Professional Documents
Culture Documents
Certified by . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dr. Lisa Withers
Associate Professor of Music
Thesis Supervisor
Certified by . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dr. John Iskra
Assistant Professor of Mathematics
Thesis Supervisor
Certified by . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dr. Mike Duffy
Professor Emeritus of Physics
Reader
Certified by . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dr. Scott Boltwood
Stuart Professor of English, Chair; Director, College Honors Program
Reader
2
On A Quote By Schoenberg:
A Mathematical Analysis of a Musical Idea
by
Sean Michael Roy Collier
Abstract
A physical, mathematical, and cultural evaluation of Arnold Schoenberg’s “Emanci-
pation of Dissonance” philosphy. (To be expanded upon)
3
4
Acknowledgments
I would like to acknowledge all those who have influenced my academic career while
at Emory & Henry. It is your willingness to foster my interdisciplinary questions that
inspired and led to the writing of this thesis. To you I owe much.
5
6
May not music be described as the mathematics of sense, mathematics as music of
the reason? The musician feels mathematics, the mathematician thinks music–music
the dream, mathematics the working life.
7
8
Contents
3 Arnold Schoenberg 37
3.1 The Man Behind The Name . . . . . . . . . . . . . . . . . . . . . . . 37
3.1.1 An Overview of His Musical Background . . . . . . . . . . . . 37
3.1.2 Research, Methodologies and Philosophy . . . . . . . . . . . . 39
3.2 A Musical Style Within a Changing Context . . . . . . . . . . . . . . 41
3.2.1 Increasing Chromaticism in Harmony and Schoenberg . . . . . 41
9
3.2.2 Twelve-Tone Serialism . . . . . . . . . . . . . . . . . . . . . . 42
3.3 Schoenberg on Consonance and Dissonance . . . . . . . . . . . . . . . 44
10
5.9.2 Differential Equations . . . . . . . . . . . . . . . . . . . . . . 77
5.9.3 The Wave Equation . . . . . . . . . . . . . . . . . . . . . . . . 78
Bibliography 105
11
12
List of Figures
13
14
List of Tables
15
16
Chapter 1
17
into a context in which a new definition is required, a distinction will be made. For
reference, I will make use of subscripts to denote the octave of a pitch class as opposed
to the sometimes utilized Helmholtz notation.
18
1.3 Consonance and Dissonance
In short, consonance is when notes sound in harmony, or stable, with one another. Its
conventional opposite is dissonance. These two ideas are interrelated in that they are
the driving forces for music, where consonance depends on the tension and sequential
resolution of dissonance.1 Throughout the evolution of Western tonal theory, this
idea of dissonance has taken on many different definitions. When Pythagoras defined
these harmonic relationships, the only consonant, and known, relationships were that
of the octave, the 4th and the 5th.2 These intervals were discovered through, as the
story goes, Pythagoras’ experience and experimentation with the sounds produced by
striking metal with hammers of different weights.3 Within his experiment, he found
that particular hammers produced intervals that were not pleasant to the ear, these
were discarded, and deemed dissonant.4 One of the first people to write about these
inherent relationships between tones was Boethius.
Boethius gave, perhaps, the first definition of consonance, being that it is “a
mixture of high and low sound falling pleasantly and uniformly on the ears.”5 He
also gave a definition for dissonance, where it is “a harsh and unpleasant percussion
of two sounds coming to the ear intermingled with each other.”6 Boethius defined
the duple, triple, quadruple, sesquialter and sesquitertian ratios, which are 1:2, 1:3,
1:4, 3:2, and 4:3 respectively, as consonant intervals. In relation, all others were
considered dissonant.7 For reference, these ratios are an octave or “diapason,” a 12th
or “diapason-plus-diapente,” a 15th or “bis-diapason,” a perfect 5th or “diapente,”
and a 4th or “diatessaron.”8 It is worthy of note that composers have not held true
to these classifications over time. An example of this is the shift of the perfect 4th.
1
Richard Parncutt, Harmony: A Psychoacoustical Approach:... Springer-Verlag Berlin An, 2012.
Print., pg. 56.
2
Ibid., pg. 56.
3
Piero Weiss and Richard Taruskin. Music in the Western World: A History in Documents.
Australia: Thomson/Schirmer, 2008. Print., pg. 5.
4
Boethius, Anicius Manlius Severinus, Calvin M. Bower, and Claude V. Palisca. Fundamentals
of Music. New Haven: Yale UP, 1989. Print., pg. 18.
5
Ibid., pg. 16.
6
Ibid., pg. 16.
7
Ibid., pg. 76.
8
Ibid., pg. 16.
19
Originally considered consonant, it has moved to being considered both, depending
on the context within which it finds itself.9
During the medieval period, church music was based on the eight modes: The Au-
thentic and Plagal versions of the Dorian, Phrygian, Lydian and Mixolydian scales.10
If an interval did not appear in one of these 8 church modes, then it was usually
considered as dissonant, except for intervals that contained the sole accidental of
B[.11 These intervals were leaps such as the tritone, and the church banned the
vast majority of them from appearing in liturgy. Slowly, the 3rd and 6th became
accepted but were considered “disjunct”, followed by many others.12 In the classi-
cal period, conventional harmony had been established through performance practice
and dissonances gained a firm definition. From this period, dissonances are considered
non-chord tones, such that in the composition of a major or minor chord consisting
of a root (fundamental), major or minor 3rd, perfect 5th, and sometimes 7th, there
exists an interval present that is not one of these notes.13 Examples of dissonance
include passing tones, neighbor tones, suspensions, appoggiaturas.14 Much of the con-
ventional harmony arising out of the western tradition stems from the compositional
styles of the great composers. Prolific writers, such as J.S. Bach, G.F. Handel, W.A.
Mozart and Antonio Vivaldi, to name a select few, had tremendous impact on what
was considered the proper way to “deal” with dissonances and create the ebb and flow
so commonly associated with them and their consonant counterparts as described by
Francis Bacon.15
The 20th century marked the culmination of increasing chromaticism leading to
what Arnold Schoenberg deems as the “emancipation of dissonance.” Moreover, he
adds the term emancipation of dissonance to the musical lexicon to represent the
9
Vincent Persichetti, Twentieth-Century Music, New York: W.W. Norton, 1961. Print.
10
Richard H. Hoppin, Medieval Music. New York: W.W. Norton, 1978. Print., 65.
11
This is an assumption made through context, but I have found no concrete information to back
up this claim as of 3/13.
12
Ibid, pg. 75.
13
Stefan M. Kostka, Dorothy Payne, and Allan Schindler. Tonal Harmony: With an Introduction
to Twentieth century Music. 7th ed. McGraw-Hill, 2013. Print., pg. 171.
14
Stefan M. Kostka, Dorothy Payne, and Allan Schindler., pg. 173.
15
Weiss., pg. 163
20
philosophical idea surrounding the real-world situation.16 The physical emancipation
was described as being the dissolution of the dichotomy between consonance and dis-
sonance, that western culture had become adjusted to a lack of resolution and was
able to progress melodically within a stagnant, atmospheric, harmonic background.17
Within the philosophy, dissonance was defined simply as something to be understood
or comprehended as a musical idea, and not as something that may or may not sound
pleasant in a particular context.18 From then on, Schoenberg and other atonal com-
posers subscribing to his philosophy attempted to break from conventional tonality
through the emancipation, but always kept in mind the direct opposition from the
formalized tonal structures they tried so hard to forget.
As musical structures began to expand both in size and in number, the need for
an expanding musical vocabulary led to the evolution of different chordal structures,
patterned leaps, melodic phrase structure and tonal tendencies. This included things
such as the rise from monophony to polyphony or the shift from counterpoint to
harmony. Culture played a large role in this evolution, as the Church’s early reign
over polyphony and homophony, such as allowed intervals and what society deemed
“acceptable” or “nuanced” dictated how much early music was allowed to expand.
Throughout history, factors for expansion have ranged from honoring God to creating
new harmonies to accommodate non-western tonalities. As the story now goes, histor-
ical evolution led to a rise in consonant structures, which in turn produced dissonant
counterparts. As composers began to extrapolate and abstract upon previously known
intervals, and thus break from convention, the emancipation of dissonance followed
suit, bringing about the end of tonal harmony in many cases. Finally, Schoenberg
presented the idea that all tonal interactions are dissonant to some degree, as even-
tually, the overtones will no longer be consonant, even if the fundamental pitches
are.19
16
Arnold Schoenberg, Style and Idea: Selected Writings from Arnold Schoenberg. New York: St.
Martins Press; London: Faber & Faber, 1975. Print., pg. 104.
17
Ibid., pg. 104.
18
Ibid., pg. 105.
19
Arnold Schoenberg, Style and Idea., pg. 104.
21
1.4 A Brief Note About Temperament
With the evolution of music comes the progression of the instruments on which it is
produced. Deeply seeded in this progression is the topic of tuning, a necessity for
instruments incapable, whether practically or otherwise, of retuning in the moment.
Today, the Western world is based on the inherently dissonant Equal temperament,
meaning things are not perfectly in tune with the pure frequency as developed by
Pythagoras and Boethius, rather everything is tuned relatively, which causes equiv-
alent pitch classes to have slightly different frequencies and thus be out of tune–
arguably a type of dissonance. The familiar, and rather young, tuning system was
highly contested and deemed less worthy than the unstandardized systems used by
Bach, Mozart, and Beethoven–for many reasons to be discussed in time–with ex-
amples emerging throughout the centuries from writers like Henry Pleasants, Stuart
Isacoff, Ross Duffin, Marin Mersenne, and CPE Bach. Equal Temperament, in par-
ticular, was the first tuning system to entirely break from the ideas first laid out by
Pythagoras, an important distinction to be explored. However disliked, it became
standardized for its applicability and ability to meet the demands which the other
systems could not.20
22
interpretation of the aforementioned texts from a mathematical standpoint as well as
a physical study of a piano tuned in ET through methods in Fourier and Harmonic
Analysis.
23
24
Chapter 2
Temperament is defined as “a way of tuning the notes of a scale using intervals that
have been modified (tempered) from their pure forms.”1 Essentially, temperament is
the way in which instruments are tuned, effectively pairing the pitch classes within
C to a given ratio from some fundamental. Though it seems logical to consider
tuning systems that follow the natural ratios discovered by Pythagoras–and those
latter added by Ptolemy and Boethius– as temperaments, it is understood that these
exist outside of this context since no actual “tempering” has occurred. As these
systems were common place until the advent of the keyboard, the appearance of the
instrument is usually considered the beginning of the need for temperament, and
thus temperament’s start in the 13-1400s.2 It is logical to ask why a tuning system
beyond these pure intervals need be created in the first place, especially if it worked
so well from antiquity to then. The answer lies within the circle of fifths and its pure
fifths counterpart, shown in Figure 2-1. Duffin provides the following example and
rationale:
If you were to start at a very low note on the piano and tune acoustically
pure [those natural ratios defined by Pythagoras] fifths up from there,
1
Duffin., pg. 38
2
Ibid.
25
eventually you’d expect to get back where you came from. For example,
starting on C, you would tune C − G − D − A − E − B − F ] − C] − G] −
E[ − B[ − F − C. Twelve fifths should equal seven octaves and come back
around the ‘circle’ to C...the twelfth fifth...actually overshoots the target
pitch and puts us in a ratio of 1.014: 1 instead of 1:1.3
The issue here is that the pure intervals overshoot the expected tone, creating what
amounts to a system which spirals out of controlliterally. This blip in the system is
referred to as comma and will play a pivotal role in the development of its solution.
More directly, the comma is the difference between the tuning system being used and
acoustically pure intervals. Temperament, then, attempts to remove the comma by
“shaving off” a bit from the fifths so that when the same experiment is repeated, the
circle is completed as expected. Since instruments, both vocal and not, lacked the
range of the newly introduced keyboard, this issue was rarely encountered prior to
their introduction into the music world.
Throughout western music, there has existed a plethora of temperaments, both
formal and personal, over the course of history. The temperaments we will consider,
such as Equal, Just and Quarter-Comma Meantone, are strongly Western, but it
3
Ibid., pgs. 44-5
26
should be noted and may be considered in future research that other temperaments
exist beyond western music. They, too, come with their own amalgamation of inter-
vals, shavings and constructions.
