notes from the

Boris Berman

pianist s bench

Yale University Press New Haven and London

Published with the assistance of the A. Whitney Griswold and the Frederick W.
Hilles Publication Funds of Yale University.
Copyright 2000 by Yale University. All rights reserved. This book may not
be reproduced, in whole or in part, including illustrations, in any form (beyond
that copying permitted by Sections 107 and 108 of the U.S. Copyright Law and
except by reviewers for the public press), without written permission from the
publishers.
Designed by Nancy Ovedovitz and set in Scala type by The Composing Room of
Michigan, Inc., Grand Rapids, Michigan. Printed in the United States of America
by Sheridan Books, Dexter, Michigan.
Library of Congress Cataloging-in-Publication Data
Berman, Boris.
Notes from the pianists bench / Boris Berman.
p. cm.
Includes bibliographical references and index.
isbn 0-300-08375-0 (alk. paper)
1. PianoInstruction and study. 2. PianoPerformance. I. Title.
MT220 .B15 2000
786.2193dc 00-036514
A catalogue record for this book is available from the British Library.
The paper in this book meets the guidelines for permanence and durability
of the Committee on Production Guidelines for Book Longevity of the Council
on Library Resources.
10 9 8 7 6 5 4 3 2 1

contents

Preface ix
Part I

In the Practice Room

Sound and Touch 3

Technique 24

Articulation and Phrasing 53

Matters of Time 75

5 Pedaling 97
6
Part II

Practicing 112
Shaping Up a Performance

Deciphering the Composers Message 139

Seeing the Big Picture 150

Technique of the Soul 169

10

At the Performance (and Prior to It) 179

11

The Art of Teaching and the Art of Learning 198

Notes 211
About the Author 215
Index 217
Music Publisher Credit 223
vii

preface

This book is a compendium of various thoughts, discussions, and experiences that I have had in the course of my work as a performer and teacher. A
reection of my personal and professional experience, it presents issues that
arise continually in my work with advanced students or that seem relevant to
me as a performer. Much of the book has been written during concert tours
as well as in the wake of numerous lessons. This has inuenced the choice
of the repertoire discussed here.
I do not purport to cover each and every problem that pianists may encounter, nor do I aspire to produce revelations; in fact, I hope that my colleagues will be able to identify with much of what I have to say. I also hope
that they will nd helpful some of the ways I suggest to address familiar
problems. My goal is not to replace the piano instructor. Rather, students
should approach the book in the same way they approach a master classas
an opportunity to be exposed to another point of view to complement the instruction of the teacher. The teacher faces a multitude of problems during
the short time of a lesson. Between correcting notes and rhythms, suggesting a better ngering, and discussing the interpretation of a particular piece,
he can seldom nd time for a general discussion of any of the issues covered
in this book.* Thus, I hope that some of my colleagues may occasionally as*In writing this book I struggled with the grammatical issue of gender parity. I ultimately felt that constant use of he or she makes the text cumbersome. Because

ix

x Preface

sign a chapter as useful supplemental reading for a student to complement

the more concrete work done during the lesson.
This is neither a how-to book nor one conned exclusively to general musical matters. My experiences as both teacher and performer have convinced
me of the fallacy of separating practical and ideal aspects of the art of playing
piano. I strongly believe that these two areas cannot be addressed independently of each other. Technical work should always have a musical goal in
sight, and lofty ideas need to be supported by know-how to be put into practice. In lessons with students, discussion of the stylistic features of a particular composer whose work is being studied can switch to the examination of
the position of the students hand or a search for the best ngering for a
dicult passage. The book reects this interdependence of the practical and
the ideal. Although it is divided into two partsthe rst dealing with more
technical issues and the second with more artistic onesthis separation is
made only for purposes of easy reference. In reality, even an accomplished
artist occasionally revisits technical issues and revises his approach. In this
sense, an inquisitive artist remains a student for a lifetime. By the same
token, a talented student, even one whose expertise is limited, is an artist
and should be treated as such.
I often had diculty deciding which material should go into which chapter, for one cannot really separate articulation from technique, or draw a dividing line between work on sound and work on technique. Cross-references abound between both parts, as well as within them, reecting my
strong conviction that we should mobilize all the tools and approaches at
our disposal to re-create a work of music in all its richness.
Sometimes, for the sake of clarity, issues in the book have been formulated and separated too neatly. The reader will do well to remember that in
real life they are often interwoven with, and infringe on, each other with fascinating complexity. Or, as Goethe said: Grau . . . ist alle Theorie / Und
grn des Lebens goldner Baum. (Grey is all theory / And green of life the

much of the writing reects my own experience, it seemed more natural to use masculine pronouns throughout the book. I hope that my female readers will not mind.

