journal of Music Theory 39(1), 1995,
pp. 101-138.
[296] ISAACSON (Eric), "Similarity of Interval-Class Content Between Pitch-Class Sets:
The IcVSIM Relation", Journal of Music Theory, 34(1), 1990, pp. 1-28.
[297] ILOMAKI (Tuukka), "Group Structures and Equivalence Classes in Extended
Twelve-Tone Operations", Proceedings of the International Computer Music Confe-
rence, Barcelona, 2005, pp. 487-490.
[298] JACKENDOFF (Ray), LERDAHL (Fred), A Generative Theory of Tonal Music, Cam-
bridge Mass., MIT Press, 1983.
[299] JEDRZEJEWSKI (Pranck), ed. La loi de la Pansonorite Ivan Wyschnegradsky (ver-
sion de 1953), Geneve, Contrechamps, 1996.
[300] JEDRZEJEWSKI (Pranck), MatMmatiques des systdmes acoustiques, Temperaments
et modeles contemporains, Paris, L'Harmattan, 2002.
[301] JEDRZEJEWSKI (Pranck), Dictionnaire des musiques microtonales, Paris, L'Har-
mattan, 2003.
[302] JEDRZEJEWSKI (Pranck), "Nceuds polychromes et entrelacs sonores. Vers de nou-
velles categories musicales", Musicae Scientiae, Special issue 2003-2004. Award
papers. Special 10th anniversary conference issue, pp. 73-84.
[303] JEDRZEJEWSKI (Pranck), "Alois Haba et Pexperimentation micro-intervallique",
Congres international de Musique Tcheque, Paris Sorbonne, 15 novembre 2002, in
L'attraction et la necessite. Musique tchdque et culture francaise au XXe sidcle.
Textes reunis par Xavier Galmiche et Lenka Stranka, Prague, Edition Barenreiter
Praha, 2004, pp. 169-176.
[304] JEDRZEJEWSKI (Pranck), ed. Une philosophie dialectique de Vart musical, Ivan
Wyschnegradsky, Paris, L'Harmattan, 2005.
[305] JEDRZEJEWSKI (Pranck), "Permutation Groups and Chord Tessellations", Procee-
dings of the International Computer Music Conference, Barcelona, 2005, pp. 231-
234.
[306] JONES (Evan), "Residue-Class Sets in the Music of Iannis Xenakis: An Analytical
Algorithm and a General Intervallic Expression", Perspectives of New Music, 39(2),
2001, pp. 229-261.
[307] JOHNSON (Tom), Self-Similar Melodies, Paris, Editions 75, 1996.
[308] JOHNSON (Tom), "Tiling the line. Self-Replicating Melodies, Rhythmic Canons,
and an Open Problem", Actes des Huiti&mes Journe'es d'Informatique Musicale,
Bourges, 2001, pp. 147-152.
[309] JOHNSON (Tom), Tiling the line in Theory and in Practice, MaMuX Seminar (Ma-
thematiques, Musique et relations avec d'autres disciplines), Ircam, February 2002.
314
Bibliography
[310] JOHNSON (Tom), Some Observations on Tiling Problems, MaMuX Seminar (Ma-
thematiques, Musique et relations avec d'autres disciplines), Ircam, Januay 2003.
[311] JOHNSON (Tom), "Tiling Melodies, Tiling Chords", Computer Modeling and Re-
trieval, 2003, LIRMM, CNRS, University of Montpellier.
[312] JOHNSON (Tom), Perfect Rhythmic Tilings, MaMuX Seminar (Mathematiques,
Musique et relations avec d'autres disciplines), Ircam, January 2004.
[313] JOHNSON (Timothy A.), "Harmonic Vocabulary in the Music of John Adams: A
Hierarchical Approach", Journal of Music Theory, 37(1), 1993, pp. 117-156.
[314] JOHNSTON (Ben), "Proportionality and Expanded Pitch Resources", Perspectives
of New Music, 5(1), 1966, pp. 112-120.
[315] JOHNSTON (Ben), "Rational Structure In Music", American Society of University
Composers Proceedings, 1-2, 1976-1977, pp. 102-118.
[316] JORGENSEN (Owen H.), Tuning, Michigan State University Press, 1991.
[317] KAMEOKA (Akio), K N O P O F F (Leon), "Consonance Theory I: Consonance of
dyads", Journal of the Acoustical Society of America, 45(6), 1969, pp. 1452-1459.
[318] KAMEOKA (Akio), K N O P O F F (Leon), "Consonance Theory II: Consonance of com-
plex tones and its calculation method", Journal of the Acoustical Society of Ame-
rica, 45(6), 1969, pp. 1460-1469.
[319] KARKOSCHKA (Erhard), Das Schriftbild der neuen Musik, Celle, Moeck, 1966.
[320] KARKOSCHKA (Erhard), "Adriaan D. F O K K E R ' Neue Musik mit 31 Tonen", Die
Musikforschung, XXII, 1969, pp. 115-116.
[321] KHRUZIN (Andrei), "Enumeration of chord diagrams", preprint ArXiv,
math.CO/0008209, 2000.
[322] KLEBAN (Peter), OZLUK (Ali), "A Farey Fraction Spin Chain", Communication in
Mathematical Physics, 203, 1999, pp. 635-647.
[323] KLEMM (Michael), Symmetrien von Ornamenten und Kristallen, Berlin, Springer,
1982.