Perhaps the most popular temperament of the 15th century and through the Re-
naissance, quarter-comma meantone attempted to fix the comma by narrowing, or
1
flattening, each fifth by 4
of the temperamental blip; in addition, it was also an at-
tempt to maintain acoustically pure major thirds, i.e. major thirds which follow the
interval 5:4, as opposed to the commonly used Pythagorean third of 81:64.4 This
temperament is the first of note to deviate from the previously upheld Pythagorean
tuning, but it is important to mention the attempt to maintain the inherent acoustic
properties in this new temperament, albeit the loss of the pure 5th. Quarter-comma
meantone, then, served two purposes: creating consonant, acoustically pure thirds
which could then be used compositionally as well as removing the comma and allow-
ing for a complete circle of fifths. A detailed description of Q.C.M. can be found in
the appendix.
Though octaves were now tempered to be consistent, quarter-comma meantone
4
Ibid., pgs. 34-5
27
was not without its faults. Due to its basis on the acoustically pure major third,
the resulting circle of fifths contained a wider, or extremely sharp, fifth for the 9th
chromatic pitch above the tuning fundamental. If one was to tune a piano in Q.C.M.
starting with any C as the fundamental, then the fifth from G] to E[ would be
considerably wide–essentially a diminished sixth. This came to be known as the wolf
due to its characteristic “howl” when played.5 A representation can be seen in Figure
2-2.
Since it was dependent on the fundamental tuning pitch, the wolf could be placed
in a fifth that was unlikely to be used. History has shown that the wolf typically ap-
peared between C] and A[ or G] and E[, though again, it could be placed anywhere.6
As one could imagine, then, if directions were not stated clearly on where to tune
the instrument for a particular piece, what was intended to be a harmonious con-
cord of fifths could quickly be seen as dissonant by the unsuspecting listener. These
implications go well beyond this, as I will show soon enough.
5
Ibid., pg. 35
6
Ibid.; Ian D. Johnston. Measured Tones: The Interplay of Physics and Music. Bristol,: IOP
Publishing, 1989. 74
28
2.2.2 Well Temperament
It has long been of popular opinion that Johann Sebastian Bach’s series of 48 preludes
and fugues within the Well-Tempered Clavier serve as evidence for Equal Tempera-
ment’s popularity within Bach’s time.7 Surely it would be the case that Bach was
using ET if he created a work that exploits the capability to transpose, modulate,
maintain sequential patterns, and so on. This, perhaps surprisingly, was not the case.
In 2005, Dr. Bradley Lehman deciphered what previously interpreted as a fanciful
drawing on the title page of Bach’s masterwork.8 The “spiraling figure” as shown in
Figure 2-3. is what Lehman describes as:
Bach’s specific and unequal method of keyboard tuning. He did not ex-
press it in our normally-expected modern formats of theory, or numbers.
Rather, he drew a diagram that looks like a practical hands-on sequence
to adjust the tuning pins, working entirely by ear. The music constrains
any solution to be nearly equal temperament, and then, the diagram’s
irregularity shows how to shape the desired subtle inequality.9
This type of tuning has come to be known as well temperament and was notably
practiced by theorist Andreas Werckmeister in 1691.10 Other composers, such as
Johann Georg Neidhardt and C.P.E. Bach are known to have used tuning systems
similar to Bach, but with their own characteristic take on just how much to temper.11
Again, we see something that is highly personal to both the composer and his music.
7
Duffin., pgs. 44-45
8
Bradley Lehman. Johann Sebastian Bachs Tuning. LaripS.com. Accessed November 5, 2017.
http://www.larips.com/.
9
Ibid.
10
Stuart Isacoff. Temperament: How Music Became a Battleground for the Great Minds of West-
ern Civilisation. London: Faber and Faber, 2007., pg. 216
11
Duffin., pgs. 148-149
29
Lehman summarizes Bach’s method as follows, “We set up the notes of the C
major scale first, and then we fit the remaining notes into carefully compromised
spots.” This is accomplished by choosing a starting pitch, in this case C, and tuning
the natural fifths and the thirds, F − C − G − D − A − E and F − A, C − E, as
described above with Q.C.M.–that is, lower the fifths and raise the thirds; now, as
opposed to Meantone, raise B − F ] − C] just a bit; then with G] − D] − A], flat
half as much as with the natural fifths.12 This, in theory, provides what amounts
to J.S. Bach’s Well Temperament. This kind of tuning is known as an irregular
tuning system. It was seen with Q.C.M. that all of the fifths were tuned in the
same way, where here some were true, some were flat and some sharp. Q.C.M.,
then, is known as a regular temperament. The two of these fought it out for the
hearts of performers until equal temperament’s subsequent adoption in 1917. Again, I
have shown that slightly different, and co-existing, temperaments can have drastically
different features, characteristics, and implications for the music they are used to
produce.
This tuning system is somewhat in a sphere of its own, and mentioned here only for
historical purposes. Just intonation involves many of the same principals as Pythago-
ras’ ratiomatic description while solving its troubled areas.13 In particular, the issue
surrounding the grating major thirds of Pythagorean tuning. In general, any tuning
system that maintains acoustically pure perfect fifths and major thirds–again, those
that match the ratios 3:2 and 5:4 respectively–is considered just according to Isacoff.
This take on tuning was first proposed by Spanish theorist Bartolomeo Ramos de
Pareja in 1482 as an attempt to solve the wolf present in meantone temperaments.14
While it provided just, or pure, intervals that were in “true consonance,” it still car-
ried issues of inconsistency. Due to the nature of tuning in this way, transposibility
12
Lehman
13
Isacoff., pg. 97
14
Ibid.
30
and quality of chords were lost, which is a major issue as counterpoint begins to take
hold during this period of music history.
Just intonation has come to be interpreted more loosely in modern times. In a
more general sense, it is understood to be a tuning system that has the ‘ability’ and
‘flexibility’ to produce harmonically pure intervals at any given time, depending on the
needs of the harmonies in which it is contextualized.15 In other words, the instrument
has the capacity of making sudden shifts in tuning in order to meet the needs of the
harmonies, which could entail raising a leading tone, making an acoustically pure
triad, and so on. Therefore, this type of tuning is only plausible on instruments such
as strings and voice. Certainly no one is going to sit at a piano with a tuning peg
with hope of achieving a harmonically pure Picardy third following their otherwise
tempered work. It’s not sensible; however, just intonation holds commonplace in
string quartets, even today, and provides our tempered ears a brief taste of what has
been taken by Equal Temperament.16
The official turning point came with the publication of William Braid White’s
Modern Piano Tuning and Allied Arts.18 The Cambridge-trained acoustician “ba-
sically founded the science of truly equal tempered piano tuning. His process of
15
Duffin, pg. 33
16
Hermann Von Helmholtz, and Alexander John Ellis. On the Sensations of Tone as a Physiolog-
ical Basis for the Theory of Music. New York: Cambridge University Press, 1954.
17
Owen Jorgensen. Tuning the Historical Temperaments by Ear. Marquette Mich.: Northern
Michigan University Press, 1977.
18
Duffin., pg. 112
31
counting beats and of using comparative intervals throughout the process made it
finally possible to ensure that the temperament was truly equal throughout.”19
In short, equal temperament is set up by dividing an octave into twelve notes,
or semitones, which share a ratio of 1:1.0595 between them.20 This number is the
√
12
2, which comes from the idea of a semitone being half of a whole tone and there
being 12 of them in an octave. Therefore, the notes for ET are built by repeatedly
multiplying pitches by this value. For example, if one were to start with A440, as is
typical in music today, then the numerical value for B[ is found by multiplying 440
by 1.10595. This pattern then continues up and down the scale. Using these values,
White could compute the number of beats that should be present within his checks
and thus produced as sensible, doable methodology for tuning a piano in ET. The
final effect of this was the narrowing of the perfect fifth and a widening of the major
third in a consistent manner. A full chart of ideal frequencies for notes in ET can be
found in the Appendix.
The idea of equal temperament was not new, though, as talk of this tuning system
can be found as early as 1640, if not before.21 Much can be said about it taking
over 300 years for the adoption to take place. The system was seen as obtrusive,
against what had been known and respected for hundreds of years.22 In particular,
ET lacked, almost entirely, Pythagorean ratios, with the only example being the
octave. The countless efforts observed over the years with regards to regular and
irregular tuning systems is a testament against what many felt was the destruction
of music. It should also be mentioned that ET was almost unachievable prior to
White’s publication. Before this, it is generally accepted that systems tuned to what
was believed to be ET were actually a close copy in the form of well temperament.23
Alexander Ellis made the comment that “[ET] is, however, so difficult to realize by
ordinary methods of tuning, that ‘equal temperament’ . . . has probably never been
19
Ibid.
20
Johnston., pg. 34
21
Duffin., pg. 40
22
Isacoff., pg. 16-18
23
Duffin., pg. 112
32
attained in this country, with any approach to mathematical precision.”24
33
These sensibilities of keys gave music an overlaid expressive element unable to be
captured in ET. Francesco Galeazzi wrote that B[ major was “tender, soft, sweet, ef-
feminate, fit to express transports of love, charm, and grace,” while E major was “very
piercing, shrill, youthful, narrow, and somewhat harsh.”29 Ludwig van Beethoven’s
biographer, Anton Felix Schindler, has discussed Beethoven’s opinions on this:
The emphasis on the piano specifically implies a consequence of well and meantone
temperaments, thus providing evidence for these key-related moods. Hundreds of
examples exist in regards to people’s disdain over the killing of sensibilities and this
is perhaps the fundamental argument against the standardization of Equal Tem-
perament in conjunction with the more philosophical departure from the respected
Pythagorean ratios.
These sensibilities, and more so this distaste for equal temperament, rests on the
fact that keys sounded fundamentally different. This, in turn, implies that certain
intervals sounded fundamentally different based on which key they found themselves
on the keyboard. For this reason, it is important to consider the sensibilities in
that consonance was used loosely. With that, its negated counterpart would also
bring with it a loose definition based highly on context and key signature. These
sensibilities imply a plethora of interpretations for these musical structures, not the
simple harmonious binary suggested by Arnold Schoenberg. This, in conjunction with
the idea of standardization of tuning systems, lead to a major flaw in the philosophy
at question within this thesis. For a more complete analysis of ET’s role in Harmony,
refer to Ross Duffin’s How Equal Temperament Ruined Harmony.
29
Bathia Churgin. Francesco Galeazzis Description (1796) of Sonata Form. Journal of the Amer-
ican Musicological Society 21, no. 2 (1968): 181-99. doi:10.2307/830853.
30
Anton Felix Schindler. Biographie von Ludvig van Beethoven (Mnster, 1860) trans. after Con-
stance S. Jolly, ed. D.W. MacArdle, Beethoven as I Knew Him, Chapel Hill, N.C., 1966., pg. 367
34
2.5 Standardization and Tuning: A major concern
It has been shown that tuning varied wildly throughout history until ET’s adoption
in 1917. Moreover, it has been shown that tuning systems were not even consistent
within the same town. Bach’s tuning was personal to him, as described by C.P.E.
Bach in his father’s obituary.31 If someone were to play a piece written for them by
Bach, say, it wouldn’t necessarily sound the same on their harpsichord as it did on
Bach’s when he wrote it. Others, such as Neidhardt, created over 24 temperaments
appropriate for different social classes.32 Alexander Ellis, who questioned the authen-
ticity of ET in general, wrote that in “1854, the first organ [was] built and tuned
originally in equal temperament” while “meantone temperament was still general in
Spain, and used in England on Greene’s three organs, at St. George’s Chapel.”33
This suggestion holds on a historical level as well. Pieces written during a period
of meantone temperament would ideally work in a well-tempered system, but would
they sound as intended? Likewise, would a piece written based solely on Pythagorean
ratios have the same intervallic relationships when played in meantone?
These inconsistencies, this clearly non-linear historical progression of tempera-
ment, shows that dissonances, as well as consonances, cannot be understood to be
consistent from a historical standpoint. The highly dissonant and “grating” major
third from the early church, produced from Pythagorean tuning, is not the same
major third produced from meantone, just, or well temperament. Likewise, a mi-
nor second, what is now considered one of the most dissonant intervals, has surely
failed to be consistent from a historical standpoint. I argue, then, that a culture of
people, particularly those in western music, cannot have become “desensitized” to
dissonances in the way that has been described by the author in question, that is to
say, it is not due to the gradual inclusion and exploration of (dissonant) intervals,
chords and tonal structures that has led to this emancipation. I aim to show that it
is due to something far more fundamental, ever present, and consistent. Schoenberg
31
Weiss
32
Isacoff; Duffin
33
Helmholtz., pg. 549
35
was almost right, but for the wrong reason.
