# Music and Mathematics

Thomas M. Fiore
fioret@umich.edu
Contents
Introduction 5
Lecture 1 Transposition and Inversion 7
1. Introduction 7
2. Mathematical Preliminaries 7
3. The Integer Model of Pitch 11
4. Fugue by Bach 12
5. Tristan Prelude from Wagner 14
Lecture 1 Homework Problems 15
1. Introduction 15
2. Mathematical Preliminaries 15
3. The Integer Model of Pitch 15
4. Fugue by Bach 16
5. Tristan Prelude from Wagner 16
1. Introduction 17
2. Fugue by Hindemith 17
Addendum to Lecture 1 Homework Problems 22
1. Introduction 22
2. Fugue by Hindemith 22
Lecture 2 The PLR Group 24
1. Introduction 24
2. Mathematical Preliminaries 24
3. The PLR Group 27
4. Elvis and the Beatles 28
5. Topology and the Torus 28
6. Beethoven’s Ninth Symphony 29
7. Conclusion 32
Lecture 2 Homework Problems 34
1. Introduction 34
2. Mathematical Preliminaries 34
3. The PLR Group 34
4. Elvis and the Beatles 34
5. Topology and the Torus 34
6. Beethoven’s Ninth Symphony 34
7. Conclusion 35
3
CONTENTS 4
Bibliography 36
Introduction
Music and Mathematics are intricately related. Strings vibrate at certain fre-
quencies. Sound waves can be described by mathematical equations. The cello has
a particular shape in order to resonate with the strings in a mathematical fashion.
The technology necessary to make a digital recording on a CD relies on mathe-
matics. After all, mathematics is the language that physicists utilize to describe
the natural world and all of these things occur in the natural world. Not only do
physicists, chemists, and engineers use math to describe the physical world, but
also to predict the outcome of physical processes.
Can one similarly ﬁnd an “equation” to describe a piece of music? Or better
yet, can one ﬁnd an “equation” to predict the outcome of a piece of music? We can
model sound by equations, so can we also model works of music with equations?
Music is after all just many individual sounds, right? Should we invest time and
money to ﬁnd these equations so that all of humankind can enjoy predictable, easily
described music?
The answer to all of these questions is predictable and easily described: a
series of emphatic “NO’s”! There is not an equation that will model all works of
music and we should not spend time looking for it. Nevertheless, there are certain
mathematical structures inherent in all works of music, and these mathematical
structures are not given by equations. The language of mathematics is a convenient
tool for comprehending and communicating this underlying structure.
In fact, one of the central concerns of music theory is to ﬁnd a good way to hear
a piece of music and to communicate that way of hearing.
1
Anyone who has ever
heard Stockhausen’s Klavierst¨ uck III (1952) knows that this is not always so easy to
do! On a higher level, the eighteenth century Scottish philosopher David Hume said
that the mind receives impressions and once these impressions become tangible and
vivacious, they become ideas. Music theory supplies us with conceptual categories
to organize and understand music. Our aural impressions become vivacious ideas
by way of these conceptual categories. To ﬁnd a good way of hearing a musical
piece means to comprehend the music in such a way as to make it tangible.
Music theorists often draw on the formidable powers of mathematics in their
creation of conceptual categories. The discrete whole numbers . . . , −2, −1, 0, 1, 2 . . .
are particularly well suited for labelling the pitches, or the keys of the piano. The
area of mathematics called combinatorics enables one to count the many ways of
combining pitches, i.e. numbers. This provides taxonomies and classiﬁcations of the
various sets that arise. Group theory, another area of pure mathematics, describes
the ways that sets and pitches relate and how they can be transformed from one to
1
John Rahn. Basic Atonal Theory. New York: Schirmer Books, 1980. See page 1.
5
INTRODUCTION 6
the other. It is in this sense that pure mathematics provides a convenient framework
for the music theorist to communicate good ways of hearing a work of music.
Music theory is not just for the listener. Music theory is also useful to the
composer. Bach, Mozart, and Beethoven were well versed in the music theory of
their respective epochs and applied it daily. This abstract discussion has relevance
for the performing artist as well. A classical pianist may play thousands upon
thousands of notes in one concert, all from memory. How does a classical pianist
do that? Is it necessary for the classical pianist to memorize each individual note?
Of course not. Music theory provides the performing artist with an apparatus for
pattern recognition, and this strengthens the musical memory. A piece of music
does not consist of many individual isolated sounds, but rather several ideas woven
together. One can ﬁnd the thread of the musical fabric with music theory. On the
other hand, this same musical memory allows the listener to conceive of a piece of
music as a whole, rather than isolated individual events. Music theory is not limited
to classical music. Jazz musicians also use music theory in their improvisations and
compositions. Non-western music also lends itself to analysis within the framework
of music theory.
In this module we investigate some of the group theoretical tools that music
theorists have developed in the past 30 years to ﬁnd good ways of hearing particular
works of music. Group theory does not provide us with equations to describe a piece
of music or predict its outcome. Instead, it is just one conceptual category that
listeners, composers, and performers can use to make sense of a work of art and to
communicate ways of hearing to others. The aural impression of a piece of music
becomes an idea in the sense of Hume with this apparatus.
In the next two lectures we will study the T/I and PLR groups and use them
to analyze works of music from Johann Sebastian Bach, Ludwig van Beethoven,
Richard Wagner, and Paul Hindemith. This group theoretical point of view will
elevate our aural impressions to the status of ideas. We will conceive of the music
not as individual sounds, but at as a whole. The mathematical framework will
provide us with a way of hearing the pieces and a means of communicating this
hearing.
These notes were prepared for a series of three guest lectures in the undergrad-
uate course Math 107 at the University of Michigan under the direction of Karen
Rhea in Fall 2003. These lectures were part of my Fourth Year VIGRE Project
aimed at introducing undergraduates to my interdisciplinary research on generalized
contextual groups with Ramon Satyendra. I extend my thanks to Ramon Satyen-
dra of the University of Michigan Music Department for helpful conversations in
preparation of this module.
Lecture 1 Transposition and Inversion
1. Introduction
Some of the ﬁrst mathematical tools that music students learn about are trans-
position and inversion. In this introductory lecture we learn about the mathemat-
ical concepts necessary to formalize these musical tools. These concepts include
set, function, and modular arithmetic. Musicians are usually come into contact
with transposition and inversion in the context of pitch. To bridge the gap between
sound and number we will conceive of the integer model of pitch as musicians nor-
mally do. After the mathematical discussion we turn to examples from Bach and
Wagner.
2. Mathematical Preliminaries
The mathematical concepts of set, function, domain, range, and modular arith-
metic will be needed for our discussion.
2.1. Sets and Functions.
Definition 2.1. A set is a collection of objects. The collection of objects is
written between curly parentheses {}. The objects of the set are called elements.
Two sets are said to be equal if they have the same elements.
For example, the set {4, 5, 10} is the set consisting of the numbers 4, 5, 10 and
nothing else. The set {A, B} is the set whose elements are the two pitch classes
A and B. The symbols }6, 100, 11, 5} do not denote the set whose elements are
the numbers 6, 100, 11, 5 because one of the parentheses is the wrong way. The
order of the elements of a set does not matter. For example,
{4, 5, 10} = {5, 4, 10} = {10, 4, 5}.
The sets {56, 70} and {56} are not equal because they do not have the same elements
(the second set is missing 70).
A function gives us a rule for getting from one set to another.
Definition 2.2. A function f from a set S to a set S

is a rule which assigns
to each element of S a unique element of S

. This is usually denoted by f : S → S

.
This symbolism is read: “f goes from S to S

.” In this situation, the set S is called
the domain of the function f and S

is called the range of the function f. The
inputs are the elements of the domain. The outputs are the elements of range.
For example, consider the function f : {1, 2, 3} → {4, 5} deﬁned by
f(1) = 4
f(2) = 5
7
2. MATHEMATICAL PRELIMINARIES 8
f(3) = 4.
Here the domain is the set S = {1, 2, 3} and the range is the set S

