GESTALT PRINCIPLES IN LIGETI’S PIANO ETUDE “DÉSORDRE”

by

David William Batchelor

Bachelor of Music
University of Houston, 2006

Master of Music
Brigham Young University, 2010

Submitted in Partial Fulfillment of the Requirements

For the Degree of Doctor of Musical Arts in

Composition

School of Music

University of South Carolina

2015

Accepted by:

Jesse Jones, Major Professor

Reginald Bain, Committee Member

John Fitz Rogers, Committee Member

Fang Man, Committee Member

Lacy Ford, Senior Vice Provost and Dean of Graduate Studies

© Copyright by David William Batchelor, 2015
All Rights Reserved.

ii
 DEDICATION
To Kristin, with all my heart. We did it.

iii
 ACKNOWLEDGEMENTS
To God, who was with me throughout the creation of this document; to my wife,

who kept our life running smoothly and loved me with great patience; to Dr. Reginald

Bain, for his help in the refinement of this document; to Dr. J. Daniel Jenkins, for

introducing me to Gestalt theory; to the other members of my dissertation committee, Dr.

John Fitz Rogers, Dr. Jesse Jones, and Dr. Fang Man, for their support and time; to

Schott Music International, for granting permission to use excerpts from the score in this

document; to the University of South Carolina, for a great education.

iv
 ABSTRACT
This document is a study of how Gestalt principles of organization are at work in

“Désordre” (1985), the first etude in the first book of piano etudes by György Ligeti

(1923–2006). After explaining how Gestalt principles can be applied to the analysis of

music, the study presents an analysis of the etude in four main parts. The first part

identifies elements of the composition that help the listener define boundaries between

phrases, phrase groups, and sections. The second part discusses how foreground and

background layers are articulated. The third part discusses the polytempo illusion. The

fourth and final part identifies elements that contribute to large-scale unity in the

composition. Finally, some pedagogical applications for teaching composition are briefly

addressed.

v
 TABLE OF CONTENTS
DEDICATION ....................................................................................................................... iii

ACKNOWLEDGEMENTS........................................................................................................ iv

ABSTRACT ............................................................................................................................v

LIST OF FIGURES ............................................................................................................... viii

LIST OF EXAMPLES ............................................................................................................. ix

CHAPTER 1: INTRODUCTION ..................................................................................................1

CHAPTER 2: GESTALT PRINCIPLES OF PERCEPTION ...............................................................4

2.1 HISTORICAL BACKGROUND ................................................................................4

2.2 PROXIMITY .........................................................................................................7

2.3 PROXIMITY IN MUSIC ..........................................................................................8

2.4 SIMILARITY .........................................................................................................9

2.5 SIMILARITY IN MUSIC .......................................................................................10

2.6 SUBLIMINAL PERCEPTIONS AND LARGE-SCALE GROUPING ...............................12

2.7 FIGURE/GROUND ARTICULATION .....................................................................14

2.8 FIGURE/GROUND ARTICULATION IN MUSIC ......................................................15

2.9 COMMON REGION .............................................................................................17

2.10 COMMON REGION IN MUSIC ...........................................................................18

2.11 CONTINUITY ...................................................................................................19

2.12 CONTINUITY IN MUSIC ....................................................................................19

vi
CHAPTER 3: A GESTALT-BASED ANALYSIS OF “DÉSORDRE” ..............................................21

3.1 FORMAL OVERVIEW..........................................................................................21

3.2 PHRASES, PHRASE GROUPS, AND SECTIONS ......................................................25

3.3 ARTICULATION OF FOREGROUND AND BACKGROUND LAYERS .........................32

3.4 PERCEPTION AND THE POLYTEMPO ILLUSION ...................................................34

3.5 PERCEPTION OF LARGE-SCALE UNITY...............................................................42

3.6 PEDAGOGICAL APPLICATIONS IN THE COMPOSITION STUDIO ............................44

BIBLIOGRAPHY ...................................................................................................................48

APPENDIX A: GESTALT PRINCIPLES OF PERCEPTION SUMMARY .........................................51

vii
 LIST OF FIGURES
Figure 2.1 Preliminary example of perceptual grouping .....................................................6

Figure 2.2 Grouping by proximity .......................................................................................7

Figure 2.3 Perception of columns because of proximity .....................................................8

Figure 2.4 Comparison of resultant groupings from proximity and similarity ....................9

Figure 2.5 Development of motivic complexity ................................................................11

Figure 2.6 Logo demonstrating subliminal grouping ........................................................12

Figure 2.7 Grouping by figure/ground articulation............................................................15

Figure 2.8 Grouping by common region against proximity ..............................................18

Figure 2.9 Grouping by continuity.....................................................................................19

Figure 3.1 Visual representation of the form of “Désordre” .............................................24

Figure 3.2 Figures occluding a ground ..............................................................................33

viii
 LIST OF EXAMPLES
Example 2.1 Grouping by proximity to demonstrate the trill threshold ..............................9

Example 2.2 Motivic development by varying degrees of similarity ................................10

Example 2.3 Melodic assimilation into the accompaniment .............................................16

Example 2.4 Role reversal of melody and accompaniment ...............................................17

Example 2.5 Continuity causes disjunct bass notes to group as a line ..............................20

Example 3.1 Initial LH and RH phrase groups of “Désordre.” .........................................22

Example 3.2 Header repetitions in the first two phrases ...................................................27

Example 3.3 Comparison of phrases to demonstrate perceived accelerando ...................29

Example 3.4 Boundary delineations at the climax.............................................................31

Example 3.5 Occlusion of the background layer ...............................................................35

Example 3.6 Perception of accents as syncopations against a traditional 4/4 meter .........37

Example 3.7 Perception of tempo phasing with elementary pulses eliminated .................37

Example 3.8 Moment when phasing is difficult to track aurally .......................................38

Example 3.9 Constancy of the eighth-note pulse...............................................................42

Example 3.10 Foreground occlusion of the background layer ..........................................43

Example 3.11 Hypothetical composition lesson: student’s original ..................................46

Example 3.12 Hypothetical composition lesson: teacher’s question .................................46

Example 3.13 Hypothetical composition lesson: teacher’s suggested revision .................47

ix
 CHAPTER 1

INTRODUCTION
György Sándor Ligeti was born on May 28, 1923 in Romania, and died on June

12, 2006 in Germany. Between 1985 and 2001, he wrote three books of etudes for piano.

“Désordre” is the first of the six etudes in Book One. It was written in 1985.

During this time period, Ligeti’s music underwent dramatic stylistic changes.

Ligeti also expanded his compositional technique in reaction to a number of outside

influences, including African music, the player piano studies of Conlon Nancarrow, and

the music of American jazz artists Thelonius Monk and Bill Evans.1 Perhaps the most

influential source of inspiration for this particular etude was sub-Saharan African Banda

Linda music. Polyrhythm and a complex balance between order and disorder characterize

both Ligeti’s etude and this African music. In the foreword to Simha Arom’s book

African Polyphony and Polyrhythm, Ligeti wrote:

Undoubtedly my interest in the music Arom has recorded stems also from
the proximity I feel exists between it and my own way of thinking with
regards to composition: that is, the creation of structures which are both
remarkably simple and highly complex. The formal simplicity of sub-
Saharan African music with its unchanging repetition of periods of equal
length, like the uniform pearls of a necklace, is in sharp contrast to the
inner structure of these periods which, because of simultaneous
superpositioning of different rhythmic patterns, possesses an extraordinary
degree of complexity… What we can witness in this music is a wonderful
combination of order and disorder, which in turn merges together
producing a sense of order on a higher level.2

1
György Ligeti, liner notes to György Ligeti, Edition 3: Works for Piano (Etudes and Musica
Ricercata), Sony Classical (DDD 62309), 1997, Pierre-Laurent Aimard, piano.
2
Simha Arom, African Polyphony and Polyrhythm: Music Structure and Methodology (Cambridge:
Cambridge University Press, 1991), foreword by Ligeti, xvii.

1
 The analysis of this etude is based upon principles of perception as found in

Gestalt theory. Chapter 2 introduces the reader to these principles and recounts their early

development in the field of psychology. Some theoretical background is required to

understand the principles at work in the author’s analysis. As such, Chapter 2 introduces

the reader to the Gestalt principles of proximity, similarity, figure/ground articulation,

common region, and continuity. Visual and musical examples demonstrate how these

principles may be used as analytical tools.

The Chapter 3 analysis identifies which Gestalt principles are involved in the

apprehension of formal divisions, compositional processes, and large-scale unity in

“Désordre.” After a brief discussion of the form, the analysis is presented in four main

parts. The first part identifies elements of the composition that help the listener perceive

boundaries between phrases, phrase groups, and sections. The second discusses which

elements articulate foreground and background layers. The third shows which elements

make the perception of polytempo possible. The fourth addresses elements that contribute

to the perception of large-scale unity in the composition. Finally, some pedagogical

applications for teaching composition are briefly addressed.

This work is essentially an extension of approaches represented by Leonard B.

Meyer’s Emotion and Meaning in Music3 and Eugene Narmour’s The Analysis and

Cognition of Basic Melodic Structures.4 This study is unique, however, in its application

of a Gestalt-based approach to the etude as a whole. John Sloboda confirms the need for

this type of approach:

3
Leonard B. Meyer, Emotion and Meaning in Music (Chicago: University of Chicago Press, 1956).
4
Eugene Narmour, The Analysis and Cognition of Basic Melodic Structures: The Implication-
realization Model (Chicago: University of Chicago Press, 1990).