Preface

xi

golden tree). Students naturally look for clear-cut answers to their problems,
and teachers understandably try to respond with catch-all solutions. Yet,
apart from working with young beginners, no recommendation or solution
should be given (or received) with the words always or never in mind. In
the green of the musics golden tree there surely is room for many exceptions to even the wisest rules.
The chapters of this book do not have to be read in order. I encourage pianists to turn to individual chapters that respond to their current needs.
Readers who are not practicing pianists (no pun intended) may be daunted
by the technical discussions in certain chapters of Part 1 (In the Practice
Room). For them, Part 2 (Shaping Up a Performance) and such chapters
of Part I as Articulation and Phrasing and Matters of Time may be more
interesting.
This book could not have been written without the help and encouragement of many people. My sincere thanks go to Michael Friedmann, my
friend and fellow faculty member at Yale School of Music. His advice
throughout every stage of my work has been invaluable. Other colleagues
read early versions of the manuscript and contributed extremely helpful
opinions. Among these individuals are Claude Frank, Peter Frankl, Annie
Frankl, Stephane Lemelin, and Janos Cegledy. Harry Haskell of Yale University Press guided me through the unfamiliar terrain of the publishing world.
Harold Meltzer, Wei-Yi Yang, Dmitri Novgorodsky, and my meticulous editor Jerey Schier all helped me to prepare the manuscript for publication.
Harold Shapero produced photos, Leora Zimmer formatted music examples. I gratefully acknowledge the assistance of the Griswold Fund and the
Frederick W. Hilles Publication Fund of Yale University.
Finally, I thank most aectionately my wife, Zina, and my children, Ilan
and Daniella (my rst reader), for their constant encouragement, particularly when the task seemed to be too immense and daunting. I doubt that the
book could have been written without their support.

PART I

in the practice room

sound and touch

Many issues are important in shaping the pianists skills. Technique, rhythm,
memorization, and repertoire are among them and will all be discussed in
this book. I would like to begin, though, with a topic that is frequently neglected by teachers and students or that receives only perfunctory attention:
sound. For music, this omission is as strange as ignoring color in visual arts,
or body movement in acting. Sound production should be considered part of
technique in a broader sense, for technique is much more than the ability to
play notes rapidly and evenly.
On the rare occasion when sound production is discussed, it is often reduced to such platitudes as It must sound beautiful, Sing! or Change
the color. Instructors rarely give advice on how to achieve a beautiful sound,
what to do with the hands or arms to make the piano sing, or what one needs
to do physically to create the sensation of a change in color. I believe that the
teacher must be specic to meet the needs of students who seek more practical guidance on these matters. Over the years, I have developed a way to
deal with this issue. Before presenting it here, I would like to oer a few
caveats:
1. Although I nd it actually quite easy to teach the basics of sound production, these skills usually do not stick to one who is indierent to quality
of piano tone, or to one whose ears do not crave a particular kind of sound.
In short, you cannot rene your touch without rening your ear. I am refer-

In the Practice Room

ring to two kinds of musical ears. One is the subjective ear, the pianists
image of the kind of sound he would like to produce. The more specic the
image, the better the results will be. The other is the objective ear, which
refers to the musicians ability to monitor the sound that actually comes
from under his ngers. Objective listening is a perennial goal, a life-long
battle, for a musician always tries to listen objectively to his own playing but
never fully succeeds. The pianist cannot do meaningful work without learning to listen intently and tirelessly to every sound he produces on the piano
(more about this in the chapter on practicing).
2. Often overlooked is the need to work on an instrument that responds
suciently to the nuances of touch. (No electronic keyboard will do, Im
afraid.) Chopin apparently had this opinion because, according to his student
Karol Mikuli, in the masters house the pupil played always on a magnicent
concert grand, and it was his duty to practice only on best quality instruments.1 As Russian pianist and writer Grigory Kogan put it: The pianist
must be able to play on any piano, but he must practice only on a good one.2
3. The pianist may be tempted to look for sound of absolute beauty that ts
all occasions. I often tell my students that there is no such thing as beautiful
sound; but there is sound appropriate to a particular style, piece, or passage.
(To be sure, there is such a thing as ugly sound, and the pianist should know
how to avoid producing it.) Sound that suits Rachmaninov would feel out of
place in Mozart, and vice versa. In fact, sound can and should be used as a tool
of stylistic denition. Stylistic awareness, expressed in the choice of tempo,
rhythm, phrasing, and articulation that the performer considers appropriate
to the style of a work, should incorporate the notion of a proper sound.
4. Even a two-year-old can produce the right sound occasionally, but it
will be a sound, a single note. Only a well-trained pianist can produce a second sound to perfectly match the qualities of the rst. It is crucial for
pianists to have the ability to sustain a certain type of sound for the length of
a passage or a phrase and to change it at will.
When pianists talk about beautiful sound, they usually mean a singing,
long-lasting tone that reveals as little as possible of the pianos inherently