= {4, 5}. These
three equations tell us the rule that assigns an output to each input. Note that
each element 1, 2, 3 of the domain gets a unique output, i.e. f(1) does not have two
diﬀerent values. In precalculus we like to say that the function passes the vertical
line test. The deﬁnition
g(1) = 4
g(1) = 5
g(2) = 4
g(3) = 4
does not deﬁne a function g : {1, 2, 3} → {4, 5} because two diﬀerent values 4 and
5 are assigned to 1. The deﬁnition
h(1) = 4
h(2) = 4
h(3) = 4
does deﬁne a function h : {1, 2, 3} → {4, 5} because a unique output is assigned to
each input. Note that the same output is assigned to each input and 5 is not even
used.
Functions can be composed, provided the domain of one is the range of the
other. For example, if j : {4, 5} → {7, 8, 9} is given by the rule
j(4) = 9
j(5) = 8
and f is as above, then we get a new function j ◦ f : {1, 2, 3} → {7, 8, 9} called the
composition of j with f deﬁned by j ◦ f(x) = j(f(x)). In this example we have
j ◦ f(1) = j(f(1)) = j(4) = 9
j ◦ f(2) = j(f(2)) = j(5) = 8
j ◦ f(3) = j(f(3)) = j(4) = 9.
Sometimes functions are given by formulas rather than tables. In the next
section we will see some functions that are given by formulas.
2.2. Modular Arithmetic.
Consider the face of a clock with the numbers 0 through 11 where 0 is in the
12 o’clock position. The day starts at midnight, so we have replaced 12 by 0 on
the usual face of a clock. Using the clock, we can determine what time it is 2 hours
after 1. We just go clockwise 2 notches after 1 and we get 3. Similarly 5 hours after
6 is 11. But what about 1 hour after 11? Well, that is 0 because we are back at
the beginning. Similarly, 2 hours after 11 is 1. This is called arithmetic modulo 12.
Summarizing, we can write
1 + 2 = 3 mod 12
6 + 5 = 11 mod 12
11 + 1 = 0 mod 12
11 + 2 = 1 mod 12.
2. MATHEMATICAL PRELIMINARIES 9
Whenever it is clear that one is working mod 12, we just leave oﬀ the suﬃx mod
12. So we have just done addition mod 12, let’s consider subtraction. If we bring
some of the numbers over to the right like usual arithmetic, we get
1 = 3 − 2
6 = 11 − 5
11 = 0 − 1
11 = 1 − 2
where we leave oﬀ mod 12 because it is clear by now that we are talking about
arithmetic modulo 12 in this paragraph and not arithmetic modulo 7. The ﬁrst
two equations appear to make sense to us from usual arithmetic. But to make
sense of the last two equations, we need to consider the face of the clock. If we are
at 0 o’clock and go back 1 hour, then we are at 11 o’clock. Similarly, if we are at 1
o’clock and go back to hours, we just go counterclockwise two notches on the face
of the clock and arrive at 11 o’clock. That is why 11 = 0 − 1 and 11 = 1 − 2.
Given any number, we can ﬁnd out what it is mod 12 by adding or subtracting
12 enough times to get a whole number between 0 and 11. For example,
−12 = 0 = 12 = 24
−13 = −1 = 11 = 23
−7 = 5 = 17 = 29.
As a result, mathematicians and musicians use the notation
Z
12
= {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11}
and call this set the set of integers mod 12.
Maybe you can guess that we are interested in integers mod 12 because there
are twelve keys on the piano from middle C to the next C not counting the last
C. All of this has an aural foundation. Our ears naturally hear pitches that are an
octave apart, i.e. pitches with 12 jumps (or intervals) between them on the piano.
Such pitches appear to be very similar for our ears. So in a sense, our human ears
are hardwired for arithmetic mod 12!
We are also interested in arithmetic modulo 7 because there are seven pitches
in the major scale, i.e. seven white keys on the piano from middle C to the next
C. For arithmetic modulo 7 we imagine there are seven hours in a day and that
the face of the clock goes from 0 to 6 instead of 0 to 11. Arguing as in the mod 12
case, we have
1 + 2 = 3 mod 7
6 + 5 = 4 mod 7
6 + 1 = 0 mod 7
6 + 2 = 1 mod 7.
We already see a diﬀerence between mod 12 and mod 7. Notice that
6 +5 =11 mod 12 but 6+5=4 mod 7. Let’s consider some examples of subtraction.
If we bring some of the numbers to the right like in usual arithmetic, we get
1 = 3 − 2 mod 7
6 = 4 − 5 mod 7
6 = 0 − 1 mod 7
6 = 1 − 2 mod 7.
2. MATHEMATICAL PRELIMINARIES 10
Subtraction can be understood by moving counterclockwise on the face of a clock
with seven hours labelled 0, 1, . . . , 6.
Given any number, we can ﬁnd out what it is mod 7 by adding or subtracting
7 enough times to get a number between 0 and 6. For example,
−7 = 0 mod 7
−9 = −2 = 5 mod 7
15 = 8 = 1 mod 7
17 = 10 = 3 mod 7.
As a result, mathematicians and musicians use the notation
Z
7
= {0, 1, 2, 3, 4, 5, 6}
and call this set the set of integers mod 7.
Next we can talk about functions f : Z
12
→ Z
12
. These are functions whose
inputs are integers modulo 12 and whose outputs are also integers modulo 12. Let’s
consider the function T
2
: Z
12
→ Z
12
deﬁned by the formula T
2
(x) = x + 2. Recall
that we considered functions given by tables in Subsection 2.2 above. We can make
a table for this function as follows.
T
2
(0) = 0 + 2 = 2 mod 12
T
2
(1) = 1 + 2 = 3 mod 12
T
2
(2) = 2 + 2 = 4 mod 12
T
2
(3) = 3 + 2 = 5 mod 12
T
2
(4) = 4 + 2 = 6 mod 12
T
2
(5) = 5 + 2 = 7 mod 12
T
2
(6) = 6 + 2 = 8 mod 12
T
2
(7) = 7 + 2 = 9 mod 12
T
2
(8) = 8 + 2 = 10 mod 12
T
2
(9) = 9 + 2 = 11 mod 12
T
2
(10) = 10 + 2 = 0 mod 12
T
2
(11) = 11 + 2 = 1 mod 12.
Another function I
0
: Z
12
→ Z
12
is given by the formula I
0
(x) = −x. For example
I
0
(1) = 11 and I
0
(6) = 6.
This concludes the introduction of sets, functions, and modular arithmetic
necessary for an understanding of transposition and inversion. Finally we can turn
to some music.
3. THE INTEGER MODEL OF PITCH 11
3. The Integer Model of Pitch
To make use of the mathematical ideas developed in the last section, we need to
translate pitch classes into numbers and introduce transpositions and inversions.
2
If you can’t read music, don’t panic, just use the integers modulo 12. If you can
read music, then the following well established dictionary shows us how to get from
pitches to integers modulo 12.
C = 0
C = D = 1
D = 2
D = E = 3
E = 4
F = 5
F = G = 6
G = 7
G = A = 8
A = 9
A = B = 10
B = 11
In this integer model of pitch, the C major chord {C, E, G} is {0, 4, 7}. This
C major chord is part of the main theme for Haydn’s Surprise Symphony. The
ﬁrst part of the main theme is C, C, E, E, G, G, E, F, F, D, D, B, B, G, which can
be written 0, 0, 4, 4, 7, 7, 4, 5, 5, 2, 2, 11, 11, 7. These angular brackets are often
used by music theorists to emphasize that the notes occur in this order. Recall that
the ordering does not matter for sets, because a set is just a collection of elements.
Unordered sets, e.g. {0, 4, 7}, are sometimes called pcsets (pitch class sets) while
ordered sets such as 0, 0, 4, 4, 7, 7, 4, 5, 5, 2, 2, 11, 11, 7 are called pcsegs (pitch class
segments).
Transpositions and inversions are functions Z
12
→ Z
12
that are useful to every
musician. There are also analogues for Z
7
. Transposition and inversion are often
applied to melodies, although they can also be applied to chords. When we hear a
melody consisting of several pitches, we hear the intervals between the individual
notes. The relationship between these intervals is what makes a melody appealing
to us. Transposition mathematically captures what musicians do all the time: the
restatement of a melody at higher and lower pitch levels in a way that preserves
intervals. Inversion is another way to create musical variation while preserving the
intervallic sound of a melody, although it does not preserve the exact intervals.
Definition 3.1. Let n be an integer mod 12. Then the function T
n
: Z
12
→ Z
12
deﬁned by the formula T
n
(x) = x + n mod 12 is called transposition by n.
2
In music theory, particularly in atonal theory, it is common to study pitch classes rather than
pitches. One can see the diﬀerence between pitch classes and pitches in the following example.
Middle C is a particular pitch, although the pitch class C refers to the aggregate of all keys on
the piano with letter name C.
4. FUGUE BY BACH 12
We already came into contact with T
2
: Z
12
→ Z
12
in the previous section.
Some examples for T
5
: Z
12
→ Z
12
are
T
5
(3) = 3 + 5 = 8
T
5
(6) = 6 + 5 = 11
T
5
(7) = 7 + 5 = 0
T
5
(10) = 10 + 5 = 15 = 3
where we have not written mod 12 because it is clear from the context.
Definition 3.2. Let n be an integer mod 12. Then the function I
n
: Z
12
→ Z
12
deﬁned by the formula I
n
(x) = −x + n is called inversion about n.
We already came into contact with I
0
: Z
12
→ Z
12
in the previous section.
Some examples for I
7
: Z
12
→ Z
12
are
I
7
(3) = −3 + 7 = 4
I
7
(7) = −7 + 7 = 0
I
7
(9) = −9 + 7 = −2 = 10.
The function I
n
is called inversion about n because it looks like a reﬂection about
n whenever one draws the number line.
Music theorists and composers like to transpose and invert entire pcsets or
pcsegs by applying the function to each element. For example, we can transpose a
C major pcset by 7 steps as in
T
7
{0, 4, 7} = {T
7
(0), T
7
(4), T
7
(7)} = {0 + 7, 4 + 7, 7 + 7} = {7, 11, 2}
by applying T
7
to each of 0,4, and 7. A musician would notice that this takes the
C major chord to the G major chord. Similarly, we could invert the pcseg for the
theme of Haydn’s Surprise Symphony about 0, although Haydn did not do this!
I
0
0, 0, 4, 4, 7, 7, 4, 5, 5, 2, 2, 11, 11, 7 = 0, 0, 8, 8, 5, 5, 8, 7, 7, 10, 10, 1, 1, 5
In this section we have introduced the integer model of pitch, which assigns to
each of the 12 pitch classes an integer mod 12. The transpositions and inversions
are functions which have inputs and outputs that are pitches. These are conceptual
categories that music theorists use to ﬁnd good ways of hearing pieces. Next we
use them to ﬁnd good ways of hearing a fugue and a prelude.
4. Fugue by Bach
Johann Sebastian Bach (1685-1750) took the art of fugue to new heights. He
composed the Well-Tempered Clavier Book I and the Well-Tempered Clavier Book
II, each of which contains 24 preludes and fugues. A fugue usually begins with a
statement of the main theme called the subject. This subject returns over and over
again in various voices and usually they are thread together in complex way. Every
fugue has occurrences of transposition and inversion. A truly fascinating website
http : //jan.ucc.nau.edu/ ∼ tas3/bachindex.html
on Bach describes in detail what a fugue is. Click on the link for movies on the
Well-Tempered Clavier. There is an animation and recording for Fugue 6 in d
minor of the Well-Tempered Clavier Book I, which we now analyze. Our analysis
4. FUGUE BY BACH 13
will be restricted to ﬁnding some transpositions and inversions, since we are only
studying some of the mathematical structure. We’ll leave detailed analysis to the
music theorists.
The subject of the fugue is the pcseg
D, E, F, G, E, F, D, C, D, B, G, A = 2, 4, 5, 7, 4, 5, 2, 1, 2, 10, 7, 9
which begins in measure 1 and lasts until the beginning of measure 3. Let’s call
this pcseg P. See the score that you got in class. Interestingly enough, this subject
consists of twelve notes! In measure 3, another voice sings the melody
A, B, C, D, B, C, A, G, A, F, D, E = 9, 11, 0, 2, 11, 0, 9, 8, 9, 5, 2, 4.
Do you see a relationship between this pcseg and P? Notice that this pcseg is
T
7
P! Just try adding 7 to each element of P and you will see it. In measure 6, the
subject returns in the exact same form as the introduction, just one octave lower.
At measure 8, a form
E, F, G, A, F, B, G, F, G, E, C, D = 4, 5, 7, 9, 5, 10, 7, 6, 7, 3, 1, 2
of the subject enters. This one doesn’t entirely match though. The ﬁrst ﬁve pitches
are almost T
2
of the ﬁrst ﬁve pitches of P, but the next 5 pitches are T
5
of the
respective pitches of P. The last pitch of the pcseg is also T
5
of the respective pcseg
of P, but the eleventh pitch doesn’t match. At measure 13 we have
A, B, C, D, B, C, A, G, A, F, D, E = 9, 11, 1, 2, 11, 1, 9, 8, 9, 5, 2, 4.
This is similar to T
7
P as in measure 3, except for the highlighted 1’s. Measures
17,18, and 21 are respectively
A, B, C, D, B, C, A, G, A, F, D, E = 9, 11, 0, 2, 11, 1, 9, 8, 9, 5, 2, 4
A, B, C, D, B, C, A, G, A, F, D, E = 9, 11, 1, 2, 11, 0, 9, 8, 9, 5, 2, 4
A, B, C, D, B, C, A, G, A, F, D, E = 9, 11, 1, 2, 11, 1, 9, 8, 9, 5, 2, 4.
These are also T
7
P except for the highlighted 1’s. The interval 7 is very important
in western music and is called the perfect ﬁfth. Here we see that transposition by
a perfect ﬁfth occurs four times before the piece is even half over. In fact, many
fugues have this property. So far we have seen that transposition plays a role in
this piece. But what about inversion?
Let’s consider measures 14 and 22. They are respectively
E, D, C, B, D, C, E, F, E, A, C, B = 4, 2, 1, 11, 2, 1, 4, 5, 4, 9, 0, 10
E, D, C, B, D, C, E, F, E, G, B, A = 4, 2, 1, 11, 2, 1, 4, 5, 4, 7, 10, 9.
They are nearly identical, except for the last three digits. Notice also that the ﬁrst
two elements E, D are the same ﬁrst two elements of P, just the order is switched.
The last three notes of 22 are even the last three notes of P, just the order is
switched. Calculating I
6
P gives
4, 2, 1, 11, 2, 1, 4, 5, 4, 8, 11, 9
which is a near perfect ﬁt with measures 14 and 22! Just the last three notes are
changed to make it sound better. So we see that inversion does indeed play a role
in the piece. The rest of the piece contains further transpositions and inversions of
the subject.
Next time we listen to the piece, we can listen for these transposed and inverted
forms of the subject. These conceptual categories make the piece more enjoyable for
5. TRISTAN PRELUDE FROM WAGNER 14
listeners because we come closer to understanding it. We have a good way of hearing
the piece. This knowledge also makes the piece easier for performers because they
recognized patterns and relationships between diﬀerent parts of the piece. However,
a music theorist would not be satisﬁed with this analysis because we have barely
scratched the surface. There is much more to this fugue than a few transpositions
and inversions. Nevertheless, this illustrates some of the mathematical features of
the piece.
5. Tristan Prelude from Wagner
Richard Wagner (1813-1883), who was born 63 years after the death of Bach, is
best known for his gigantic operas. Wagner’s compositions are drastically diﬀerent
from Bach’s. We take the prelude to the famous opera Tristan and Isolde as an
example for transposition and inversion. This particular passage is notorious for
3
More modern methods of atonal analysis,
which use transposition and inversion, are more fruitful. In this analysis we con-
sider unordered sets, i.e. pcsets, although we worked with pcsegs in the previous
example.
4
Consider the piano transcription of the ﬁrst few measures of the prelude. The
piano transcription entitled “Wagner: Tristan Prelude” is in the packet of music I
handed out in class. Let P
i
denote the set of pitch classes that are heard during
the circle i on the piano transcription. For example, P
2
= {F, B, D, A}. Then we
notice the following pattern after looking very carefully.
P
1
P
2
P
3
P
4
P
5
P
6
P
7
P
8
P
9
P
12
P
13
{0, 2, 5, 8} {0, 2, 6, 8} {0, 2, 6, 8} {0, 2, 5, 8}
This table means that all of the pcsets in the ﬁrst column can be transposed
or inverted to {0, 2, 5, 8}, all of the pcsets in the second column can be transposed
or inverted to {0, 2, 6, 8}, etc. Notice that the ﬁrst and last column are essentially
the same, while the middle two columns are essentially the same! Here, essentially
means they can be transposed or inverted to the same thing.
Notice also that everything is done according to the groups of circled notes in
the music, and we almost have three groups of four, which would give us 12 again!
The ﬁrst and last pitches of each four note group, namely G − B, B − D, and
D − F, also form a set that can be transposed or inverted to {0, 2, 5, 8}!
In other words, we mathematically see and musically hear a self similarity on
diﬀerent levels. When we listen to the piece again, we can listen for these features.
The conceptual categories of transposition and inversion provide us with a good
way of hearing these introductory measures to Wagner’s opera Tristan and Isolde.
Mathematics is the tool that we use to communicate this way of hearing to others.
3
This analysis is obtained from John Rahn page 78, who in turn quotes Benjamin Boretz.
4
What works for one piece of music may not work for another. In the Bach fugue it was
better to use pcsegs because the pcsets would tell us very little in that case. However, pcsets are
more appropriate for the Tristan prelude than pcsegs.
Lecture 1 Homework Problems
1. Introduction
1. Reread the introduction to this music module. Is music part of the physical
world? Write a short paragraph on this topic.
2. Mathematical Preliminaries
2.1. Sets and Functions.
2. Give three examples of sets that are not listed in the text.
3. Give two functions whose domain is {5, 4, 7} and whose range is {1, 2}. You
will probably want to use tables.
4. Is there a function with f(4) = 5 and f(4) = 7?
2.2. Modular Arithmetic.
5. Do the following calculations mod 12. Your answers should be numbers
between 0 and 11. The numbers 0 and 11 may also be answers!
7 + 5
1 + 4
8 + 8
6 + 6
9 − 7
7 − 9
2 − 8.
3. The Integer Model of Pitch
6. Use the integer model of pitch to rewrite the following melody in “Heavenly
Aida” in Act I of Verdi’s opera Aida: G, A, B, C, D, G, G.
7. Use the integer model of pitch to rewrite the following melody in the “Tore-
ador Song” in Act II of Bizet’s opera Carmen: C, D, C, A, A, A, G, A, B, A, B, G, C.
8. Calculate T
4
(3), T
1
(2), T
8
(7), I
4
(6), I
4
(8).
15
5. TRISTAN PRELUDE FROM WAGNER 16
9. Transpose the melody above from “Heavenly Aida” a perfect ﬁfth by apply-
ing T
7
to each element.
10. Invert the melody above in the “Toreador Song” about 6 by applying I
6
to
each element.
11. Calculate T
5
◦ I
3
(4) and I
3
◦ T
5
(4). Are they the same?
4. Fugue by Bach
12. Look at the website on Bach listed in the text.
5. Tristan Prelude from Wagner
Challenge: The unordered pcsets P
9
and P
13
on the piano transcription of the
Tristan Prelude are
P
9
= {C, F, G, D}
P
13
= {B, D, A, F}.
Translate these pcsets to integers mod 12. Find an integer n such that I
n
(P
9
) = P
13
.
You may have to change the ordering of the elements of the set.
1. Introduction
We consider a further application of transposition and inversion in music theory.
Thus far we have considered only composers who have lived before the 20th century,
so it is high time we consider someone who is closer to our time.
2. Fugue by Hindemith
Paul Hindemith (1895-1963) was known as a champion of contemporary music
and a promoter of early music. In 1941, just 5 years before becoming a U.S.
citizen, he composed Ludus Tonalis. This is a collection of 12 fugues with eleven
interludes, framed by a prelude and a postlude. Such a collection of fugues reminds
one immediately of Bach’s Well-Tempered Clavier, although Hindemith also had
more modern ideas of symmetry and symmetry breaking in mind. The title Ludus
Tonalis means Tonal Game in Latin, and this can be seen in the symmetry and
asymmetry of individual pieces as well as in the collection as a whole. One of the
most striking features of the collection is that the postlude is exactly the same as
the prelude except upside down and backwards!
Paul Hindemith also had an interesting life. An excellent website
http : //www.hindemith.org/
on Hindemith describes his trips to Egypt, Turkey, and Mexico, his ﬂight from the
Nazis, and his emigration to the United States. The website also has historical
photos and references to literature.
Hindemith’s Fugue in G provides us with further examples of transposition
and inversion, although we will encounter diﬃculties. This example will illustrate
some of the diﬃculties that music theorists encounter and how they get around
some of these diﬃculties. Hindemith’s fugue will also be a warm-up for the next
lecture on the PLR group. Professor Ramon Satyendra and I have recently created
a theoretical apparatus to treat musical diﬃculties such as the one we are about to
study.
5
The Fugue in G begins with a statement of the subject as in Bach’s Fugue in
d-minor. The subject is
G, G, G, G, G, G, G, C, D, G, C, F = 7, 7, 7, 7, 7, 7, 7, 0, 2, 7, 0, 5
and consists of eleven notes and ﬁve pitches in two measures which have ﬁve beats
each (all prime numbers!). The subject is very prominent because the repeated G
at the beginning tells us the voice is entering. When we listen to the piece, we
5
Thomas M. Fiore and Ramon Satyendra. “Generalized Contextual Groups.” To appear in
Music Theory Online.
17
2. FUGUE BY HINDEMITH 18
can listen for that repeated staccato note and we will easily ﬁnd occurrences of
the subject. For example, one quickly hears that measures 3, 8, and 15 contain
repeated notes which begin occurrences of the subject.
Instead of comparing occurrences of the subject as in Bach, we will look at a
smaller unit. The two three-note groups at the end of the subject are also very
prominent to our ears. In this analysis we will consider the relationships between
these three-note groups in the various occurrences of the subject. These relation-
ships are given by the conceptual categories transposition, inversion, and contextual
inversion, which will enable us to ﬁnd a good way of hearing the piece.
Let’s consider the set S of all transposed and inverted forms of the pcseg
G, C, D = 7, 0, 2
which is the ﬁrst occurrence of the three note pattern we want to study. Some
examples of elements of S are
T
0
7, 0, 2 = 7, 0, 2
T
1
7, 0, 2 = 8, 1, 3
T
2
7, 0, 2 = 9, 2, 4
etc.
I
0
7, 0, 2 = 5, 0, 10
I
1
7, 0, 2 = 6, 1, 11
I
2
7, 0, 2 = 7, 2, 0
etc.
Notice that the elements of the set S are ordered sets, namely pcsegs. Two pcsegs
are diﬀerent if they have diﬀerent orders. Thus, the pcsegs 7, 0, 2 and 7, 2, 0 are
diﬀerent elements of S.
6
The elements of S that are transpositions of 7, 0, 2 are
called prime forms and the elements of S that are inversions of 7, 0, 2 are called
inverted forms. So for example, 8, 1, 3 is a prime form and 6, 1, 11 is an inverted
form.
There are several functions S → S that are important for our analysis. The
transpositions T
n
: Z
12
→ Z
12
and inversions I
n
: Z
12
→ Z
12
induce functions
S → S which we again denote by T
n
and I
n
. These “induced” transpositions and
inversions are obtained by just applying T
n
and I
n
to each entry as we did above, so
these are nothing new. Now we introduce a new function J : S → S as follows. We
deﬁne J(x) to be that form of the opposite type as x which has the same ﬁrst two
pitch classes as x but in the opposite order.
7
So for example J7, 0, 2 = 0, 7, 5
because 7, 0, 2 and 0, 7, 5 are opposite types of forms and they have the ﬁrst two
pitch classes in common but in switched order. Similarly, J0, 7, 5 = 7, 0, 2. The
function J : S → S is an example of a contextual inversion. It is called contextual
6
This is another diﬃculty that arises. The pcset {7, 0, 2} is inversionally symmetric, so
it is impossible to deﬁne the contextual inversion J on the unordered sets. One must deﬁne
the contextual inversion J on the ordered pcsegs as we are about to do. Don’t worry if you
understanding it.
7
One might wonder why this is a well deﬁned function, but that can be proved mathematically.
You will just have to believe me that in this setup there is exactly one form of the opposite type
as x that has the same ﬁrst two pitch classes in common with x but in the opposite order.
2. FUGUE BY HINDEMITH 19
because it inverts depending on the context of the ﬁrst two pitch classes in the
pcseg.
Let’s see what the musical meaning of this J is. At ﬁrst it seems to be deﬁned
in an unnecessarily complicated way, but actually it is a very musical function.
The output is that opposite form that is closest (but not equal) to the input in the
sense that they overlap in the ﬁrst two pitch classes. This is quite audible. In the
subject of the fugue for example, we have two three-note groups. Guess how they
are related! Well, the three note groups are
G, C, D = 7, 0, 2
C, G, F = 0, 7, 5
and they overlap by two notes. We see also that they are of opposite types, namely
7, 0, 2 is a prime form and 0, 7, 5 is an inverted form. So J7, 0, 2 = 0, 7, 5 and
J0, 7, 5 = 7, 0, 2. But there is just one catch. When we look at the actual music,
we see that the second three-note group is G, C, F and not C, G, F. The order
isn’t exactly right. This is one of the diﬃculties that the music theorist encounters.
Nevertheless, the unordered pcsets {G, C, F} and {C, G, F} are the same. In other
words, J is a good enough approximation for us to use and we can control the error
by just looking at the unordered sets when we compare with the actual piece.
Now for the surprising part. Consider the following diagram.
(1)