2
 Cognitive psychology has...achieved its greatest successes in advancing
the understanding of processes that span seconds rather than minutes or
hours. For instance, a great deal is known about how human beings
process words and a lot about how they process sentences, but almost
nothing about how they process extended discourse, as found in books or
plays. Similarly, music research has yielded immense dividends at the
level of notes, chords and phrases, but very little at the level of complete
works.5

It is also important to mention that the author has benefited from a large number of

excellent analyses of “Désordre” including: Guthrie (1989), Kinzler (1991), Licata

(1992), Tsong (2001), and especially Haapamaki (2012) and Uranker (1998). It is the

author’s hope that the perceptual approach taken here will contribute to the understanding

of Ligeti’s etude.

5
In Diana Deutsch, et al., “Psychology of Music,” Grove Music Online, Oxford Music Online, Oxford
University Press, accessed February 4, 2015,
http://www.oxfordmusiconline.com/subscriber/article/grove/music/42574pg1.

3
 CHAPTER 2

GESTALT PRINCIPLES OF PERCEPTION

This chapter provides the historical and theoretical foundation necessary for

understanding the following chapter’s Gestalt-based analysis of “Désordre.” The

principles of perception that are introduced are proximity, similarity, figure/ground

articulation, common region, and continuity. (Other Gestalt principles are introduced in

Chapter 3.) Part of this introduction is dedicated to explaining the role of subliminal

perceptions in the apprehension of large-scale formal structures. Visual and musical

examples accompany the definition of each Gestalt principle.

2.1 Historical Background

Gestalt theory began in the early 20th century as a formal branch of psychology.

Its origins, however, may be traced back to the work of 19th-century psychologists Carl

Stumpf and Christian von Ehrenfels.6 In 1883, Carl Stumpf published a theory about the

relationship between the parts of a whole and the whole in his book Tonpsychologie.7

Stumpf described consonance in terms of a “perceptual fusion of the component tones

into a single impression….”8 He noticed that there was something intrinsically different

about the perception of a dyad. Ehrenfels took Stumpf’s theory a step further with his

investigations into why the mind was still able to recognize a melody after it was

6
Stephanie Sabar, “What’s a Gestalt?” Gestalt Review 17, no. 1 (2013): 7, accessed March 11, 2015,
http://www.gisc.org/gestaltreview/documents/whatsagestalt.pdf.
7
Carl Stumpf, Tonpsychologie, 2 vols. (Leipzig: Hierzel, 1883–90).
8
Deutsch, et al., “Psychology of Music.”

4
transposed to a new key.9 In 1938, Ehrenfels’s student Max Wertheimer, described the

initial problem:

Psychology had said that experience is a compound of elements: we hear a

melody and then, upon hearing it again, memory enables us to recognize
it. But what is it that enables us to recognize the melody when it is played
in a new key? The sum of the elements is different, yet the melody is the
same; indeed, one is often not even aware that a transposition has been
made.10

Ehrenfels concluded that something about the relationships between the tones of a

melody remained the same upon transposition. Because of these relationships (i.e., the

intervals between the pitches), the melody has a complex identity, which the mind can

remember even after transposition. Conversely, the tones of the same melody presented

in a different order would take on a new identity because of the altered relationships

between the pitches. In the first few decades of the 20th century, Ehrenfels’s students

Wolfgang Köhler and Kurt Koffka developed his theories further. Their work eventually

led to the formal acceptance of Gestalt theories into the discipline of psychology.

These founders of Gestalt theory identified certain “principles of perceptual

organization” that seem to describe the ways that objects are perceived.11 The principles,

or laws as the founders called them,12 correlate to “rules of the organization of perceptual

senses.”13 Stephanie Sabar describes how these principles affect our perception of the

visual world. Sabar explains that the founders of Gestalt theory learned that:

9
Christian von Ehrenfels, Über Gestaltqualitäten [On Gestalt Qualities] (Leipzig: Reisland, 1890),
trans. Barry Smith in Foundations of Gestalt Theory (Munich: Philosophia, 1988), 83–85.
10
Max Wertheimer, “Gestalt Theory,” in Source Book of Gestalt Psychology, ed. Willis D. Ellis (New
York: Harcourt, Brace, 1938), 4.
11
Max Wertheimer, “Principles of Perceptual Organization,” in Readings in Perception, ed. David
Beardslee and Michael Wertheimer (Princeton, NJ: Van Nostrand, 1958), 115–35.
12
See Meyer, Emotion and Meaning in Music; Narmour, Basic Melodic Structures; and Narmour, The
Analysis and Cognition of Melodic Complexity: The Implication-realization Model (Chicago: University of
Chicago Press, 1992).
13
Dejan Todorovic, “Gestalt Principles,” Scholarpedia.org, 2008, accessed March 11, 2015,
doi:10.4249/scholarpedia.5345.

5
 …our eyes are not like a camera or a window. We do not see the world
objectively. Rather, what we see is interpreted and given meaning by the
observer….”14

The founders discovered that “perception involved more than just the stimuli that entered

the eye.”15 When perceived, the stimuli become something more, or different as Köhler

famously worded it: “The whole is different than the sum of the parts.”16 According to

Gestalt theory, the mind groups stimuli into what Ehrenfels called Gestalts (Gestalten),17

or percepts of higher significance than the parts that caused them to form. The German

word Gestalt means form or shape. When writing about Gestalt theory, English writers

typically substitute the word “group” for Gestalt as the author does in this study.

Visual examples can demonstrate the concept of perceptual grouping. In Figure

2.1, letters of like case (upper vs. lower) tend to be grouped together. This is grouping by

the principle of similarity.

G X a Q S B V e L F C

Figure 2.1. Preliminary example of perceptual grouping.

Other letters in the figure tend to group because of their proximity. The lowercase letters

seem to act as spatial dividers between the uppercase letters, causing the following three

groups of uppercase letters to form: GX, QSBV, and LFC.

14
Sabar, “What’s a Gestalt?”, 8.
15
Russell A. Dewey, “Psychology: An Introduction,” chap. 4, part 1, in Psych Web (psywww.com),
accessed March 11, 2015, http://www.intropsych.com/.
16
Wolfgang Köhler, The Task of Gestalt Psychology, intro. Carroll C. Pratt (Princeton, NJ: Princeton
University Press, 1969) 9–10. This quote is commonly mistranslated as “The whole is greater than the sum
of the parts.” According to Pratt, Köhler disliked this translation.
17
Ehrenfels, Gestaltqualitäten, various pages. Note that Gestalten is the German plural form of Gestalt.

6
 Similarity and proximity are two important principles of perception. Many other

principles have been discovered and named. Others have yet to be discovered.18

Groupings can be explained using the principles of similarity and proximity. These

principles are similarly applicable to the perception of music. For example, Alf

Gabrielsson writes: “Melodies are usually dominated by small intervals (principle of

proximity) and performed using the same timbre (similarity).”19

The following sections present detailed definitions of the principles of proximity,

similarity, figure/ground articulation, common region, and continuity—first as they

pertain to vision, and then as they pertain to musical experience. These are the main

Gestalt principles that will be utilized in the Chapter 3 analysis.

2.2 Proximity

The principle of proximity states that elements tend to be grouped if they are near

to each other.20 In Figure 2.2, the dots nearest to one another tend to group, while others,

such as the second and third dots, do not.

Figure 2.2 Grouping by proximity.21

In general, the first group the mind forms upon seeing stimuli is what Wertheimer calls

the “natural grouping.”22 Different groupings may be forced into perception, but:

18
See Stephen E. Palmer, “Common Region: A New Principle of Perceptual Grouping,” Cognitive
Psychology 24 (July 1992): 436–47.
19
In Deutsch, et al., “Psychology of Music.”
20
See Todorovic, “Gestalt Principles,” and Palmer, Vision Science, 257–58.
21
The visual examples in this chapter are often based on the examples in Wertheimer’s “Laws of
Organization.”
22
Wertheimer, “Laws of Organization in Perceptual Forms,” In A Source Book of Gestalt Psychology,
ed. Willis D. Ellis, 71–88 (New York: Harcourt, Brace, 1938), accessed on March 10, 2015,
http://psychclassics.asu.edu/Wertheimer/Forms/forms.htm.

7
 …this cannot be maintained for long. The natural grouping soon returns as
an overpowering “upset” of the artificial arrangement.23

For example, in Figure 2.3, the natural grouping tends to yield a perception of four

columns of dots, but four rows of dots can be forced into focus temporarily.

Figure 2.3 Perception of columns because of proximity.

2.3 Proximity in Music

The perception of elements in pitch space appears to be determined at times by the

principle of proximity. For example, a note of a melody that follows another by stepwise

motion tends to group with it. As documented in the work of George Miller and George

Heise,24 the greater the distance between the tones, the greater the tendency to hear more

than one line. Scott Lipscomb reports on their findings: “When two pitches are far

enough apart, fission (a perceptual splitting apart) occurs. Likewise, when two pitches are

close together, they tend to fuse.”25 The author-composed melody in Example 2.1

demonstrates the effect of the so-called trill threshold.26 The melody begins with conjunct

motion. Starting in m. 5, with the leap of a P4, it widens into two separate strands. By the

end of m. 10, when the intervals are again a third or smaller, its strands have merged back

together.