[324] KLUMPENHOUWER (Henry), A Generalized Model of Voice Leading for Atonal Mu-
sic, Ph.D. diss., Harvard University, 1991.
[325] KLUMPENHOUWER (Henry), "The Cartesian Choir", Music Theory Spectrum,
14(1), 1992, pp. 15-37.
[326] KLUMPENHOUWER (Henry), "Some Remarks on the Use of Riemann Transforma-
tions", Music Theory Online, 0(9), 1994.
[327] KLUMPENHOUWER (Henry), "The Inner and Outer Automorphisms of Pitch-Class
Inversion and Transposition: Some Implications for Analysis with Klumpenhouwer
Networks", Integral, 12, 1998, pp. 81-93.
[328] KLUMPENHOUWER (Henry), "Dualist Tonal Space and Transformation in
Nineteenth-Century Musical Thought", in Christensen, The Cambridge History
of Western Music Theory, Cambridge University Press, 2002, pp. 456-476.
[329] KOLMAN (Oren), "Transfert principles for generalized interval systems", Perspec-
tives of New Music, 41(2), 2003.
[330] KORNERUP (Thorwald), Das Goldene Tonsystem, Aschehoug Forlag, Kopenhagen,
1935.
315
Bibliography
316
Bibliography
317
Bibliography
[373] MAZZOLA (Guerino), Gruppen und Kategorien in der Musik. Entwurf einer ma-
thematischen Musiktheorie, Berlin, Heldermann Verlag, 1985.
[374] MAZZOLA (Guerino), "ObertSne oder Symmetrie: Was ist Harmonie?", Mi-
krotone 1, Innsbruck, Helbling, 1986, pp. 191-210.
[375] MAZZOLA (Guerino), "Konsonanz-Dissonanz und verborgene Symmetrien", Mi-
krotone 2, Innsbruck, Helbling, 1988, pp. 95-104.
[376] MAZZOLA (Guerino), Geometrie der Tone, Basel, Birkhauser, 1990.
[377] MAZZOLA (Guerino), Mathematical Music Theory: An Informal Survey, Locarno,
Edizioni Cerfim, 1993.
[378] MAZZOLA (Guerino), "Semiotic Aspects of Musicology: Semiotics of Music" in
R. Posner et al. A Handbook of the Sign-Theoretic Foundations of Nature and
Culture, Berlin, Walter de Gruyter, 1998.
[379] MAZZOLA (Guerino), The Topos of Music, Basel, Birkhauser, 2002.
[380] MAZZOLA (Guerino), "Mathematical Music Theory - Status Quo 2000", in Maz-
zola, Noll and Lluis-Puebla, eds. Perspectives in Mathematical and Computational
Music Theory, Epos, 2004, pp. 42-77.
[381] MAZZOLA (Guerino), "Towards a Galois Theory of Concepts", in Mazzola, Noll and
Lluis-Puebla, eds. Perspectives in Mathematical and Computational Music Theory,
Epos, 2004, pp. 78-88.
[382] MAZZOLA (Guerino), NOLL (Thomas), LLUIS-PEBLA (Emilio), eds. Perspectives in
Mathematical and Computational Music Theory, Electronic Publishing Osnabrtick,
2004.
[383] MCLAREN (Brian), "General Methods of Generating Musical Scales", Xenharmo-
nikdn, 13, 1990, pp. 20-44.
[384] MCLAREN (Brian), "Those Difficult Teens (and Other Microtonal Scales)", Xen-
harmonikdn, 14, 1991, pp. 3-6.
[385] MCLAREN (Brian), "A Nostalgic Look At the 20s - and Other Musical Scales",
Xenharmonikdn, 14, 1991, pp. 6-8.
[386] MCLAREN (Brian), "The Low Profile of Those Higher Divisions of the Octave",
Xenharmonikdn, 14, 1991, pp. 8-11.
[387] MCLAREN (Brian), "The Uses and Characteristics of Non-Octave Scales", Xenhar-
monikdn, 14, 1993, pp. 12-22.
[388] M C L A R E N (Brian), "The Uses and Characteristics of Non-Just Non-Equal-
Tempered Scales", Xenharmonikdn, 15, 1993, pp. 27-41.
[389] MEAD (Andrew W.), "Some Implications of the Pitch-Class Order-Number Iso-
morphism Inherent in the Twelve-Tone System", Perspectives of New Music, 26(2),
1988, pp. 96-163.
[390] MEAD (Andrew W.), An Introduction to the Music of Milton Babbitt, Princeton
University Press, 1994.
[391] MEEOS (Nicolas), Heinrich Schenker. Une introduction, Bruxelles, Mardaga, 1993.
[392] MEEtJS (Nicolas), "Vecteurs harmoniques", Musurgia , 10(3-4), 2003, pp. 7-34.
[393] MENAGE (Marcel), "La Set Complex Theory: de quels enjeux s'agit-il?", Analyse
musicale, 17, 1989, pp. 87-90.
318
Bibliography
[394] MENAGE (Marcel), "Sur la modalisation des partitions musicales", Analyse musi-
cale, 22, 1991, pp. 31-46.
[395] MENAGE (Marcel), R I O T T E (Andre). "Modelisation informatique des partitions,
analyse et composition assistee", Les Cahiers de VIRCAM, 3, 1993, pp. 43-59.
[396] MENDEL (Arthur), "Pitch in the 16 th and early 17 th Centuries", Musical Quarterly,
34, 1948, pp. 28-45, pp. 199-221, pp. 336-357, pp. 575-593.