J
//
T
5

T
5

J
//
Let’s ﬁll in the top left circle with 7, 0, 2 and see what happens.

7, 0, 2
J
//
T
5

T
5

J
//
2. FUGUE BY HINDEMITH 20
Next we apply J and T
5
to the top left entry and ﬁnd the results for the next two
entries.

7, 0, 2
J
//
T
5

0, 7, 5
T
5

0, 5, 7
J
//
There is just one circle left. We can get at it from above or from the left. From
above we get T
5
0, 7, 5 = 5, 0, 10. From the left we get J0, 5, 7 = 5, 0, 10. So
we ﬁll it in.

7, 0, 2
J
//
T
5

0, 7, 5
T
5

0, 5, 7
J
//
5, 0, 10
We see that the pathway doesn’t matter! But does the pathway matter if we ﬁlled
in the ﬁrst circle of diagram (1) with a diﬀerent element of S? No! Even if we ﬁll
in the circle with another element of S the pathway still does not matter. That can
be proved mathematically. Since the pathway doesn’t matter, we say that diagram
(1) commutes.
But what in the world does this commutative diagram have to do with Hin-
demith’s fugue? Let’s look at the ﬁrst two instances of the subject, one starts in
measure 1 and the next one starts in measure 3. Applying T
5
to the ﬁrst one gives
us the second one. But in this fugue it is better to look at smaller units than the
subject. These smaller units are what we are focusing on. So let’s look at the four
three-note groups in measures 2 and 4. Those are precisely the ones we ﬁlled in
the diagram!! Even the temporal aspect matches! As time ticks, we move from the
upper left circle to the lower right circle in the direction of the arrows. In fact, this
diagram occurs in at least four diﬀerent places of the piece! Compare your score
from class. The measures are
2 and 4
9 and 16
37 and 39
55 and 57.
In all cases except one the measures are just two apart. Does Hindemith break the
symmetry that one time to be playful in his tonal game? We can only wonder.
2. FUGUE BY HINDEMITH 21
In summary, we have found a good way of hearing Hindemith’s Fugue in G
using the conceptual categories of transposition, inversion, contextual inversion,
and commutative diagrams. This analysis was diﬀerent than the analysis from
Bach because we looked at units smaller than the entire subject. We investigated
relationships between three-note groups in the subject and its many occurrences.
The contextual inversion was important for these three-note groups because all
adjacent ones overlap in a very audible way. The mathematical structure we found
was not given by equations, but by relationships between three-note groups given in
terms of commutative diagrams. In the next lecture we will study more three-note
groups, namely major and minor chords. Transposition, inversion, and contextual
inversion will make another appearance. The Beatles and Beethoven are expecting
us!
Addendum to Lecture 1 Homework Problems
1. Introduction
1. How many years passed between Bach’s birth and Hindemith’s death?
2. Fugue by Hindemith
2. In which city in Pennsylvania did Hindemith conduct a symphony in 1959?
Hint: use the link on the Hindemith website for Life.
3. Calculate
J3, 8, 10
J5, 10, 0
J9, 2, 4
J8, 3, 1
J10, 5, 3
J2, 9, 7.
Compare the answers to the ﬁrst three with the answers for the second three. Do
you see a pattern? Use this pattern to ﬁgure out J ◦ J(x).
4. Put the pcseg 2, 7, 9 from measure 9 of Hindemith’s fugue into the upper
left circle of diagram (1) and calculate what goes in the other circles. Do both
paths give you the same answer?
5. Put the pcseg 10, 3, 5 from measure 37 of Hindemith’s fugue into the upper
left circle of diagram (1) and calculate what goes in the other circles. Do both paths
6. Put the pcseg 0, 5, 7 from measure 55 of Hindemith’s fugue into the upper
left circle of diagram (1) and calculate what goes in the other circles. Do both
paths give you the same answer?
Challenge: We talked about commutative diagrams in Hindemith’s tonal game.
Let’s consider a commutative diagram in our mathematical game. Look at the title
of this page and compare it to the ﬁrst three section headings in the table of con-
tents: “Lecture1 Transposition and Inversion,” “Lecture 1 Homework Problems,”
and “Addendum to Lecture 1.” Which of the following is the meaning of the title
(Addendum to) (Lecture 1 Homework Problems)
22
2. FUGUE BY HINDEMITH 23
(Addendum to Lecture 1) (Homework Problems)
In other words, does the playful diagram
Lecture 1
//