36
Chapter 3
Arnold Schoenberg
Arnold Franz Walter Schoenberg was born to a lower, middle-class family in Vienna
in 1874.1 Due to his familial status, Schoenberg lacked a formal education in the
sense that early on, he was expected to drop out of school to serve as a clerk for a
small bank following the death of his father.2 This, unsurprisingly, meant he did not
follow the typical route of many of his contemporaries. Instead, his musical journey
was largely self-maintained with only a few scholars to provide guidance, instruction
and motivation.
Musically, his journey started with violin lessons around the age of eight.3 These
then led to his first compositions for the violin and viola. During his time as a clerk,
Schoenberg spent his evenings exploring disciplines such as “music, literature and
philosophy.”4 Of his two friends made during this time, Oskar Alder, a proficient vi-
olinist, became what was essentially Schoenberg’s first music teacher. As Schoenberg
1
Neighbour, O. W. Schoenberg, Arnold. Schoenberg, Arnold
Grove Music. January 20, 2001. Accessed January 10, 2018.
http://www.oxfordmusiconline.com/grovemusic/view/10.1093/gmo/9781561592630.001.0001/
omo-9781561592630-e-0000025024.
2
Ibid.
3
Ibid.
4
Ibid.
37
began to teach himself cello, Alder supplemented his journey with some elementary
harmony and the capability to perform chamber literature.5 This in turn led to com-
positions and continued to fire Schoenberg’s passion for music in what could have
easily seemed a desperate situation.
He soon joined an amateur orchestra where he met Alexander von Zemlinsky.
This friendship would open the door for Schoenberg’s journey into the music world as
well as providing him with the necessary and so deeply missed instruction needed to
continue. A student at the Viennese Conservatory, Zemlinsky was able to provide a
helpful musical viewpoint of Schoenberg’s works in conjunction with his connections
he had made with big composers such as Brahms and Mahler.6 In this sense, Zem-
linksy was Schoenberg’s first, and potentially only, legitimate teacher. Though his
influence is “questionable” on Schoenberg’s academic life, it is clear that the budding
composer deeply respected Zemlinsky’s opinions and tutelage.7 Apart from Zemlin-
sky’s mentorship, Schoenberg is considered to have been “essentially self-taught.”8
He “mastered traditional technique and remained all his life involved in the study
and teaching not only of the classical discipline but also of the profoundest and most
universal aspects of the tradition.”9
As his life progressed, Schoenberg became an influential figure in his field, where
his “artistic philosophy and its influence has by no means disappeared.”10 Musically,
his periods can be broken down, similar to Beethoven, into four groups: early tonal,
expressionistic, serial, and an occasional return to tonality.11 The third period is what
is commonly portrayed as representative of Schoenberg, in particular, his methodol-
ogy of twelve-tone serialism preceded by his definitive break from tonality in 1908.12
Some examples from each period include: “Six Songs,” Op. 3, Erwartung, Op. 17,
5
Ibid.
6
Ibid.
7
Ibid.
8
Salzman, Eric. Twentieth-century Music: An Introduction. Upper Saddle River, NJ: Prentice
Hall, 2002., pgs. 33-4
9
Ibid.
10
Ibid., pg. 34
11
Neighbour
12
Salzman., pg. 35
38
Variations for Orchestra, Op. 31, and “3 Folksongs,” Op. 49. Moving forward, he
eventually accepted a professorship at the University of California in Los Angeles
where he taught theory and composition. Following a sharp decline in his health,
he taught sparingly, including periods at the University of Chicago and Santa Bar-
bara. In 1951, Arnold Schoenberg died after a tiring battle with diabetes, asthma
and cardiac issues. Shortly before his death, he revised a small selection from his vast
accumulation of largely unpublished essays and published Style and Idea.13 Much of
his main philosophical thoughts on music can be found within this text, including
that which is in question here.
Unlike hundreds of theorists that came before him, especially those born into a deep
reverence for classical tonality, Schoenberg was “never primarily interested in working
out the harmonic implications of these kinds of [tonal] structures,” but was rather
focused on researching and seeking out the meaning of “harmonic structures in their
relationship to line.”14 That is to say, he was interested in the structures themselves,
their boundaries, their characteristics and what could be inferred from extrapolation.
A brief example of this is his introduction of set theory into musical harmonic analysis,
where one can understand a chord in a new framework that provides algorithmic
methodologies to form new structures from that single chord.
His technique and development are described by Salzman as follows:
39
ideas, older and more universal than classical tonality itself, that we can
understand his development.15
His fascination with the theoretical foundations of music was deeply rooted in
tonal tradition. This time is considered to be the point in music history where com-
posers began to treat music as a self-sufficient entity in its own right. They began
to understand music as a means for itself rather than existing to please the listener.
It was a time in which composers began to question what had been treated as law
for the past 1900 years. This is evident in Schoenberg’s music as well as his teaching
philosophy. Erwin Stein, a pupil of Schoenberg, reflects on his teacher’s philosophy,
“Schoenberg could not be satisfied with showing the pupil how something is done;
the teacher and pupil should search behind and beyond the how’, trying to find what
it is, and why it is so, and indeed, whether it need be so.”16 Almost all of his theo-
retical formulations and teachings on twentieth century tonal harmony can be found
in his Harmonielehre or Theory of Harmony–a name despised by Schoenberg–written
during his professorship at UCLA. This work is an amalgamation of his philosoph-
ical views mixed with a systematic exploration of harmony near its breaking point,
providing valuable insight into his theories on music.
Schoenberg is considered by some to be thoroughly Hegelian, where he “believed
that music, like all aspects of human life, is part of a process of change and historical
evolution. For Schoenberg the classical tradition of form–the concept of all-pervasive
intellectual organization–represented a universal principle, while the development of
chromaticism represented a principle of change and evolution.”17 This is paired with
his views “on more complex relationshipson the similarities and relationships between
artistic creation and other human activities, on the connections between the natural
world outside ourselves and the participating or observing subject.”18 Music, to
Schoenberg, was the lens by which he considered and analyzed nature, art, culture,
and irrational aesthetics. These ideologies Schoenberg associated with the study
15
Ibid.
16
Schoenberg, Arnold, and Roy E. Carter. Theory of Harmony. Berkeley, CA ; Los Angeles, CA
; London: University of California Press, 1978., pg. xv
17
Salzman., pg. 34
18
Scheonberg,Theory of Harmony., xv
40
of music created a sense of interconnectedness between music and the world that
surrounds it. Music is heavily influenced by the world and remains influential within
the culture that produced it, creating a cyclical relationship that transcends time.
In understanding Schoenberg’s views on music, and more specifically, music the-
ory and the structures foundational to it, one can more thoroughly understand his
philosophies considered here as well as the context surrounding their propositions.
His writings within Style and Idea provide a more complete framework to under-
standing the mental workings that provided his philosophy, and should be read by
all. I think it is important to understand the context that fostered the “emancipation
of dissonance” if we are to fully comprehend its implications.
41
The musical journey of Schoenberg followed an abbreviated version of the history
of western tonality. His early works are considered highly tonal, but he quickly
moved towards a more Wagnerian sense of composition, influenced also in part by
Bach, Brahms and Beethoven.19 By his twenties, he had moved to expressionism,
more commonly referred to as atonality. Familiar and notable pieces include the
Second String Quartet and Pierrot Lunaire, Op. 21 characterized by “the impulse
toward the highest chromatic use of all twelve notes on an even and revolving basis.”20
This evolved during his forties into twelve-tone serialism– the third period. The use
of a tone row and matrix characterized this movement that remained Schoenberg’s
composing method for sixteen years.21 He regarded serialism as a way to “maintain
the continuity of the great tradition while carrying it forward into the future.”22 The
final period marks something of a return to tonality.
It is almost poetic that Schoenberg goes on this intense journey through music
only to end up at a point similar to that where he began. I think something can
be said in terms of comfort in the familiar. His health was suffering, and music
may have moved from being an intellectual endeavor to an expression of the emotion
within him. Alone for much of the latter part of his life, his musical compositions
served as an extension of his sense of self. Perhaps, then, this return to tonality, or
something similar, was an attempt to reconnect to a world he felt he had left behind in
pursuit of a greater understanding. Nevertheless, this journey affected his music and
philosophical views in considerable ways. In particular, focusing on his third period,
that of serialism and twelve-tone music, provides valuable insight into Schoenberg’s
understanding of consonance, dissonance, and temperament.
Briefly, twelve-tone serialism involves the bijective mapping of the pitch classes in
the chromatic set C to a twelve-tone row in a structured formulation which is then
19
Salzman., pg. 34
20
Ibid., pg. 35
21
Neighbour
22
Salzman., pg. 196
42
manipulated to by modular arithmetic to provide all combinations, inversions, trans-
positions, and retrogrades of the original row through the tone matrix.23 This idea
rests heavily on the idea of enharmonics–the idea that an A[ is equivalent to a G] in
every context. This idea came along with equal temperament, as before this the idea
of enharmonics was rarely used; consequently, these pitches were often regarded as
separate entities with different tonal functions within particular settings. This went
so far as to having separate keys for them on instruments. If this was still the case, C
would contain more than 12 pitch classes with different implications and connotations
for themselves and the intervals they help create. From this, the entire structure of
the twelve tone matrix would break down as the uniformity of the system would no
longer exist. This brings up an important dependency: Schoenberg needed equal
temperament for his grand idea to function properly, as other tuning systems, such
as the ones mentioned here, did not always follows this mindset.
It is thought, then, that “the growth of equal temperament, chromaticism, and
modulation had made possible the historical rise of functional tonality in the sev-
enteenth and eighteenth centuries and destroyed it in the twentieth; thus tonality
contained within itself from the start the seed of its own destruction.”24 Coming of
age alongside ET, Schoenberg surely was aware of the criticisms of such a system,
and due to his formulation, unintentionally formulated a case for its adoption. This
dependency gives more context for his philosophies on consonance and dissonance,
in particular, it tells us that he was thinking of harmony and its components from
an equal temperament frame of mind. The implications of this will soon become
apparent.
As an interesting aside, the twelve-tone system forms a group with transposition
as its operation and the tone matrix its Caley table, i.e. it is a complete, closed
system that is highly structured and can be understood through this context.
23
Salzman., pgs. 111-6; Stefan Kostka. Materials and Techniques of Post-Tonal Music. 4th ed.
Englewood Cliffs, NJ: Prentice Hall, 2012. Print
24
Salzman., pg. 34
43
3.3 Schoenberg on Consonance and Dissonance
Schoenberg’s views on the theory of consonance and dissonance can be found in
his Harmonielehre, inappropriately translated as the Theory of Harmony. Writing
specifically on the juxtaposition between the two classifications, Schoenberg makes the
claim that the upheld expressions of consonance and dissonance are “unwarranted,”
and that “the evolution of harmony will, in a short time, prove the inadequacy of
this classification.”25 Due to the centuries of music theory against him, he settles on
a classification of the two in terms of the overtone series. He provides new definitions
of the antithesis as follows:
In short, if, given a fundamental note C, another note can be found within the
first few overtones, then these pitches are considered consonant. Considering the
contrapositive, if the pitch cannot be found within the first few overtones, he defines
them as “more remote consonances (today called ‘dissonances’).”27 For example: let
the fundamental be C2 , then the first few overtones C3 , G3 , C4 , and E4 would be
considered consonant; moving forward, further overtones such as B[ and D would be
considered dissonant, or more remote consonances.
Exploring this idea of overtones more directly, Schoenberg defines, as is usual in the
music theory mindset, “the following as dissonances: major and minor seconds, the
major and minor sevenths, the ninths, etc.; in addition, all diminished and augmented
25
Schoenberg, Theory of Harmony., pg. 21
26
Ibid.
27
Ibid., pg. 22
44
intervals, thus, diminished and augmented octaves, fourths, fifths, etc.”28 This then
implies that all other intervals are consonant; however, he makes the observation that
the minor third and major and minor sixths do not appear in the overtone series as
well as the “imperfect consonance” of the fourth, as it can be found in the inverted
overtone series, but not directly. The same can be said for the preceding outliers, and
thus he forms a complete system of categorization.