Sound and Touch

percussive nature. Even in the relatively infrequent instances when composers highlight the instruments percussiveness (examples that come to
mind are Bartks First Concerto and Stravinskys Les Noces), the pianist
should not be indierent to the quality of sound; he should aspire to emulate
the brassy resonance of a gong or the powerful combination of dryness and
resonance of African drums, rather than a clatter of kitchen pots.
Each professional pianist has (or should have) endured long and often
frustrating hours in the practice room looking for his own way of producing
this beautiful, long-lasting sound. We are all dierent physically, and for this
reason every pianist develops his own strategy. The multitude of approaches
and their combinations, however, can be reduced to two generic types. My
late teacher the wonderful Russian pianist and pedagogue Lev Oborin
dened the polarity of these physical approaches as sostenuto and leggiero. I
prefer using the English words in and out. Both of these ways of playing,
as we will see, share a common goal: to mask the most treacherous, dangerously telling momentthat of the actual attack, when the hammer hits the
string.
Eloquent imagery has been used to describe the in kind of sound production. Rachmaninov talked about ngers growing roots in the keyboard.
Joseph Homann said that the sound should be produced as if there were a
very ripe strawberry sitting on a key and you had to push through it. These
images imply two important features of the in type. One is the deliberate
speed of the process: the slow pace at which the roots grow, and the unhurried tempo at which the strawberry must be penetrated to avoid a messy keyboard. The other is the continuous quality of the process; the roots grow
without stopping at a certain point. The in type, then, is based on a slow
immersion in the keyboard: the action continues even after the sound has
been produced, as if the moment of attack were ignored. The weight
brought into the key stays there without being released; it is then poured
into the next note of the phrase.
The out type is quite the opposite. The sound is produced by a quick
stroke, as if the nger left the key even before the sound could be heard. Obviously, if the note has to be sustained or connected to the next, the nger

6 In the Practice Room

does not leave the key. But most of the weight is gone; only a vestige remains
to hold the key down. This type of action is similar to playing the harp (is not
the piano essentially a horizontally placed harp?). The harpist strikes and
then escapes the strings almost before the notes are produced, otherwise the
sound is dampened by the ngers. Or think about the way the percussionist
plays the tam-tam: the performer never leaves the mallet pressed against the
instrument. After striking the tam-tam he pulls the mallet out of the way,
allowing the instrument to resonate without being obstructed.
Playing this way, the pianist should not direct the movement downward
into the keyboard. Rather, he should employ a circular (tangential) motion,
as if passing through the key but not stopping. Once again, the action is similar to the circular motion used to pluck the harp, strike the tam-tam, crash a
pair of cymbalsor play baseball or tennis. In the sports analogy the tennis
racquet and baseball bat pass the point at which they strike the ball, continuing in a circular motion (called follow-through). The pianist directs the motion toward himself as if he were grabbing the sound from the keyboard
and bringing it out.*
Some pianists prefer to move the hand forward rather than toward themselves. Konrad Wol describes Artur Schnabels playing in this manner.3
Schnabel may have learned it from Theodore Leschetitzky, whose other student (and one-time wife), the legendary Russian pianist and teacher Anna
Esipova, recommended: Place your hand on the keys, form the chord, and
move the hand as if pushing a drawer into a desk.4 What is important to me
is that both this motion and the out way as discussed above do not aim vertically downward but touch the key at an angle. Both motions can be described as caressing; they both allow the nger to glide along the key.** I
nd it more practical to move the hands toward me instead of away from the
body because in the latter case the the pianos name-board restricts the
*The circular movement approach is executed by either nger, hand, or forearm,
depending on which lever the pianist decides to employ at any particular time. The
reasons for choosing a particular lever are discussed below.
**The gliding movement can be very helpful in achieving smoothness of legato,
though strictly speaking it is irrelevant for producing the sound.

Sound and Touch 7

movement on the far end. More room exists between the keyboard and the
pianists body.)
These two types of sound production, the in and the out, almost never
appear in their pure form; rather, there are countless combinations of the
two. Dierent national schools have shown preferences for one or the other:
pianists of the Russian school have favored the in approach, while those of
French or German musical descent seem to have preferred the out way of
playing. (I use the perfect tense here because the current cross-fertilization
of traditions has left hardly any national school untouched by other inuences.)
For me, it is important to use dierent kinds of sound for dierent types
of music. A work of introverted character, such as Brahmss Intermezzo
op. 119, no. 1 (Ex. 1.1), may benet from an in approach, while more outspoken, extroverted music, like the beginning of Chopins C-Minor Nocturne (Ex. 1.2), asks for the out stroke. Many pieces can be presented
equally convincingly using either of these approachesChopins F-sharp
Major Nocturne, for example (Ex. 1.3). The pianist who is conversant with
both may choose the one that seems more appropriate.