Lecture 1 Homework Problems

//
Addendum to Lecture 1 Homework Problems
commute? Recall that a diagram commutes when both paths give you the same
result. You may interpret the horizontal arrows as assigning (Homework Problems)
Lecture 2 The PLR Group
1. Introduction
A group is a mathematical object of central importance to music theorists.
A group is yet another conceptual category that music theorists draw upon in
order to make music more tangible. We have been working towards the concept
of a group with our numerous examples of transposition and inversion. So far,
we have looked at various instances of transposition and inversion in works by
Bach, Wagner, and Hindemith. Now we look one level deeper and consider how
the functions transposition and inversion interact with each other. This is a big
step from considering just individual instances of transposition and inversion. We
will see that the collection of transpositions and inversions form a group in the
mathematical sense of the word. This group is called the T/I group.
Another group of musical relevance is the PLR group. This is a set of functions
whose inputs are major and minor chords and whose outputs are major and minor
chords. These musical functions go back to the music theorist Hugo Riemann (1849-
1919). As a result, the PLR group is sometimes called the neo-Riemannian group.
The PLR group and the T/I group are related in many theoretically interesting
ways. Nevertheless, we will focus on musical examples. If you know how to play
guitar, you might know the Elvis Progression I-VI-IV-V-I from 50’s rock. Any song
with this progression provides us with a musical example as we’ll see below. We’ll
also look at a song from the Beatles on the Abbey Road album. A more striking
example however is the second movement of Beethoven’s Ninth Symphony. We will
see that a harmonic progression in the Ninth Symphony traces out a path on a
torus!
2. Mathematical Preliminaries
In this section we introduce the mathematical concept of a group and give some
examples. A group is basically a set with a way to combine elements similar to the
way that one multiplies real numbers. It will be a bit formal at ﬁrst, but don’t let
that stop you from reading! Even if you don’t fully understand it, keep going! It’s
not meant to be easy.
Definition 2.1. A group G is a set G equipped with a function ∗ : G×G → G
which satisﬁes the following axioms.
(1) For any three elements a, b, c of G we have (a ∗ b) ∗ c = a ∗ (b ∗ c), i.e. the
operation ∗ is associative.
(2) There is an element e of G such that a ∗ e = a = e ∗ a for every element a
of G, i.e. the element e is the unit of the group.
(3) For every element a of G, there is an element a
−1
such that a∗ a
−1
= e =
a
−1
∗ a, i.e. every element a has an inverse a
−1
.
24
2. MATHEMATICAL PRELIMINARIES 25
That is the abstract deﬁnition of a group. Now let’s consider some examples
of groups to see what the deﬁnition really means.
Example 1. Let G be the set of real numbers greater than zero. Some el-
ements are .5, 100, π, 7
100
, and .000000008 for example. Next let ∗ be the usual
multiplication of numbers. From school we know that the multiplication of real
numbers is associative. Let e = 1. From school we also know that any nonzero
number times 1 is just that number again, no matter if we multiply on the left or
right. For example, 1 × π = π = π × 1. So G has a unit. If a is a real number
greater than 0, then we can deﬁne a
−1
= 1/a in order to satisfy the last axiom.
For example,
π
−1
× π = 1 = π × π
−1
.
We have just veriﬁed the axioms (1), (2), and (3) in the deﬁnition above. Hence
the set of real numbers greater than 0 equipped with ∗ = × is a group.
Example 2. Let G be the set of whole numbers, i.e. G = {. . . , −2, −1, 0, 1, 2, . . . }.
Let ∗ be the usual addition of whole numbers. Then e = 0 deﬁnes a unit because
0 + a = a = a + 0 for any whole number a. If we take a
−1
to be −a, then one can
check that G equipped with ∗ = + satisﬁes the axioms (1),(2),(3) above and is thus
a group.
Example 3. Suppose S = {1, 2, 3}. Then let G denote the set of invertible
functions S → S. Don’t worry about the exact meaning of invertible. Let ∗ be
the function composition described in Lecture 1. Then two functions f : S → S
and g : S → S can be composed to give g ◦ f = g ∗ f. From school we know that
function composition is associative, i.e. (h ◦ g) ◦ f = h ◦ (g ◦ f). An example of an
element of G is the unit e : S → S deﬁned by
e(1) = 1
e(2) = 2
e(3) = 3.
Another example of an element h of G is h : S → S deﬁned by
h(1) = 3
h(2) = 2
h(3) = 1.
Its inverse h
−1
: S → S is h. You can check that h ◦ h(x) = x. One can check that
G satisﬁes the axioms (1),(2),(3) above and is thus a group.
Even if you didn’t understand every step of these examples, the main point is
that a group is a mathematical object that consists of set G and an operation ∗
which gives us a way to combine elements. This combination of elements is similar
to the usual multiplication of numbers in the sense that it is associative, has a
unit, and has inverses. The three examples above give three diﬀerent examples of
groups. In each example we speciﬁed a set G and an operation ∗ on that set, and
then checked that it satisﬁed the axioms of a group.
8
The last example was a warm
up for the T/I group.
8
Note that G, ∗, and e have diﬀerent meanings in each example.
2. MATHEMATICAL PRELIMINARIES 26
Example 4. Let S be the set of transposed and inverted forms of the C major
chord 0, 4, 7. The elements of S can be listed as prime forms and inverted forms.
For future reference we also record the letter names of the prime and inverted forms.
Prime Forms Inverted Forms
C = 0, 4, 7 0, 8, 5 = f
C = D = 1, 5, 8 1, 9, 6 = f = g
D = 2, 6, 9 2, 10, 7 = g
D = E = 3, 7, 10 3, 11, 8 = g = a
E = 4, 8, 11 4, 0, 9 = a
F = 5, 9, 0 5, 1, 10 = a = b
F = G = 6, 10, 1 6, 2, 11 = b
G = 7, 11, 2 7, 3, 0 = c
G = A = 8, 0, 3 8, 4, 1 = c = d
A = 9, 1, 4 9, 5, 2 = d
A = B = 10, 2, 5 10, 6, 3 = d = e
B = 11, 3, 6 11, 7, 4 = e
A musician might notice from this labelling that S is the set of 24 major and minor
chords.
9
We use capital letters to denote the letter names of major chords and lower
case letters to denote the letter names of minor chords like musicians normally do.
Transposition and inversion “induce” functions T
n
: S → S and I
n
: S → S by
applying the function to each entry of the input pcseg. Then let G consist of the
24 functions T
n
: S → S and I
n
: S → S where n = 0, 1, 2, . . . , 11. We also let ∗ be
function composition as in the previous example. It can be mathematically veriﬁed
that the transposition and inversion compose according to the following rules.
T
m
◦ T
n
= T
m+n
T
m
◦ I
n
= I
m+n
I
m
◦ T
n
= I
m−n
I
m
◦ I
n
= T
m−n
Here the indices are read mod 12. We see that the result of composing transpositions
and inversions is also a transposition or inversion, so that ∗ really is an operation
on G. Similarly, one can verify the axioms of a group and show that G forms a
group. This group is called the T/I group.
We have taken a big step now to consider how the transpositions and inver-
sions interact with each other. They interact with each other to form a group.
This is deeper than just considering individual transpositions and inversions by
themselves. This abstract structure of a mathematical group is of tremendous im-
portance because it allows us to see things in music that we otherwise would not
see.
9
Here chord just means a collection of pitch classes that are played simultaneously. The
major and minor chords are special chords that are very prominent in Western music. Don’t
worry about where they come from or why they have these letter names, you can just take this
list of prime forms and inverted forms for granted.
3. THE PLR GROUP 27
3. The PLR Group
We now introduce the PLR group as a group of functions S → S like the
T/I group in Example 4. From now on we let S denote the set of prime forms
and inverted forms of the C major chord 0, 4, 7 as in Example 4. First we deﬁne
functions P, L, and R with domain and range S. These three functions will be
contextually deﬁned just like J in the Addendum. Let P(x) be that form of opposite
type as x with the ﬁrst and third notes switched. For example
P0, 4, 7 = 7, 3, 0
P3, 11, 8 = 8, 0, 3.
Let L(x) be that form of opposite type as x with the second and third notes
switched. For example
L0, 4, 7 = 11, 7, 4
L3, 11, 8 = 4, 8, 11.
Let R(x) be that form of opposite type as x with the ﬁrst and second notes switched.
For example
R0, 4, 7 = 4, 0, 9
R3, 11, 8 = 11, 3, 6.
We also say that these functions are contextually deﬁned because they are not
deﬁned on the individual constituents of the pcseg like T
n
and I
n
are.
These functions are highly musical. A musician would notice that P is the
function that takes a chord and maps it to its parallel minor or major, e.g. P
applied to C major gives us c minor and P applied to c minor gives us C major.
The function L is a leading tone exchange for more theoretical reasons. It takes C
major to e minor for example. The function R takes a chord to its relative minor
or major, for example R applied to C major is a minor and R applied to a minor
is C major. These three functions are also musical in the sense that they take a
chord to another one that overlaps with the original one in two notes.
Definition 3.1. The PLR group is the group whose set consists of all possible
compositions of P,L, and R. The operation is function composition.
For example, some elements of the PLR group are, P, L, R, L ◦ R, R ◦ L,
P ◦ L ◦ P, L ◦ L, and R ◦ L ◦ P. At ﬁrst you might think that there are inﬁnitely
many ways to combine P, L, and R. But that is not true! In fact there are only 24
elements of the PLR group. For example, the elements L ◦ L and R ◦ R and the
unit are the same, namely L ◦ L(x) = x = R ◦ R(x) for all elements x of S. It can
be mathematically proven that there are only 24 elements.
10
But this is all very
abstract, so we need to get down to some actual musical examples.
10
they are both isomorphic to the dihedral group of order 24. Of course I haven’t told you what
isomorphic or dihedral means, but philosophically it means that the T/I group and the PLR
group are abstractly the same as the dihedral group! Now that is a surprise, since they ﬁrst
appear to be very diﬀerent in their deﬁnitions. Another mathematical surprise is the following.
Consider the group of all invertible functions S → S. Then the T/I group is the centralizer of
the PLR group in this larger group and vice-a-versa! This means philosophically that the two
are “dual” in a musical sense described by David Lewin in his seminal work Generalized Musical
Intervals and Transformations. This is very deep, and takes a long time to understand. So don’t
be discouraged at ﬁrst!
5. TOPOLOGY AND THE TORUS 28
4. Elvis and the Beatles
The Elvis Progression I-VI-IV-V-I from 50’s rock can be found in many popular
songs. It provides us with an example because it is basically the following diagram.
'`
R
// '`
L
// '`
R◦L◦R◦L
// '`
L◦R
// '`
If we put the C major chord 0, 4, 7 into the left most circle and apply the functions,
we get the progression C major, a minor, F major, G major, C major. This
progression can be found for example in the 80’s hit “Stand by Me.”
The inﬂuential Beatles made their U.S. debut in 1964 on the Ed Sullivan show,
just one year after the death of Paul Hindemith. From a temporal point of view,
it is good to do an example from the Beatles too. The main progression of “Oh!
Darling” from the album Abbey Road is E major, A major, E major, f minor, D
major, b minor, E major, b minor, E major, and A major.
11
Have a look at the
handout for Lecture 2. Let’s just focus on the inner four chords f minor, D major,
b minor, and E major. This is the progression we get when we insert 1, 9, 6 = f
minor into the ﬁrst circle of the following diagram as in your homework!
'`
L
// '`
R
// '`
P◦R◦L
// '`
This shows that mathematical analysis can also be used for popular songs,
not just “classical” music. Next we work our way towards the culmination of this
module: Beethoven’s Ninth Symphony and the path it traces out on a torus.
5. Topology and the Torus
Topology is a major branch of mathematics which studies qualitative ques-
qualitative questions that a topologist would ask about a geometric object are the
following. Is the geometric object connected? Does it have holes? Does it have a
boundary? For example, a circle and square are qualitatively the same because one
can be stretched to the other.
12
Neither has a boundary and both have a hole in
the center. Both are connected. A line segment with endpoints on the other hand,
is qualitatively diﬀerent from the circle. We cannot obtain a line segment from a
circle by quantitative changes such as stretching, twisting, or shrinking. But we
can obtain a line segment from a circle by the qualitative change known as cutting.
A line segment with endpoints is qualitatively diﬀerent from a circle because it has
a boundary (two endpoints) and it has no holes. The circle and the line segment
are each connected. Topology is not concerned with quantitative properties such
as area, angles, and lengths, so topologists consider two objects the same if they
only diﬀer in quantitative ways. The square and the circle diﬀer quantitatively, but
are the same qualitatively. For this reason, topologists consider the square and the
11
Actually there are some seventh chords in here and the ﬁrst E major chord has an added
C pitch class, but we’ll just ignore that for the sake of simplicity. These seventh chords do help
us make our point about overlapping chords though.
12
Here we are considering the circle and the square without their insides. They are not
shaded in. For example, the square we are talking about only consists of the four marks that
make up the sides of the square.
6. BEETHOVEN’S NINTH SYMPHONY 29
Figure 1. The Square Sheet.
circle to be the same. Topology is basically rubber-sheet geometry: we imagine two
objects are made of rubber and consider them the same if one can be stretched,
shrunk, or twisted into the shape of the other.
13
The website
http : //www.lehigh.edu/dmd1/public/www − data/essays.html
has several links to essays about the subject matter of topology. Some of the essays
are more technical than others. Another great website
http : //www.math.ohio − state.edu/ ∼ fiedorow/math655/yale/
is entitled “Math That Makes You Go Wow.” This interactive website talks about
philosophical, literary, musical, and artistic implications of topology.
The torus is an example of a mathematical object of interest to topologists.
To make a torus we start oﬀ with the square sheet in Figure 1. Although it is not
shaded in, we mean the whole rectangular region in Figure 1 belongs to the square
sheet. The arrowheads indicate how we will glue. First we glue the horizontal lines
according to the single arrowheads and obtain the cylinder in Figure 2. Next we
glue the two circles at the end of the cylinder according to the double arrowheads
and make the torus in Figure 3. The torus looks just like an inner tube ﬁlled with
air.
The website “Math That Makes You Go Wow” mentioned above has an interac-
tive torus. Go to the website and click on the link in Chapter 2 entitled “Orientable
Surfaces: Sphere, Torus.” Scroll down to the torus and move it around with your
mouse to visualize it better. But what in the world does the torus have to do with
the PLR group and Beethoven’s Ninth Symphony?
6. Beethoven’s Ninth Symphony
Ludwig van Beethoven (1770-1827) composed his Ninth Symphony during the
years 1822-1824, roughly 80 years before Henri Poincar´e initiated the study of
13
Topology is diﬀerent from topography, which is the study of the nature and the shape of
terrain.
6. BEETHOVEN’S NINTH SYMPHONY 30
Figure 2. The Cylinder.
Figure 3. The Torus.
topology. In measures 143-176 of the second movement of the Ninth Symphony one
can ﬁnd the extraordinary sequence of 19 chords.
C, a, F, d, B, g, E, c, A, f, D, b, G, e, B, g, E, c, A
Here again capital letters refer to major chords and lower case letters refer to minor
chords.
14
The letter names of the chords can be converted back to numbers using
the table of prime forms and inverted forms in Example 4. Note that the entire
sequence can be obtained by applying to C the functions R, then L, then R, then
L, and so on. In other words, we have the diagram below where the arrows are
alternately labelled by R and L. Beethoven did not include the last ﬁve chords
below in his composition, but we’ll see why I wrote them below in a minute. Notice
that all 24 major and minor chords appear below and none are repeated. This
patten in itself is surprising.
14
This sequence was ﬁrst observed by Cohn in a series of articles dating back to 1991,1992,
and 1997.
6. BEETHOVEN’S NINTH SYMPHONY 31
'`
C
R
// '`
a
L
// '`
F
R
// '`
d
L
ttj
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
'`
B
R
// '`
g
L
// '`
E
R
// '`
c
L
ttj
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
'`
A
R
// '`
f
L
// '`
D
R
// '`
b
L
ttj
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
'`
G
R
// '`
e
L
// '`
B
R
// '`
g
L
ttj
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
'`
E
R
// '`
c
L
// '`
A
R
// '`
f
L
ttj
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
'`
D
R
// '`
b
L
// '`
G
R
// '`
e
Let’s consider the graph on the handout.
15
A graph is just a collection of
dots called vertices and line segments called edges which connect some vertices.
In this graph the vertices are labelled by the major chords and minor chords, i.e.
elements of S. The edges are labelled by the functions R, P, and L. The edges
labelled by R, P, and L are represented by dashed lines, solid lines, and dotted
lines respectively. Two chords (vertices) are connected by a dashed line (i.e. by an
edge labelled by R) if we can get from one to the other with the R function. The
same goes for P and L. This graph is highly musical because the neighbors of a
chord are exactly those three other chords that are maximally close to it, i.e. those
15
This graph and the torus below it are a reproduction of the graph and torus in Jack
Douthett and Peter Steinbach. “Parsimonious Graphs: A Study in Parsimony, Contextual Trans-
formations, and Modes of Limited Transposition.” Journal of Music Theory 42/2 (1998): 241-263.
7. CONCLUSION 32
three chords which overlap with the original one in two pitch classes. For example,
the neighbors of D are b, f, d. In numbers, the neighbors of 2, 6, 9 are
6, 2, 11
1, 9, 6
9, 5, 2,
all of which agree with 2, 6, 9 in two pitch classes. That’s not really a surprise
because P, L, and R were designed to do precisely this.
progression from Beethoven is a path on it! Let’s see this. Notice that we can
glue the top and the bottom because the top two rows of vertices match up with
the bottom two rows. See the ﬁgures in the section on topology. Next we glue
the circles on the resulting cylinder after twisting them a third of the way around.
Here we are gluing those chords from the left side to those chords on the right side
that are the same. For example, the a on the left side is glued to the a on the
right side. The E on the left side is glued to the E on the right side and so on.
We have to twist to get them to match up. So now we have our torus. To see
that Beethoven’s Ninth Symphony traces out a path on it, we just connect the dots
which are labelled by the chord progression. Notice the pattern. If Beethoven had
continued the pattern, it would trace out all of the 24 major and minor chords.
Well, 19 out of 24 is pretty close though!
This example is perhaps the most striking of all our musical examples because
it relates group theory, topology, and Beethoven all in one! These conceptual
categories provide us with tools and a language to ﬁnd an entirely new way of
hearing this piece. Without mathematics we would never have heard a torus in
Beethoven’s Ninth Symphony. You might be interested to know that there are
other examples of music on topological objects. Bach’s Musical Oﬀering contains a
passage which is music on a M¨obius strip. Schoenberg and Slonimsky also provide
us with examples. See the website “Math That Makes You Go Wow” for further
details.
7. Conclusion
In this module we have investigated some of the group theoretical tools that
music theorists have developed to ﬁnd good ways of hearing particular works of
music. These tools provide us with conceptual categories to make our ﬂeeting im-
pressions of music into vivacious ideas. Our ﬁrst conceptual categories were supplied
by the transpositions and inversions. Bach’s Fugue in d minor from the ﬁrst book
of the Well-Tempered Clavier had several examples of transposition and inversion.
The subject made an appearance in various forms and these forms were describe
by transposition and inversion. The Tristan Prelude in Wagner’s opera Tristan
and Isolde gave examples of transposed and inverted chords in the century after
Bach’s death. Hindemith’s Fugue in G from the twentieth century had examples
of contextual inversion within the subject. In that analysis it was fruitful to look
at a unit smaller than the subject and to look for overlapping three note groups.
Commutative diagrams also appeared in this context.
Our next example of a conceptual category was the concept of a group. The
concept of a group is deeper than individual instances of transposition and inver-
sion because it allows us to see a structure on the collection of transpositions and
7. CONCLUSION 33
inversions. At that point, we ﬁxed the notation S to mean the set of 24 major
and minor triads and considered various invertible functions S → S. Transposition
and inversion induce such functions for example. The functions P, L, and R are
also important functions S → S. All possible compositions of these three functions
give us the 24 elements of the PLR group. The PLR group makes an appearance
in the Elvis Progression and in the Beatles song “Oh! Darling” from the Abbey
Road album. The most striking of our musical examples however was in the second
movement of Beethoven’s Ninth Symphony. Repeated application of R and L to the
C major chord generates a chord progression in measures 143-176 and this chord
progression traces out a path on a torus.
I hope that the conceptual category of this module has made your impressions
of mathematics, music, and music theory into vivacious ideas!
Lecture 2 Homework Problems
1. Introduction
2. Mathematical Preliminaries
1. Let G = Z
12
. Let e = 0 and ∗ = +. Let’s consider if this deﬁnes a group
integers mod 12 gives us another integer mod 12? Is it true that 0 +x = x = x +0
for any integer x mod 12? Is it true that x−x = 0 = −x+x for any integer x mod
12? Try out x = 1, 2, 3 for example. We already know that addition is associative.
Is G = Z
12
a group with the above deﬁnitions?
3. The PLR Group
2. Calculate P1, 5, 8, L10, 6, 3, and R9, 1, 4.
4. Elvis and the Beatles
3. Insert 1, 9, 6 into the left circle of the diagram displayed in the paragraph
about “Oh! Darling” from the Beatles. Calculate the other three circles by applying
the functions. Convert the pcseg numbers back to letters using the table of prime
forms and inverted forms from Example 4. Does it match with the inner four chords
of the chord progression for “Oh! Darling”?
5. Topology and the Torus
4. Are the triangle and the circle qualitatively the same? In other words, can
we obtain one from the other by shrinking, stretching, or twisting?
5. Give one way in which a triangle and a line segment are qualitatively diﬀer-
ent. Hint: How many holes does each have?
6. Beethoven’s Ninth Symphony
6. Calculate each of the following.
R0, 4, 7
L ◦ R0, 4, 7
R ◦ L ◦ R0, 4, 7
L ◦ R ◦ L ◦ R0, 4, 7
Next translate the results into chord names using the letters in the table in Exam-
ple 4. How does this relate to the chord progression in the second movement of
Beethoven’s Ninth Symphony?
34
7. CONCLUSION 35
7. Conclusion
7. Which of our musical examples was your favorite and why? Write at least
four sentences.
Bibliography
[1] Benjamin Boretz. “Meta-Variations, Part IV: Analytic Fallout (I).” PNM 11/1 (1972): 146-
223.
Model for Seventh Chords.” Journal of Music Theory 42/2 (1998): 191-193.
[3] David Clampitt. “Alternative Interpretations of Some Measures From Parsifal.” Journal of
Music Theory 42/2 (1998): 321 -334.
[4] John Clough. “A Rudimentary Geometric Model for Contextual Transposition and Inversion.”
Journal of Music Theory 42/2 (1998): 297-306.
[5] Richard L. Cohn. “Neo-Riemannian Operations, Parsimonious Trichords, and Their Tonnetz
Representations.” Journal of Music Theory 41/1 (1997): 1-66.
[6] Jack Douthett and Peter Steinbach. “Parsimonious Graphs: A Study in Parsimony, Contex-
tual Transformations, and Modes of Limited Transposition.” Journal of Music Theory 42/2
(1998): 241-263.
[7] Thomas M. Fiore and Ramon Satyendra. “Generalized Contextual Groups.” To appear in
Music Theory Online.
[8] Edward Gollin. “Some Aspects of Three-Dimensional Tonnetze.” Journal of Music Theory
42/2 (1998): 195-206.
[9] Julian Hook. Uniform Triadic Transformations. Ph.D. diss., Indiana University, 2002.
[10] Julian Hook. “Uniform Triadic Transformations.” Journal of Music Theory 46/? (2002): ?-?.
[11] Henry Klumpenhouwer. “Some Remarks on the Use of Riemann Transformations.” Music
Theory Online 0/9 (1994).
[12] Jonathan Kochavi. “Some Structural Features of Contextually-Deﬁned Inversion Operators.”
Journal of Music Theory 42/2 (1998): 307-320.
[13] David Lewin. Generalized Musical Intervals and Transformations. New Haven: Yale Univer-
sity Press, 1987.
[14] David Lewin. Musical Form and Transformation: 4 Analytic Essays. New Haven: Yale Uni-
versity Press, 1993.
[15] David Lewin. “Generalized Interval Systems for Babbit’s Lists, and for Schoenberg’s String
Trio.” Music Theory Spectrum 17/1 (1995): 81-118.
[16] David Lewin. “Transformational Considerations in Schoenberg’s Opus 23, Number 3.” un-
published, 2003.
[17] John Rahn. Basic Atonal Theory. New York: Schirmer Books, 1980.
[18] Ramon Satyendra. “An Informal Introduction to Some Formal Concepts from Lewin’s Trans-
formational Theory.” To appear in Journal of Music Theory.
36