It is important to note, though, that Schoenberg only provides this system to
classify consonance and dissonance for the understanding of the student learning
from his Harmonielehre in a curriculum that favors Western tonal music theory.
Throughout the work, he interjects questions and statements that imply his differing
view on the subject. Some examples of this include “which dissonances are to be
dealt with?” and further still, “dissonance requires this careful handling only so long
as it is still a striking occurrence in our harmonic life.”29 These views are again
represented in his later work Art and Style, where twelve-tone serialism leads to the
following conclusions from Schoenberg:
[W]hile a “tonal” composer still has to lead his parts into consonances
or catalogued dissonances, a composer with twelve independent tones ap-
parently possesses the kind of freedom which many would characterize by
saying: “everything is allowed.” “Everything” has always been allowed to
two kinds of artists: to masters on the one hand, and to ignoramuses on
the other.30
His true mindset is reflected in his proposal that today’s dissonances should be consid-
ered more remote consonances, and moreover, that the mere classification of intervallic
relationships is inadequate at categorizing the plethora of situations possible.
The implied distinction between these two classifications reflects the philosophical
journey Schoenberg found himself in the early 1900s. No longer was it about simple
part writing rules for well-taught composition students; no longer was it how the
composer ought to treat dissonances, rather it was why these rules existed at all.
28
Ibid.
29
Ibid., pgs. 92-100
30
Schoenberg, Art and Style., pg. 130
45
What was it that formed them and should they be followed at all? These ideas
became enveloped into to the growing musical conflict surrounding Schoenberg, and
more specifically, the pretense of what would become his radical view of dissonance
encompassed in his “emancipation of dissonance” philosophy.
46
Chapter 4
These ideas are: (1) The proof that instruction in musical composition
does enough if it is purely instruction in the handicraft without regard
for a natural system or for aesthetics. (2) The recognition of the merely
gradual distinction between consonance and dissonance. (3) The proof
that the three alleged laws of dissonance treatment–descend, ascend, or
1
Schoenberg, Art and Style., pgs. 102-146
2
Schoenberg, Theory of Harmony., pg. 323
3
Ibid.
47
sustain–were long ago overtaken by the hoary reality of a fourth law: skip
away from the dissonance. (4) The thesis: There are no such things as
non-harmonic tones, tones foreign to harmony, but just tones foreign to
the harmonic system.4
48
beauty to being about comprehensibility. That due to the prevalence of the first few
overtones, they are more audible, more familiar, and thus more comprehensible. The
opposite can be said about dissonances, according to Schoenberg. That the “sharply
contradictory terms” of concord and discord lies not in question of aesthetic pleasure,
but in the level of acquaintance, however intimate.8 He argues, then, that as physical
manifestations of these higher overtones have been introduced into tonal harmony,
a “closer acquaintance with the more remote consonances–the dissonances, that is–
gradually eliminated the difficulty of comprehension and finally admitted not only
the emancipation of dominant and other seventh chords, diminished sevenths and
augmented triads, but also the emancipation of Wagner’s, Strauss’, Moussorgsky’s,
Debussy’s, Mahler’s, Puccini’s, and Reger’s more remote dissonances.”9
The term emancipation of dissonance is thus referring to the comprehensibility of
dissonance, and by transitivity, the comprehensibility of consonance. To summarize,
Schoenberg makes the claim that as music has developed through history, western
culture has interjected dissonances of varying degree and frequency, increasing as
time has moved forward. His argument follows that as the physical dissonances were
introduced, their initially-hidden counterparts, or more remote consonances, within
the overtone series became more comprehensible and understood within harmony.
From their introduction, their repeated use led to a numbing of their “striking nature”
as musical history progressed. As chromaticism reached its apex, so too did this
pattern of introduction, and thus dissonance had reached a point of comprehensibility.
This is the emancipation of dissonance as it is understood today.
A couple of important related matters should be discussed here. Schoenberg is
approaching overtone theory from an equal temperament standpoint, and although
he makes the comment that context should, in theory, do very little to alter our
understanding of consonance and dissonance, he is quoted as saying, “Music depends
not only on acoustics but upon logic and upon those particular laws which result
from the combination of tone and tune.”10 Essentially, context does matter to some
8
Ibid.
9
Ibid., pgs. 104-5
10
Henry Pleasants. The Agony of Modern Music. New York: Simon and Schuster, 1955., pg. 99
49
degree in Schoenberg’s system. Finally, it is important to point out the linear nature
of music history assumed within his theory.
4.2.1 Musically
50
is the consequence of a “century of abuse” and spells the end of “serious music.” This
century is the rise of chromaticism, another of Schoenberg’s keystones. He goes on to
categorize dissonance in the 20th century as “robbed of its tonal properties of tension
and suspense; and left merely a tiresome ugliness.”13
Paul Hindemith claims that “harmony has become thoroughly known; no undis-
covered chord can be found,” moreover, if one were to suppose such harmony did
exist, this assumption “is extravagant, born of fancy, and defies reasonable investiga-
tion.”14 These views imply that harmony is complete, and as dissonances are a subset
of harmony, implies the awareness and comprehensibility of all possible dissonances.
As this is the hypothesis of the emancipation, Hindemith provides strong support
for Schoenberg’s claim, however indirectly. A plethora of supporting examples come
from other composers such as Charles Ives and Arthur Honegger, which should not
surprise the reader. As a testament to the prolific nature of his claim, it is still taught
as one of the major shifting points of music history in the 20th century and continues
to appeal to many due to its natural argument. As with most physical activities, the
more repetition one does, the more familiar the action becomes; it becomes mundane.
The same idea applies to dissonance, Schoenberg suggests, and for this reason it is
easy to agree with him based solely on the philosophy’s similarity to other situations
involving repeated exposure to an idea, physical sensation or activity.
13
Ibid., pg. 102
14
Paul Hindemith. A Composers World: Horizons and Limitations. Garden City, NY: Anchor
Books, Doubleday and Company, 1961., pg. 93
51
evoke.”15 It should be noted that the study of psychoacoustics is based on musical
structures removed from context, as not having them extrapolated could be problem-
atic.16 Perception of sound is studied in a two-step process, beginning with sensation,
which is what psychoacoustics is concerned with, and then cognition. In a 1987 pa-
per by DeWitt and Crowder, a distinction was made between rational music theories
based on acoustical principles and empirical theories based on context and learning.17
The umbrella that encompasses these two ideas of innate and acquired perception
is referred to as sensory.18 It is then understood by psychoacousticians that “mu-
sic may be regarded as a multi-layered structure of more or less familiar patterns,
and its perception seen as a multi-layered process of pattern recognition, based on
19
familiarity.“ This definition lies more in line with Arnold Schoenberg’s, mentioned
above, as opposed to a pure, music theory point of view; again, the idea that music
is something that evolves as the people around it do the same.
Another possible definition for the nature vs. nurture aspects of sound recognition
is that bioacoustical processes are what make up the sensory component, whereas the
cultural component is the aspects of perception associated with consistencies and
differences in the human environment.20 Prenatal conditioning is a highly studied
field, where certain phonically related associations are explored as well as the period
at which the auditory systems are developed within the womb in relation to the
sounds the fetus is experiencing.
52
from sensory hair cells in contact with a thin membrane (the basilar mem-
brane).21
Though this is the normal process most experience every day, there also lies the
possibility to experience sound by directly placing a vibrating object on the bones
of the head, which can then transfer the vibration to the cochlea.22 Low frequencies
can also be felt by nerve induces, such as low pipe organ pitches, lending itself to this
idea of internal frequencies.23
Looking into the Harvard article, the experiment was “designed specifically to test
the hypothesis of an innate preferential bias favoring consonance over dissonance.”24
This study, as opposed to the body of work related to infant reaction to music, focuses
not on single, successive instances of consonance and dissonance, but within a melodic
context. This article is a focus of this thesis because it satisfies Schoenberg’s required
context for understanding music. The results from the study suggested that by four
months, infants have a pyschoacoustical predisposition to consonance as opposed to
dissonance, as measured by visual and motor reactions to the consonant and dissonant
melodies.25 An important comment is made, however, that this correlation should
not be generalized to all consonance and dissonance, but only those used within the
study.26
In an article interviewing researchers about the study, psychologist Jerome Kagan
says, “I would bet this is not learned. It’s so early.”27 The implications of this are
two-fold: first, to be able to move beyond this predisposition would suggest that the
repeated exposure to helps to lessen the stigma innately associated with dissonances,
21
Neville H. Fletcher,, and Thomas D. Rossing. The Physics of Musical Instruments. New York:
Springer, 2010. Print., pg. 155.
22
Ibid., pg. 155
23
Ibid.
24
Marcel R. Zentner, and Jerome Kagan. “Infants Perception of Consonance and Dissonance
in Music.” Infant Behavior and Development 21, no. 3 (1998): 483-92. doi:10.1016/s0163-
6383(98)90021-2.
25
Ibid.
26
Ibid.
27
Richard A. Knox. “Babies Love Lullabies, Show a Real Aversion to Dissonance.” De-
seretNews.com. September 27, 1996. Accessed February 9, 2017.
https://www.deseretnews.com/article/515587/BABIES-LOVE-LULLABIES-SHOW- A-REAL-
AVERSION-TO-DISSONANCE.html
53
but perhaps contradictorily, this implies that there could exist a case where a disso-
nance is so striking that it could not be tolerated, and thus emancipated. In the case
of the former, there again suggests support for Schoenberg’s emancipation. Moreover,
Carol Krumhansl, professor at Cornell University, believes that “the explanation for
it has to do with the physics of sound - the overtone structure of harmonic tones.”28
This begs the question of when humans start to shift the aural preferences, and to
what degree it is possible to do so; however, that is a question for a later time.
A more complete, yet arguably dated, conversation on psychoacoustics and music
theory can be found in Hermann von Helmholtz’s On the Sensations of Tone as a
Physiological Basis for the Theory of Music. Briefly, Helmholtz believed that humans
were born with innate preferences towards particular dissonances and consonances,
and that some are born with more of a tolerance than others.29 While it falls in line
somewhat with the aforementioned study, Helmholtz seemed less open to the idea of
a shifting preference than the 1998 study.
4.2.3 Counterarguments
With this being said, as with most arguments, counterarguments are in abundance.
One such example comes from Pleasants. In the midst of supporting Schoenberg
indirectly, he proposes the argument that “the tonal framework they [Schoenberg
and his followers] so hopefully destroyed proved to have been the very substance of
their creative language.”30 That is, regardless of how tirelessly the composer tried
to move beyond a tonal center into this newly emancipated plane, one would always
find themselves falling back into the habits of tonality. Pleasants expresses this idea
more eloquently:
54
attempting to produce a consistent effect of tonality, but forced by fear
of imitation to adhere to the modern concept of dissonance, renders the
atonal effect of his failures doubly conspicuous.31
This implies that one can never truly escape tonal, harmonic functionality, and thus
cannot truly emancipate dissonance in its entirety. That, ever so often, a tonal
stimulant will find its way into the emancipated structure calling back to a time
meant to be forgotten.
This is not the only counterargumnet to Schoenberg’s theory, some of which we
have already seen within this thesis. In the following section, I aim to prove why I
believe Schoenberg’s justification is flawed as well as introduce what I believe to be
a more concrete justification for the emancipation. I invite the reader to evaluate
the presented information and formulate their own opinion on the authenticity of his
argument.
In chapter 2, I showed that the inconsistencies and the non-linearity of historical pro-
gression of temperament provided justification for the idea that dissonances, as well
as consonances, cannot be understood to be consistent from a historical standpoint;
that a culture of people, particularly those in western music, cannot have become
“desensitized” to dissonances in the way that has been described by Schoenberg’s
emancipation philosophy. That, again, it is not due to the gradual inclusion and
exploration of (dissonant) intervals, chords and tonal structures that has led to this
emancipation. This is a major point of contention in his argument. In a sense, this
is the entire pretense of his justification.
His philosophy is heavily based on the assumption that what is now considered
dissonant is the same structure as that which was considered dissonant 500 years ago,
which is simply untrue in most cases outside of the octave. Moreover, it relies on
31
Pleasants., pg. 109.
55
the assumption that entire generations of communities of people heard the same dis-
sonance in the same context in the same temperament, which a counterexample can
be provided immediately when considering Neidhardt’s 24 different temperaments or
the introduction of equally tempered organs when meantone-tempered organs were
still being used. Furthermore, in considering just intonation throughout the evolu-
tion of tuning, it becomes clear that, regardless of the mainstream temperament for
keyboard players, the dissonances that people were hearing were inconsistent based
on what instrumental ensembles they were listening to. This is just another example
of the inconsistent and nonlinear evolution of dissonances throughout music history.