Ex. 1.1 Brahms, Intermezzo in B Minor, op. 119, no. 1

Lento

mezza voce

Ex. 1.2 Chopin, Nocturne in C Minor, op. 48, no. 1

8 In the Practice Room

Ex. 1.3 Chopin, Nocturne in F-sharp Major, op. 15, no. 2

At times the pianist may wish to imitate the sound of other instruments,
especially when trying to realize an orchestral reduction at the piano. The
airy sound of the French horn, as in the beginning of the orchestral part of
the B-at Major Concerto by Brahms (Ex. 1.4), will be better rendered by the
out stroke. The warmth of the strings in the excerpt from Liszts First Concerto (Ex. 1.5), on the other hand, calls for the in approach.
3

Ex. 1.4 Brahms, Concerto no. 2 in B-at Major, op. 83, mvt. 1

Ex. 1.5 Liszt, Concerto no. 1 in E-at Major

So far, we have contemplated ways of producing the sound while dealing

with relatively soft music. For loud playing, I am afraid, the in approach almost never succeeds. Imagine a long crescendo: as we increase the speed of
immersion into the key to produce the louder sound, the time between the
moment of attack and the imaginary goal of the movement becomes shorter
and shorter, until the two coincide. As a result, instead of masking the moment of attack we are highlighting it; the sound becomes unpleasantly hard
and harsh, and immersion turns into pressure.

Sound and Touch

Allegro deciso

Ped.

Ex. 1.6 Liszt, Concerto no. 2 in A Major

My solution to avoid harshness of sound is to switch to the out way using a hit-and-run approach. The louder the dynamic level is, the faster the
movement should be. The chords in Ex. 1.6, for instance, are played as if being torn from the piano. (Naturally, the pedal will prolong their duration and
enhance the resonance.) If the notes must be sustained, the ngers do not
leave the keys, but the weight of the hands is used for the attack only, and
they do not sink into the keys even for a moment. A good example of this approach is the rst subject of Beethovens Fourth Concerto (rst movement)
as it appears forte in the beginning of the recapitulation (Ex. 1.7).

Ex. 1.7 Beethoven, Concerto no. 4 in G Major, op. 58, mvt. 1

Earlier, I was talking about the need for the pianist to have sound imagination, the rened subjective ear. This is not enough, however, for the performer must also possess the technical ability to realize the sonorities he
hears in his head. The pianist needs to know what physical actions inuence
sound and in what way. Here are several variables that are used in both in
and out types of sound production. Some also aect other aspects of playing, such as articulation or velocity. (The real life of a practicing pianist can-

10

In the Practice Room

not be neatly compartmentalized.) But rst I would like to mention one

physical constant that is indispensable for producing rich, nuanced tone:
the exible wrist. Josef Lhevinne compared its role to that of shock absorbers
in a car.5 The wrist cushions the sound and absorbs the excess force. (Frequently the pianist uses the elbow as an additional shock absorber, as
described in the next chapter.)
1. Weight. The more weight that is applied to the key, the fuller (and/or
louder) the sound. The pianist needs to be able to use the full weight of his
ngers, hand, forearm, and upper arm. Equally important is knowing how
not to use weight when a lighter sonority is required. I often ask my students
to experiment with making their ngers heavy by letting the weight of the
bigger joints pour into the ngers (not to be confused with applying pressure), and then gradually withdrawing the weight to regain the lightness.
(For the latter, imagine a vacuum cleaner being applied to your shoulder
blade, sucking the weight from the hand. This image can be particularly
helpful for larger-sized pianists who nd it dicult to prevent the weight of
their arms from participating when the music requires lightness of touch.)
These experiments are important for learning one of the most necessary
skills for the pianist: to let in just as much or as little of the weight as is
needed for a particular passage.
Pianists who possess a delicate physique sometimes feel that they cannot
muster enough weight to play a loud passage. They try to compensate by
pressing into the key using the in touch, which is generally not suitable for
forte, as discussed above. The pressure usually produces a hard, forced
sound. In such cases I strongly recommend that the pianist resist the temptation to apply additional pressure. Instead, I would switch to the out
approach and increase the speed of the stroke.
2. Mass. This variable concerns how much of the body is involved in
sound production. The sound can be produced with the nger alone, or with
the nger supported by the hand, or with nger supported by the hand plus
the forearm, or with the nger supported by the hand plus the forearm plus
the upper arm. The bigger the participating jointthat is, the greater the
massthe fuller the sound.