Contents
Introduction Lecture 1 Transposition and Inversion 1. Introduction 2. Mathematical Preliminaries 3. The Integer Model of Pitch 4. Fugue by Bach 5. Tristan Prelude from Wagner Lecture 1 Homework Problems 1. Introduction 2. Mathematical Preliminaries 3. The Integer Model of Pitch 4. Fugue by Bach 5. Tristan Prelude from Wagner Addendum to Lecture 1 1. Introduction 2. Fugue by Hindemith Addendum to Lecture 1 Homework Problems 1. Introduction 2. Fugue by Hindemith Lecture 2 The P LR Group 1. Introduction 2. Mathematical Preliminaries 3. The P LR Group 4. Elvis and the Beatles 5. Topology and the Torus 6. Beethoven’s Ninth Symphony 7. Conclusion Lecture 2 Homework Problems 1. Introduction 2. Mathematical Preliminaries 3. The P LR Group 4. Elvis and the Beatles 5. Topology and the Torus 6. Beethoven’s Ninth Symphony 7. Conclusion
3

5 7 7 7 11 12 14 15 15 15 15 16 16 17 17 17 22 22 22 24 24 24 27 28 28 29 32 34 34 34 34 34 34 34 35

CONTENTS 4 Bibliography 36 .

Not only do physicists. Music theory supplies us with conceptual categories to organize and understand music. The discrete whole numbers . Sound waves can be described by mathematical equations. there are certain mathematical structures inherent in all works of music. . describes the ways that sets and pitches relate and how they can be transformed from one to 1John Rahn. Strings vibrate at certain frequencies. or the keys of the piano. another area of pure mathematics. and engineers use math to describe the physical world. To ﬁnd a good way of hearing a musical piece means to comprehend the music in such a way as to make it tangible. . The area of mathematics called combinatorics enables one to count the many ways of combining pitches. mathematics is the language that physicists utilize to describe the natural world and all of these things occur in the natural world. so can we also model works of music with equations? Music is after all just many individual sounds. they become ideas. . Nevertheless. The cello has a particular shape in order to resonate with the strings in a mathematical fashion. In fact. the eighteenth century Scottish philosopher David Hume said that the mind receives impressions and once these impressions become tangible and vivacious. easily described music? The answer to all of these questions is predictable and easily described: a series of emphatic “NO’s”! There is not an equation that will model all works of music and we should not spend time looking for it. 1980. −2. After all. one of the central concerns of music theory is to ﬁnd a good way to hear a piece of music and to communicate that way of hearing. Basic Atonal Theory. Can one similarly ﬁnd an “equation” to describe a piece of music? Or better yet. See page 1. . The technology necessary to make a digital recording on a CD relies on mathematics. i.Introduction Music and Mathematics are intricately related.1 Anyone who has ever heard Stockhausen’s Klavierst¨ck III (1952) knows that this is not always so easy to u do! On a higher level. numbers. 2 . −1. can one ﬁnd an “equation” to predict the outcome of a piece of music? We can model sound by equations. Our aural impressions become vivacious ideas by way of these conceptual categories. are particularly well suited for labelling the pitches. right? Should we invest time and money to ﬁnd these equations so that all of humankind can enjoy predictable. 0. but also to predict the outcome of physical processes. The language of mathematics is a convenient tool for comprehending and communicating this underlying structure. . This provides taxonomies and classiﬁcations of the various sets that arise. New York: Schirmer Books. 5 .e. and these mathematical structures are not given by equations. chemists. 1. Group theory. Music theorists often draw on the formidable powers of mathematics in their creation of conceptual categories.

Jazz musicians also use music theory in their improvisations and compositions. Ludwig van Beethoven. A classical pianist may play thousands upon thousands of notes in one concert. . The aural impression of a piece of music becomes an idea in the sense of Hume with this apparatus. Group theory does not provide us with equations to describe a piece of music or predict its outcome. In this module we investigate some of the group theoretical tools that music theorists have developed in the past 30 years to ﬁnd good ways of hearing particular works of music. and performers can use to make sense of a work of art and to communicate ways of hearing to others. but rather several ideas woven together. composers. The mathematical framework will provide us with a way of hearing the pieces and a means of communicating this hearing. I extend my thanks to Ramon Satyendra of the University of Michigan Music Department for helpful conversations in preparation of this module. Music theory is also useful to the composer. Music theory is not just for the listener. all from memory. This abstract discussion has relevance for the performing artist as well. it is just one conceptual category that listeners. Music theory provides the performing artist with an apparatus for pattern recognition. These notes were prepared for a series of three guest lectures in the undergraduate course Math 107 at the University of Michigan under the direction of Karen Rhea in Fall 2003. but at as a whole. and Paul Hindemith. rather than isolated individual events. Mozart. In the next two lectures we will study the T /I and P LR groups and use them to analyze works of music from Johann Sebastian Bach. It is in this sense that pure mathematics provides a convenient framework for the music theorist to communicate good ways of hearing a work of music.INTRODUCTION 6 the other. This group theoretical point of view will elevate our aural impressions to the status of ideas. One can ﬁnd the thread of the musical fabric with music theory. and this strengthens the musical memory. and Beethoven were well versed in the music theory of their respective epochs and applied it daily. Instead. Music theory is not limited to classical music. These lectures were part of my Fourth Year VIGRE Project aimed at introducing undergraduates to my interdisciplinary research on generalized contextual groups with Ramon Satyendra. Bach. A piece of music does not consist of many individual isolated sounds. How does a classical pianist do that? Is it necessary for the classical pianist to memorize each individual note? Of course not. Richard Wagner. Non-western music also lends itself to analysis within the framework of music theory. this same musical memory allows the listener to conceive of a piece of music as a whole. We will conceive of the music not as individual sounds. On the other hand.

The set {A . This is usually denoted by f : S → S . B } is the set whose elements are the two pitch classes A and B . This symbolism is read: “f goes from S to S . 5}. 100. A function f from a set S to a set S is a rule which assigns to each element of S a unique element of S .” In this situation. Definition 2. 5} deﬁned by f (1) = 4 f (2) = 5 7 . 11. and modular arithmetic will be needed for our discussion. consider the function f : {1. function. 10} = {5.1. 5} do not denote the set whose elements are the numbers 6. function. For example. The outputs are the elements of range. The symbols }6. Definition 2.1. After the mathematical discussion we turn to examples from Bach and Wagner. 70} and {56} are not equal because they do not have the same elements (the second set is missing 70). and modular arithmetic. 4. domain.2. 100.Lecture 1 Transposition and Inversion 1. The order of the elements of a set does not matter. These concepts include set. The objects of the set are called elements. 3} → {4. 5. Musicians are usually come into contact with transposition and inversion in the context of pitch. Introduction Some of the ﬁrst mathematical tools that music students learn about are transposition and inversion. 2. 5. {4. the set {4. A set is a collection of objects. 4. Two sets are said to be equal if they have the same elements. 11. In this introductory lecture we learn about the mathematical concepts necessary to formalize these musical tools. the set S is called the domain of the function f and S is called the range of the function f . To bridge the gap between sound and number we will conceive of the integer model of pitch as musicians normally do. The collection of objects is written between curly parentheses {}. 10} is the set consisting of the numbers 4. range. 2. 2. The inputs are the elements of the domain. A function gives us a rule for getting from one set to another. 5 because one of the parentheses is the wrong way. Mathematical Preliminaries The mathematical concepts of set. Sets and Functions. The sets {56. 5. 10} = {10. 10 and nothing else. For example. For example.

2. 5}. 2. 3 of the domain gets a unique output. MATHEMATICAL PRELIMINARIES 8 f (3) = 4. 3} → {4. But what about 1 hour after 11? Well. 2.2. 5} because a unique output is assigned to each input. 3} → {4. . In precalculus we like to say that the function passes the vertical line test. The deﬁnition g(1) = 4 g(1) = 5 g(2) = 4 g(3) = 4 does not deﬁne a function g : {1. Functions can be composed. For example. 8. Consider the face of a clock with the numbers 0 through 11 where 0 is in the 12 o’clock position. 3} and the range is the set S = {4. Summarizing. if j : {4. Using the clock. Sometimes functions are given by formulas rather than tables. 5} because two diﬀerent values 4 and 5 are assigned to 1. 2. 8. provided the domain of one is the range of the other. 5} → {7. 9} called the composition of j with f deﬁned by j ◦ f (x) = j(f (x)). In this example we have j ◦ f (1) = j(f (1)) = j(4) = 9 j ◦ f (2) = j(f (2)) = j(5) = 8 j ◦ f (3) = j(f (3)) = j(4) = 9. i. Note that each element 1. These three equations tell us the rule that assigns an output to each input. In the next section we will see some functions that are given by formulas. we can write 1 + 2 = 3 mod 12 6 + 5 = 11 mod 12 11 + 1 = 0 mod 12 11 + 2 = 1 mod 12. Modular Arithmetic. Here the domain is the set S = {1. The day starts at midnight. 2. Note that the same output is assigned to each input and 5 is not even used. Similarly 5 hours after 6 is 11. so we have replaced 12 by 0 on the usual face of a clock. that is 0 because we are back at the beginning.e. This is called arithmetic modulo 12. 2. f (1) does not have two diﬀerent values. we can determine what time it is 2 hours after 1. 3} → {7.2. then we get a new function j ◦ f : {1. We just go clockwise 2 notches after 1 and we get 3. 9} is given by the rule j(4) = 9 j(5) = 8 and f is as above. The deﬁnition h(1) = 4 h(2) = 4 h(3) = 4 does deﬁne a function h : {1. Similarly. 2 hours after 11 is 1.

we need to consider the face of the clock. let’s consider subtraction. 11} and call this set the set of integers mod 12. But to make sense of the last two equations. i. 10. Arguing as in the mod 12 case. Given any number. 1. Our ears naturally hear pitches that are an octave apart. we get 1 = 3 − 2 mod 7 6 = 4 − 5 mod 7 6 = 0 − 1 mod 7 6 = 1 − 2 mod 7. 5. If we are at 0 o’clock and go back 1 hour. 8. 9. we can ﬁnd out what it is mod 12 by adding or subtracting 12 enough times to get a whole number between 0 and 11. we get 1=3−2 6 = 11 − 5 11 = 0 − 1 11 = 1 − 2 where we leave oﬀ mod 12 because it is clear by now that we are talking about arithmetic modulo 12 in this paragraph and not arithmetic modulo 7. Maybe you can guess that we are interested in integers mod 12 because there are twelve keys on the piano from middle C to the next C not counting the last C. We already see a diﬀerence between mod 12 and mod 7. Similarly. For example. Such pitches appear to be very similar for our ears. The ﬁrst two equations appear to make sense to us from usual arithmetic. If we bring some of the numbers to the right like in usual arithmetic. we have 1 + 2 = 3 mod 7 6 + 5 = 4 mod 7 6 + 1 = 0 mod 7 6 + 2 = 1 mod 7. we just leave oﬀ the suﬃx mod 12.2. Let’s consider some examples of subtraction. So in a sense. All of this has an aural foundation. 6. our human ears are hardwired for arithmetic mod 12! We are also interested in arithmetic modulo 7 because there are seven pitches in the major scale. . If we bring some of the numbers over to the right like usual arithmetic.e. i. That is why 11 = 0 − 1 and 11 = 1 − 2.e. 3. mathematicians and musicians use the notation Z12 = {0. 7. we just go counterclockwise two notches on the face of the clock and arrive at 11 o’clock. −12 = 0 = 12 = 24 −13 = −1 = 11 = 23 −7 = 5 = 17 = 29. Notice that 6 +5 =11 mod 12 but 6+5=4 mod 7. So we have just done addition mod 12. As a result. seven white keys on the piano from middle C to the next C. if we are at 1 o’clock and go back to hours. 4. pitches with 12 jumps (or intervals) between them on the piano. then we are at 11 o’clock. 2. For arithmetic modulo 7 we imagine there are seven hours in a day and that the face of the clock goes from 0 to 6 instead of 0 to 11. MATHEMATICAL PRELIMINARIES 9 Whenever it is clear that one is working mod 12.