Schoenberg’s argument begins to become inconsistent within the historical pretext it
so helplessly depends.
Another point of contention is non-Western tonalities that are fundamentally
based on dissonance. While one could make the argument that these are systems
in which the emancipation of dissonance has already, and completely, occurred, I ar-
gue that it supports the opposite. If one has a system defined by dissonance, can
an emancipation truly occur? The dissonances didn’t need to be emancipated, as
they existed in a system which began with dissonance as a focal point; the need for
emancipation is null, and so too is the effect of repeated exposure. So then, is the
emancipation based on the historical inclusion of dissonances, as Schoenberg suggests,
or is it related more so to the culture surrounding the tonality?
Similar to the infant study done by the Harvard psychologists, MIT professor Josh
McDermott performed a study in 2016 examining the perception of consonance and
dissonance of an Amazonian tribe, “the Tsimane’–a native Amazonian society with
minimal exposure to Western culture.”32 This cross-cultural exploration explored
three groups from three locals–United States, Bolivia, and the Tsimane’–with specific
preference given to this tribe due to “harmony, polyphony, and group performances
32
Josh H. McDermott, Schultz, A.F., Unduragga, E.A., Godoy, R.A. “Indifference to dissonance
in native Amazonians reveals cultural variation in music perception.” Nature, 535, 547550, 2016.
56
are by all accounts absent from their music.”33 That is not to say that they do
not have music. The lack of harmonic structure within their musical context raised
the question of “whether they would exhibit aesthetic responses to consonance and
dissonance despite not having prior exposure to them.”34 The results found that
while the tribe was able to distinguish between consonance and dissonance, as well as
harshness (periodic sound fluctuations) and other parameters, they had no preference
from one to the other, while the other two populations showed clear preference for
the more consonant examples.35
They concluded that:
The idea that there exists an innate indifference between consonance and dis-
sonance undermines the conclusions given by the 1998 infant study, and moreover,
implies that humans are influenced by the musical culture surrounding them, and thus
form musical biases, prior to four months old. Therefore, one could conclude that
the need for the emancipation of dissonance was created from the same system that
created its need. This implies a tautology and an inherent issue within the argument
itself if we follow Schoenberg’s historical progression justification. There are some
issues associated with this study, however, as this has only studied one non-Western
culture. There also lies issue in the fact that this study contradicts much of the major
33
Ibid.
34
Ibid.
35
Ibid.
36
Ibid.
57
conclusion of the infant study; surely, then, both cannot be true. Nevertheless, both
cases end in contradiction with Arnold Schoenberg’s evidence for his philosophy, and
therefore we will not worry about this too much. Thus, biologically, we have come to
a contradiction to his philosophy.
Thus, Arnold Schoenberg’s reasoning for his emancipation of dissonance philoso-
phy both historically (culturally) and biologically has been disproven. Though this
has been done, I still believe his argument is valid for a different reason, removing
culture as the main point of contention while still remaining within Western harmony.
Briefly I mention the literal rise of tuning pitch over the years. Today, instruments
in western culture are typically tuned to A440. This number has not been consistent
throughout history, however, and has been on a gradual rise since the implementation
of tuning systems in general. Along with this, it is not consistent across countries.
Viennese orchestras in the east typically tune their instruments to A444, but this
number has varied throughout history, lying somewhere between A415 and A466.
The fact that strings must then be tuned tighter (looser) could imply discrepancies
for intervallic relationships, but this is just a thought.
37
Note: Since we are looking at a Western philosophy, I will remain within the context of Western
music
58
Before we see this, some mathematical knowledge is required. If one is familiar with
summations, elementary integral calculus, and Fourier analysis, feel free to proceed
to chapter 6.
59
60
Chapter 5
We will now examine the overtone (harmonic) series from a physical standpoint as
opposed to the musical standpoint.
Historically, the first academic attention to overtones was given by Ren Descartes
and Marin Merseene, who noted that when a pitch is played, they could hear a
fifth and third, and other “little sounds.”1 Shortly after, Frenchman Joseph Sauveur
published his 1704 treatise On Harmonic Sounds, in which he “collected all that was
known at the time: that a musical note from any natural vibrator has overtones
in it; that for a string and an air column these overtones form a perfect harmonic
series; that bells and other resonant bodies have overtones also, but they don’t form
a harmonic series; that a string sounding one of its overtones vibrates with nodes and
antinodes.”2 Daniel Bernoulli proved that while the amplitude (height) of particular
harmonics approach 0 as one goes higher into the overtone series, the energy is never
fully zero and thus the overtone series is infinite in nature. Furthermore, he showed
that higher overtones are dissonant shortly after Robert Hooke’s work on simple
harmonic motion.3 The overtones were fully defined by Jean-Baptiste Joseph Fourier
1
Peter Pesic. Music and the Making of Modern Science. Cambridge: MIT Press, 2014., pg. 116
2
Johnston., pg. 90
3
Isacoff., 196
61
in 1768 with his introduction of Fourier Analysis, which we will explore soon.4
Physically, what is occurring is that given a fundamental pitch with a given
waveform–the most familiar example being a sine or cosine curve as shown in Figure
5-1–all of the waves that fit symmetrically inside the boundary of the main wave form
are overtones, as seen in Figure 5-2.
These are defined mathematically using the relation fn = nf1 , where f1 is the fun-
damental frequency and fn the nth overtone. Therefore, overtones are referred to
as integer multiples of the fundamental. Here, frequency is the number of complete
oscillations (cycles) for a given unit of time, often one second, given in units of Hertz
(Hz), or one cycle per one second. For A440, or 440 Hz, this means there are 440
complete oscillations within one second. The overtone series, then, is the infinite
series defined by f1 , f2 , f3 , f4 , ... This multiplicity allows us to prove Pythagoras’ and
Boethius’ intervals as well as derive others.
62
Figure 5-2: Overtones
63
Given f1 , we see
f1 = f1 := U nison ⇒ 1 : 1
f2 = 2f1 := 8va ⇒ 2 : 1
3f1
f3 = 3f1 := P 5 8va ⇒ = P5 ⇒ 3 : 2
2
5f1
f 5 = 5f1 := M 3 2(8va) ⇒ = M3 ⇒ 5 : 4
(2)(2)
Other intervals can be proved in a similar way; in terms of intervals not found in the
harmonic series, such as the perfect fourth, P 4, these can be found by multiplying by
its inversion and setting them equal to an octave. For the perfect fourth, this looks
like
3 4f1
2f1 = (f4 ) ⇒ = f4 ⇒ 4 : 3
2 3
As an aside, we see that Pythagoras’ ratio for a major third, M 3 := 81 : 64, is not
what is derived from the harmonic series as seen above.
fB = | f1 − f2 |
Where fB is the frequency of the beats and f1,2 the frequencies of the original pitches
measured in Hz.6 This is how White was able to formulate an exact methodology for
5
John S. Rigden. Physics and the Sound of Music. 2nd ed. New York: J. Wiley, 1985., pgs.
185-193
6
Ibid.
64
tuning pianos in equal temperament, as he was able to find what the frequency of the
beats should be for an equally tempered system.
Helmholtz found that the average human is able to distinguish beats in the range
1 ≤ fB <∼ 30
with fB ≥ 30 being distinguished as a new pitch, as opposed to the same pitch being
out of tune.7 Moreover, the human ear is capable of hearing frequency up to 12,000
Hz. This ability, as well as the acoustical nature of vibrating mechanisms, means that
if one can find the frequency of beats for fundamental pitches, it is possible to do a
similar process with their overtones.
The other physical principles inherently related to sound are the ideas of super-
position and interference. A familiar example is ripples in a pond. If two rocks are
dropped into a pond at different points, the ripples produced by these will eventually
meet and the waves will superimpose to form another ripple as a combination of the
two. Superposition, briefly, is when waves meet and interact with each other. More
specifically, “the principle of superposition states that, when two or more waves of the
same type cross at some point, the resultant displacement at that point is equal to
the sum of the displacements due to each individual wave.”8 If the peaks and valleys
of the waves align exactly, the superimposed wave, the resultant, is a wave that is
doubled in altitude (loudness). If the opposite occurs, the resultant is a wave that is
canceled out, the altitude goes to zero, and the sound immediately decays. This is
illustrated in Figure 5-3. The way that the waves interfere with each other is aptly
defined as interference.
The importance of these processes come into play when considering the inter-
actions between fundamentals, their overtones, and their interpretation within the
physiological mechanics that allow us to hear sound. An example of this are aural
harmonics, where the ear adds non-present harmonics due to the biomechanical mech-
7
Helmholtz., pg. 236
8
Isaac Physics
65
Figure 5-3: Constructive and Destructive Interference
A cornerstone to Fourier Analysis is the idea of the infinite sum, typically denoted by
Σ∞
x=0 f (x) = f (0) + f (1) + f (2) + · · ·
Where you take your function, f (x), and, in a sense, plug in each value between
the first value–here x = 0–to the last–here ∞–and add all of them together. This
9
Rigden., pgs. 185-193
10
B. H. Suits. Physics of Music-Notes. Accessed April 17, 2017.
https://pages.mtu.edu/suits/Physicsofmusic.html.
66
string of sums is referred to as a series, either finite or infinite. Obviously, one cannot
add to infinity; in this case, one tries to find a pattern and generalizes upon this to
get an idea of what the sum would look like if we were to somehow manage to add
indefinitely.
An example of a finite sum is
Σ5x=0 x = 0 + 1 + 2 + 3 + 4 + 5 = 15
Σ∞ 2 2 2 2
x=0 x = 0 + 1 + 2 + 3 + · · ·
or
1 1 1 1
Σ∞
x=1 2
= 2 + 2 + 2 +···
x 1 2 3
Along with this idea of series comes the theories of convergence and divergence.
Briefly, this means that either the sum eventually “settles out” around a value, and
thus converges like example 3, or the sum keeps getting bigger as more terms are
added with no settling in sight, and thus diverges like in example 2. This is all to
say that Fourier series have been proven to converge upon whatever function they are
attempting to impersonate, which allows us to say that our approximation is ‘good
enough.’ This idea will become more obvious in section 5.6.
67
5.5 The Calculus
Differential and Integral Calculus are the two halves which make up The Calcu-
d
lus. Differential Calculus, which utilizes the differential dx
, is concerned with rate
R
of change within a system while integral calculus, which uses the integral dx, is
concerned with area bounded by (underneath) a curve. It is intuitive to think of
an integral as an infinite sum of areas, in particular, the infinite sum of the area
of rectangles of smaller and smaller width underneath the curve who’s area is being
approximated, as shown by Figure 5-4.
Within this thesis, integral calculus serves the role of facilitator in finding the
coefficients of the Fourier series, which I will define soon. As such, I provide the
following integrals (and derivatives) which are necessary in order to understand the
proceeding section.
68
5.5.1 Cosine and Sine
Z
a
a cos(bx)dx = sin(bx) + C
b
Z
−a
a sin(bx)dx = cos(bx) + C
b
and the derivatives
d d
a cos(bx) = a cos(bx) = −ab sin(bx)
dx dx
d d
a sin(bx) = a sin(bx) = ab cos(bx)
dx dx
where a, b and C are arbitrary constants and x the independent variable that we
integrate (differentiate) with respect to.
To truly utilize these integrals, we need the fundamental theorem of Calculus. This
theorem comes in two parts; we are concerned with Part II.
Theorem: Fundamental Theorem of Calculus (Part II)
Let f be a continuous function on [a, b]. If F is an antiderivative of f on [a.b], then
Z b
f (x)dx = F (b) − F (a)
a
Here, continous losely means that the function has no gaps or pointy edges, [a, b]
refers to the boundary being examined, and the antiderivative is another word for
integral. Bear with me here, it’s not as complicated as it seems and I have given you
all the tools you need!