[397] MENDEL (Arthur), "On the Pitches in use in Bach's Time", Musical Quarterly, 41,
1955, pp. 332-354, pp. 466-480.
[398] MENDEL (Arthur), "Pitch in Western Music since 1500: a re-examination", Acta
Musicologica, 50, 1978, pp. 1-93.
[399] MERSENNE (Marin), Correspondances, Paris, CNRS, Edited by C. de Waard, 1932-
1945.
[400] MERSENNE (Marin), Harmonie Universelle, Sebastien Cramoisy, Paris 1636-37.
Reprinted by CNRS 1975.
[401] MERSENNE (Marin), Cogitata physico-mathematica, Paris, Bertier, 1644.
[402] MERSENNE (Marin), Harmonicorum Libri XII, Paris, Bauldry, 1648.
[403] MESSIAEN (Olivier), Techniques de mon langage musical, Paris, Alphonse Leduc,
1944.
[404] MESSIAEN (Olivier), Traite de Rythme, de couleur et d'omitologie, Paris, Alphonse
Leduc, 7 vols, 1992.
[405] MlLHAUD (Darius), "Polytonalite et atonalite", Revue Musicale, 4, 1923, reprinted
in Notes sur la musique: essais et chroniques, textes reunis et presentes par Jeremy
Drake, Paris, Flammarion, 1982, pp. 173-189.
[406] MORENO (Enrique), Expanded Tunings in Contemporary Music: Theoretical Inno-
vations and Practical Applications, Edwin Mellen Press, 1992.
[407] MORRIS (Robert), "Set Groups, Complementation, and Mappings Among Pitch-
Class Sets", Journal of Music Theory, 26(1), 1982, pp. 101-144.
[408] MORRIS (Robert), Composition with pitch-classes: a theory of compositional design,
New Haven, Yale University Press, 1987.
[409] MORRIS (Robert), "Pitch-Class Complementation and its Generalizations", Jour-
nal of Music Theory, 34(2), 1990, pp. 175-245.
[410] MORRIS (Robert), "Equivalence and Similarity in Pitch and their Interaction with
Pcset Theory", Journal of Music Theory, 39(2), 1995, pp. 207-243.
[411] MORRIS (Robert), "K, Kh and Beyond", in J. Baker et al, Music Theory in Concept
and Practice, 1997.
[412] MORRIS (Robert), "Voice-Leading Spaces", Music Theory Spectrum, 20(2), 1998,
pp. 176-208.
[413] MORRIS (Robert), Class Notes for Advanced Atonal Music Theory, 2 vol., Prog
Peak Music, 2001.
[414] MORRISON (Gary), "Decimal Music", Xenharmonikdn, 7-8, 1979, pp. 1-4.
[415] MORRISON (Gary), "Ten Equal: The Other Side of the Coin", Interval, 2(1), 1979,
pp. 7-8.
319
Bibliography
[416] MORSE (Philip), INGARD (K. Uno), Theoretical Acoustics, Mc Graw Hill, 1968,
reprinted by Princeton University Press, 1986.
[417] MUZZULINI (Daniel), "Leonhard Eulers Konsonanztheorie", Musiktheorie 9(2),
Laaber Verlag, Laaber, Germany, 1994.
[418] MUZZULINI (Daniel), "Musical Modulation by Symmetries", Journal of Music
Theory, 39(2), 1995, pp. 311-325.
[419] NAUERT (Paul), "Field Notes: A Theory of Fixed-Pitch Formations", Perspectives
of New Music, 41(1), 2003, pp. 180-239.
[420] NEDERVEEN (Cornelius Johannes), Acoustical Aspects of Woodwind Instruments,
Northern Illinois Press, 1998.
[421] NEIDHARDT (Johann Georg), Gdntzlich erschopfte mathematische Abtheilungen
des diatonisch-chromatischen, temperirten Canonis Monochordi, Kftnisgberg, 1732.
[422] NEIDHARDT (Johann Georg), Systema generis diatonici chromatid exnumeris serie
naturali procedentibus, Ktoigsberg, 1734.
[423] NETSKE (Andreas), NOLL (Thomas), "Inner metric analysis", Tatra Mountains
Mathematical Publications, 23, 2001, pp. 91-111.
[424] NEUWIRTH (Erich), Musical Temperaments, Springer Verlag, 1997.
[425] NEWMAN (D. J.), "Tesselation of integers", Journal of Number Theory, 9, 1977,
pp. 107-111.
[426] NICOLAS (Frangois), "Le feuillete du tempo. Essai sur les modulations metriques",
Paris, Entretemps, 9, 1990, pp. 51-77.
[427] NICOLAS (Frangois), "Quand Palgebre mathematique aide a penser (et pas seule-
ment & calculer) la combinatoire musicale", MaMuX Seminar (Mathematiques,
Musique et relations avec d'autres disciplines), Ircam, 2003.
[428] NOLAN (Catherine), "Structural Levels and Twelve-Tone Music: A Revisionist
Analysis of the Second Movement of Webern's Piano Variations Opus 27", Journal
of Music Theory, 39(1), 1995, pp. 47-76.
[429] NOLL (Thomas), Morphologische Grundlagen der abendldndischen Harmonik, Doc-
toral Thesis, TU Berlin, 1995, publ. in Musikometrika, 7, 1997.
[430] NOLL (Thomas), "Harmonische Morpheme", Musikometrika 8, 1997, pp. 7-32.