Next we can talk about functions f : Z12 → Z12 .2. 2. . 4. functions. For example I0 (1) = 11 and I0 (6) = 6. Let’s consider the function T2 : Z12 → Z12 deﬁned by the formula T2 (x) = x + 2. We can make a table for this function as follows. . . 6. As a result.2 above. 5. Finally we can turn to some music. T2 (0) = 0 + 2 = 2 mod 12 T2 (1) = 1 + 2 = 3 mod 12 T2 (2) = 2 + 2 = 4 mod 12 T2 (3) = 3 + 2 = 5 mod 12 T2 (4) = 4 + 2 = 6 mod 12 T2 (5) = 5 + 2 = 7 mod 12 T2 (6) = 6 + 2 = 8 mod 12 T2 (7) = 7 + 2 = 9 mod 12 T2 (8) = 8 + 2 = 10 mod 12 T2 (9) = 9 + 2 = 11 mod 12 T2 (10) = 10 + 2 = 0 mod 12 T2 (11) = 11 + 2 = 1 mod 12. 1. we can ﬁnd out what it is mod 7 by adding or subtracting 7 enough times to get a number between 0 and 6. This concludes the introduction of sets. Given any number. For example. Another function I0 : Z12 → Z12 is given by the formula I0 (x) = −x. These are functions whose inputs are integers modulo 12 and whose outputs are also integers modulo 12. 3. 1. . −7 = 0 mod 7 −9 = −2 = 5 mod 7 15 = 8 = 1 mod 7 17 = 10 = 3 mod 7. mathematicians and musicians use the notation Z7 = {0. Recall that we considered functions given by tables in Subsection 2. and modular arithmetic necessary for an understanding of transposition and inversion. . 6} and call this set the set of integers mod 7. MATHEMATICAL PRELIMINARIES 10 Subtraction can be understood by moving counterclockwise on the face of a clock with seven hours labelled 0.

When we hear a melody consisting of several pitches. 5. This C major chord is part of the main theme for Haydn’s Surprise Symphony. F. 4. The relationship between these intervals is what makes a melody appealing to us. 0. 2. The Integer Model of Pitch To make use of the mathematical ideas developed in the last section. These angular brackets are often used by music theorists to emphasize that the notes occur in this order. 11. 7. the C major chord {C. B. B. F. we hear the intervals between the individual notes. Inversion is another way to create musical variation while preserving the intervallic sound of a melody. Middle C is a particular pitch. 7. One can see the diﬀerence between pitch classes and pitches in the following example. 4. we need to translate pitch classes into numbers and introduce transpositions and inversions. E. 7. Transposition mathematically captures what musicians do all the time: the restatement of a melody at higher and lower pitch levels in a way that preserves intervals. are sometimes called pcsets (pitch class sets) while ordered sets such as 0. 2. it is common to study pitch classes rather than pitches.g. G . although it does not preserve the exact intervals. E. 7 . 11. 7}. although they can also be applied to chords.1. don’t panic. then the following well established dictionary shows us how to get from pitches to integers modulo 12. 7. 2. although the pitch class C refers to the aggregate of all keys on the piano with letter name C. THE INTEGER MODEL OF PITCH 11 3. Recall that the ordering does not matter for sets. C=0 C =D =1 D=2 D =E =3 E=4 F =5 F =G =6 G=7 G =A =8 A=9 A = B = 10 B = 11 In this integer model of pitch. e. 7 are called pcsegs (pitch class segments). 5. 2. 4. 5. G. particularly in atonal theory. which can be written 0. Let n be an integer mod 12. D. Transpositions and inversions are functions Z12 → Z12 that are useful to every musician. G. 11. 4.3. D. C. 0. E. Then the function Tn : Z12 → Z12 deﬁned by the formula Tn (x) = x + n mod 12 is called transposition by n. 2In music theory. G} is {0. 7}. If you can read music. E. The ﬁrst part of the main theme is C. There are also analogues for Z7 . {0. because a set is just a collection of elements. Definition 3. 5. 4. just use the integers modulo 12. 4.2 If you can’t read music. Transposition and inversion are often applied to melodies. 4. 4. Unordered sets. . 11.

0. We already came into contact with I0 : Z12 → Z12 in the previous section. Next we use them to ﬁnd good ways of hearing a fugue and a prelude.nau. 8. 2} by applying T7 to each of 0. T7 (4). 7 = 0. 5 In this section we have introduced the integer model of pitch. Some examples for I7 : Z12 → Z12 are I7 (3) = −3 + 7 = 4 I7 (7) = −7 + 7 = 0 I7 (9) = −9 + 7 = −2 = 10. 10.4. Let n be an integer mod 12. 5. Every fugue has occurrences of transposition and inversion. FUGUE BY BACH 12 We already came into contact with T2 : Z12 → Z12 in the previous section. and 7. each of which contains 24 preludes and fugues. 8. 4. 5. 7. 5. 4. 1.html on Bach describes in detail what a fugue is. Some examples for T5 : Z12 → Z12 are T5 (3) = 3 + 5 = 8 T5 (6) = 6 + 5 = 11 T5 (7) = 7 + 5 = 0 T5 (10) = 10 + 5 = 15 = 3 where we have not written mod 12 because it is clear from the context.edu/ ∼ tas3/bachindex. The transpositions and inversions are functions which have inputs and outputs that are pitches. 7} = {T7 (0). 7. which we now analyze. 10. Music theorists and composers like to transpose and invert entire pcsets or pcsegs by applying the function to each element. 1. although Haydn did not do this! I0 0. 4.4. Click on the link for movies on the Well-Tempered Clavier. Definition 3. Similarly. T7 (7)} = {0 + 7.2. Then the function In : Z12 → Z12 deﬁned by the formula In (x) = −x + n is called inversion about n. 7. 11. Fugue by Bach Johann Sebastian Bach (1685-1750) took the art of fugue to new heights. 4. A musician would notice that this takes the C major chord to the G major chord. 11.ucc. 4. There is an animation and recording for Fugue 6 in d minor of the Well-Tempered Clavier Book I. A fugue usually begins with a statement of the main theme called the subject. 0. This subject returns over and over again in various voices and usually they are thread together in complex way. 7 + 7} = {7. 2. 5. 2. These are conceptual categories that music theorists use to ﬁnd good ways of hearing pieces. He composed the Well-Tempered Clavier Book I and the Well-Tempered Clavier Book II. we could invert the pcseg for the theme of Haydn’s Surprise Symphony about 0. 7. Our analysis . which assigns to each of the 12 pitch classes an integer mod 12. we can transpose a C major pcset by 7 steps as in T7 {0. For example. A truly fascinating website http : //jan. 8. 4 + 7. The function In is called inversion about n because it looks like a reﬂection about n whenever one draws the number line. 11.

2. 2. 11. G . 2. G . 9. 11. Let’s call this pcseg P . 9. 2. D. C . But what about inversion? Let’s consider measures 14 and 22. we can listen for these transposed and inverted forms of the subject. 2. C. 4 . 1. D. 0. D. 2. 11. D are the same ﬁrst two elements of P . 4 . 11. 9. C. 11. 9. 6. 5. 9. 2. G. C . 4 A. C. Next time we listen to the piece. 8. A = 2. 4. D. 9 which is a near perfect ﬁt with measures 14 and 22! Just the last three notes are changed to make it sound better. 8. 4. 9. 10. F . E = 9.18. 1. B. D = 4. 2. FUGUE BY BACH 13 will be restricted to ﬁnding some transpositions and inversions. So we see that inversion does indeed play a role in the piece. 8. since we are only studying some of the mathematical structure. E. They are respectively E. C. At measure 8. 4 A. 7. 9. C. A. A. 5. 0. B. At measure 13 we have A. G . A. A = 4. A. G. A. In measure 6. A. E. 11. F. 1. 8. B. G. many fugues have this property. 2. 10 E. 1. C . 11. 0. D. B. B. C . F. 9. 2. D. F. These are also T7 P except for the highlighted 1’s. 11. except for the last three digits. E. A. except for the highlighted 1’s. 11. This one doesn’t entirely match though. The last three notes of 22 are even the last three notes of P . B. B . They are nearly identical. G . B. just one octave lower.4. 5. 5. 5. Do you see a relationship between this pcseg and P ? Notice that this pcseg is T7 P ! Just try adding 7 to each element of P and you will see it. The subject of the fugue is the pcseg D. B . 11. 4. G. 4 . C . 7. 2. In fact. 5. 2. D. Measures 17. B . F. 1. 11. 9. D. A. A. 5. just the order is switched. E. 2. A. F. B = 4. 2. E = 9. 7. 7. 1. 9. F. 0. 4. 3. E = 9. 5. C . 9. F. 5. 1. C . C . 11. 1. 8. but the eleventh pitch doesn’t match. D. D. 4. B. 7. Notice also that the ﬁrst two elements E. Here we see that transposition by a perfect ﬁfth occurs four times before the piece is even half over. 11. C . E . The ﬁrst ﬁve pitches are almost T2 of the ﬁrst ﬁve pitches of P . A. See the score that you got in class. F. E = 9. 1. 5. F. 2. a form E. 2. this subject consists of twelve notes! In measure 3. 10. 9. 10. 1. G . 2. 2. 8. 1. 9 which begins in measure 1 and lasts until the beginning of measure 3. 4. B. 1. D. The last pitch of the pcseg is also T5 of the respective pcseg of P . 4. C . 4. E. the subject returns in the exact same form as the introduction. The interval 7 is very important in western music and is called the perfect ﬁfth. just the order is switched. F. D. G. another voice sings the melody A. 5. 9 . So far we have seen that transposition plays a role in this piece. 2 of the subject enters. We’ll leave detailed analysis to the music theorists. E = 9. C . 7. but the next 5 pitches are T5 of the respective pitches of P . D. B. B. The rest of the piece contains further transpositions and inversions of the subject. A. and 21 are respectively A. Calculating I6 P gives 4. 11. D. G. 1. D. E. 5. These conceptual categories make the piece more enjoyable for . D. F. Interestingly enough. This is similar to T7 P as in measure 3. B. 0. 1. C .

We take the prelude to the famous opera Tristan and Isolde as an example for transposition and inversion. pcsets. B. also form a set that can be transposed or inverted to {0. When we listen to the piece again. We have a good way of hearing the piece. 8}. 8}. 4What works for one piece of music may not work for another. Notice that the ﬁrst and last column are essentially the same. 2. 6. namely G − B. although we worked with pcsegs in the previous example. i. 2. 8} {0. For example. This knowledge also makes the piece easier for performers because they recognized patterns and relationships between diﬀerent parts of the piece. The piano transcription entitled “Wagner: Tristan Prelude” is in the packet of music I handed out in class. while the middle two columns are essentially the same! Here. Mathematics is the tool that we use to communicate this way of hearing to others. 3This analysis is obtained from John Rahn page 78. 5. all of the pcsets in the second column can be transposed or inverted to {0. 8} This table means that all of the pcsets in the ﬁrst column can be transposed or inverted to {0. However.4 Consider the piano transcription of the ﬁrst few measures of the prelude. Tristan Prelude from Wagner Richard Wagner (1813-1883). Notice also that everything is done according to the groups of circled notes in the music. 8}! In other words. A}. we mathematically see and musically hear a self similarity on diﬀerent levels. 5. and D − F . this illustrates some of the mathematical features of the piece. 2. 2.5. pcsets are more appropriate for the Tristan prelude than pcsegs. 2. and we almost have three groups of four. Wagner’s compositions are drastically diﬀerent from Bach’s. 6. is best known for his gigantic operas. However. P2 = {F. D . 5. etc. who was born 63 years after the death of Bach. are more fruitful. In the Bach fugue it was better to use pcsegs because the pcsets would tell us very little in that case. 8} {0. Let Pi denote the set of pitch classes that are heard during the circle i on the piano transcription. 2. we can listen for these features. 6. which would give us 12 again! The ﬁrst and last pitches of each four note group. which use transposition and inversion. 5. Then we notice the following pattern after looking very carefully. This particular passage is notorious for its resistance to traditional analysis. a music theorist would not be satisﬁed with this analysis because we have barely scratched the surface. 2. 8} {0. Nevertheless. The conceptual categories of transposition and inversion provide us with a good way of hearing these introductory measures to Wagner’s opera Tristan and Isolde. There is much more to this fugue than a few transpositions and inversions. who in turn quotes Benjamin Boretz.3 More modern methods of atonal analysis.e. . P1 P2 P3 P4 P5 P6 P7 P8 P9 P12 P13 {0. 5. In this analysis we consider unordered sets. essentially means they can be transposed or inverted to the same thing. B − D. TRISTAN PRELUDE FROM WAGNER 14 listeners because we come closer to understanding it.

B . 7. 2}. Reread the introduction to this music module.1. C. B . I4 (6). C. Mathematical Preliminaries 2. G. 8. A. Introduction 1. G. Do the following calculations mod 12. A. 7} and whose range is {1. Sets and Functions. 3.2. Give three examples of sets that are not listed in the text. G. A. Is music part of the physical world? Write a short paragraph on this topic. 15 . Calculate T4 (3). 4. The Integer Model of Pitch 6. Is there a function with f (4) = 5 and f (4) = 7? 2. 2.Lecture 1 Homework Problems 1. Use the integer model of pitch to rewrite the following melody in the “Toreador Song” in Act II of Bizet’s opera Carmen: C. Use the integer model of pitch to rewrite the following melody in “Heavenly Aida” in Act I of Verdi’s opera Aida: G. T1 (2). A. I4 (8). A. You will probably want to use tables. 5. D. 2. T8 (7). Modular Arithmetic. B. C . The numbers 0 and 11 may also be answers! 7+5 1+4 8+8 6+6 9−7 7−9 2 − 8. D. A. Give two functions whose domain is {5. 4. 3. Your answers should be numbers between 0 and 11. G .

Calculate T5 ◦ I3 (4) and I3 ◦ T5 (4). F }. Fugue by Bach 12. G . Transpose the melody above from “Heavenly Aida” a perfect ﬁfth by applying T7 to each element. Find an integer n such that In (P9 ) = P13 . 10. D . 11. . 5. Invert the melody above in the “Toreador Song” about 6 by applying I6 to each element. A. TRISTAN PRELUDE FROM WAGNER 16 9. Are they the same? 4. D} P13 = {B. F. You may have to change the ordering of the elements of the set. Look at the website on Bach listed in the text. Translate these pcsets to integers mod 12. Tristan Prelude from Wagner Challenge: The unordered pcsets P9 and P13 on the piano transcription of the Tristan Prelude are P9 = {C.5.

G. 7.org/ on Hindemith describes his trips to Egypt. Such a collection of fugues reminds one immediately of Bach’s Well-Tempered Clavier. just 5 years before becoming a U. framed by a prelude and a postlude.5 The Fugue in G begins with a statement of the subject as in Bach’s Fugue in d-minor. Fiore and Ramon Satyendra. 7. 7. G. he composed Ludus Tonalis. Introduction We consider a further application of transposition and inversion in music theory. The subject is G. Fugue by Hindemith Paul Hindemith (1895-1963) was known as a champion of contemporary music and a promoter of early music. “Generalized Contextual Groups. D. 5 and consists of eleven notes and ﬁve pitches in two measures which have ﬁve beats each (all prime numbers!).Addendum to Lecture 1 1. G. C. This is a collection of 12 fugues with eleven interludes. The subject is very prominent because the repeated G at the beginning tells us the voice is entering. The title Ludus Tonalis means Tonal Game in Latin. C. 7. we 5Thomas M. so it is high time we consider someone who is closer to our time. Hindemith’s Fugue in G provides us with further examples of transposition and inversion. Professor Ramon Satyendra and I have recently created a theoretical apparatus to treat musical diﬃculties such as the one we are about to study. Hindemith’s fugue will also be a warm-up for the next lecture on the P LR group. G. and this can be seen in the symmetry and asymmetry of individual pieces as well as in the collection as a whole. 0. although Hindemith also had more modern ideas of symmetry and symmetry breaking in mind. An excellent website http : //www.” To appear in Music Theory Online. 7. 2. citizen. 7. Turkey. although we will encounter diﬃculties. 17 . Thus far we have considered only composers who have lived before the 20th century. and Mexico. F = 7. his ﬂight from the Nazis. 0. G. and his emigration to the United States. The website also has historical photos and references to literature. 2.S. 7.hindemith. This example will illustrate some of the diﬃculties that music theorists encounter and how they get around some of these diﬃculties. G. One of the most striking features of the collection is that the postlude is exactly the same as the prelude except upside down and backwards! Paul Hindemith also had an interesting life. When we listen to the piece. In 1941. G.