23
Wertheimer, “Laws of Organization.”
24
George A. Miller and George A. Heise, “The Trill Threshold,” Journal of the Acoustical Society of
America 22 (September 1950), 637–38.
25
Scott D. Lipscomb, “The Cognitive Organization of Musical Sound,” in Handbook of Music
Psychology, ed. Donald A. Hodges (San Antonio, TX: IMR Press, 1996), 148.
26
The musical examples in this chapter were composed by the author for this paper.

8
Example 2.1 Grouping by proximity to demonstrate the trill threshold.

2.4 Similarity

The principle of similarity states that elements tend to be grouped if they are

similar to each other.27 The dots in Figure 2.4a28 are homogenously colored and spaced

and give no tendency to group in one direction more than another.

a b c d e

Figure 2.4. Comparison of resultant groupings from proximity and similarity.

In 2.4b where the spacing remains the same, the similar coloring between certain dots

yields a grouping into columns. In 2.4c, proximity yields a natural grouping into three

rows.29 This same spacing in 2.4d, however, is less forceful. Here, the principles of

similarity and proximity compete against one another, yielding two possible natural

groupings between which the mind is able to switch back and forth with relative ease. In

2.4e, the principles work together to produce a natural grouping into rows. The figure

27
See Todorovic, “Gestalt Principles,” and Palmer, Vision Science, 258.
28
The identifying letters a, b, c, d, and e are purposefully off center to avoid affecting the grouping.
29
When the eye traces from 2.2b to 2.2c, the grouping from the first may temporarily affect the
grouping in the second.

9
demonstrates how certain groupings can be made to dominate when two or more

principles work toward the same grouping.

2.5 Similarity in Music

According to John Sloboda,

The most basic form of recognition is the experience of similarity. This

simply requires that something heard is experienced as being identical to,
or sharing important characteristics with, something heard before. Without
the ability to recognize similarity between elements within a piece, the
apprehension of form would be impossible.30

In A Generative Theory of Tonal Music (1983), Fred Lerdahl and Ray Jackendoff further

explain that similarity groupings made over the course of an entire piece form

hierarchical relationships.31 Lipscomb summarizes their findings:

Initially, local patterns (e.g., motives and themes) are identified, then they
are placed hierarchically within the context of the entire composition.32

Motives are perceived as gaining complexity as the composer develops them. Example

2.2 demonstrates the development of motivic complexity.

a a a'

a" a" a"

a a a'''

Example 2.2 Motivic development by varying degrees of similarity.

30
In Deutsch, et al., “Psychology of Music.”
31
Fred Lerdahl and Ray Jackendoff, A Generative Theory of Tonal Music (Cambridge: MIT Press,
1983), 13–17.
32
Lipscomb, “Cognitive Organization,” 165.

10
Segments marked a indicate the motive that undergoes development in the example.

Because of transpositional and durational differences, motive a' shares close but inexact

similarity with motive a. Motive a'' adds another degree of variation through interval

expansion. Motive a''' varies the pitch content for yet another degree of variation.

Motives a', a'', and a''' all relate to motive a through close but inexact similarities. These

slight variations enable grouping processes to form complex relationships between the

motives. Each new variation of the original adds a level of complexity to the group.

Figure 2.5 represents the process (as it happens in time) by which motive a gains

complexity.

a a' a'' a'''

Figure 2.5 Development of motivic complexity.

Tracing from left to right in the figure shows the progression from motive a through a'''.

The arrows reflect how grouping processes continually compare new material with

material heard earlier in order to determine the level of similarity they share. When

variations are similar enough to the original, they can be grouped with it. In the melody

of Example 2.2, each variation of the motive is compared with the original. As Figure 2.5

shows, the variations are also compared with each other. Each variation adds a level of

complexity to the preceding motive. By the time the melody ends, three levels of

complexity have been added to motive a, two to a', and one to a''.

11
2.6 Subliminal Perceptions and Large-scale Grouping

The melodic analysis of Example 2.2 demonstrates how complex groupings can

form over large spans of time. Some of these groupings may even occur subliminally

without a listener being able to pinpoint exactly which structural elements are responsible

for the perception. Stephen Palmer, in his book Vision Science,33 addresses the issue of

whether perceptions can be made below the threshold of awareness:

By far the most dramatic and important claim about unconscious

perception in normal observers is for the existence of subliminal
perception: the ability to register and process information that has been
presented under conditions in which it is below the threshold of
awareness.34

Palmer supports this theory by documenting scientific research that provides strong

evidence for it.35

Figure 2.6 shows a visual example created by the author to demonstrate

subliminal grouping.

Figure 2.6. Logo demonstrating subliminal grouping.

33
Stephen E. Palmer, Vision Science: Photons to Phenomenology (Cambridge, MA: MIT Press, 1999).
34
Ibid., 639. Palmer clarified the difference between the words subliminal and subconscious with the
following explanation: “The term ‘subliminal’ is derived from the prefix sub, meaning ‘below,’ and the
root limen, meaning ‘threshold’.”
35
Ibid., 639–44.

12
In this logo, grouping immediately occurs between highly similar elements (e.g., triangles

with triangles and circles with circles). Grouping also occurs because of two other factors

that may be less obvious, and are perhaps only noticeable after a few seconds of scrutiny.

Cohesion of the whole is strengthened by the arrangement of sun symbols into a circle.

This larger circle groups by similarity with each smaller circle in the sun symbols. The

second somewhat hidden ingredient of cohesion has to do with the central octagon in the

word Son. While the surrounding sun symbols form a circle, the interior angles created by

their connecting triangles form a large octagon. The smaller octagon in the word groups

by similarity with the larger, slightly hidden octagon, giving the name of the company a

second sense of cohesion to the figures surrounding it (the first being the triangles that

form the letters S and n). The observer may not consciously notice the connection

between the octagon and the surrounding symbols, and yet, because of the grouping

influence of repetitious elements, the observer’s mind may notice it anyway and the small

octagon in the center will not seem entirely out of place.

Much of what contributes to a listener’s perception of overall cohesiveness in a

composition is a result of perceptions made below the threshold of awareness. In the

1982 article “Pattern, Poetry, and Power in the Music of Frederic Chopin,” Douglas

Hofstadter presents an analysis of Chopin’s music in which he speaks of subliminal

groupings that may be responsible for certain perceptions:

[Musical] phenomena perceived to be magical are always the outcome of

complex patterns of nonmagical activities taking place at a level below
perception.36

36
Douglas R. Hofstadter, Metamagical Themas: Questing for the Essence of Mind and Pattern (New
York: Basic Books, 1985): 177–78. The article “Pattern, Poetry, and Power” is reprinted in the book.

13
The development of motivic complexity described in the discussion of Example 2.2, or

the groupings discussed in the Figure 2.6 logo, may not all be conscious perceptions. Yet,

the ease with which listeners and observers form groups when presented with some

degree of repetition suggests that they do not need to consciously notice each degree of

similarity in order to perceive unity. It is not necessary, for example, to consciously

detect every motivic variation in the first movement of Beethoven’s Symphony no. 5 to

be able to walk away from a performance of it having felt that its structure had a strong

sense of unity.

2.7 Figure/Ground Articulation

In 1923, Max Wertheimer described the requirements for the perception of a

figure upon a ground:

When an object appears upon a homogeneous field there must be stimulus

differentiation (inhomogeneity) in order that the object may be perceived.
A perfectly homogeneous field appears as a total field [Ganzfeld]
opposing subdivision, disintegration, etc. To effect a segregation within
this field requires relatively strong differentiation between the object and
its background.37

The principle of figure/ground articulation states that an element surrounded by a

dissimilar element tends to be grouped as a figure upon a ground if the surrounding

element is large and homogenous enough to be perceived as distinct from the smaller

element.38 This is demonstrated in Figure 2.7, where three objects (figures) seem to lie

upon a ground (the gray rectangular area). Dejan Todorovic describes the phenomenal

characteristics of this principle this way:

37
Wertheimer, “Laws of Organization.” Brackets in original.
38
Palmer names three additional requirements for the perception of figures upon grounds, but they do
not appear to be immediately applicable to this analysis. See Palmer, Vision Science, 281–83.

14
Figure 2.7. Grouping by figure/ground articulation.

The areas of the figure and the ground usually do not appear juxtaposed in
a common plane, as in a mosaic, but rather as stratified in depth: there is a
tendency to see the figure as positioned in front, and the ground at a
further depth plane and continuing to extend behind the figure, as if
occluded by it. Furthermore, the border separating the two segments is
perceived as belonging to the figure rather than to the ground, and as
delineating the figure’s shape as its contour, whereas it is irrelevant to the
shape of the ground.39

In other words, the natural grouping suggests the figures are sitting on top of the gray

area rather than suggesting they are holes or cutouts exposing a darker background.

2.8 Figure/Ground Articulation in Music

The perception of a melody over an accompaniment is an example of grouping

according to the principle of figure/ground articulation. For example, Classical Alberti

bass accompaniments can be thought of as backgrounds on which melodies can sit.

Lipscomb confirms the indispensability of figure/ground articulation in this kind of

musical texture. He writes that the principles of perception:

...assist in the process of recognizing the most important events and

abstracting them perceptually from a less significant background of
activity. This is often referred to as a figure-ground relationship, an ability
essential to musical hearing. How would it be possible to understand the
intricacies of sonata form if we were incapable of abstracting the primary
theme (figure) from its accompaniment (ground)?40

Example 2.3 demonstrates how, if a melody takes on enough of the attributes of its

accompaniment, the two will perceptually fuse together.