69
5.5.3 An Example
The most common example we will encounter in this thesis are integrals evaluating
sines and cosines over some interval. I provide the following to help illustrate these
ideas. Say we want to find the integral of cos(2x). Then using the equation for cosine,
we see
Z
1
cos(2x)dx = sin(2x) + C
2
And say we wanted to find the integral over [0, 2π]. Then using the fundamental
theorem, we see
Z 2π
1 1 1 1
cos(2x)dx = sin(2(2π) − sin(2(0)) = sin(4π) − sin(0) = 0
0 2 2 2 2
I have assumed some knowledge of trigonometry, but if this feels unfamiliar, the
following relations should prove to be helpful. In these, n is an integer Z–1, 2, 3, ...
nπ
0 n = 0 or n odd
cos( )=
2 1 n even
−1 n odd
cos(nπ) =
1 n even
and
nπ
−1 n = 3, 7, 11...
sin( )=
2 1
n = 1, 5, 9...
0 n odd
sin(nπ) =
0 n even
While the discussion on integrals and derivatives is not exhaustive, having this qual-
itative background provides some ease of mind for the coming sections and sparks
70
what I hope to be an interest in the subject.
Definition: Let f (x) be a periodic function from (−l, l). Then f (x) can be
represented by a Fourier Series of the form
a0 nπx nπx
f (x) = + Σ∞
n=1 [an cos( ) + bn sin( )]
2 l l
Where
1Z l
a0 = f (x)dx
l −l
1Z l nπx
an = f (x) cos( )dx
l −l l
71
1Z l nπx
bn = f (x) sin( )dx
l −l l
With the following properties:
2 Rl
A function is even if f (−x) = f (x); if f is even, then bn = 0 and an = l 0
f (x) cos( nπx
l
)dx
2 Rl
A function is odd if f (−x) = −f (x); if f is odd, then an = 0 and bn = l 0
f (x) sin( nπx
l
)dx
Using Fourier series, it is possible to decompose a periodic sound wave into its funda-
mental and harmonics (overtones)–the individual sine and cosine terms. The intensity
and prevalence of the harmonics is then found by squaring the Fourier coefficients of
those terms, though this is not exact as the ear is a non-uniform amplifier, as men-
tioned in 5.3. By Parseval’s Theorem, the total energy of a sound wave is equal to
the sum of the energies associated with the various harmonics. Consequently, the
intensity of a sound wave (average energy striking unit area of your ear per second) is
proportional to the average of the square of the excess pressure, or acoustic pressure,
which is the deviation from atmospheric pressure caused by a sound wave.Moreover,
the piano can be well understood through Fourier Analysis, and will be used fre-
quently in the following chapter.
72
Figure 5-6: A Physical Situation
keyboard of the piano. This will change calculation but not methodology. I omit
the derivations in interest of succinctness, but the reader should have the ability to
perform the integrations with the information given in 5.5.1-3 and 5.6.2. We can
describe f (x, 0) as
2h x
0<x< l
l 2
f (x, 0) =
2h (l − x)
l
<x<l
l 2
l
2Z l nπx 4h Z 2 nπx 4h Z l nπx 8h nπ
bn = f (x) sin( )dx = 2 x sin( )dx+ 2 l (l−x) sin( )dx = 2
sin( )
l 0 l l 0 l l 2 l (nm) 2
8h nπ 8h
sin( ) = ±
(nm)2 2 (nm)2
8h xπ 1 3xπ 1 5xπ
f (x, 0) = [sin( ) − sin( ) + sin( )−···
m2 l 32 l 52 l
8h
Where each (nm)2
sin( nπx
l
) represents a harmonic (overtone) produced by the piano
string, and it is in this way that we have deconstructed the complex sound wave
that emits from the vibrating string. It is important to note that when solving the
wave equation analytically for a plucked or struck string, the solution is a bit more
73
involved and complicated. However, the basic principle provided by our simplified
model produces a solution that differs only by a constant.
By definition, the Fourier Transform transforms a function of one variable, time (say),
which lies in the time domain to a second function which lies in the frequency domain
and changes the basis of the functions into sines and cosines.11 More intuitively,
Fourier Transforms are the integral analog of Fourier series. In particular, if we take
our interval (−l, l) and let l approach infinity, we arrive at what amounts to the
Fourier Transform.
Since these are typically applied to real-world problems, the logical development
of the transform began to focus on discreet intervals of time. From this we get the
Discreet Fourier Transform–or DFT–whose primary use is taking digital signals and
breaking them into their components. This should sound familiar, as this is exactly
what Fourier series tackle with a complex tone. From this, we can use the DFT to
analyze sound samples in terms of their overtones; however, the DFT is not efficient
in this process. For this reason, the Fast Fourier Transform, an algorithmic child of
the DFT, was developed by in. It is the Fast Fourier Transform, or FFT, that will be
used in the following chapter in performing the digital signal processing of overtones
with respect to the piano.
It should be noted that entire books are devoted to this subject. My aim is to
simply introduce the idea of what is happening behind the signal processing, and allow
the reader to explore the mathematics further. If one would like a more thorough
treatment of the Fourier Transform within this setting of music, math, and physics,
one can be found in Musimathics by Gareth Loy.
11
Haberman
74
5.9 Some Context Within PDEs
What follows is a more direct and mathematical treatment of the derivation of Fourier
series within the context of Partial Differential Equations (PDEs). This material is
most accessible to readers with a background in calculus; however, sections 5.9.1-2
attempt to provide a qualitative understanding of the principals used to allow for a
general grasp of 5.9.3. This is for those who really wish to understand the motivating
mathematics, and again, if ones interests are peaked, Gareth Loy’s book is a fantastic
introductory resource for the interdisciplinarian or Mary Boas’ Mathematical Methods
for the Physical Sciences for those wanting a bit more of a rigorous treatment.
Recall the idea of the derivative introduced in section 5.5, where finding the derivative
df (x)
of a function f (x), written as dx
, gives the rate of change of the function at a
particular time. This rate of change is maybe more familiar as the slope of a function,
with the normal example being about the slope of a line.
A partial derivative is an abstraction of this idea, where instead of finding the
rate of change in a singular direction, we find the change in as many directions as we
have in our multivaribale function f (x, y, ...), where we look at each one separately.
∂
These are denoted as the partial analog of the differential, ∂x
. Again, these work
similarly to the differentials for one variable. Moreover, when taking the derivative
of a multivariable function with respect to a particular variable, one ‘pretends’ that
all functions not of that variable are constants. This is illustrated below.
Consider the function f (x, y) = sin(x) cos(y). Say we want to find the partial
derivative with respect to x. Then we take the derivative of all of the functions of x,
in this case sin(x) just as before while treating cos(x) as a constant. Then we find
the partial derivative with respect to x to be
∂f (x, y) ∂ ∂ d
= [sin(x) cos(y)] = cos(y) sin(x) = cos(y) sin(x) = cos(y) cos(y)
∂x ∂x ∂x dx
75
where we let a = cos(y) so to treat it as a constant and the derivatives follow just as
they do in 5.5.1.
We can do a similar process with respect to y, in which we treat sin(x) as a
constant and get the following partial derivative
∂f (x, y) ∂ ∂ d
= [sin(x) cos(y)] = sin(x) cos(y) = sin(x) cos(y) = − sin(x) sin(y)
∂y ∂y ∂y dy
Double integrals work in the same way, where in we integrate with respect to a
particular variable and treat all functions of other variables as constants. Much like
the partial derivative, double integrals find the area under the curve in one direction
at a time, looking at specific variables and not the whole thing at one time. This is
simply an abstraction of the integration seen in 5.5, and is truly no more complicated
aside from added steps.
Suppose we want to take the double integral of our function above. Then we see
Z Z Z Z
f (x, y)dxdy = sin(x) cos(y)dxdy
So we take the integral of our fundtion first with respect to x, where we treat cos(y)
as a constant. This gives
Z Z Z
sin(x) cos(y)dxdy = − cos(x) cos(y)dy
Z
− cos(x) cos(y)dy = − cos(x) sin(y) + C
which gives the double integral, where essentially we performed the two integrals
presented in 5.5.1.
76
5.9.2 Differential Equations
Recall from high school algebra the idea of a polynomial, ax2 + bx + c = 0, where
one would the factor the polynomial (if possible) to find what x’s would satisfy the
equation. Those x’s are solutions to the algebraic equation. To help recall, consider
the polynomial x2 + 5x + 6 = 0. This can be factored into (x + 3)(x + 2) = 0 which
gives solutions x = −2, −3.
Differential equations can be considered as, perhaps unsurprisingly, an abstrac-
tion of this idea; however, instead of looking for numbers to satisfy x’s, one looks for
functions f (x, y, ...) to satisfy derivatives. They are a bit more complicated, though,
as there are certain restrictions, referred to as boundary conditions, that must be
satisfied. Since there are two types of derivatives–ordinary (single variable) and par-
tial (multivariable)–there are two classifications of differential equations–ordinary and
partial. Consequently, since ordinary derivatives are a special case of partial deriva-
tives, partial differential equations (PDEs) can sometimes be reduced to multiple
ordinary differential equations (ODEs). In particular, this reduction is the only way
to solve one of these equations analytically.
One of the simplest examples of a differential equation is
df (x)
= cos(x)
dx
So one asks what function has cos(x) as a derivative? From 5.5.1, we know this to
be sin(x) + C. Form here, one would use the boundary conditions to find constants
and other things like this. The more complicated the differential equation, the more
involved the solution becomes, but this example illustrates the basic ideas going on
behind differential equations.
These equations are used to represent physical situations, such as heat transfer or
electricity and magnetism, and for our purposes, acoustic waves are represented by
the wave equation, given in the following section.
77
5.9.3 The Wave Equation
Physically, in perhaps its most basic form, sound is an oscillatory process taking
place in relation to time and space. When looking at sound in reference to numerous
changing variables such as time and position in a uniform stagnant fluid, it is definable
by the wave equation,
1 ∂ 2p
∇2 p − =0
c2 ∂t2
∂2 ∂ 2 ∂2
where ∇2 = ∂x2
+ ∂y 2 + ∂z 2 . This is a second order partial differential equation.
12
Un-
der specific boundary conditions, this will simplify to an ordinary differential equation
that can be solved, producing the solution,
p = p0 sin(ωt ± kx)
Where ω is the angular frequency and k is the spring constant, which describes the
behavior of an acoustic wave through matter.13 In reference to something like the
vibrations of the ear drum in response to a given frequency, it is important to note
that the wave equation will also serve to model the motion of others things, such as a
vibrating string or membrane, where p changes from pressure to displacement, y, and
c from the speed of sound in a given substance to ν for wave velocity.14 This gives,
1 ∂ 2y
∇2 y − =0
ν 2 ∂t2
This differential equation can be modified to define the certain timbres and tones of
instruments, and in general results in a solution in the form of a Fourier series.15 These
were derived by Joseph Fourier in order to solve another PDE, the heat equation. As
12
Haberman., pg. 70.
13
Ibid., pg. 80
14
Boas, Mary L. Mathematical Methods in the Physical Sciences 2nd ed.: Ill.: Bib.: Wiley, 1983.,
pg. 554.
15
Ibid., pgs. 555-6.
78
has already been shown, these series are especially useful in mathematically defining
a wave that is not necessarily sinusoidal, such as a triangle waves or saw tooth waves,
which are more common wave forms for musical instruments.16
These series are mathematical structures analogous to the naturally occurring
overtone series, or harmonic series–hence the naming of Harmonic Analysis of which
Fourier Analysis is a subdiscipline. I do not need to emphasize how indispensable
the series are when applying Fourier Analysis to determine the effects of overtones
in dissonance and dissonance relations. They also help to deconstruct frequencies,
allowing for better understanding of the overtones in play as well as the superposition.
From here, one can use this idea of series to make an analogy of chords, since a Fourier
series is a combination of sine waves at different frequencies, which is essentially what
a chord is. This is exactly what will be done in the following chapter.
16
Ibid., pg. 300.
79
80
Chapter 6
6.1 Methodology
6.1.1 Theoretical
Using the theoretical frequencies for given temperaments, I will compare pitches
within overtone series of common tonal structures–pitch classes, intervals, chords,
etc.–in both the singular case, as has been thoroughly researched by many including
Arnold Schoenberg, but then extended beyond the singular case to encompass more
of the musical structures typically encountered while listening to a piece of music. In
particular, I will be comparing particular intervals that appear within the systems
of overtone series as well as the relationship between common pitches between them.
A complete list of the frequencies being used for each temperament can be found
within the Appendix. These comparisons will include observations such as frequency
of beats, functionality, and periodicity of a particular kind of occurrence.