0. the pcsegs 7. These “induced” transpositions and inversions are obtained by just applying Tn and In to each entry as we did above. There are several functions S → S that are important for our analysis. 2. 11 I2 7. We deﬁne J(x) to be that form of the opposite type as x which has the same ﬁrst two pitch classes as x but in the opposite order. one quickly hears that measures 3. Instead of comparing occurrences of the subject as in Bach. FUGUE BY HINDEMITH 18 can listen for that repeated staccato note and we will easily ﬁnd occurrences of the subject. 0. 2 are called inverted forms. 0. C. 2 = 8. 4 etc. 8. 2. Don’t worry if you don’t understand this. Similarly. 1. 2 = 5. 0. The two three-note groups at the end of the subject are also very prominent to our ears. . 7. 0. 2 and 7. 2} is inversionally symmetric. 2 are called prime forms and the elements of S that are inversions of 7. 1. we will look at a smaller unit.7 So for example J 7. These relationships are given by the conceptual categories transposition. 2 = 7. 1. One must deﬁne the contextual inversion J on the ordered pcsegs as we are about to do. 7. and contextual inversion. 2 = 6. 3 is a prime form and 6. 0. 2 = 7. 10 I1 7. J 0. namely pcsegs. So for example. 11 is an inverted form. 7One might wonder why this is a well deﬁned function. 0. Notice that the elements of the set S are ordered sets. For example. 2 T1 7. 0. 0. Thus. We thought about this problem for a while before understanding it. 2 = 0. Now we introduce a new function J : S → S as follows. You will just have to believe me that in this setup there is exactly one form of the opposite type as x that has the same ﬁrst two pitch classes in common with x but in the opposite order. 0. 0 are diﬀerent elements of S. 5 = 7. D = 7. and 15 contain repeated notes which begin occurrences of the subject. Two pcsegs are diﬀerent if they have diﬀerent orders. 8. 5 because 7. I0 7. 2 which is the ﬁrst occurrence of the three note pattern we want to study. 5 are opposite types of forms and they have the ﬁrst two pitch classes in common but in switched order. so it is impossible to deﬁne the contextual inversion J on the unordered sets. inversion. The transpositions Tn : Z12 → Z12 and inversions In : Z12 → Z12 induce functions S → S which we again denote by Tn and In . 2 and 0. 0. 0. 1. 0. The function J : S → S is an example of a contextual inversion. In this analysis we will consider the relationships between these three-note groups in the various occurrences of the subject. but that can be proved mathematically. 2. 3 T2 7. 0. 7. so these are nothing new. 2 = 9. which will enable us to ﬁnd a good way of hearing the piece. 0. because you shouldn’t.6 The elements of S that are transpositions of 7. 0. The pcset {7. Some examples of elements of S are T0 7.2. Let’s consider the set S of all transposed and inverted forms of the pcseg G. 0 etc. 2 . It is called contextual 6This is another diﬃculty that arises.

0. G. Consider the following diagram. We see also that they are of opposite types. In other words. 2 and see what happens. F . 5 and J 0. 2 = 0. (1) gfed `abc J `abc / gfed T5 T5   J gfed `abc `abc / gfed Let’s ﬁll in the top left circle with 7. 0. C. J is a good enough approximation for us to use and we can control the error by just looking at the unordered sets when we compare with the actual piece. This is quite audible. F } are the same. 2 . The output is that opposite form that is closest (but not equal) to the input in the sense that they overlap in the ﬁrst two pitch classes. At ﬁrst it seems to be deﬁned in an unnecessarily complicated way. 2 is a prime form and 0. 7. Nevertheless. gfed `abc 7. The order isn’t exactly right. but actually it is a very musical function. Let’s see what the musical meaning of this J is. Now for the surprising part. we have two three-note groups. D = 7. F } and {C. But there is just one catch. 7. 0. When we look at the actual music. 0. C. the three note groups are G. 5 = 7. This is one of the diﬃculties that the music theorist encounters. 7. G. 5 is an inverted form. 0. In the subject of the fugue for example. 2 C. F = 0.2. 7. 5 and they overlap by two notes. G. 0. Guess how they are related! Well. 2 J `abc / gfed T5 T5  gfed `abc J  `abc / gfed . So J 7. F and not C. the unordered pcsets {G. C. FUGUE BY HINDEMITH 19 because it inverts depending on the context of the ﬁrst two pitch classes in the pcseg. we see that the second three-note group is G. namely 7.

we say that diagram (1) commutes. 10 . FUGUE BY HINDEMITH 20 Next we apply J and T5 to the top left entry and ﬁnd the results for the next two entries. 5. gfed `abc 7. So we ﬁll it in.2. 2 J `abc / gfed 0. The measures are 2 and 4 9 and 16 37 and 39 55 and 57. one starts in measure 1 and the next one starts in measure 3. From above we get T5 0. From the left we get J 0. But what in the world does this commutative diagram have to do with Hindemith’s fugue? Let’s look at the ﬁrst two instances of the subject. 7. But in this fugue it is better to look at smaller units than the subject. Since the pathway doesn’t matter. 0. So let’s look at the four three-note groups in measures 2 and 4. 7 = 5. we move from the upper left circle to the lower right circle in the direction of the arrows. 0. 0. 7 J  `abc / gfed There is just one circle left. 10 We see that the pathway doesn’t matter! But does the pathway matter if we ﬁlled in the ﬁrst circle of diagram (1) with a diﬀerent element of S? No! Even if we ﬁll in the circle with another element of S the pathway still does not matter. Those are precisely the ones we ﬁlled in the diagram!! Even the temporal aspect matches! As time ticks. We can get at it from above or from the left. 5 T5 T5  gfed `abc 0. Applying T5 to the ﬁrst one gives us the second one. gfed `abc 7. Does Hindemith break the symmetry that one time to be playful in his tonal game? We can only wonder. . 7. 7 J  `abc / gfed 5. 5 T5 T5  gfed `abc 0. this diagram occurs in at least four diﬀerent places of the piece! Compare your score from class. 5. 5. That can be proved mathematically. 7. In all cases except one the measures are just two apart. In fact. These smaller units are what we are focusing on. 10 . 0. 0. 2 J `abc / gfed 0. 5 = 5.

inversion. Transposition. contextual inversion. This analysis was diﬀerent than the analysis from Bach because we looked at units smaller than the entire subject. The contextual inversion was important for these three-note groups because all adjacent ones overlap in a very audible way. The mathematical structure we found was not given by equations. and contextual inversion will make another appearance. In the next lecture we will study more three-note groups. The Beatles and Beethoven are expecting us! . but by relationships between three-note groups given in terms of commutative diagrams. we have found a good way of hearing Hindemith’s Fugue in G using the conceptual categories of transposition. inversion. and commutative diagrams. We investigated relationships between three-note groups in the subject and its many occurrences.2. namely major and minor chords. FUGUE BY HINDEMITH 21 In summary.

. You may interpret the horizontal arrows as assigning (Homework Problems) and the vertical arrows as adding (Addendum to). does the playful diagram / Lecture 1 Homework Problems Lecture 1  Addendum to Lecture 1  / Addendum to Lecture 1 Homework Problems commute? Recall that a diagram commutes when both paths give you the same result. FUGUE BY HINDEMITH 23 (Addendum to Lecture 1) (Homework Problems) In other words.2.

Another group of musical relevance is the P LR group. i. A group is yet another conceptual category that music theorists draw upon in order to make music more tangible. This is a big step from considering just individual instances of transposition and inversion. keep going! It’s not meant to be easy. b. every element a has an inverse a−1 . (3) For every element a of G. i. It will be a bit formal at ﬁrst. we have looked at various instances of transposition and inversion in works by Bach. and Hindemith. A group is basically a set with a way to combine elements similar to the way that one multiplies real numbers. As a result. c of G we have (a ∗ b) ∗ c = a ∗ (b ∗ c). (2) There is an element e of G such that a ∗ e = a = e ∗ a for every element a of G. you might know the Elvis Progression I-VI-IV-V-I from 50’s rock.e. the operation ∗ is associative. Mathematical Preliminaries In this section we introduce the mathematical concept of a group and give some examples. We will see that a harmonic progression in the Ninth Symphony traces out a path on a torus! 2. we will focus on musical examples. If you know how to play guitar. We’ll also look at a song from the Beatles on the Abbey Road album.Lecture 2 The P LR Group 1. Definition 2. i. Wagner. Now we look one level deeper and consider how the functions transposition and inversion interact with each other.e. A more striking example however is the second movement of Beethoven’s Ninth Symphony. We have been working towards the concept of a group with our numerous examples of transposition and inversion. Nevertheless. (1) For any three elements a. the element e is the unit of the group. Introduction A group is a mathematical object of central importance to music theorists. So far. Any song with this progression provides us with a musical example as we’ll see below. the P LR group is sometimes called the neo-Riemannian group. there is an element a−1 such that a ∗ a−1 = e = a−1 ∗ a. The P LR group and the T /I group are related in many theoretically interesting ways. 24 . This is a set of functions whose inputs are major and minor chords and whose outputs are major and minor chords. We will see that the collection of transpositions and inversions form a group in the mathematical sense of the word. A group G is a set G equipped with a function ∗ : G × G → G which satisﬁes the following axioms.1.e. These musical functions go back to the music theorist Hugo Riemann (18491919). This group is called the T /I group. but don’t let that stop you from reading! Even if you don’t fully understand it.

}. (h ◦ g) ◦ f = h ◦ (g ◦ f ).e. 3}. . One can check that G satisﬁes the axioms (1). and then checked that it satisﬁed the axioms of a group. −2. Example 3.(2). 0. . Let e = 1. and e have diﬀerent meanings in each example. G = {. .e. . Then two functions f : S → S and g : S → S can be composed to give g ◦ f = g ∗ f . the main point is that a group is a mathematical object that consists of set G and an operation ∗ which gives us a way to combine elements. Let ∗ be the function composition described in Lecture 1. MATHEMATICAL PRELIMINARIES 25 That is the abstract deﬁnition of a group. 1. 8Note that G.8 The last example was a warm up for the T /I group.(3) above and is thus a group. 100. Even if you didn’t understand every step of these examples. then one can check that G equipped with ∗ = + satisﬁes the axioms (1). Then e = 0 deﬁnes a unit because 0 + a = a = a + 0 for any whole number a. 7100 . Suppose S = {1. π. For example. Now let’s consider some examples of groups to see what the deﬁnition really means. Then let G denote the set of invertible functions S → S. and (3) in the deﬁnition above. Example 2. . You can check that h ◦ h(x) = x. Some elements are . 2. . For example. Hence the set of real numbers greater than 0 equipped with ∗ = × is a group.(2). then we can deﬁne a−1 = 1/a in order to satisfy the last axiom. We have just veriﬁed the axioms (1). If a is a real number greater than 0. From school we know that function composition is associative. i. From school we know that the multiplication of real numbers is associative. The three examples above give three diﬀerent examples of groups. Let G be the set of whole numbers. Next let ∗ be the usual multiplication of numbers. Let G be the set of real numbers greater than zero. ∗. Example 1.2. . Its inverse h−1 : S → S is h. and has inverses. Don’t worry about the exact meaning of invertible. −1.5. Another example of an element h of G is h : S → S deﬁned by h(1) = 3 h(2) = 2 h(3) = 1. So G has a unit. and . no matter if we multiply on the left or right.000000008 for example. has a unit. An example of an element of G is the unit e : S → S deﬁned by e(1) = 1 e(2) = 2 e(3) = 3. If we take a−1 to be −a. π −1 × π = 1 = π × π −1 . Let ∗ be the usual addition of whole numbers. 2. 1 × π = π = π × 1. In each example we speciﬁed a set G and an operation ∗ on that set. From school we also know that any nonzero number times 1 is just that number again.(3) above and is thus a group. This combination of elements is similar to the usual multiplication of numbers in the sense that it is associative. i. (2).

Prime Forms Inverted Forms C = 0. 11. 7 . 11 = b G = 7. 3 = d = e B = 11. 4. 11. Similarly. We see that the result of composing transpositions and inversions is also a transposition or inversion. one can verify the axioms of a group and show that G forms a group. 1 = c = d A = 9. 6. 0 5. 1. 4 9. . MATHEMATICAL PRELIMINARIES 26 Example 4. 8. This group is called the T /I group. 9Here chord just means a collection of pitch classes that are played simultaneously. 6. For future reference we also record the letter names of the prime and inverted forms. 4 = e A musician might notice from this labelling that S is the set of 24 major and minor chords. 3. Transposition and inversion “induce” functions Tn : S → S and In : S → S by applying the function to each entry of the input pcseg. 5 10. 7 = g D = E = 3. This is deeper than just considering individual transpositions and inversions by themselves. 9. It can be mathematically veriﬁed that the transposition and inversion compose according to the following rules. 6 = f = g D = 2. Then let G consist of the 24 functions Tn : S → S and In : S → S where n = 0. 2. 0. 2 7. . 4. We also let ∗ be function composition as in the previous example. 3 8. The elements of S can be listed as prime forms and inverted forms. 2. 7. 4. 3. 1 6. We have taken a big step now to consider how the transpositions and inversions interact with each other. 10. 5.9 We use capital letters to denote the letter names of major chords and lower case letters to denote the letter names of minor chords like musicians normally do. 9 = a F = 5. 7 0. 5 = f C = D = 1. 2. 8 = g = a E = 4. 1. They interact with each other to form a group. 0. . Don’t worry about where they come from or why they have these letter names. 10 3. 0 = c G = A = 8. 8 1. This abstract structure of a mathematical group is of tremendous importance because it allows us to see things in music that we otherwise would not see. 11 4. 6 11. The major and minor chords are special chords that are very prominent in Western music. you can just take this list of prime forms and inverted forms for granted. 9. 9 2. 5. 1. 2 = d A = B = 10. so that ∗ really is an operation on G. 11. 8. 10 = a = b F = G = 6. Tm ◦ Tn = Tm+n Tm ◦ In = Im+n Im ◦ Tn = Im−n Im ◦ In = Tm−n Here the indices are read mod 12.2. . 7. . Let S be the set of transposed and inverted forms of the C major chord 0. 10.