39
Todorovic, “Gestalt Principles.”
40
Lipscomb, “Cognitive Organization,” 145–46.

15
Example 2.3 Melodic assimilation into the accompaniment.

This brief analysis of Example 2.3 will make it apparent that the principle of

figure/ground articulation is less dependent on the similarity of elements than it is on

their dissimilarity. Without enough dissimilarity, objects could not be distinguished as

figures.

In the example above, the part played by the right hand (RH) begins by presenting

a melody (figure) over a drone (ground) in the left hand (LH). To make this distinction,

durational and registral differences work together to distinguish the separate layers. Two

layers are perceived whose respective elements have been grouped by similarity.

The distinction between melody and accompaniment becomes impossible as this

melody progresses. Measures 4 and 10 present situations in which the melody takes on

the characteristics of its accompaniment, thereby fusing with it. First, the melody

descends into the accompaniment’s range, increasing the likelihood of grouping by

16
proximity. Then, it sheds its rhythmic distinctions and adopts those of the

accompaniment, increasing the likelihood of grouping by durational similarity. The

melody, in effect, transforms into the accompaniment and becomes part of it.

Example 2.4 demonstrates that the opposite is also possible.

Example 2.4 Role reversal of melody and accompaniment.

The melody in this example begins to take on accompanimental characteristics at m. 4

and fully assimilates as part of the accompaniment by the second beat of m. 5. At the

transition from mm. 5–6, the LH accompanimental material eases into the character of a
^ ^
melody through a stepwise 7 to 1 resolution. From mm. 6–9, the LH material takes on all

of the characteristics of the previous RH melody. At this point, the roles have reversed.

2.9 Common Region

The principle of common region, a relatively new principle of perception

discovered by Palmer in 1992,41 states that elements within a closed region of space tend

41
See Palmer, “Common Region,” 436–47.

17
to be grouped together.42 In Figure 2.8a, the dots that are set closest together tend to

group (proximity), i.e., the second with the third and the fourth with the fifth, leaving the

outer two dots relatively ungrouped.

a b

Figure 2.8. Grouping by common region against proximity.

In 2.8b, because of common regions delineated by rectangular boundaries, the dots group

into three sets of two, even though proximity relationships remain the same from 2.8a

to 2.8b.

2.10 Common Region in Music

A musical analogue for the principle of common region requires a listener to

perceive boundaries. One way to create a boundary in music is to replace the pattern a

listener is experiencing with a new one. When this happens, the listener perceives a

boundary at the seam. According to the principle of common region, musical elements

preceding the boundary tend to be perceived as belonging together. These elements group

by similarity because they share the same pattern and because they fall within the same

boundary. Earlier in this chapter, Sloboda was quoted as having said: “Without the ability

to recognize similarity between elements within a piece, the apprehension of form would

be impossible.”43 It should be added that the perception of musical boundaries would not

be possible without the ability to recognize dissimilarity between elements. The ability to

notice dissimilarity makes the perception of boundaries possible. This concept has

42
Palmer, Vision Science, 260.
43
In Deutsch, et al., “Psychology of Music.”

18
significant applications in the next chapter’s analysis when it will be shown that Ligeti

disrupts these grouping processes by blurring the edges of a significant boundary between

sections.

2.11 Continuity

The principle of continuity, or good continuation,44 states that elements that can

be seen as smooth continuations of each other tend to be grouped together.45 For

example, Figure 2.9a tends to group as two intersecting lines. The natural grouping is not

likely to be seen as two right angles whose vertices meet, as shown in 2.9b.

a b

Figure 2.9 Grouping by continuity.

2.12 Continuity in Music

Notes of a melody moving in stepwise motion tend to group by proximity, but it is

grouping by continuity that creates the perception of a single line. In reality, the notes are

not connected at all. They are simply articulated moments in time. In fact, melodies, or

musical lines, are often disjunct arpeggiations, moving across intervals of 3rds and 4ths,

and as is often the case in bass lines, across 5ths or even 8ves. And yet, their individual

notes are typically perceived as continuations of each other. Example 2.5 demonstrates a

bass line that would yield such a perception.

44
See Wertheimer, “Laws of Organization.”
45
Palmer, Vision Science, 259.

19
Example 2.5 Continuity causes disjunct bass notes to group as a line.

In his book Explaining Music, Leonard B. Meyer uses the word criticism

synonymously with analysis, writing that “criticism attempts to understand and explain

the choices made by the composer in a particular work.”46 Many of Ligeti’s

compositional choices may be understood in terms of the Gestalt principles of perception

introduced in this chapter. The Chapter 3 analysis will apply these Gestalt principles to

Ligeti’s “Désordre.”

46
Leonard B. Meyer, Explaining Music: Essays and Explorations (Berkeley: University of California
Press), 1973: 16–17.

20
 CHAPTER 3

A GESTALT-BASED ANALYSIS OF “DÉSORDRE”

After a brief discussion of the form, the analysis in this chapter is presented in

four parts. The first part identifies elements of the composition that help the listener

define boundaries between phrases, phrase groups, and sections. The second discusses

how foreground and background layers are articulated. The third discusses issues of

perception and Ligeti’s single-performer polytempo illusion. The fourth identifies

elements that contribute to large-scale unity in the composition. A summary of the

Gestalt principles explained in Chapter 2 can be found in Appendix A for reference

during this chapter’s analysis.

Prior to a performance of “Désordre” in Gütersloh, Germany on May 5, 1990,

Ligeti said:

That which is eminently new in my piano etudes is the possibility of a

single interpreter being able to produce the illusion of several
simultaneous layers of different tempi.47

In the analysis that follows, the author uses the term “perceived tempo” to refer to these

individual layers of different tempos. This term will be used extensively in the formal

overview in section 3.1 and in the discussion of Ligeti’s single-performer polytempo

illusion in section 3.4.

3.1 Formal Overview.

Example 3.1 shows the first 20 measures of the piece.

47
Ligeti, Erato.

21
 5

13

13

17

17

*) Use the pedal sparingly throughout. Play the melody legato in both hands.

György Ligeti ÉTUDES POUR PIANO No. 1 “Désordre,” LH mm. 1–20

Copyright © 1986 by Schott Music GmbH & Co. KG
All Rights Reserved.
Used by permission of European American Music Distributors Company,
sole U.S. and Canadian Agent for Schott Music GmbH & Co. KG

Example 3.1 Initial LH and RH phrase groups of “Désordre.”

22
Circles in the example indicate beginnings of phrases. Three phrases in a particular hand

constitute a phrase group. The process of repeating and transposing the two (LH and RH)

phrase-group patterns defines much of this etude’s structure.

Although this analysis will primarily focus on the perception of rhythm, it is

important to notice that the original patterns shown in Example 3.1 consist of RH

material that is presented exclusively on the white keys against LH material that is

presented exclusively on the black keys. This quasi-bimodal tension is maintained

throughout the etude (except for the last note) and is one of the many special features of

“Désordre.” Figure 3.1 provides a diagram of the etude’s large-scale form to serve as a

visual guide for analytical discussions that follow.

Analysts seem to generally agree that the etude is divided into three main

sections.48 For example, Mayron K. Tsong writes that the etude “follows a basic ternary

structural pattern of A B A'.”49 Thus Figure 3.1 divides into three sections labeled: I, II,

and III, respectively. Because of Ligeti’s misalignment of barlines in the score (e.g., see

Example 3.1, RH m. 4), the vertical pitch axis divides into RH (top) and LH (bottom)

measures. The range of the pitch axis is C1–C8, where C4 is middle C. A horizontal

dotted-line grid marks the location of each C: i.e., C1, C2, C3, etc. The dots in the figure

indicate the beginnings of phrases. Line segments extending from these dots indicate the

approximate phrase lengths. Phrase groups are indicated using rectangular boxes and

each phrase group is numbered. Because the phrase-group lengths differ in each hand,

48
See Sampo Elias Haapamaki, “Order in ‘Désordre’: Rhythmic and Melodic Structure in György
Ligeti’s Piano Etude No. 1,” DMA diss. (Columbia University, 2012), 5; Mayron K. Tsong, “Études pour
piano, premier livre of György Ligeti: Studies in Composition and Pianism,” MM thesis (Rice University,
2001), 24; and Mark Uranker, “Ligeti’s ‘Désordre’ from Études pour piano—premier livre: An Analysis
and Performance Guide,” MM thesis (California State University, Long Beach, 1998), 27.
49
Tsong, “Études pour piano,” 24.

23
24

Figure 3.1 Visual representation of the form of “Désordre” (not to scale).

the RH contains 14 phrases, whereas the LH contains only 11. Each phrase group is

related to the original pattern (Example 3.1) by diatonic transposition Tn, where n is an

integer representing the transposition distance and direction, measured in steps along

respective white-key and black-key collections, from the original pattern. For example, in

the RH, T0 represents the original white-key pattern beginning on B4, T1 represents the

pattern beginning on C5, T2 represents the pattern on D5, and so on. In the LH, T0

represents the original black-key pattern beginning on D#4, T-2 represents the pattern

beginning on A#3, T-4 represents the same pattern beginning on F#3, and so on. The lines

connecting the dots in Figure 3.1 are designed to visually highlight and track the

movement of the RH and LH pitch materials over time. Figure 3.1 makes clear certain

overall trends: e.g., the large-scale pitch ascent in the RH, and the large-scale descent in

the LH, as well as a gradual registral widening between the hands.