[431] NOLL (Thomas), "Geometry of Chords", in Puebla, Emilio Lluis, ed. Electronic
Bulletin of the Sociedad Matematica Mexicana, 1, 2001, to appear.
[432] NOLL (Thomas), "Tone Apperception and Weber-Fechner's Law", MaMuX Semi-
nar (Mathematiques, Musique et relations avec d'autres disciplines), Ircam, 2002.
[433] NOLL (Thomas), "Vade Mecum to Mathematical Music Theory", in Mazzola, Noll
and Lluis-Puebla, eds. Perspectives in Mathematical and Computational Music
Theory, Epos, 2004, pp. 14-28.
[434] NOLL (Thomas), BRAND (Monica), "Morphology of Chords", in Mazzola, Noll and
Lluis-Puebla, eds. Perspectives in Mathematical and Computational Music Theory,
Epos, 2004, pp. 366-398.
[435] NOLL (Thomas), GARBERS (J5rg), "Harmonic Path Analysis", in Mazzola, Noll
and Lluis-Puebla, eds. Perspectives in Mathematical and Computational Music
Theory, Epos, 2004, pp. 399-431.
320
Bibliography
[436] NORDEN (N. Lindsey), "A New Theory of Untempered Music", The Musical Quar-
terly 22(2), 1936, pp. 217-233.
[437] NORDEN (N. Lindsey), "A New Study of Intervals", Organ Institute Quarterly,
5(1), 1955, p. 31.
[438] NULL (Edwin von der), Moderne Harmonik, Leipzig, Pr. Kistner und C.F.W. Siegel,
1932.
[439] OBOUHOV (Nicolas), "Le livre de vie", Paris, La Revue Musicale, 290-291, 1972,
pp. 1-73.
[440] OETTINGEN (Arthur von), Harmoniesystem in Dualer Entwicklung: Studien zur
Theorie der Musik, Leipzig, W. Glaser, 1866.
[441] OTTERSTROM (Thorvald), A Theory of modulation, University of Chicago, 1935.
[442] OzONAM (Jacques), MONTUCLA (Jean-Etienne), Recreations mathematiques et
physiques, Paris, 1790.
[443] PAPADOPOULOS (Athanase), "Mathematics Music and J.S. Bach", Rend. Sem. Fac.
Sci. Univ. Caligari, 70(2), 2000, pp. 1-20.
[444] PAPADOPOULOS (Athanase), "Mathematics and music theory: from Pythagoras to
Rameau", Math. Intelligencer, 24(1), 2002, pp. 65-73.
[445] PAPADOPOULOS (Athanase), "The Mathematics of the Music of Olivier Messiaen",
Lett. Mat. Pristem., 47, 2003, pp. 27-41.
[446] PAPADOPOULOS (Athanase), "Consonance musicale et complexity mathematique",
LVuvert, 112, 2005, pp. 47-68.
[447] PARTCH (Harry), Genesis of a Music: An Account of a Creative Work, its Roots
and its Fulfillment, 2nd enl. edition, Da Capo Press, New York, 1974. Reprint,
1979. First edition 1949.
[448] PARZYSZ (Bernard), Musique et mathematique, Publication de PA.P.M.E.P., 53,
Paris, 1984.
[449] PECK (Robert W.), "Toward an Interpretation of Xenakis's Nomos Alpha", Pers-
pectives of New Music, 41(1), 2003, pp. 106-157.
[450] PECK (Robert W.), "Klein-Bottle Tonnetze", Music Theory Online, 9(3), 2003.
[451] PERLE (Georg), Serial Composition and Atonality, Los Angeles, University of Ca-
lifornia Press, 1962.
[452] PERLE (Georg), "Pitch Class Set Analysis: An Evaluation", Journal of Musicology,
8(2), 1990, pp. 151-172.
[453] PERRON (Oskar), "Grundlagen fur eine Theorie des Jacobischen Kettenbruchalgo-
rithmus", Math. Ann., 64, 1907, pp. 1-76.
[454] P I E R C E (John R.), The Science of Musical Sound, New York, Scientific American
Books, 1983.
[455] P I E R C E (John R.), MATHEWS (Max), "Control of Consonance and Dissonance with
Nonharmonic Overtones", Music by Computers, von FOrster and Beauchamp eds,
New York, 1969, pp. 129-132.
[456] PlKLER (Andrew G.), "History of Experiments on the Musical Interval Sense",
Journal of Music Theory, 10(1), 1966, pp. 54-95.
[457] PlRRO (Andr6), Descartes et la musique, Paris, 1907. New edition, 1974.
321
Bibliography
[458] PLOMB (R.), LEVELT (W.), "Tonal Consonance and Critical Bandwith", Journal
Acoustical Society of America, 38, 1965, pp. 548-560.
[459] POLANSKY (Larry), "Notes of a Lousy Carpenter", One/one, 1(1), 1985.
[460] POLANSKY (Larry), "The HMSL Intonation Environment", One/one, 3(1), 1987,
pp. 4-12.
[461] POLANSKY (Larry), "Paratactical Tuning: An Agenda for the Use of Computers
in Experimental Intonation", Computer Music Journal, 11(1), 1987, pp. 61-68.
[462] POLE (William), The Philosophy of Music, London, TrUbner and Co, 1879.
[463] P6LYA (George), "Kombinatorische Anzahlbestimmungen fur Gruppen, Graphen
und chemische Verbindungen", Acta Mathematica, 68, 1937, pp. 145-254.