and R.3. These functions are highly musical. First we deﬁne functions P. Another mathematical surprise is the following. P ◦ L ◦ P . 3. and R ◦ L ◦ P .g. 7 = 4. 3. 8 = 8. It takes C major to e minor for example. It can be mathematically proven that there are only 24 elements. the elements L ◦ L and R ◦ R and the unit are the same. Let P (x) be that form of opposite type as x with the ﬁrst and third notes switched. A musician would notice that P is the function that takes a chord and maps it to its parallel minor or major. 10Many mathematical things can be proven about this group and the T /I group. 0. For example. 11. 6 . 0 P 3. 3 . For example L 0. P . L ◦ L. L ◦ R. These three functions are also musical in the sense that they take a chord to another one that overlaps with the original one in two notes. 8. some elements of the P LR group are. 11 . This is very deep. 4.1. L. The function R takes a chord to its relative minor or major. The operation is function composition. The P LR group is the group whose set consists of all possible compositions of P . e. For example P 0. 11. For example. R ◦ L. 7 as in Example 4. We also say that these functions are contextually deﬁned because they are not deﬁned on the individual constituents of the pcseg like Tn and In are. So don’t be discouraged at ﬁrst! . Of course I haven’t told you what isomorphic or dihedral means. 8 = 11. At ﬁrst you might think that there are inﬁnitely many ways to combine P. The function L is a leading tone exchange for more theoretical reasons. Let L(x) be that form of opposite type as x with the second and third notes switched. But that is not true! In fact there are only 24 elements of the P LR group.L. 7. 7 = 11. 9 R 3. so we need to get down to some actual musical examples. 4. for example R applied to C major is a minor and R applied to a minor is C major.10 But this is all very abstract. 8 = 4. For example. Consider the group of all invertible functions S → S. 11. L. Then the T /I group is the centralizer of the P LR group in this larger group and vice-a-versa! This means philosophically that the two are “dual” in a musical sense described by David Lewin in his seminal work Generalized Musical Intervals and Transformations. R. and R. they are both isomorphic to the dihedral group of order 24. Definition 3. Let R(x) be that form of opposite type as x with the ﬁrst and second notes switched. 4 L 3. 4. From now on we let S denote the set of prime forms and inverted forms of the C major chord 0. 4. For example R 0. 7 = 7. P applied to C major gives us c minor and P applied to c minor gives us C major. These three functions will be contextually deﬁned just like J in the Addendum. L. and R with domain and range S. 0. and takes a long time to understand. The P LR Group We now introduce the P LR group as a group of functions S → S like the T /I group in Example 4. THE P LR GROUP 27 3. since they ﬁrst appear to be very diﬀerent in their deﬁnitions. but philosophically it means that the T /I group and the P LR group are abstractly the same as the dihedral group! Now that is a surprise. namely L ◦ L(x) = x = R ◦ R(x) for all elements x of S.

ohio − state. .math.html has several links to essays about the subject matter of topology.” Scroll down to the torus and move it around with your mouse to visualize it better. which is the study of the nature and the shape of terrain. To make a torus we start oﬀ with the square sheet in Figure 1. roughly 80 years before Henri Poincar´ initiated the study of e 13Topology is diﬀerent from topography. or twisted into the shape of the other. Beethoven’s Ninth Symphony Ludwig van Beethoven (1770-1827) composed his Ninth Symphony during the years 1822-1824. Some of the essays are more technical than others. But what in the world does the torus have to do with the P LR group and Beethoven’s Ninth Symphony? 6. The website “Math That Makes You Go Wow” mentioned above has an interactive torus. we mean the whole rectangular region in Figure 1 belongs to the square sheet. Next we glue the two circles at the end of the cylinder according to the double arrowheads and make the torus in Figure 3. Go to the website and click on the link in Chapter 2 entitled “Orientable Surfaces: Sphere. The torus looks just like an inner tube ﬁlled with air. literary. Torus. First we glue the horizontal lines according to the single arrowheads and obtain the cylinder in Figure 2.lehigh.edu/ ∼ f iedorow/math655/yale/ is entitled “Math That Makes You Go Wow. Topology is basically rubber-sheet geometry: we imagine two objects are made of rubber and consider them the same if one can be stretched.” This interactive website talks about philosophical. and artistic implications of topology. shrunk. circle to be the same.6.13 The website http : //www. The arrowheads indicate how we will glue. Although it is not shaded in. The Square Sheet.edu/dmd1/public/www − data/essays. The torus is an example of a mathematical object of interest to topologists. Another great website http : //www. musical. BEETHOVEN’S NINTH SYMPHONY 29 Figure 1.

B . G . E . In measures 143-176 of the second movement of the Ninth Symphony one can ﬁnd the extraordinary sequence of 19 chords. e . Notice that all 24 major and minor chords appear below and none are repeated. C. Beethoven did not include the last ﬁve chords below in his composition. but we’ll see why I wrote them below in a minute. we have the diagram below where the arrows are alternately labelled by R and L. f.6. The Cylinder. g . D . . 14This sequence was ﬁrst observed by Cohn in a series of articles dating back to 1991. BEETHOVEN’S NINTH SYMPHONY 30 Figure 2. d. A Here again capital letters refer to major chords and lower case letters refer to minor chords. then L. and so on. b . c. and 1997. In other words. A . The Torus. g. This patten in itself is surprising.1992.14 The letter names of the chords can be converted back to numbers using the table of prime forms and inverted forms in Example 4. then L. Figure 3. c . F. then R. a. topology. B. E. Note that the entire sequence can be obtained by applying to C the functions R.

and Modes of Limited Transposition. i. P.” Journal of Music Theory 42/2 (1998): 241-263. The same goes for P and L. solid lines. and L. “Parsimonious Graphs: A Study in Parsimony. The edges are labelled by the functions R. BEETHOVEN’S NINTH SYMPHONY 31 XYZ[ / _^]\ j d jjjj jjj jjjj jjjj j L jjjj j jjjj jjj jjjj jjjj jj tjjjjR L R _^]\ XYZ[ XYZ[ XYZ[ XYZ[ / _^]\ / _^]\ / _^]\ g c B E jjjj jjj jj jjjj jjjj L jjj jjjj jjjj jj jjjj jjjj j tjjjjR L R XYZ[ XYZ[ XYZ[ _^]\ XYZ[ / _^]\ / _^]\ / _^]\ f A D b jj jjjj jj jjj jjjj jjjj L jj jjjj jjjj jj jjjj jjjj tjjjjR L R _^]\ XYZ[ XYZ[ XYZ[ XYZ[ / _^]\ / _^]\ / _^]\ B G e j g jjjj jj jjjj jjjj jj L jjjj jjjj jjj jjjj jjjj jjj tjjjjR L R _^]\ XYZ[ XYZ[ XYZ[ XYZ[ / _^]\ / _^]\ / _^]\ c f E A jjjj jj jjj jjjj jjjj jj L jjjj jjjj jj jjjj jjjj jj tjjjjR L R _^]\ XYZ[ XYZ[ XYZ[ XYZ[ / _^]\ / _^]\ / _^]\ e D G b Let’s consider the graph on the handout.e.e. This graph is highly musical because the neighbors of a chord are exactly those three other chords that are maximally close to it. by an edge labelled by R) if we can get from one to the other with the R function. and dotted lines respectively. i.15 A graph is just a collection of dots called vertices and line segments called edges which connect some vertices. The edges labelled by R. In this graph the vertices are labelled by the major chords and minor chords. _^]\ XYZ[ C R XYZ[ / _^]\ a L XYZ[ / _^]\ F R . Contextual Transformations.6. elements of S.e. and L are represented by dashed lines. P . Two chords (vertices) are connected by a dashed line (i. those 15This graph and the torus below it are a reproduction of the graph and torus in Jack Douthett and Peter Steinbach.

In numbers. CONCLUSION 32 three chords which overlap with the original one in two pitch classes. 7.7. In that analysis it was fruitful to look at a unit smaller than the subject and to look for overlapping three note groups. You might be interested to know that there are other examples of music on topological objects. If Beethoven had continued the pattern. we just connect the dots which are labelled by the chord progression. Here we are gluing those chords from the left side to those chords on the right side that are the same. Commutative diagrams also appeared in this context. The Tristan Prelude in Wagner’s opera Tristan and Isolde gave examples of transposed and inverted chords in the century after Bach’s death. Without mathematics we would never have heard a torus in Beethoven’s Ninth Symphony. So now we have our torus. See the website “Math That Makes You Go Wow” for further details. For example. See the ﬁgures in the section on topology. the neighbors of D are b. The concept of a group is deeper than individual instances of transposition and inversion because it allows us to see a structure on the collection of transpositions and . 5. d. Conclusion In this module we have investigated some of the group theoretical tools that music theorists have developed to ﬁnd good ways of hearing particular works of music. Bach’s Fugue in d minor from the ﬁrst book of the Well-Tempered Clavier had several examples of transposition and inversion. Notice the pattern. For example. Our next example of a conceptual category was the concept of a group. 2 . all of which agree with 2. the neighbors of 2. L. That’s not really a surprise because P. topology. Our ﬁrst conceptual categories were supplied by the transpositions and inversions. and R were designed to do precisely this. Bach’s Musical Oﬀering contains a passage which is music on a M¨bius strip. The E on the left side is glued to the E on the right side and so on. 2. 9. Well. 19 out of 24 is pretty close though! This example is perhaps the most striking of all our musical examples because it relates group theory. 6. To see that Beethoven’s Ninth Symphony traces out a path on it. What is surprising about this graph is that it makes a torus and the chord progression from Beethoven is a path on it! Let’s see this. Next we glue the circles on the resulting cylinder after twisting them a third of the way around. 9 are 6. the a on the left side is glued to the a on the right side. Schoenberg and Slonimsky also provide o us with examples. 6 9. The subject made an appearance in various forms and these forms were describe by transposition and inversion. These tools provide us with conceptual categories to make our ﬂeeting impressions of music into vivacious ideas. We have to twist to get them to match up. 11 1. and Beethoven all in one! These conceptual categories provide us with tools and a language to ﬁnd an entirely new way of hearing this piece. f . it would trace out all of the 24 major and minor chords. 6. Hindemith’s Fugue in G from the twentieth century had examples of contextual inversion within the subject. 9 in two pitch classes. Notice that we can glue the top and the bottom because the top two rows of vertices match up with the bottom two rows.

The most striking of our musical examples however was in the second movement of Beethoven’s Ninth Symphony. music. Repeated application of R and L to the C major chord generates a chord progression in measures 143-176 and this chord progression traces out a path on a torus. CONCLUSION 33 inversions. The P LR group makes an appearance in the Elvis Progression and in the Beatles song “Oh! Darling” from the Abbey Road album. The functions P. L. At that point. I hope that the conceptual category of this module has made your impressions of mathematics. and music theory into vivacious ideas! .7. we ﬁxed the notation S to mean the set of 24 major and minor triads and considered various invertible functions S → S. All possible compositions of these three functions give us the 24 elements of the P LR group. and R are also important functions S → S. Transposition and inversion induce such functions for example.

How does this relate to the chord progression in the second movement of Beethoven’s Ninth Symphony? 34 . Let’s consider if this deﬁnes a group by answering the following questions about the axioms: Is it true that adding two integers mod 12 gives us another integer mod 12? Is it true that 0 + x = x = x + 0 for any integer x mod 12? Is it true that x − x = 0 = −x + x for any integer x mod 12? Try out x = 1.Lecture 2 Homework Problems 1. Convert the pcseg numbers back to letters using the table of prime forms and inverted forms from Example 4. 3 for example. 6. R 0. stretching. 7 R ◦ L ◦ R 0. Insert 1. 2. Is G = Z12 a group with the above deﬁnitions? 3. Let e = 0 and ∗ = +. Hint: How many holes does each have? 6. Elvis and the Beatles 3. or twisting? 5. Calculate each of the following. 7 Next translate the results into chord names using the letters in the table in Example 4. 6 into the left circle of the diagram displayed in the paragraph about “Oh! Darling” from the Beatles. L 10. 4. 4. Give one way in which a triangle and a line segment are qualitatively diﬀerent. 4. 7 L ◦ R 0. 9. 4 . 1. can we obtain one from the other by shrinking. 5. and R 9. Does it match with the inner four chords of the chord progression for “Oh! Darling”? 5. We already know that addition is associative. 8 . Are the triangle and the circle qualitatively the same? In other words. Mathematical Preliminaries 1. Let G = Z12 . 4. 4. Introduction 2. Beethoven’s Ninth Symphony 6. 7 L ◦ R ◦ L ◦ R 0. The P LR Group 2. Topology and the Torus 4. 3 . Calculate the other three circles by applying the functions. Calculate P 1.

Conclusion 7.7. . Which of our musical examples was your favorite and why? Write at least four sentences. CONCLUSION 35 7.

” Journal of Music Theory 42/2 (1998): 195-206. Musical Form and Transformation: 4 Analytic Essays.” PNM 11/1 (1972): 146223. 1980. and Modes of Limited Transposition. [6] Jack Douthett and Peter Steinbach.” To appear in Journal of Music Theory. “Alternative Interpretations of Some Measures From Parsifal.” unpublished.” Journal of Music Theory 42/2 (1998): 321 -334. and for Schoenberg’s String Trio. [4] John Clough. [12] Jonathan Kochavi. “Some Aspects of Three-Dimensional Tonnetze.” Journal of Music Theory 42/2 (1998): 241-263. “An Informal Introduction to Some Formal Concepts from Lewin’s Transformational Theory.D. [7] Thomas M. “Neo-Riemannian Operations. New Haven: Yale University Press.. diss. 2002. and Their Tonnetz Representations. [10] Julian Hook. Cohn. New York: Schirmer Books.” Journal of Music Theory 41/1 (1997): 1-66. Parsimonious Trichords. “Transformational Considerations in Schoenberg’s Opus 23. “A Rudimentary Geometric Model for Contextual Transposition and Inversion.Bibliography [1] Benjamin Boretz. Number 3. “Generalized Contextual Groups. [13] David Lewin. “Meta-Variations. Generalized Musical Intervals and Transformations. [9] Julian Hook. 1993. “Some Remarks on the Use of Riemann Transformations. [3] David Clampitt.” Music Theory Spectrum 17/1 (1995): 81-118. [17] John Rahn. [11] Henry Klumpenhouwer. 36 .” Journal of Music Theory 42/2 (1998): 191-193. Uniform Triadic Transformations.” Journal of Music Theory 42/2 (1998): 297-306. Childs. Basic Atonal Theory. [18] Ramon Satyendra.” To appear in Music Theory Online. [2] Adrian P. “Generalized Interval Systems for Babbit’s Lists. “Uniform Triadic Transformations. “Moving Beyond Neo-Riemannian Triads: Exploring a Transformational Model for Seventh Chords. Part IV: Analytic Fallout (I). Indiana University. [16] David Lewin. [15] David Lewin. [8] Edward Gollin. Fiore and Ramon Satyendra.” Journal of Music Theory 46/? (2002): ?-?.” Music Theory Online 0/9 (1994). [5] Richard L. 2003. New Haven: Yale University Press.” Journal of Music Theory 42/2 (1998): 307-320. 1987. “Parsimonious Graphs: A Study in Parsimony. [14] David Lewin. Ph. “Some Structural Features of Contextually-Deﬁned Inversion Operators. Contextual Transformations.