Table 3.1 provides a formal overview and a bullet-point summary of the most

important aurally perceivable characteristics in each section of the etude. The table also

identifies three new subsections: Transition 1, Transition 2, and Ending. Familiarization

with these characteristics will be helpful in the sections that follow. Like Figure 3.1, this

table is designed to serve as an overview and guide for the analytical discussions that

follow.

3.2 Phrases, Phrase Groups, and Sections

This analysis seeks to identify compositional elements that trigger grouping

processes and enable the listener to distinguish between the following three components

of formal delineation: (1) phrases, (2) phrase groups, and (3) sections.

25
Table 3.1 Aurally Perceivable Characteristics in “Désordre”

Section Measures Aurally Perceivable Characteristics

I RH 1–56 ˗ Melodies are doubled at the octave.
˗ Phrase-group patterns are established.
˗ The etude begins in the middle register of the piano, but the
registral distance between the left and right hands gradually
widens with the presentation of each successive phrase group.
˗ Contraction of space between melodic notes (phrase contraction)
creates the illusion of polytempo by breaking the synchronization
of accents.
Transition 1: RH 44–56 ˗ More frequent phrase contractions lead to a perceived tempo
acceleration.
II RH 57–98 ˗ Melodies are still doubled at the octave.
˗ Phrase groups line up for the first (and last) time since the first
measure.
˗ Perceived tempos move approximately three times faster than at
the beginning of Section I.
˗ The widening of the registral distance between the hands
continues from where it left off in Section I and the LH reaches
the lowest black keys.
˗ The polytempo illusion persists because of phrase misalignment.
Transition 2: RH 88–98 ˗ A second perceived tempo acceleration is achieved by phrase
contraction, and builds to a maximum in preparation for the
climax at m. 99.
III RH 99–153 ˗ Octave doublings discontinue in favor of 2nds, 3rds, 4ths, 5ths,
6ths, and 7ths, beginning in two-voices and then building to four-
voices in the RH and three in the LH.
˗ The LH shifts its pattern from the lowest register to a higher
register one octave below the RH.
˗ The perceived tempos abruptly revert to those from Section I.
˗ Phrase expansion slows the perceived tempo of the LH, while the
RH’s perceived tempo remains constant to the end.
Ending : RH 115–52 ˗ When the LH pattern returns to T0, phrase expansion begins to
drastically slow down the perceived tempo until it comes to a
stop in the penultimate measure.
˗ As the LH’s perceived tempo slows down, and the space between
melodic notes increases, the eighth-note background layer begins
to emerge as the main focal point, after which, it makes an
upward run to C8.

The etude’s phrases are designed in a way that makes their boundaries clear to the

listener. The phrase boundaries in “Désordre” are defined in two ways: header repetitions

and ending elongations. Header repetitions, the author’s term, mark the beginnings of

phrases. Example 3.2 shows the opening eight measures of the etude with header

repetitions indicated using arrows.

26
 5

György Ligeti ÉTUDES POUR PIANO No. 1 “Désordre,” mm. 1–8

Copyright © 1986 by Schott Music GmbH & Co. KG
All Rights Reserved.
Used by permission of European American Music Distributors Company,
sole U.S. and Canadian Agent for Schott Music GmbH & Co. KG

Example 3.2 Header repetitions in the first two phrases.

Rhythms that occur at the end of each phrase also stand out in the texture and help

introduce the header repetitions. In mm. 4 and 8, Ligeti lengthens the final melodic

duration of the phrase in both hands to fill the measure, as circled in the example. This

elongation helps signal the end of the phrase, and functions as a kind of musical breath

mark. When this breath precedes a header repetition, it allows the repetition to stand out

of the texture. Header repetitions and ending elongations work together to define phrase

boundaries that trigger elements within a phrase to group by common region (section

2.10). In this way, the listener is able to define the phrase.

Boundaries between phrase groups are likewise delineated in a clear manner.

Because header repetitions and ending elongations occur in every phrase, in order to

27
make a distinction between phrases and phrase groups, additional elements must be

added to delineate a more complex group to occur every three phrases. Ligeti does this by

lengthening every third phrase, thus delaying the header repetition that follows.

Throughout the etude, every third phrase is lengthened in this way, establishing a large-

scale pattern that affects the grouping structure, if only subliminally. As shown in section

2.6, highly similar elements are easily detected by the mind’s subliminal processes.50

Hartmuth Kinzler’s 1991 analysis of “Désordre” concluded that Ligeti’s sequencing of

phrase group transposition patterns could be easily perceived:

The sequencing ensues without any exception and, taken for itself,
together with the corresponding rhythmical procedures, is easily
perceivable….51

Just as changes are needed to distinguish phrase groups from phrases, changes are

needed to distinguish sections from phrase groups. In Figure 3.1, perceived “tempo

changes” may be seen by comparing the width of the rectangular boxes that identify the

phrase groups. In Section II, phrase groups progress at a faster rate than in the outer

sections (I and III). It is not that the phrases in Section II have fewer melodic notes;

rather, the space between them has been shortened, as it would be with standard tempo

changes. However, perceived tempo changes do not suffice as a means to delineate the

boundaries between the sections. If one takes a close look at the Section I phrase lengths

as they approach the boundary of Section II, it is apparent that the phrase lengths slowly

approach those found in Section II. Therefore, perceiving the boundary between Sections

I and II based on these tempo changes alone can be difficult. This gradual speeding up of

50
See Palmer, Vision Science, 258.
51
Hartmuth Kinzler, “György Ligeti: Decision and Automatism in ‘Désordre,’ 1re étude, premier
livre,” Interface 20 (1991): 102.

28
the texture subtly hides the section boundary in a manner not unlike camouflage.52 At this

moment in the etude, Ligeti uses the principles of similarity and continuity as cloaking

devices: that is, to deflect grouping processes, hiding much of the dissimilarity between

the sections that would otherwise suggest a boundary to a listener. Ligeti does this by

effecting a perceived accelerando. In Transitions 1 and 2 (Table 3.1), Ligeti contracts the

space between melodic notes little by little in a way that makes them appear to speed up.

In the Ending (Table 3.1), he does the opposite in the LH, creating the illusion of a

ritardando where none is marked in the score. Example 3.3 shows the perceived

accelerando in Transition 1 by comparing the second phrase of RH Phrase Group 1

(3.3a) with the same phrase of RH Phrase Group 4 (3.3b).

Example 3.3 Comparison of phrases to demonstrate perceived accelerando.

Note that Ligeti has placed accents on each melodic note to bring them out of the texture.

The melody in Example 3.3b is a transposition (T3) of the original shown in 3.3a. All of

the notes of the original melody are present in the transposition. The differences between

52
See Palmer, Vision Science, 261, for an explanation of how camouflage “foils” grouping processes.

29
the melodies have to do with the number of eighth notes between the accents. In every

case, there are fewer eighths between accents in 3.3b than in 3.3a. And yet, the rhythms

in 3.3b do not seem altered, only accelerated. Ligeti achieves this effect of an apparent

accelerando by contracting the space between accents in a way that maintains the

rhythmic proportions from the original. In other words, every instance of a short-long or

long-short rhythm in 3.3a is kept in 3.3b, as shown in the example. When the melody in

3.3b finishes 8 eighth notes sooner than its original, it appears to be faster.

As mentioned in Chapter 2, elements that can be heard as smooth continuations of

each other tend to be grouped together. As a result of this phrase-length contraction

process described above, a listener may hear Sections I and II as a contiguous group

without a dividing boundary.

No such devices conceal the boundary between Sections II and III. In fact, just the

opposite is true. Four powerful compositional devices work together to ensure that this

boundary will be perceived. This moment in the music is shown in Example 3.4. The first

compositional device is an abrupt shift in perceived tempo. The fast pace of Section II

ends abruptly at the arrival of the RH’s m. 99. It then reverts to the perceived tempo that

had been present in most of Section I. The second device is an abrupt change in the LH’s

register. At the climax that overlaps the LH’s mm. 96 and 97, the LH’s descent to the

bottom of the keyboard strikes the lowest black key at the triple sforzando, then rests for

a few eighth notes as the pianist shifts the hand up four octaves to resume its normal

activity. The third device is textural. At RH m. 99 where the music reaches a second

climactic triple sforzando, Ligeti gives us a chord containing four pitches. Up until that

point, no more than two pitches (excluding octave doublings) have been heard in a single

30
 93

91

99

97

György Ligeti ÉTUDES POUR PIANO No. 1 “Désordre,” LH mm. 91–100

Copyright © 1986 by Schott Music GmbH & Co. KG
All Rights Reserved.
Used by permission of European American Music Distributors Company,
sole U.S. and Canadian Agent for Schott Music GmbH & Co. KG

Example 3.4 Boundary delineations at the climax.

RH chord. The fourth and final device is the discontinuation of the octave doubling of the

two melodies that has been present since m. 1, as shown in Example 3.4. Instead, Ligeti

employs other intervals in a coloristic manner. This change in the texture combines with

the force of the first three devices (abrupt change in the perceived tempo, registral

changes, and the introduction of chords) to suggest that a new section has arrived.