[464] P6LYA (George), READ (R. C), Combinatorial Enumeration of Groups, Graphs,
and Chemical Compounds, Springer Verlag, 1987.
[465] POSNER (Roland) et al, Semiotics: A Handbook of the Sign-Theoretic Foundations
of Nature and Culture, Berlin, Walter de Gruyter, 1998.
[466] QuiNN (Ian), "Listening to Similarity Relations", Perspectives of New Music, 39,
2001, pp. 108-158.
[467] RAHN (John), Basic Atonal Theory, New York, Longman, 1980.
[468] RAHN (John), "On Some Computational Models of Music Theory", Computer Mu-
sic Journal, 4(2), 1980, pp. 66-72.
[469] RAHN (John), "Notes on Methodology in Music Theory", Journal of Music Theory,
33(1), 1989, pp. 143-154.
[470] RAHN (John), "New Research Paradigms", Music Theory Spectrum, 2(1), 1989,
pp. 84-94.
[471] RAHN (John), Music Inside Out Going too far in musical essays, Overseas Publi-
sher Association, 2001.
[472] RAICHEL (Daniel R.), The Science and Applications of Acoustics, Amer. Inst, of
Physics, 2000.
[473] RAMEAU (Jean Philippe), Traits de Vharmonie rMuite a ses principes naturels,
Paris, 1722.
[474] RAMEAU (Jean Philippe), Nouveau systeme de musique tMorique et pratique, pour
servir d'introduction au traite de VHarmonie, Paris, Ballard, 1726.
[475] RAMEAU (Jean Philippe), Generation Harmonique ou traite de musique tMorique
et pratique, Paris, Prault, 1737.
[476] RAMEAU (Jean Philippe), Demonstration duprincipe de Vharmonie, Paris, Durand
et Pissot, 1750.
[477] RAPOPORT (Paul), "The Structural Relationships of Fifths and Thirds in Equal
Temperaments", Journal of Music Theory, 37(2), 1993, pp. 351-389.
[478] RAPOPORT (Paul), "Just Shape, Nothing Central", MusicWorks, 60, 1994,
pp. 42-49.
[479] RAPOPORT (Paul), "Gimme Fives", MusicWorks 61, 1995, pp. 40-45.
[480] RASCH (Rudolph), "Wohltemperit en Gelijkzwevend", Mens and Melodie, 36,
Utrecht, 1981, pp. 264-273
322
Bibliography
[481] RASCH (Rudolph), "Gelijkzwevend stemmen", Mens and Melodie, 37, Utrecht,
1982, pp. 14-30.
[482] RASCH (Rudolph), "Description of Regular Twelve-Tone Musical Tunings", Jour-
nal of the Acoustical Society of America, 73(3), 1983, pp.1023-1035.
[483] RASCH (Rudolph), "Priedrich Suppig and His Labyrinthus musicus of 1722", The
Organ Yearbook, XV, 1984, pp. 33-53.
[484] RASCH (Rudolph), "Does Well-Tempered Mean Equal-Tempered ?" in Bach, Han-
del, Scarlatti - Tricentenary Essays, Peter Williams, Cambridge University Press,
1985, pp. 293-310.
[485] RASCH (Rudolph), "Relations between Multiple Divisions of the Octave and the
Traditional Tonal System", Interface, 14, 1985, pp. 75-108.
[486] RASCH (Rudolph), "Integer Approximation of Musical Scale Frequencies", Inter-
face, 15, 1986, pp. 25-34.
[487] RASCH (Rudolph), ed. Corpus Microtonale, vol. 1: Adriaan Daniel Fokker - Selected
Compositions, with introduction by Rudolf Rasch, Utrecht, 1987.
[488] RASCH (Rudolph), "The Theories of Adriaan Fokker, Part One", One/one, 4(1),
1988.
[489] RASCH (Rudolph), "The Theories of Adriaan Fokker, Part Two", One/one, 4(2),
1988.
[490] RAYLEIGH (John William), The Theory of Sound, London, 2 vols., Macmillan,
Second Edition, 1894.
[491] READ (Ronald C ) , "Combinatorial problems in the theory of music", Discrete
Mathematics, 167-168(1-3), 1997, pp. 543-551.
[492] REGENER (Eric), "The Number Seven in the Theory of Intonation", Journal of
Music Theory, 19(1), 1975, pp. 140-135.
[493] REINER (David), "Enumeration in Music Theory", Amererican Mathematical
Monthly, 92, 1985, pp. 51-54.
[494] R E T I (Rudolph), Tonality, Atonality, Pantonality: a study of some trends in Twen-
tieth Century music, New York, Macmillan, 1958.
[495] RlCHTER-HERF (Franz), ed. Mikrotone: Bericht ilber das 1. Internationale Sym-
posium Mikrotonforschung, Musik mit Mikrotonen, Innsbruck, Helbling, 1986.
[496] R I C H T E R - H E R F (Franz), ed. Mikrotone II: Bericht iiber das 2. Internationale Sym-
posium Mikrotonforschung, Musik mit Mikrotonen, Innsbruck, Helbling, 1988.
[497] R I C H T E R - H E R F (Franz), ed. Mikrotone III: Bericht iiber das 3. Internationale Sym-
posium Mikrotonforschung, Musik mit Mikrotonen, Innsbruck, Helbling, 1990.