Sound and Touch 11

When we want to increase the volume we activate the bigger joints. When
we want to stay at the same dynamic level, however, and at the same time,
achieve a fuller sound, we seek the support of the bigger joints, rather than
their overt participation. To help my students accomplish this goal, I often
mention that they should develop the feeling of a long nger or imagine
that all the juices from the arm are owing into the nger. Another useful
image is that of a long neck to help feel uninterrupted succession of muscles from behind ones ear to the neck, to the upper arm and so forth down
to the nger tip. (Compare this with the extension principle discussed in
the next chapter.) When the pianist feels the need to add air to his sound,
which he experiences as excessively thin and bony, a exible elbow can be
particularly helpful. Imagining a parachute attached to ones elbow can be
useful.
It is important to distinguish between mass and weight. One can use the
whole arm and still make a very light sound, or use just the nger, which
would have the weight of bigger joints poured into it.
3. Speed. The speed with which the nger strikes the key contributes to
changes not only in volume but also in articulation. Elementary physics
teaches us that speed can be developed only over a certain distance. This
means that a nger needs to fall from a certain height, unless we want to
play very softly. If the pianist tries to play loudly while his ngers remain
glued to the surface of the key, all he can do is to push the key down, producing a very pressured sound. The greater the speed we want to develop, the
higher we should position the nger. At a certain point, however, the pianist
cannot rely merely on the energy of the falling nger; he must involve the
hand as well. (This is discussed further in the chapter on technique).
In my discussion of the out way of playing, I stated that its basic movement must be fast. Yet gradations of the speed can add much variety to this
type of touch, and in my lessons I use dierent images to underline the
dierences. I talk about taking the sound out of the piano as opposed to
pulling it or, on the other end of the spectrum, tearing it out as opposed
to plucking.
Speed not only can compensate for insucient weight, it can also be in-

12

In the Practice Room

terchangeable with mass. A similar dynamic level can be achieved by using

a larger joint with lesser speed, or a smaller joint with greater speed. The decision regarding the course of action to take depends on the pianists feeling
of the sound that is best suited or most appropriate stylistically for a particular piece of music. A mezzo-forte singing line in Mozart, for example, may
require using ngers to activate keys fairly quickly, probably supported by
the hand, possibly with just a little participation of the forearm. In contrast,
a similar dynamic level in a Rachmaninov work will be best produced by
bringing the weight of the arm into the key with a relatively slow speed. The
lean sound appropriate for Mozart is quite dierent from the full or thick
sound suitable for Rachmaninov.
4. Perception of depth. More than the other variables, this depends on the
pianists imagination, because the depth of the key has very little leeway per
se. Yet every properly trained pianist is able to hear the dierence between
deep and shallow touch. One usually plays deep into the key to achieve a
singing tone. (The depth should not be exaggerated, though, as it invites
pressure, which in turn produces a forced, strangulated tone. I have seen
pianists who played as if they were intent on making a hole in the bottom of
the keyboard.)
A common mistake is to aim deep while playing loudly, but to use a much
shallower stroke while playing softly. For developing the deep touch in
piano, Esipovas advice is very useful: First practice the phrase (which is to
be played pianissimo) slowly, feeling the bottom of the key and in the dynamics of mezzo-forte. Afterward repeat it with the same feeling of depth,
but very softly. Keep switching from one way of playing to the other.6
The depth of the touch should remain the same for the duration of a
phrase or passage. Inconsistent touch, when adjacent notes are produced in
dierent layers of depth, is clearly noticeable to the trained ear and testies
to the pianists poor sound control. However, not all lines in the musical texture need to sing; for those that do not a shallower touch may be more appropriate. For each particular line or phrase we aim for a certain depth in the
keys. It is crucial to sustain this depth until the nature of the material
changes. Very often we have two or more elements, played simultaneously
or in quick succession, each performed with a dierent depth. Obviously,