The Gestalt principle at work in the delineation of the boundary in Example 3.4 is

similarity (section 2.5). However, it is important to keep in mind that while each section

has a certain degree of similarity within itself, the perception of boundaries between

sections is made possible because of dissimilarities. As discussed in section 2.10,

similarity groups elements of a section together, but dissimilarity delineates boundaries.

31
The presence of musical boundaries helps articulate common regions and delineate which

items belong to which section. Thus, the principles of similarity and common region can

explain how a listener perceives musical elements as belonging to phrases, phrase groups,

or sections.

3.3 Articulation of Foreground and Background Layers

In discussing the articulation of foreground and background layers in “Désordre,”

Sampo Haapamaki writes:

…[the] basic sound elements in this composition are (1) two individual
loud, accented and irregular melodic lines…in the forte foreground and (2)
soft and regular eighth-note runs in the piano background.53

Kinzler extends this idea:

In addition to the first and second planes [the melodies], a single-voiced

eighth-note series appears in piano, which defines itself as a third plane by
the contrast of the dynamics, the constant duration of the individual
pitches, and by its single-voicedness, as a background to the foreground of
the octaves plane.54

Both authors seem to take for granted that the distinguishing characteristic is the

dynamics. But other issues come into play as well, and these, together with the forte-to-

piano differences, work to support a common objective. Kinzler touched upon one called

the “octaves plane.” Doubling the two melodies at the octave reinforces the dynamic

markings, which are further reinforced by the notated accents. Still more factors are at

work. The eighth-note ascents, Ligeti’s “elementary pulses,” are fast and quasi-robotic in

their ever-ascending stepwise motions.55 The two melodies are slow and singable; Tsong

says they are “reminiscent of a Hungarian folk melody.”56 The contrast between the two

53
Haapamaki, “Order in ‘Désordre’,” 6. Italics in the original.
54
Kinzler, “Decision and Automatism,” 91.
55
Ligeti, liner notes to Ligeti: Piano Etudes, Book 1, Erato (ECD 75555), 1990, Pierre-Laurent Aimard,
piano.
56
Tsong, “Études pour piano,” 24.

32
melodies and the eighth-note ascents is apparent from the outset as can be seen in

Example 3.1.

Contrast delineates boundaries in this music. With respect to figure/ground

articulation, boundaries are more commonly referred to as edges of a figure upon a

ground. The use of accents, octave doublings, dynamic contrasts, and the decision to

make the two melodies slow and singable over quick and mechanical eighth-note ascents

delineates the edges of the melodic figure and distinguish it from its accompanimental

ground. Ligeti could have made this distinction by fewer means, but doing so may have

complicated the task of perceiving these melodic edges. He chose, instead, to take

advantage of multiple methods to serve the same purpose. The tactic to employ many

techniques to achieve the same goal was something Ligeti made use of when delineating

the boundary between Sections II and III, as described earlier in this chapter.

Figure 3.2 represents something that begins in Transition 1 and culminates in

Transition 2 (Table 3.1).

Figure 3.2 Figures occluding a ground.

The figure resembles Figure 2.7, which was used to explain the principle of figure/ground

articulation. However, in Figure 3.2, the objects previously surrounded by the gray area

are now so large that they almost cover it. And yet, even though the gray area does not

33
surround them, it still appears to be behind them. The principle of figure/ground

articulation, which requires one area to surround another, no longer determines the

grouping in this instance. Rather, the active principle is one not addressed in Chapter 2,

that of occlusion.57 If an object appears to occlude part of another, the first tends to be

foregrounded. During Transition 2 (Table 3.1), Ligeti causes the figures (melodic

accents) to occlude nearly all of the ground (eighth-note ascents). Example 3.5 compares

the music a few measures before Transition 1 begins (LH mm. 37–40) with the music

after the transition ends (LH mm. 54–57), and then with the tail end of Transition 2 when

the ground virtually disappears (LH mm. 91–96).

Kinzler makes a similar observation regarding the occlusion of the background

layer:
An ever present foreground plane is contrasted by a background plane,
which—in changing proportion—gradually disappears and suddenly
comes back in the middle of the piece.58

At the start of Section III, the relationship between figures and their ground return to

what they had been at the start of the etude, and perception is again influenced by

figure/ground articulation.

3.4 Perception and the Polytempo Illusion

The topic of simultaneous layers of different tempos in “Désordre” is a complex

one. Aside from the tempo specified in the score (63 whole-notes per minute—76 in

Ligeti’s original), the two most obvious active tempo layers in “Désordre” are: (1) the

perceived tempo of the RH; and (2) the perceived tempo of the LH. The following

discussion will attempt to explain why each layer may legitimately be heard as a different

tempo.
57
Palmer, Vision Science, 236.
58
Kinzler, “Decision and Automatism,” 122.

34
 38

37

VS.
56

54

VS.

93

91

György Ligeti ÉTUDES POUR PIANO No. 1 “Désordre,” LH mm. 37–40, 54–57, and 91–96
Copyright © 1986 by Schott Music GmbH & Co. KG
All Rights Reserved.
Used by permission of European American Music Distributors Company,
sole U.S. and Canadian Agent for Schott Music GmbH & Co. KG

Example 3.5 Occlusion of the background layer.

As we saw in Chapter 2, visual illusions are an effective way to demonstrate how

the mind can sometimes be tricked. For example, the phenomenon of apparent motion

can convince the mind that a stationary object is in motion.59 Vection, or as Palmer calls

it, “induced self-motion,”60 can convince a person that their body is in motion when it

59
Palmer, Vision Science, 468.
60
Ibid., 504–05.

35
actually is not. In order for these phenomena to take effect, the mind must receive the

same stimuli it would receive if those experiences were real. Later in this chapter, the

discussion will show that the correct stimuli for the perception of polytempo are present

in the etude.

It should be noted that it would be highly unlikely for performers to execute a

proper performance of the etude if they focus exclusively on these stratified perceived

tempo layers. For example, Uranker explains that pianists should focus their attention on

the eighth-note background layer.61 When a pianist focuses on this layer, Uranker says

the accents feel more like syncopations than downbeats:

The rhythmic relationships in “Désordre” become much easier to hear,

play, and comprehend if the pianist regards these relationships as a
compound rhythm—not as two separate streams of music of ever-
changing relationships…. It is much easier to feel the music as a downbeat
with syncopation.62

To show the view suggested by Uranker’s compound-rhythm approach, Example 3.6

gives the opening two melodies as notated using 4/4 measures. Notice how the LH’s

rhythms in these measures all fit into a 4/4 framework, enabling the performer to keep a

steady beat. A listener, on the other hand, has the option of ignoring the background

eighth-note pulse through selective hearing. Example 3.7 shows this approach. The

numbers between the staves indicate the offset interval in eighth notes. The example

places a barline before every note in the two melodies to better visualize the offset.

Notice that a distinct phasing becomes apparent. Because the phasing takes place at such

a slow pace, a listener will have little trouble gauging the distance between offset accents.

The LH will sound as though it lags behind the RH by greater and greater durations as

61
Uranker, “Ligeti’s ‘Désordre’,” 31.
62
Ibid.

36
Example 3.6 Perception of accents as syncopations against a traditional 4/4 meter.

Example 3.7 Perception of tempo phasing with elementary pulses eliminated.

37
the music progresses.63 Kinzler writes:

It is immediately evident that the shifting by 1 or 2 eighth notes in our

model is perceived as an “after-striking.” But beginning at the shifting of 3
eighth notes, the “correct” relationship becomes insecure…. With further
shiftings a “pre-striking” perception finally asserts itself.64

By the time the phasing has brought the LH behind by three eighth notes, the phasing

becomes difficult to track. As the LH falls further and further behind, Ligeti explains:

…the metric relationship is gradually blurred until we reach a point where

we are unable to discern which hand leads and which lags behind. A state
of order is in due course restored as the two successions of accents shift
closer and closer to one another, eventually falling simultaneously in the
two hands, at which point the cycle begins anew.65

Example 3.8 shows a moment in Section I when the two melodies are so far apart

that the offset will be difficult to track aurally.

21

-5 -6
-5 -5 -5

21

György Ligeti ÉTUDES POUR PIANO No. 1 “Désordre,” LH mm. 21–24

Copyright © 1986 by Schott Music GmbH & Co. KG
All Rights Reserved.
Used by permission of European American Music Distributors Company,
sole U.S. and Canadian Agent for Schott Music GmbH & Co. KG

Example 3.8 Moment when phasing is difficult to track aurally.

Arrows indicate accents that may cause a listener to lose track of the offset distance. For

example, the first accent in LH m. 22 has fallen behind the first accent in RH m. 22 by a

63
Ligeti, “On My Études,” 7.
64
Kinzler, “Decision and Automatism,” 110.
65
Ligeti, “On My Études,” 7.

38
distance of five eighth notes. However, because the last accent of LH m. 21 occurs after

the first accent of RH m. 22, the LH accent in m. 21 will probably appear to be the accent

that has fallen behind. The alignment of LH and RH accents at RH m. 24 may even

convince a listener that a new phrase alignment has taken place.