[498] R I C H T E R - H E R F (Franz), ed. Mikrotone IV: Bericht iiber das 4- Internationale Sym-
posium Mikrotonforschung, Musik mit Mikrotonen (2-5 May 1991 in Salzburg),
Musikedition Nymphenburg, 1993.
[499] RlEMANN (Hugo), Musikalische Logik: Hauptziige der physiologischen und psycho-
logischen Begriindung unseres Musiksystems, Leipzig, C. Kahnt, 1873.
[500] RlEMANN (Hugo), Musikalische Syntaxis: Grundrifi einer harmonischen Satzbil-
dungslehre, Leipzig, Breitkopf und Hartel, 1877.
[501] RlEMANN (Hugo), Skizze einer Neuen Methode der Harmonielehre, Leipzig, Breit-
kopf und Hartel, 1880.
323
Bibliography
[502] RlEMANN (Hugo), Grosse Kompositionslehre, vol 1: Der homophone Satz (Melo-
dielehre und Harmonielehre), Berlin, Max Hesse, 1902.
[503] RlEMANN (Hugo), Handbuch der Harmonielehre, Leipzig, Breitkopf und Hartel,
5/1912, 6/1918.
[504] RlOTTE (Andre), Generation des cycles equilibres, Ispra, Euratom, Rapport interne
353, 1963.
[505] RlOTTE (Andre), Formalisation des structures musicales, PhD Thesis, Universite
Paris VIII, Departement d'informatique, Paris, 1979.
[506] RlOTTE (Andre), Formalisation des structures musicales, Lectures, Universite Pa-
ris VIII, Departement d'informatique, Paris, 1978-1990.
[507] RlOTTE (Andre), "Formalisation des echelles de hauteurs en analyse et en compo-
sition", Musique et assistance informatique, MIM, Marseille, 1990.
[508] RlOTTE (Andre), "L'utilisation des modeles mathematiques en analyse et en com-
position musicale, Quadrivium, Musique et Sciences, IPMC, Paris, 1992.
[509] RlOTTE (Andre), "Organisations du temps autour de la serie, de Webern a Boulez",
Les cahiers de philosophic, 20, La loi musicale, Lille, 1996.
[510] RlOTTE (Andre), "Mathematique du son, musique du nombre", in Genevois and
Orlarey, Musique et mathematiques, Aleas, 1997, pp. 11-25.
[511] RlOTTE (Andre), MESNAGE (Marcel), Formalismes et modules, Collection Mu-
sique/Sciences, Editions Delatour/Ircam, 2006.
[512] RlSSET (Jean-Claude), "Paradoxes de hauteurs", Proceedings of the 1th Interna-
tional Congress of Acoustics, Budapest, 1971.
[513] RlSSET (Jean-Claude), "Hauteur et timbre", Revue d'Acoustique, 42,1977, pp. 263-
268.
[514] RlSSET (Jean-Claude), "Musique, recherche, theorie, espace, chaos", Paris, Ircam,
Inharmoniques, 8-9, 1991, pp. 272-316.
[515] ROSCH (Eleanor), "Family resemblances: studies in internal structure of catego-
ries", Cognitive Psychology, 7, 1975, pp. 573-605.
[516] ROEDER (John), A Theory of Voice Leading for Atonal Music, Ph.D. diss., Yale
University, 1984.
[517] ROEDER (John), "A Geometric Representation of Pitch-Class Series", Perspectives
of New Music, 25, 1987, pp. 362-409.
[518] ROEDER (John), "A MaMuTh Achievement", Perspectives of New Music, 31(2),
1993, pp. 294-312.
[519] ROGERS (David W.), "A Geometric Approach to Pcset Similarity", Perspectives
of New Music, 37(1), 1999, pp. 77-90.
[520] ROMANOWSKA (Anna), "Algebraic Structure of the Tone System", Demonstratio
Mathematica, 7(4), 1974, pp. 525-542.
[521] ROMANOWSKA (Anna), "Algebraic Language in the Theory of Harmony", Tatra
Mt. Math. PubL, 23, 2001, pp. 113-123.
[522] ROSSING (Thomas D.), The Science of Sound, Addison-Wesley, Second Edition,
1990.
[523] ROSSING (Thomas D.), Science of Percussion Instruments, World Scientific, 2000.
324
Bibliography
[524] ROTHENBERG (David), "A Mathematical Model for Perception Applied to the
Perception of Pitch", in G. Goos and J. Harmanis, Lecture Notes in Computer
Science: Formal Aspects of the Cognitive Process, 1975, pp. 127-141.
[525] ROUSSEAU (Jean-Jacques), Dictionnaire de musique, Paris, 1767.
[526] SALMON (Thomas), A Proposal to Perform Musick in Perfect and Mathematical
Proportions, London, 1688.
[527] SALMON (Thomas), "The Theory of Music reduced to aritmetical and geometrical
proportions", Philosophical Transactions of the Royal Society of London, 24(302),
1705, pp. 2072-2100.
[528] SAMSON (Jim), Music in Transition: a study of tonal expansion and atonality,
1900-1920, New York, W.W. Norton, 1977.
[529] SANDS (Arthur D.), "On the factorization of finite abelian groups", Acta Math.
Acad. Sci. Hungar., 8, 1957, pp. 65-86.
[530] SANDS (Arthur D.), "The Factorisation of Abelian Groups", Quart. J. Math. Ox-
ford, 2(10), 1959, pp. 81-91.