Sound and Touch

13

this situation requires great ability to control the touch, especially when
these dierent elements appear in the part of the same hand. (Dierentiation in articulation, very much linked with the subject of depth, is discussed
in the chapter Articulation and Phrasing.)
When the out way of playing is used, the dierence in depth can be as
eective as with the in touch. For the out approach I encourage students
to imagine taking (or pulling, tearing, and so on, as described above) the
sound from a deeper or shallower layer within the keyboard.
5. The shape of the fingers. Josef Lhevinne observed: It is almost an axiom
to say that the smaller the surface of the rst joint of the nger touching the
key, the harder and blunter the tone; the larger the surface, the more ringing
and singing the tone.7 The dierence in sound is made by touching the key
with either the eshier part of the nger or with the tip. To play music that
requires clarity of articulation, the pianist often curves his ngers so that the
smallest joint is almost perpendicular to the keys. On the other hand, eorts
to create a singing sound of great warmth will succeed if the ngers assume
a atter position, shaping the phrase as if molding warm clay. To avoid muscular tension, ngers should never be outstretched more than is natural.
The physiologically correct position can be checked by letting the arm hang
freely alongside the body; the ngers will naturally assume their proper
curved position (Fig. 1.1).
Whether one uses atter or more rounded ngers, the sensitivity of the
ngertips is of supreme importance. The tips of the ngers have to be
alert and active even in the softest and most delicate passages. To quote
Natan Perelman, the doyen of piano professors in St. Petersburg, Russia,
The soul of pianists is located in their ngertips.8
Having described these concrete tools of piano playing, I should emphasize that they serve to achieve the most important goal: the ability to create an
illusion. What one does is innitely less important than the sound that
emerges from the instrument. Thus, for instance, it is not always necessary
to play physically legato to create the legato sound. In fact, eorts to connect
notes physically may make the melodic line less smooth than by playing it
non legato (naturally, with the help of pedaling). In the passage from De-

14

In the Practice Room

Fig. 1.1 Naturally curved position

of the hand

bussys La lle aux cheveux de lin (Ex. 1.8), the pianist who tries to connect
the chords with his ngers can easily become stuck in the keyboard; the
lightly gliding melodic line will be better served by gentle non legato playing, assisted by a frequently changed pedal. The key to success in this and
similar cases is an ideal matching of the attacks of successive chords. Here,
too, very close listening is crucial.
(sans lourdeur)

Ex. 1.8 Debussy, La lle aux cheveux de lin, from Preludes, book 1

Cdez

Sound and Touch

15

Control over dynamics is an important manifestation of cultivated touch.

Heightened awareness of dierentiation of dynamics can be traced to the
piano literature of the early twentieth century. Baroque music used just two
indications, forte and piano, almost exclusively (only occasionally can we
nd examples of pianissimo in J.S. Bachs music). In keyboard music, these
dynamic markings served merely as an indication for a louder or softer keyboard on harpsichord or organ. Classical period composers used a greater
number of dynamic markings; Romantic composers used still more. Even
so, these indications were not meant to be interpreted literally. In fact, very
often we would play an expressive passage in Romantic music in the mezzoforte dynamic range, though the composers indication may be piano.
It was Claude Debussy who revolutionized our perception of the scale of
dynamics. In the development of piano playing his importance may be compared only to that of Liszt. Whereas Liszt reached new horizons in matters
of velocity, Debussy raised the level of awareness of touch control to an unprecedented height. His indications require precise changes of minute dynamic gradations. See, for example, the end of Pagodes from Estampes, in
which the dynamic indications are ff, dim, p, dim, pp, pi pp, encore plus pp,
and aussi pp que possible. Even more demanding are the occurrences when
he uses several layers of texture within the same dynamic level, each requiring a touch and articulation of its own (Ex. 1.9). Touch control was carried
further in serial music, which assigns every note its own indication of
dynamics and articulation, as in the groundbreaking Modes de valeurs et
dintensits by Messiaen (Ex. 1.10).
I am convinced that a contemporary pianist simply cannot function without acquiring the precision of control over the scale of dynamics. We all
know that every pianists dynamic scale is dierent. Besides, when a pianist
switches from one instrument to another, from one hall to another, he adjusts the dynamics accordingly. And yet, within the conditions of a specic
performance, one is frequently required to establish a more or less absolute
dynamic scale. To clarify this concept, I often ask students to play a short
melodic phrase repeatedly, changing the dynamics on my request. Starting
with mf, for example, we proceed to pp, to f, to mp, to p, and then back to mf.

16

In the Practice Room

un peu en dehors
Lent

doucement sonore

3
3

m.g.

3
3

3
3

Ex. 1.9 Debussy, Les cloches travers les feuilles, from Images, book 2
Modr
8

Ex. 1.10 Messiaen, Modes de valeurs et dintensits

When the student returns to mf, I insist that it sound neither louder nor
softer than it did before. The objective of this exercise is to establish a scale
of dynamics to teach the pianist, for example, that mf is not merely somewhat louder than p, but exactly two steps above it.*
*Establishing the dynamic scale does not absolve the pianist from mastering
dierences of sound within each category. Forte pesante should sound dierent from
forte leggiero; piano espressivo from piano misterioso.