At the beginning of the etude, the two melodies are easy to track, even when out

of synchronization. At times, they sound as if they are in canon with each other.66 This

seems to be an intentional decision to ensure the listener has time to become accustomed

to the necessity of tracking the offset between the two melodies. At one point in

Uranker’s analysis, he turned to the perspective of the listener and commented that:

…if the melody were more disparate and less tautly constructed, the point
of the phasing might be lost on the listener.67

There are at least four Gestalt principles that contribute to the perception of

polytempo in “Désordre”: similarity, proximity, figure/ground articulation, and

synchrony. The principle of synchrony was not introduced in Chapter 2, but will be

defined in a moment.68

First, similarity between the rhythms of the two competing melodies helps the

listener decide which notes should be heard as lagging behind others. For example, the

canon-like imitation can almost be thought of as presenting a near copy of the RH

melody in the LH, but slightly offset in time. Similarities between the two melodies are

so strong that in no instance do the long-short, or short-long, rhythmic proportions of the

66
See Haapamaki, “Order in ‘Désordre’,” 21, for a supporting discussion of canon-like qualities in the
melodies of “Désordre.”
67
Uranker, “Ligeti’s ‘Désordre’,” 24.
68
Palmer, Vision Science, 259–60.

39
melodies differ from one hand to the other. This is discussed in detail by Uranker69 and

Haapamaki.70

Similarities are also present in the contours of the two melodies. Haapamaki

further comments that they appear to be virtual copies of one another, but stated in

different scalar collections, i.e., black-keys vs. white-keys. Thus, the melodies’

similarities in both contour and rhythm will most likely convince the listener that the

lines progress in canon with each other. Because of this, the mental task of tracking the

distance of offset is greatly facilitated.

Second, proximity (section 2.3) is crucial to a listener’s ability to track and

compare the perceived tempo differences. As Example 3.8 demonstrates, the task of

tracking the offset is considerably more difficult when the offset is large, especially

because intermittent accents begin to compete. When the offset is only a factor of one or

two eighth notes, however, it is easy to follow.

Third, while the principle of figure/ground articulation is not always necessary to

enable the perception of polytempo, it is necessary in this etude. Without it, the two

melodies would be indistinguishable from the busy eighth-note background. The

melodies would likely be heard as irregular accents in reference to the eighth-note scalar

ascents, which, according to Uranker, is how performers will probably have to perceive

them.71 Without their “edges,” the melodies could not be perceived as objects that one

could track. Rather, they would assimilate into the homogeneity of the background

accompaniment.

69
Uranker, “Ligeti’s ‘Désordre’,” 9–12.
70
Haapamaki, “Order in ‘Désordre’,” 21.
71
Uranker, “Ligeti’s ‘Désordre’,” 31.

40
 Fourth and finally, the principle of synchrony states that elements occurring at the

same time tend to group. Because the two melodies begin in synchronization, the first

instance of asynchrony (m. 5 where the barlines reflect this misalignment) is unsettling.

A listener may expect (and even hope for) the synchronization to return. This desire

could be thought of as a kind of magnetism. As the two melodies draw farther and farther

apart, the magnetism will resist the direction of separation and try to hold them together,

which can help a listener keep track of the offset. After a certain amount of time, Ligeti

systematically increases the distance of separation, at which point the offset can no longer

be tracked and the magnetism dissolves. But later, the magnetism returns when the

accents grow nearer again. When this happens, a listener may experience the expectation

that the synchronization will return and a certain degree of satisfaction will probably be

felt when it does.72

In sum, in order to ensure that the polytempo illusion is heard by the listener,

Ligeti made use of at least four perceptual tools. (1) He wrote two melodies that are

nearly exact replicas of one another. The contrast between them is mainly a result of their

differing scalar collections (white-key vs. black-key). By doing this, Ligeti maintained

contour and rhythm without oversaturating the texture with too much repetition.73 (2) The

phasing moves slow enough that a listener can track the offset. (3) Ligeti doubled each

melodic note at the octave, reinforced the doublings with accents and dynamics, and

included drastic contrasts between melody and accompaniment, ensuring that the “edges”

72
This bears a certain relation to Joseph Straus’s ideas of “pattern completion” in Stravinksy’s music.
See Joseph Straus, “A Principle of Voice Leading in the Music of Stravinsky,” Music Theory Spectrum 4
(Spring 1982), 106–24.
73
See Meyer, Emotion and Meaning in Music, 135–38, for a discussion of saturation and expectation in
music.

41
could be perceived to distinguish his figures from his grounds. And finally (4), through

the principle of synchrony, he created a magnetism that holds the phasing together for as

long as possible.

3.5 Perception of Large-scale Unity

As described in Chapter 2, a listener has the ability to group compositional

elements across large spans of music and even across entire compositions. To conclude

the analytical portion of this study, three main elements will be discussed that contribute

to the perception of large-scale unity in “Désordre”: (1) the unvaried phrase-group

pattern, (2) the consistent texture, and (3) scalar ascents.

As shown in Example 3.1, once mm. 1–20 establish the original phrase-group

pattern, this pattern is maintained throughout the etude, changing only in terms of rhythm

and transposition. The constant presence of the original pattern acts as a cohesive element

binding the three sections together through similarity.

Three elements of the texture remain constant throughout the composition. First,

the eighth-note pulse never once lets up, even at the climax shown in Example 3.9.

99

97

György Ligeti ÉTUDES POUR PIANO No. 1 “Désordre,” LH mm. 97–100

Copyright © 1986 by Schott Music GmbH & Co. KG
All Rights Reserved.
Used by permission of European American Music Distributors Company,
sole U.S. and Canadian Agent for Schott Music GmbH & Co. KG

Example 3.9 Constancy of the eighth-note pulse.

42
When the LH rests for three eighth notes, the RH relentlessly maintains the pulse.

Second, melodic accents may be found in every measure of the etude. As shown in

Example 3.10, this is even the case at the LH’s mm. 91–96 when it appears that the

background layer has overtaken the melodic foreground.

93

91

György Ligeti ÉTUDES POUR PIANO No. 1 “Désordre,” LH mm. 91–96

Copyright © 1986 by Schott Music GmbH & Co. KG
All Rights Reserved.
Used by permission of European American Music Distributors Company,
sole U.S. and Canadian Agent for Schott Music GmbH & Co. KG

Example 3.10 Foreground occlusion of the background layer.

As described earlier, just the opposite is true. Here, the two melodies of the foreground

layer, which are now progressing at almost one accent per eighth, will likely be perceived

as occluding the background. Third, the scalar ascents in the background layer become

powerfully cohesive elements of the composition when compared to the composition-

length scalar ascent present in the RH phrase groups, as shown in Figure 3.1. The figure

labels the beginning pitch of every phrase group. Those in the RH combine to form a

steady scalar ascent over the course of the entire etude.

It should be noted that despite the general registral descent of the LH phrase-

group patterns over the course of the etude, its background scalar material consistently

ascends. Because of this, it seems safe to assign greater hierarchical significance to the

43
idea of scalar ascent in “Désordre” than to the idea of scalar descent. Over the course of

the etude, the small ascents seem to push little by little upward toward an eventual goal of

C8, the highest note on the piano. The music seems to be pulled upward by a kind of

upside-down musical gravity leading to the upper register of the piano. C8 on the piano

seems like the natural destination for the music, not A0. In a sense, it is as if the scalar

ascents point toward, or rather foreshadow, the ending. Scalar ascents, presented

gradually or quickly, seem to be one of the pillars of the structure of the composition and

are powerful contributors to the overall sense of unity in it.

3.6 Pedagogical Applications in the Composition Studio

It is the author’s hope that these Gestalt principles have helped to clarify how

Ligeti’s etude is organized, and that the techniques presented may also prove beneficial to

other composers. Composers like Ligeti seem to exhibit an innate understanding of how

listeners perceive their music. In 1989, James Guthrie, who appears to have been the first

to formally analyze this etude, supplied clues that “Désordre” might have been inspired

by perceptual phenomena:

The title page of the original manuscript contains descriptions of various

rhythmic dimensions. These descriptions, scribbled in the margins, are
discarded titles: “Étude polyrhythmique,” “Pulsation,” “Stroboscope,”
“Mouvement irregulier,” “Contraction-Dilation,” “Pulsation irregulier,”
“Deplacment,” and “Ordre-Désordre.”74 75

One of these words (stroboscope) names a perceptual device used for creating illusions.

A stroboscope is a mechanism that was used to make something moving fast appear to be

moving slowly, in the same way people can blink their eyes quickly when staring at a

74
Ligeti, Études pour piano - premier livre, facsimile edition 7428 (Mainz: Schott, 1986), cover.
75
James Martin Guthrie, “Études pour piano—premier livre, by György Ligeti, and The Song of Glory,
an Original Opera in One Act,” DMA diss. (Louisiana State University and Agricultural and Mechanical
College, 1989).

44
slowly spinning bicycle wheel and perceive the spokes as moving backward, or even

standing still. This has parallels in “Désordre.” Because of the spacing of the accents on

top of the fast-moving eighth-note runs, the illusion is created that the irregular accents

are actually slower melodies.

To the author’s knowledge, Ligeti never spoke of using Gestalt principles to guide

his choices in this etude. However, his compositional techniques seem to suggest that he

knew how listeners might perceive it. Just as a person can communicate with another

without knowing how to analyze the speech in terms of Chomskian linguistics, Ligeti

does not need to be consciously aware of the Gestalt principles behind his music.