6.1.2 Experimental
These ideas will then be analyzed experimentally through digital signal processing
using the FFT. This will be accomplished by studying recordings of the Memorial
Chapel’s Steinway “D” piano through the sound analyzing software Audacity. When
analyzing the sound samples, I will again be looking at the frequencies of particular
81
pitch classes within a given overtone series, their prevalence (intensity) by Fourier
coefficients, their relation with common pitches within the system, and the intervallic
relationships between particular overtones. In short, this is an experimental version
of the theoretical analysis, and as such, I will compare and contrast the theoretical
implications against the observed phenomena. As it is exceedingly difficult to find an
instrument not tuned in equal temperament, this is the only temperament examined
experimentally. Based on these results and the theoretical differences observed, one
should be capable of extrapolating these conclusions onto the other temperaments
not included here.
In most things, the theoretical models do not mirror reality in an exact sense. The
same can be said about overtones, in particular, the overtones emitting from pitches
produced by a piano. As mentioned in Chapter 2, equal temperament places every
note equidistant from every other, and as such, should follow a consistent progression
moving up and down the range of the instrument; however, this is not the case in real
time. There exist particular consequences due to the nature of the piano’s construc-
tion. More specifically, these include topics such as inharmonicity and the Railsback
Curve, both of which profoundly alter the theoretical model of equal temperament
presented earlier. Before I begin showing the relations between overtones, we will take
a brief interlude through these inconsistencies in order to develop a more complete
picture of the context surrounding my theory.
6.2.2 Inharmonicity
82
piano, the inharmonicity causes the partials to deviate towards higher frequencies.”1
Richard Feynman was the first to note some of the root causes of these effects in his
1961 letter on his newly developed “Stiff-string” theory.2 In this, he determined the
true frequency of a fundamental for a non-copper-wrapped string is defined by
B
f1 = f10 (1 + )
2
πEA2 µ(f10 )2
Where f10 is the ideal frequency of the string ignoring stiffness and B = T2
with T being tension, A the area of the sting cross section, E as Young’s modulus of
steel (the stiffness of the wire), and µ the weight of the wire.3 This was later used to
derive
fn ≈ f1 (1 + αn2 )
Which gives the ‘true’ value for the harmonic considering the stiffness of the string
where α is proportional to Young’s modulus from above.4 Feynman also notes the
effects of the soundboard, the interrelations between the three strings used to produce
a single note, and the material of the strings as other causes of potential changes in
frequency. However, he did not realize, at least in these papers, that these explicitly
resulted in inharmonicity.
When one studies simple harmonic motions, or vibrations of strings, the stiffness
of strings is typically ignored in simplified situations, such as the one we considered
in chapter 5, due to its quasi-negligible effect on the result; however, this stiffness
is exactly the cause of the inharmonicity within the piano. So then “the enormous
tension to which the strings are subject makes them extremely rigid,” where stiffness
affects the string by acting as a dispersive medium.5 This means that the sound
1
Burred, Juan Jos. The Acoustics of the Piano. Arturo Soria,
Madrid: Professional Conservatory of Music, 2004. Accessed March 3, 2017.
https://pdfs.semanticscholar.org/ba44/07fd779bb3541b56ff56a056aed5fada163e.pdf.
2
Richard Feynman, and John C. Bryner. Stiff-string Theory: Richard Feynman on Piano Tuning.
Physics Today, December 2009. Accessed January 12, 2018. doi:10.18411/a-2017-023.
3
Ibid.
4
N. Giordano. Explaining the Railsback Stretch in Terms of the Inharmonicity of Piano Tones and
Sensory Dissonance. The Journal of the Acoustical Society of America 138, no. 4 (2015): 2359-366.
doi:10.1121/1.4931439.
5
Burred., pg. 22
83
waves propagate at different velocities that are dependent on frequency. A higher
frequency causes faster propagation. Therefore, in the case of a complex musical
sound, the higher partials propagate faster than the fundamental and lower partials.”6
This effect can be seen in Figure 6-1, where it is clear that as one examines higher
overtones, the more divergent it becomes from the ideal harmonic series. Moreover,
the effect of inharmonicity is not consistent throughout the instrument. If one were
to look at the middle range of the instrument, the effects are almost unnoticeable.
This also plays into the fact that the intensities of particular overtones over others is
not consistent throughout. In general:
84
Likewise, B increases in the treble because the short length and high
tension cause the strings to act more like bars.7
So how then is this issue of inharmonicity resolved within the piano? The answer
lies with something referred to as the Railsback curve (stretch). Shown in Figure
6-2, the experimental curve was first derived by O. L. Railsback in 1938.8 This
was studied in more detail by Schuck and Young in 1943, who were the first to
7
Ibid., pgs. 23-24
8
Giordano
85
measure spectral inharmonicity. The curve essentially shows the deviation of standard
piano tuning from its theoretical counterpart within equal temperament. This type
of tuning is referred to as having a “stretched octave,” where the pitches in the
bass are tuned slightly flatter than ideal and pitches in the treble slightly sharper.9
Perhaps unsurprisingly, this deviation is inextricably linked to inharmonicity, and
in particular, the Railsback curve is essentially the integral of the inharmonicity at
an octave.10 More fundamentally, the Railsback curve is a direct consequence of
inharmonicity and is the piano tuners attempt at rectifying this intrinsic property.
Thus, in general, a piano is never truly tuned in equal temperament, though it comes
quite close, and just as with inharmonicity, the effect is less noticeable within larger
pianos.
It is also worth mentioning that this process is largely based on the instrument
being tuned and the person tuning the instrument. It would be nice to assume that
each piano tuner tunes every piano in exactly the same way; and conversely, it would
be nice to assume that each piano being tuned behaves in the same way for each
tuner. This is simply not the case, and while the results are similar in character, it is
a variant parameter within this thesis and worthy of mention before moving forward.
One may ask then, why do piano makers and tuners continue to deal with what seems
to be a flawed system instead of re-engineering the piano to solve them? While no
direct answer can be given, one suggestions lies in that some seem to prefer the slight
inharmonicity to a piano tuned to the exact proportions laid out by ET.11 That the
piano seems “duller” when tuned this way. Another idea lies in the fact that tuning
is not fixed; that is, it shifts over time. Some speak about pianos “settling” over the
days following a fresh tuning, where the tuning moves into a well-balanced system.
This results from the strings stretching ever so slightly, and, depending on the age
9
Ibid.
10
Ibid.
11
Burred., pgs. 24-25
86
and tenacity of the particular strings, varies to some degree. Nevertheless, this is
essentially the instrument and system Arnold Schoenberg was familiar with when
writing his statement on emancipation, and as such, will be what we study now.
87
6.3 Overtones in the Singular Case
6.3.1 An Example
We will first consider the theoretical case. Suppose a single pitch, say C3 , played on
a piano. Theoretically, we assume a sequence of notes, just as in figure 1-1, which
follows
C3 , C4 , G4 , C5 , E5 , G5 , · · ·
Comparing these first few values to those predicted by ET, we find that the
.01 ≤ fB ≤ 5.20
With the smallest discrepancy coming from the first overtone (octave), f2 , and the
greatest from the fifth overtone (major third), f5 . These small discrepancies are calm
12
*I make note of the fact that I have assumed an ideal string, rather than considering stiffness
as with Feynman. I do this for multiple reasons: We are in the lower middle octave of the piano,
which is the least affected by inharmonicty and as such the effect is almost negligible, as well as the
difficulty and time it would take to find the information to satisfy his equations. While an important
consequence exists for lower (upper) octaves, I do not consider it here. This may be examined in
further study of this topic when I have the necessary tools.
88
examples of inharmonicity! Something to consider here is the prevalence and intensity
of each of these overtones. As mentioned in Chapter 5, not every overtone is created
equal, where some are distinctly audible while others may be lost in the background.
This is something that is easier to see in our physical example.
Now consider the experimental case. Through Audacity, I performed a digital
signal analysis on C3 as produced by the Steinway “D” in the Memorial Chapel. This
process involved using the FFT to decompose the complex signal into its character-
istic overtones (harmonics), which I could then get the frequency of each particular
overtone. The resultant graph can be seen in Figure 6-3. Here, each overtone is
represented by a distinct color, showing the composite nature of a complex tone.
We expect the same overtone series as with the theoretical case, so I will go on to the
frequencies derived. Experimentally, the frequencies of the overtone series for C3 are
89
Again comparing these to the theoretical values of ET, we find
1.63 ≤ fB ≤ 9.25
Consistent from the theoretical sample, we find the slightest discrepancy between the
octave and the most between the major third. Again, there exists this example of
inharmonicity. Moreover, now we are able to see that the overpowering overtones are
E5 and G4 in this example. This implies that for C2 , this highly flattened major third
is heavily pronounced.
It should be of no surprise that the octave is the smallest discrepancy and the third
the most, as the octave is the least affected and the major third–and their inversion
the minor sixth–the most by equal temperament. Along with this, though the major
third was the most prevalent overtone in this example, this is not always the case;
furthermore, this intensity seemed to vary widely based on the pitch being played.
While this is only one example of the 88 notes on a keyboard, this pattern of the octave
being the least affected and major third the most (within the first 5 overtones) is
consistent throughout. Moreover, if one continues in the overtone series, it becomes
clear that the phenomenon is riddled with intervallic dissonances beyond this fifth
overtone. These include major and minor seconds, augmented fifths, and more.
This effect is precisely what Arnold Schoenberg was referring to when he made
his comment that if one goes far enough, everything is dissonant. The caveat to
this, then, is that it can be argued that these more distant harmonics are not easily
registerable by the cochlea, at least in a distinct sense, and thus it is more of a colored
sensation, a roughness, as opposed to distinct dissonances. The sensible question to
ask then becomes, “What does this imply when two pitches are played at the same
time?”
90
6.3.3 Implications for Cases of n ≥ 2
Based on what has been seen from 6.3.2, each overtone series produced by a particular
note is inherently dissonant as one progresses higher through the series, and moreover,
the intervallic relationships between pitch classes within the series are not necessarily
equivalent to their ideal ET counterparts. If this is the case, what happens if the
fundamentals are tuned to an interval according to ET? Furthermore, since overtones
do not follow the ratiomatic description of ET, are relationships preserved among
common tones within overtone series for combinations of two or more notes? Most
basically, is a unison still a unison? This abstraction has received little research, the
only other examples being a side research project by Richard Feynman in the 1960s
and in Spectral Analysis of Musical with Emphasis on the Piano (2015) by David M.
Koenig. In Feynman’s letter, however, though he defined the ‘true’ frequency, he
neglected the effects of inharmonicity when he explored this abstraction.13 Moreover,
Koenig notes that these interactions exist, but does not explore their implications.14
Though our studies were developed independently, we have some common overlap.
In particular, some of the conclusions I draw in the following section are consistent
with theirs, and are thus supported and verified by an outside, unbiased examiner.
Consider now the case of two pitch classes played together. This, as defined in Chapter
1, is an interval, with examples such as seconds, thirds, fourths and fifths. Proceeding
in a similar fashion as in 6.3, we first consider the theoretical case.
Consider two pitch classes, say A2 and C3 , which produce a minor third. Theo-
retically, the computed overtone frequencies are
Taking note of the two overtone series in conjunction, there are a few things to note.
13
Feynman
14
Koenig
91
Overtone Series for A2 & C3
A2 C3
A2 = f1 = 110 C3 = f1 = 130.81
A3 = f2 = 220 C4 = f2 = 261.62
E4 = f3 = 330 G4 = f3 = 392.43
A4 = f4 = 440 C5 = f4 = 523.24
C]5 = f5 = 550 E5 = f5 = 654.05
E5 = f6 = 660 G5 = f6 = 784.86
Firstly, the most noticeable are the common tones between the two series, in this
case E5 , denoted in purple. Since these are the same pitch, we can calculate the fre-
quency of the beats between them. This values gives fB = 5.05. This is well within
the interval of distinction and would thus be considered out of tune. Secondly, there
is an abundance of major and minor seconds present, denoted in green. These are
dissonances Schoenberg notes as being higher in the overtone series for a particular
pitch, but are not audible due to their placement and intensity. As is clearly seen,
however, these dissonances appear in what is otherwise considered a consonant inter-
val, relatively close to the fundamental, and would theoretically be audible. There
also exists a major third against a minor third, as well as two perfect fifths that are
slightly different from one another. Again, these are only the first five overtones, but
consider the interactions of higher overtones. The abundance of dissonances is more
than doubled from the singular example.