Sound and Touch

17

Playing two or more notes simultaneously, we must prioritize, or voice,

the dynamics between them, even when the composer is not making
specic demands. The more notes that are struck simultaneously, the more
important the issue of voicing becomes, particularly in loud playing. Nothing on the piano sounds more vulgar than a loud chord in which all notes
shout indiscriminately. Each chord must have a certain leading note, while
the other tones in the chord should be voiced down. To accomplish this, the
pianist needs to decide which sound in the chord (or which line in the
chordal progression) to highlight. Then he must have enough control over
his ngers and enough nger independence to execute it. The exercise in
Ex. 1.11 helps the fingers to develop these qualities. The white notes of the
chords shown in the example are to be played forte, the rest of the notes piano. Practice with both hands playing separately as well as together. Later
choose dierent, less convenient chords.

Ex. 1.11

In the past several hundred years the leading melodic line was entrusted
with increasing frequency to the top voice. As a result, our ear habitually
craves the clarity of the top line. In chordal textures, the highest note of the
chord is almost always the melodic one and needs to be highlighted.
Insucient voicing of the top notes of chords (or of octaves) makes them
sound blind, lacking prole and clarity. Sometimes we voice to reveal a
hidden melody, like the bass line of a passage from Beethovens Eroica
Variations (Ex. 1.12). On other occasions the pianist achieves dierent
shades of color by choosing to highlight certain notes within the chord, or a
3

Ex. 1.12 Beethoven, 15 Variations with a Fugue (Eroica), op. 35, nale

18 In the Practice Room

certain line in the chordal progression. In the beginning of Debussys La

cathdrale engloutie the dierent choice in voicing indicated by  and 
in the two examples changes the color of the passage (Ex. 1.13).
When a melodic line is doubled in one or two octaves, whether played by
one or two hands, we usually bring out the upper line. However, changing
the balance in favor of the lower voice can be very eective in giving a darker
color to the passage. In all matters of voicing, one should take care to maintain consistency of balance between lines or within chords.
A related issue is the balance among dierent elements of a texture, for
instance among melody, accompaniment, and secondary voices. Even the
less-complicated texture, for instance, of many Chopin nocturnes, valses,
or mazurkas requires careful treatment. What many inexperienced players
regard as two-part writing (melody and accompaniment) is, in fact, threepart writing: melody, low bass, and chords in the middle register, or the harmonic stung (Ex. 1.14). Beautiful sonority is achieved only if the bass

a)

+
+

+
+

b)

+
+

+
+

+
+

Ex. 1.13 Debussy, La cathdrale engloutie from Preludes, book 1

Sound and Touch

19

Ex. 1.14 Chopin, Nocturne in F Minor, op. 55, no. 1

has a resonant, airy quality and the stung is played very lightly and sensitively.* Balance among these components will determine the way one uses
the pedal (see the chapter Pedaling). The separation of sonorities among
these strata will help the performer to reveal the inner life of each of these
layers and enhance the clarity of the voice leading.
In accompaniment presented in the form of rolled arpeggios (Ex. 1.15),
one should not play each note of an arpeggio with the same degree of loud-

Andante

Ex. 1.15 Chopin, Nocturne in E Minor, op. 72, no. 1

*In such instances, the bass is usually played out from a deep layer of the keyboard. A comparison with the resonant pizzicato of the double bass may be helpful.
Clearly, the loudness as well as the speed of the stroke change depending on the
circumstances.

20

In the Practice Room

ness. Usually, after the full and resonant low bass the few notes that follow it
are played softer, gradually emerging with a slight dynamic increase.
I have paid so much attention to the issue of producing the sound, of the
beginning of the tone, that it threatens to overshadow the even more important factor, that of the follow-up, of listening through the note. Phrases
cannot sing without the pianist listening between the notes. Heinrich
Neuhaus suggests a very good exercise to become aware of this concept (Ex.
1.16a).9 It can be made more complicated by including additional dynamic
gradations (Ex. 1.16b). It is essential to match the dynamic level of the new
sound to that of the preceding note, not at the point of the attack but at the
very end of it.* If a pianist is using the in touch while listening between the
notes, the pressure his nger exerts on the key will keep diminishing as
the ear follows the decay of the sound, so that it is ready to match the touch
required to produce the dynamics of the ensuing note.
a)

b)

Ex. 1.16

Yet a pianist cannot be too dependent on the natural decay of the piano
sound, otherwise all phrases will have to be played diminuendo. If his intention is to make a crescendo, he does not follow what he hears but, rather,
what he imagines. (In the in sound the weight of the nger staying on the
key will grow accordingly.) This technique may help to make credible
crescendi on a sustained note or chord, both of which are found in
*Neuhaus also oers the very good suggestion of practicing a melodic passage
much slower than it is going to be played, as if in slow motion.