This raises an interesting question. Could composers benefit from an

understanding of Gestalt principles? Diana Deutsch ends her chapter in The Psychology

of Music observing that:

In treatises on music theory, we encounter a number of rules that instruct

the student in the art of composition. Among these are the “law of
stepwise progression,” which states that melodic progression should be by
steps (i.e., a half step or a whole step) rather than by skips (i.e., more than
a whole step) because stepwise progression is considered to be in some
way “stronger” or “more binding.” Another law prohibits the crossing of
voices in counterpoint. What is left unspecified is why these precepts
should be obeyed: It is assumed that the reader will either follow them
uncritically or recognize their validity by introspection. The findings that
we have been reviewing provide such laws with rational bases by
demonstrating the perceptual effects that occur when they are violated.
This in turn enables musicians to make more informed compositional
decisions.76

Just as it has been shown that “Désordre” may be explained in terms of Gestalt principles,

the careful study of perceptual principles can improve a composer’s ability to structure

musical ideas.

76
Diana Deutsch, “Grouping Mechanisms in Music,” in The Psychology of Music, 2nd ed., ed. Diana
Deutsch (San Diego: Academic Press, 1999), 342. Italics added.

45
 Gestalt principles can also help teachers of composition explain to students why

certain compositional choices seem to “work” and why others clearly do not. In a

hypothetical composition lesson, assuming both the teacher and the student are familiar

with Gestalt principles, Example 3.11 represents compositional material that the student

brings to the teacher.

Example 3.11 Hypothetical composition lesson: student’s original.

The teacher might ask if the student’s intention was to have the first measure sound as

notated in Example 3.12.

Example 3.12 Hypothetical composition lesson: teacher’s question.

If the student were to respond that the notes assigned to the LH were meant to be heard as

a pedal-point ostinato, the teacher could point out that, according to the principle of

proximity, the notes in m. 1 would probably fuse into a single line as shown in Example

3.12. The teacher might continue by drawing upon the principle of figure/ground

46
articulation, explaining that unless edges are properly articulated in order to distinguish

the LH notes as a separate line, an ostinato would probably not be perceived until m. 2

when the notes finally cross the trill threshold and are no longer able to be grouped by

proximity. The teacher then might suggest the revision shown in Example 3.13.

Example 3.13 Hypothetical composition lesson: teacher’s suggested revision.

Here the teacher would explain that, according to the principle of continuity, the presence

of the ostinato prior to the entrance of the RH material would more likely cause a listener

to hear the LH content as a line continuing through the RH notes. As is clearly evident in

Example 3.13, the teacher could further suggest a uniform articulation in the LH material,

which, according to the principle of similarity, would strengthen its distinction from the

RH material, thereby better articulating the ostinato’s “edges.”

47
 BIBLIOGRAPHY
Arom, Simha. African Polyphony and Polyrhythm: Musical Structure and Methodology.
Foreword by György Ligeti, xvii–xviii. Cambridge: Cambridge University Press,
1985.

Deutsch, Diana. “Grouping Mechanisms in Music.” In The Psychology of Music, 2nd

edition, edited by Diana Deutsch, 299–348. San Diego: Academic Press, 1999.

Deutsch, Diana, et al. “Psychology of Music.” Grove Music Online. Oxford Music Online.
Oxford University Press, accessed February 4, 2015,
http://www.oxfordmusiconline.com/subscriber/article/grove/music/42574pg1.

Ehrenfels, Christian von. Über Gestaltqualitäten [On Gestalt Qualities]. Leipzig:

Reisland, 1890. Translated by Barry Smith in Foundations of Gestalt Theory, 82–123.
Munich: Philosophia, 1988.

Guthrie, James Martin. “Etudes pour piano—premier livre, by György Ligeti, and The
Song of Glory, an Original Opera in One Act.” DMA diss., Louisiana State University
and Agricultural and Mechanical College, 1989.

Haapamaki, Sampo Elias. “Order in Désordre: Rhythmic and Melodic Structure in

György Ligeti’s Piano Etude No. 1.” DMA diss., Columbia University, 2012.

Hanninen, Dora A. A Theory of Music Analysis: On Segmentation and Associative

Organization. Rochester, NY: University of Rochester Press, 2012.

Hartman, James Barclay. “A Gestalt Theory of Musical Perception.” PhD diss.,

Northwestern University, 1959.

Hodges, Donald A. Handbook of Music Psychology. San Antonio, TX: IMR Press, 1996.

Hofstadter, Douglas R. Metamagical Themas: Questing for the Essence of Mind and
Pattern. New York: Basic Books, 1985.

Howell, Peter, et al., eds. Musical Structure and Cognition. London: Academic Press,
1985.

Kinzler, Hartmuth. “György Ligeti: Decision and Automatism in ‘Désordre,’ 1re étude,
premier livre.” Interface 20 (1991): 89–124.

Koffka, Kurt. Principles of Gestalt Psychology. New York: Harcourt, 1935.

48
Köhler, Wolfgang. Gestalt Psychology. New York: Liverlight, 1929.

———. Gestalt Psychology: An Introduction to New Concepts in Modern Psychology.

New York: Liverlight, 1947.Köhler, Wolfgang. The Task of Gestalt Psychology.
Princeton, NJ: Princeton University Press, 1969.

Lerdahl, Fred, and Ray Jackendoff. A Generative Theory of Tonal Music. Cambridge:
MIT Press, 1983.

Licata, Thomas. “A Rhythmic Analysis of György Ligeti’s Piano Etude ‘Désordre’.” MM

thesis, University of Maryland at College Park, 1992.

———. “On My Études for Piano,” Sonus 9 (Fall 1998): 4–7.

———. Liner notes to György Ligeti, Edition 3: Works for Piano (Etudes and Musica
Ricercata). Sony (DDD 62309), 1997. Pierre-Laurent Aimard, piano.

———. Liner notes to Ligeti: Piano Etudes, Book 1. Erato (ECD 75555), 1990. Pierre-
Laurent Aimard, piano.

Lipscomb, Scott D. “The Cognitive Organization of Musical Sound.” In Handbook of

Music Psychology edited by Donald A. Hodges, 133–76. San Antonio, TX: IMR
Press, 1996.

Meyer, Leonard B. Emotion and Meaning in Music. Chicago: University of Chicago

Press, 1956.

———. Explaining Music: Essays and Explorations. Berkeley: University of California

Press, 1973.

Miller, George A., and George A. Heise. “The Trill Threshold.” Journal of the Acoustical
Society of America 22 (September 1950), 637–38.

Narmour, Eugene. The Analysis and Cognition of Basic Melodic Structures: The
Implication-realization Model. Chicago: University of Chicago Press, 1990.

———. The Analysis and Cognition of Melodic Complexity: The Implication-realization

Model. Chicago: University of Chicago Press, 1992.

Palmer, Stephen E. “Common Region: A New Principle of Perceptual Grouping.”

Cognitive Psychology 24 (July 1992): 436–47.

———. “Hierarchical Structure in Perceptual Representation.” Cognitive Psychology 9,

no. 4 (1977): 441–74.

———. Vision Science: Photons to Phenomenology. Cambridge, MA: MIT Press, 1999.

Sabar, Stephanie. “What’s a Gestalt?” Gestalt Review 17, no. 1 (2013): 6–34, accessed
March 11, 2015. http://www.gisc.org/gestaltreview/documents/whatsagestalt.pdf.
49
Sloboda, John A. The Musical Mind: The Cognitive Psychology of Music. Oxford:
Clarendon Press, 1985.

Stumpf, Carl. Tonpsychologie. 2 vols. Leipzig: Hierzel, 1883–90.

Todorovic, Dejan. “Gestalt Principles.” Scholarpedia.org, 2008, accessed March 11,

2015. doi:10.4249/scholarpedia.5345.

Tsong, Mayron K. “Études pour piano, premier livre of György Ligeti: Studies in
Composition and Pianism.” MM thesis, Rice University, 2001.

Uranker, Mark. “Ligeti’s ‘Désordre’ from Études pour piano—premier livre: An Analysis
and Performance Guide.” MM thesis, California State University, Long Beach, 1998.

Wertheimer, Max. Experimentelle Studien über das Sehen von Bewegung [Experimental
Studies on the Seeing of Motion]. Leipzig: J.A. Barth, 1912.

———. “Gestalt Theory.” In A Source Book of Gestalt Psychology. Edited by Willis D.

Ellis, 1–11. New York: Harcourt, Brace, 1938.

———. “Laws of Organization in Perceptual Forms.” In A Source Book of Gestalt

Psychology. Edited by Willis D. Ellis, 71–88. New York: Harcourt, Brace, 1938,
accessed on March 10, 2015.
http://psychclassics.asu.edu/Wertheimer/Forms/forms.htm.

———. “Principles of Perceptual Organization.” In Readings in Perception, edited by

David C. Beardslee and Michael Wertheimer, 115–35. Princeton, NJ: Van Nostrand,
1958.

50
 APPENDIX A

GESTALT PRINCIPLES OF PERCEPTION SUMMARY

Proximity: Objects near to one another tend to be grouped.

One group of eight Four groups of two

Similarity: Objects seen as similar to one another tend to be grouped together.

One group Three groups

Figure/Ground Articulation: An area surrounded by a dissimilar area tends to be

grouped as a figure upon a ground if the surrounding area is large and homogenous
enough to be perceived as being distinct from the smaller area.

Three figures on a ground

Common Region: Objects within the same boundaries tend to be grouped.

Two groups of two Three groups of two

Continuity: Objects that can be seen as smooth continuations of each other tend to be
grouped.

is not

51