Considering the physical example, we see,
The same patterns exist to an equivalent degree. In the physical example, we find
fB = 5.00 for E5 . Likewise, the same patterns of dissonance emerge, but to slightly
92
varying degree as the frequencies shift slightly. For example, C5 and C#5 forms a
minor second. In the theoretical example, they are 26.76 Hz apart, where in the
physical example they are only 25 Hz. This implies that they are ever so much
closer, and the dissonance would be proportionally ever so much more grating to the
ear. Though the amount of dissonances has not increased from the theoretical to the
physical, the degree at which they exist has. Moreover, just as we saw with the single
case, certain overtones are more present than others, and in this specific example,
they happen to be exactly those that create the most dissonant relations.
Even though this is only an example, due to the inclusive nature of overtones,
these relationships can be found in almost all of the intervals found in Western tonal
harmony. I invite the reader to try the this on their own. Let us continue the
abstraction.
We now move to the triad, or a chord with three distinct pitches, often a root and a
major (minor) third and perfect fifth above it. As we have been building up the A
minor triad, we will continue in this way. Consider, then, A2 , C3 , and E3 . Theoreti-
cally, we have
From here, one sees that there exist all of the dissonances from the interval, but now
in conjunction with the newly added overtone series. There now exists another set
of common tones, E4, and more dissonant intervals. In particular, there now exits a
tone cluster between G, G#, and A, where a tone cluster is typically defined as the
simultaneous playing of three or more consecutive chromatic pitches, i.e. they are
93
right next to each other.
Now Considering the physical example, the frequencies are
Here we see an example of the Railsback curve and the stretched octave coming into
play. As the number of E5’s within this system of overtone superposition increases,
the more audible their interactions become. The fact that the E’s have become closer
in frequency reflects the tuner’s attempt to minimize the number of beats occurring.
While still out of tune to some degree, it is lesser than the theoretical example. Other
instances, such as dissonances becoming wider or narrower are still present as well.
Perhaps a pattern is beginning to show to the reader. This is a ruthlessly additive
cycle where the plethora of dissonances is growing exponentially to the amount of
pitches added.
Let us consider one more case, for completeness if nothing else. Given the minor
seventh chord A2 , C3 , E3 , and G3 , we see
The chart is now almost entirely consumed by common tones that are slightly out
94
of tune with one another, their common intervals that follow suit, tone clusters,
and almost every dissonant interval known to traditional Western Harmony. The
abstraction, and thus inclusion of natural dissonance, doesn’t end with the seventh.
At the height of classical tonality, more and more intervals were being added to a
chord to extend them. This started with the seventh chord and moved on to ninths,
thirteenths, and on as far as the piano could manage. Parallel to this extension
involved pitches added within these backbone structures–such as Eric Whitaker’s
famous major second inclusion in the otherwise traditional major triad.
6.4.4 Conclusions
There are a few conclusions to note. First, the number of dissonances present is
exponentially related to the number of fundamental pitches. Secondly, beginning
with the seventh chord, over 80% of the system of overtones is dissonant in some
way; this percentage only encompasses up to the fifth overtone and rises as more
are added. Thirdly, all dissonant intervals as defined by Schoenberg in Chapter 3
are present within the first five overtones for a seventh chord, with the number of
intervals decreasing as the number of pitch classes becomes less. In short, those who
partake of Western music played within an equal temperament context experience it
through a cacophony of dissonance. Truly there is only one conclusion: the overtones
are out of tune.
95
the nature of the overtone series persist. For continuity, I provide the following brief
examples demonstrating my claim: Bach’s well temperament and an example of just
intonation.
Consider again the seventh chord A4 , C5 , E5 , and G5 . Using the relationships pro-
posed by Dr. Bradley Lehman and A415 as our (baroque) tuning pitch, we reconstruct
this chord as follows.15
By the natural laws governing overtones, the particular set of pitches involved will
always be the same. What will change is the degree at which particular dissonances
clash as well as the relationship between common tones and common intervals within
the series. This is clear from the preceding chart illustrating Bach’s temperament. In
this particular example, fB varies widely between pairs of common tones. For exam-
ple, when considering E7, fB = 1.32, 21.1, 22.4 depending on the pair being compared.
We see here all that we saw when looking in our equal temperament context. More-
over, this increased disparity between common tones (and intervals) is reflective of
the tonal palette, or the sensibilities spoken about in Chapter 2.16 It should be noted
that as Bach’s tuning was irregular, the exact proportion between each note is any-
thing but. Therefore, the given frequencies are approximations within the boundary
provided by Lehman.
15
Duffin., pg. 163
16
D. Gareth Loy. Musimathics: The Mathematical Foundations of Music. Cambridge, MA: MIT
Press, 2011., pg. 69
96
6.5.2 Just Intonation Revisited
Suppose now we are tuning this chord, A4 , C5 , E5 , and G5 , in the case of a string
orchestra, as it would be here that just intonation would play a pivotal role. As such,
we take A415 as our (baroque) tuning pitch, we reconstruct this chord as follows.17
Again, the pitches have not changed. It is only the degree of dissonance that has
changed. However, now the common tones are perfect unisons. Therefore, the small
portion of dissonances cause by this has been removed. Nevertheless, the dissonance
remains, forever transcending the notion of temperament. Thus it is shown that
regardless of temperament, the overtonic system is consistently dissonant, to varying
degree, and for this reason, the conclusion presented in 6.4 holds for all temperaments.
97
dance to the temperament being used, and thus their distinction as pointed out in
Chapters 2 and 4, they were omnipresent in some form.18 This repeated, continuous,
long term exposure, then, and not the gradual exposure and inclusion of particular
intervals and dissonances presented by Schoenberg, is what I believe, and have shown
to be, the justification for his Emancipation of Dissonance philosophy.
18
McDermott
98
Chapter 7
Schoenberg was right about the freedom of dissonance, but not for the right rea-
sons. It has been shown that Arnold Schoenbergs argument for the emancipation of
dissonance is inconsistent and false due to the non-linear nature of the evolution of
temperament, the inconsistency of temperament in terms of place and composer, and
the psychoacoustical nature of the human hearing mechanism. It has also been shown
that given any interval or extension, there exists a plethora of dissonances inherently
present within the system that our ears are sensitive enough to detect. These inherent
dissonances are ever present and apparent.
This would imply that those in Western tonality were always aware of the intervals
that have come about through history, such as the major third or sixth, but it took
our culture some time to deconstruct the complex signals in our brains and use them
in fundamental notes. Whether this be subconscious or the result of hearing overtones
in a grand space and attempting to mimic them, their first physical representation was
not the first instance in which they were registered and analyzed by the human ear.
This suggests that the concepts of harmonic consonance and dissonance has simply
evolved in a physical sense, as in, it evolved as overtonic relations became common
within fundamentals, and not in a spontaneous, guessing game methodology that is
implied by music historians. It has also been shown that much of Western culture’s
opinion on what dissonance is stems from the tuning system experienced, and further-
more, that the dissonances themselves are only changed by degree, and not nature,
99
in regards to temperament. Temperament may change how grating (or pleasant) a
particular dissonance may seem, but it does not affect its inherent presence.
All of this would imply that it cannot be the case that our physical perception of
dissonance has evolved over time or that it was this historical inclusivity of dissonance
which finally led to its emancipation. This is due to the fact that it has always been
there in one form or another, and it is only the manner through which it was expressed
that has changed. And again, temperament was unstandardized on even the personal
compositional and performance level. This means music didnt sound the same to
everyone, everywhere. That music couldnt have followed this linear, step by step
evolution that is portrayed. It follows that it is unlikely that this choppy, non-linear
evolution is directly responsible for Schoenbergs Emancipation of Dissonance.
It is possible, as with any theory, to have objections. One may be the mere
attempt to categorize dissonance quantitatively. This is claimed to be impossible by
both Gareth Loy and Ian Johnston. Moreover, the idea that I have removed these
intervals from their context could be another point of contention; however, I argue
that the context enhances the topics discussed here, and adds to rather than hindering
the argument. Moving forward, the next logical step is to use this analysis for tuning
systems in other cultures and see if common themes arise. From there, seeing how
those themes relate to the acceptance and forward use of dissonance, as has been
shown here for Western culture.
Thus, Schoenberg was right that we can let dissonance go unresolved, but because
of the inherent relationships between overtones, tonal structures, and tuning systems
that we as a culture have always experienced, more so than the historical progression
of music and the culture surrounding it as suggested. In short, we have always known
dissonance, and it is in this inherently cognizant understanding that Schoenbergs
philosophy took root. This understanding that in the most dissonant of surround-
ings, consonance perseveres, and concordance is found within a discordant context.
Music is still beautiful, powerful, and thought provoking, and in no way have I aimed
to deny this. It is my hope that I have simply stretched the understanding of the
musical context we find so familiar in order to achieve a deeper understanding of the
100
art form so precious to myself and others. To understand one of the most important
philosophical explorations of the twentieth century; one that is fundamentally rooted
in the fact that just because something is dissonant does not mean it is not beautiful.
This analogy extends far beyond this thesis, and with that, dear reader, I leave
you to your thoughts.
101
102
Appendix A
From these intervals one is able to replicate the studies performed in Chapter 6.
In order to use the cents, one can follow the following procedure:
Given a fundamental frequency f ,
Chose y = value of interval in cents
y
Find x by x = 1200
103
104
Bibliography
Boas, Mary L. Mathematical Methods in the Physical Sciences 2nd ed.: Ill.: Bib.:
Wiley, 1983
Bonds, Mark Evan. A History of Music in Western Culture. Upper Saddle River,
NJ: Prentice Hall, 2006.
Burred, Juan Jos. The Acoustics of the Piano. Arturo Soria, Madrid: Professional
Conservatory of Music, 2004. Accessed March 3, 2017.
https://pdfs.semanticscholar.org/ba44/07fd779bb3541b56ff56a056aed5fada163e.pdf.
Duffin, Ross W. How Equal Temperament Ruined Harmony: and Why You Should
Care. New York: W.W. Norton, 2008.
105
Ellis, Alexander John. On Musical Duodenes, or the Theory of Constructing
Instruments with Fixed Tones in Just or Practically Just Intonation. London:
Taylor and Francis, 1874.
Helmholtz, Hermann Von, and Alexander John Ellis. On the Sensations of Tone
as a Physiological Basis for the Theory of Music. New York: Cambridge
University Press, 1954.
Hoppin, Richard H. Medieval Music. New York: W.W. Norton, 1978. Print
106
Isacoff, Stuart. Temperament: How Music Became a Battleground for the Great
Minds of Western Civilisation. London: Faber and Faber, 2007
Johnston, Ian D. Measured Tones: The Interplay of Physics and Music. Bristol,:
IOP Publishing, 1989.
Koenig, David M., and Delwin D. Fandrich. Spectral Analysis of Musical Sounds
with Emphasis on the Piano. Oxford: Oxford University Press, 2015.
Kostka, Stefan M., Dorothy Payne, and Allan Schindler. Tonal Harmony: With
an Introduction to Twentieth century Music. 7th ed. McGraw-Hill, 2013.
107
Marenbon, John. The Cambridge Companion to Boethius. Cambridge, UK:
Cambridge University Press. 2009.
McDermott, Josh H., Schultz, A.F., Unduragga, E.A., Godoy, R.A. “Indifference
to dissonance in native Amazonians reveals cultural variation in music
perception.” Nature, 535, 547550, 2016.
Pleasants, Henry. The Agony of Modern Music. New York: Simon and Schuster,
1955.
108
Rigden, John S. Physics and the Sound of Music. 2nd ed. New York: J. Wiley,
1985.
Schindler, Anton Felix. Biographie von Ludvig van Beethoven (Münster, 1860)
trans. after Constance S. Jolly, ed. D.W. MacArdle, Beethoven as I Knew Him,
Chapel Hill, N.C., 1966.
Tolstov, Georgi P., and Richard A. Silverman. Fourier Series. Englewood Cliffs:
Prentice-Hall, 1962.
Weiss, Piero and Richard Taruskin. Music in the Western World: A History in
Documents. Australia: Thomson/Schirmer, 2008.
Zentner, Marcel R., and Jerome Kagan. “Infants Perception of Consonance and
Dissonance in Music.” Infant Behavior and Development 21, no. 3 (1998): 483-92.
doi:10.1016/s0163-6383(98)90021-2.
109