[531] SANDS (Arthur D.), "On the Factorisation of Abelian Groups II", Acta Math. Acad.
Sci. Hungar., 13, 1962, pp. 153-159.
[532] SANDS (Arthur D.), "On Keller's conjecture for certain cyclic groups", Proc. Edin-
burgh Math. Soc., 22(2), 1977, pp. 17-21.
[533] SAUVEUR (Joseph), TraitS de la theorie de la musique, Paris, 1697.
[534] SAUVEUR (Joseph), "Sur la determination d'un son fixe", Histoire de VAcademic
Royale des Sciences, Paris 1700. Reprinted by R. Rasch, The Diapason Press,
Utrecht, 1988.
[535] SAUVEUR (Joseph), "Rapport des sons des cordes des instruments de musique, aux
fleches des cordes et nouvelle determination des sons fixes", Histoire de VAcad&mie
Royale des Sciences, Annee 1713, Paris 1716. Reprinted by R. Rasch, The Diapason
Press, 1988, pp. 324-349.
[536] SAUVEUR (Joseph), "Sur Papplication des sons harmoniques aux jeux de POrgue",
Histoire de VAcademie Royale des Sciences, Paris, Hachereau, 1720. Reprinted by
R. Rasch, The Diapason Press 1988.
[537] SAUVEUR (Joseph), "Methode generate pour former des systemes temperes de mu-
sique", Histoire de VAcademie Royale des Sciences, Annee 1707, Paris, Martin,
1730, pp. 203-222. Reprinted by R. Rasch, The Diapason Press 1988.
[538] SAUVEUR (Joseph), "Table generate des systemes temperes de musique", Histoire
de VAcademie Royale des Sciences", Annee 1711, Paris 1730, pp. 307-316. Reprin-
ted by R. Rasch, The Diapason Press, Utrecht, 1988.
[539] SAUVEUR (Joseph), "Systeme general des intervalles et des sons et son applicalion
a tous les systemes et a tous les instruments de musique", Histoire de VAcad&mie
Royale des Sciences, Annee 1701, Paris 1743, pp. 299-966. Reprinted by R. Rasch,
The Diapason Press, 1988.
[540] SCABOROUGH (Don L.), "Connectionist Models for Tonal Analysis", Computer
Music Journal, 13, 1989, pp. 49-55.
[541] SCHAUB (Stephan), Algebraic Structures and Symmetry in Western Music, Master
Thesis, Department of Mathematics, University of Arizona, 1996.
325
Bibliography
326
Bibliography
[563] STARR (Daniel), MORRIS (Robert), "A General Theory of Combinatoriality and
the Aggregate", Perspectives of New Music, 1977-78, 16(1), pp. 364-389 and 16(2),
pp. 50-84.
[564] STEIN (Sherman K.), SZAB6 (Sandor), Algebra and Tiling, The Carus Mathema-
tical Monographs, 25, 1994.
[565] STOEKER (Philip), "Klumpenhouwer Networks, Trichords and Axial Isography",
Music Theory Spectrum, 24(2), 2002, pp. 231-245.
[566] STRAUB (H.), Beitrage zur modultheoretischen Klassifikation musikalischer Motive,
Diplomarbeit ETH Zurich, 1989.
[567] STRAUS (Joseph Nathan), Introduction to post-tonal theory, Englewood Cliffs,
Prentice Hall, 1990.
[568] STUMPF (Carl), "Konsonanz und Dissonanz", Beitr. Akust. Musikwissenshaft, 1,
1898, pp. 1-108.
[569] SWIFT (Richard), "Some Aspects of Aggregate Composition", Perspectives of New
Music, 14(2) and 15(1), 1976, pp. 236-248.
[570] SZABO (Sandor), "A type of factorization of finite abelian groups", Discrete Math,
54, 1985, pp. 121-124.
[571] TAMBA (Akira), La th6orie et Vesthetique musicale japonaises, Paris, Publications
Orientalistes de Prance, 1988.
[572] TANAKA (Shohe), "Studien im Gebiete der reinen Stimmung", Vierteljahrsschrift
fur Musikwissenschaft, 6, 1890, pp. 1-90.
[573] TANGIAN (Andranik), "The Sieve of Erathosthene and Johnson's Problem on Ryth-
mic Canons", Discussion paper No. 309, Fern University of Hagen, 2001.
[574] TANNER (Robert), "Le r6le de la fusion dans Pappreciation de la justesse des ac-
cords", in Colloque Acoustique musicale, Paris, CNRS, 1959, pp. 49-59.
[575] TANNER (Robert), La musique antique grecque, expliquie par une consequence de
la tMorie psycharithmetique, Paris, Richard Masse, 1961.
[576] TANNER (Robert), "Critique de la theorie de la resonance", in La resonance dans
les €chelles musicales, Colloque du CNRS, Paris, 1963, pp. 63-100.
[577] TANNER (Robert), "Le probleme des deux tierces (Pythagore et Zarlino). Sa solu-
tion psycharithmetique et experimentale", Acustica, 27(6), 1972, pp. 335-352.
[578] TANNER (Robert), "La fonction et la justesse melodique des intervalles", Acustica,
34(5), 1976, pp. 259-268.
[579] TEITELBAUM (Richard), "Intervallic Relations in Atonal Music", Journal of Music
Theory, 9(1), 1965, pp. 72-127.
[580] TENNEY (James), A History of Consonance and Dissonance, New York, Excelsior
Music Pub., 1988.