Sound and Touch 21

Beethovens works and which, strictly speaking, are not performable on the
piano.
At any point in this chapter someone could ask, Are all these minute details really noticeable to most listeners? My answer would have to be, Probably not. The ear of somebody who works constantly with the nuances of
the sound of his instrument becomes much more discerning than that of an
outsider, even when the outsider is himself a musician. I remember being
present at a lesson where a very good percussion player demonstrated to a
student dierences in sound produced by dierent strokes on the triangle.
To my great embarrassment, I could not tell much of a dierence, but both
the teacher and the student clearly recognized it. When working with string
players, I often witness their deliberations (and arguments) over whether to
play a certain note on the A string or the D string. To me, the dierence does
not seem signicant.
Are musicians splitting hairs to worry about such matters? For me, the
importance of this work lies as much in the practical result it achieves as in
the dedication to the music it manifests. Together with Arnold Schoenberg,
I marvel: How high the development of spirit that could nd pleasure in
such subtle things!10 I am also reminded of the story about Michelangelo,
working with great care on the back of one of his sculptures that was to be
placed in a corner of a church. He was asked why he spent so much eort on
the part that no one would see. His reply was, God will see it.
From time to time, a composer reveals that he was thinking in orchestral
terms: such indications as quasi tromboni or quasi corni appear in the
score. Even without such an explicit suggestion, the pianist often feels that
a certain phrase would sound wonderful being played, for example, by the
oboe or the cello. He then may wish to try to create the sound evocative of
that particular instrument. Performing transcriptions of orchestral works,
the pianist often nds that creating the illusion of a specic instruments
sonority is indispensable. In the last four bars of The Young Juliet,
Prokoevs own transcription from his ballet Romeo and Juliet (Ex. 1.17), a

22

In the Practice Room

Meno mosso

assai rit.

tranquillo
Harp

Lento

Saxophone

Ex. 1.17 Prokoev, The Young Juliet from Ten Pieces from Romeo and Juliet, op. 75

pianist cannot adequately represent the sonority of the original without trying to imitate its orchestration. The rst scale in the left hand is scored for
the harp; it is answered by the saxophone. In the orchestra, the two scales
sound dramatically dierent; the pianist is challenged to re-create this
dierence.
I can anticipate a sober critic saying here: No matter how hard you try, the
piano will always sound like a piano. I agree entirely with this statement but
nd these attempts extremely important nevertheless, as they will yield a
much richer and more varied piano sonority. Or, as Perelman put it: The
best results are reached when one demands an impossible thing of the student: vibrato! pizzicato! tutti! French horn! drum! sing! quartet! clarinet! . . .
until the student starts getting hallucinations of sound color.11
It is dicult to nd practical advice on how to make the piano sound like
other instruments. In fact, I have encountered only one such example: Alfred Brendels essay, Turning the Piano into an Orchestra (Liszts transcriptions and paraphrases), published in his book Musical Thoughts and Afterthoughts.12 Here are some of his suggestions:
The sound of the oboe I achieve with rounded, hooked-under, and, as it
were, bony ngers, in poco legato. . . .

Sound and Touch

23

The ute . . . whenever possible, I play every note with the help of a separate arm movement. . . .
The bassoon . . . the touch is nger-staccato. . . .
The noble, full, somewhat veiled, romantic sound of the horn demands a loose arm and a exible wrist. Although its dynamics extend
from pp to f, the sordino pedal should always be used. . . .
Do not forget that the harp is a plucked instrument! The pianist should
play harp notes with round, tensed ngerssempre poco staccato
within the sustained pedal. In rapid, sharply ripped-o arpeggios, the
nger-play is assisted by movements of the wrist.
I suggest that every pianist try to apply Brendels recommendations,
though I predict that the rate of success will not be very high. This does not
mean that the suggestions of this wonderful and experienced pianist are
wrong or imprecise. Rather, this advice may not work as successfully for
anyone else because the delicate area of piano coloring depends so much on
individual physique or on each pianists innate approach to playing.
Still, it is extremely important for the teacher to be able to suggest to the
student what to do physically if he wants to imitate a bassoon, harp, or other
instrument. Even more important is to encourage students to seek their
own approach, to nd what would work best for each person. I see it as
building ones vocabulary of physical motions, a personal pianistic toolbox. The larger the vocabulary, the more eloquent our musical speech
becomes; the better equipped the toolbox, the more eective and ecient
the pianists work will be. And the better the pianist controls sound, the
more eectively he is able to communicate musical expression to an audience.