[581] TERHARDT (Ernst), "Pitch, Consonance and Harmony", Journal of the Acoustical
Society of America, 55, 1974, pp. 1061-1069.
[582] TlJDEMAN (Robert), "Decomposition of the integers as a direct sum of two sub-
sets", Number Theory (Paris 1992-1993), London Math. Soc. Lecture Note Ser.
215, Cambridge University Press, 1995, pp. 261-276.
[583] TREU (Abdias), Directorium mathematicum ad cuius ductum et informationem
tota mathesis Harmonica, Altdorf, 1657.
327
Bibliography
328
Bibliography
[602] WILCOX (Howard J.), "Group tables and the generalized hexachord theorem",
Perspectives of New Music, 21(1-2), 1983, pp. 535-539.
[603] WILD (John), "Tessellating the chromatic", MaMuX Seminar (Mathematiques,
Musique et relations avec d'autres disciplines), 2002.
[604] WlLLE (Rudolf), "Mathematische Sprache in der Musiktheorie", Feedback Papers
24, 1981, pp. 6-10.
[605] WlLLE (Rudolf), "Mathematik und Musiktheorie", in Musik und Zahl, Verlag fur
systematische Musikwissenschaft GmbH, Bonn, 1976, pp. 233-264.
[606] WILSON (Ervin M.), "The Bosanquetian 7-rank Keyboard After Poole and Brown",
Xenharmonikdn, 1, 1972.
[607] WILSON (Ervin M.), "On the Development of Intonational Systems by Extended
Linear Mapping", Xenharmonikdn, 3, 1975.
[608] WILSON (Ervin M.), "Pythagorean Intervals", Interval, 1(3), 1979.
[609] WILSON (Ervin M.), "The Purvi Modulations", Xenharmonikdn, 10, 1982.
[610] WILSON (Ervin M.), "The Marwa Permutations", Xenharmonikdn, 9, 1986.
[611] WILSON (Ervin M.), "The Generalized Marimba Keyboard", Interval, 1986-1987,
pp. 8-9.
[612] WILSON (Ervin M.), "D'Alessandro: Like A Hurricane", Xenharmonikdn, 12, 1990,
pp. 1-39.
[613] WILSON (Ervin M.), "Partch's Scale on the Bosanquet Generalized Keyboard",
One/one, 1(4), 1985.
[614] WiTTSTEiN (Theodor), Grundzuge der mathematisch-physikalischen Theorie der
Musik, Hannover, Hahn'sche Buchhandlung, 1888.
[615] W O L F (Daniel), "Four Tunings for Sound Environments", One/one, 6 (3), 1990.
[616] WURSCHMIDT (Joseph), "Logarithmische und graphische Darstellung der musika-
lischen Intervalle", Zeitschrift filr Physik 3, 1920, pp. 89.
[617] WURSCHMIDT (Joseph), "Viertel und Sechsteltonmusik, eine kritische Studie",
Neue Musik-Zeitung, 42, 1921, pp. 183.
[618] WURSCHMIDT (Joseph), "Die 19-stufige Skala", Neue Musik-Zeitung, 42, 1921,
pp. 215.
[619] WURSCHMIDT (Joseph), "Die rationellen Tonsysteme in Quinten-Terzengewebe",
Zeitschrift filr Physik, 41, 1928, p. 527.
[620] WURSCHMIDT (Joseph), "Tonleitern, Tonarten und Tonsysteme", Sitzungsberichte
der Physikalisch-medizinischen Sozietdt zu Erlangen, Band 63, 1931.
[621] WYSCHNEGRADSKY (Ivan), "Etude sur Pharmonie des quartes superposees", Le
MSnestrel, Paris, 12 and 19 april 1935.
[622] WYSCHNEGRADSKY (Ivan), "Preface a un Traite d'harmonie par quartes superpo-
sees", Polyphonie, Paris, 1944, pp. 56-62.
[623] WYSCHNEGRADSKY (Ivan), "L'ultrachromatisme et les espaces non-octaviants",
Paris, La Revue Musicale, 290-291, 1972, pp. 73-130.
[624] WYSCHNEGRADSKY (Ivan), La loi de la Pansonorite', Edited by Pranck Jedrze-
jewski with the collaboration of Pascale Criton, Contrechamps, Genova, 1996.
329
Bibliography
330
Index
332
Index
333
Index
Tangles, 121
Tcherepnin Alexander, 46, 49
Tilings, 180, 195
Tippett Michael, 53
Tonal Gravity, 21
Tonality Diamond, 163
Tonic Pole, 25
334
Dep6t legal: 2 e trimestre 2006
Imprime en Prance
•••••••••••••
ill SBifii
£1
:ss jjj[:jjjjjjjj:jj[j![[[:! II
J •••••••••••••••••IIIIIIIIIIilillHIl
JML
!•••••••••••••• ••••••••••••••••••••
(•••••Ml
i?• • • • • • • • • • • • • • • • • • a ••«••••••••
(•••••••I
•••• •••••
SI!!!!!![[ !•••
••••••••
Franck Jedrzejewski
With the participation of Ircam and the support of CNRS and SFAM
:s
•••••••••i
!iB
CENTRE NATIONAL
DE LA RECHERCHK
SCIENTIFIOUE
• •••••••!
w
Miiiiii m
.. •••••*••••*••••••
Nxpu^'W
•••••••••••••••••.
IIISHI:
Dep
!••••