On Playing the Piano


One of the first things a child is taught when learning the piano is to play a C-major scale. We always begin with the simple fingering 1 2 3 1 2 3 4 5, and we are shown how to exploit the special character of the human hand and the mobile thumb by crossing the thumb under the third finger as we play the scale. In other scales (E-flat for example), we cross the thumb even more awkwardly under the fourth finger. This is a basic part of piano technique as it is conceived in conservatories the world over. Nevertheless, it is a mark of the extraordinary variability of approaches to playing the piano that this fundamental practice is not as useful for some pianists as piano teachers think. A pupil of the late Dinu Lipatti, one of the most interesting pianists of this century, told me that Lipatti once remarked: “You know, it has been at least ten years since I last crossed my thumb under the third finger.”

I was pleased to hear this, because I too have discovered that this position is in fact very uncomfortable. Perhaps that is because my thumb is relatively short, not even reaching up to the middle joint of my second finger. I therefore find that wiggling my thumb into an awkward position moves the hand into an inconvenient angle and it is better for me to keep my hand at a steady angle and displace the arm quickly to the right when shifting from the third finger to the thumb, and I have learned how to accomplish this legato.

Everything depends, of course, on the shape of the hand, and it must be stressed that there is no type of hand that is more suited to the piano than another. Josef Hofmann, one of the greatest pianists that I have ever heard—certainly the most remarkable in his control of the widest possible range and variety of tone color—had a hand so small that he could reach no more than the eight notes of an octave, and Steinway built him a special piano in which the ivories were slightly narrower so that he could reach a ninth. His friend Sergei Rachmaninoff had a very large hand, as did Rudolf Serkin, and Sviatoslav Richter could not only reach a twelfth but could play the last chord of the Schumann Toccata all at once without arpeggiation—an effect that would certainly have astonished the composer. My teacher Moriz Rosenthal, famous for his technique, had a small hand with stubby fingers; Vladimir Horowitz’s fingers were exceptionally long, while Robert Casadesus had fingers so thick that he had trouble fitting them in between the black keys. There is no such thing as an ideal pianist’s hand.

In addition, there is no agreement on how to hold the hand at the piano: most children are taught to curve their fingers and place the wrist in a middle position, neither too low nor too high, but of course playing rapid parallel octaves generally demands a higher position for wrist and arm. Horowitz played with his fingers stretched flat, and José Iturbi used to hold his wrist below the level of the keyboard.

This variety is the reason that almost all books on how to play the piano are absurd, and that any dogmatic system of teaching technique is pernicious. (Most pianists, in fact, have to work to some extent in late adolescence to undo the effects of their early instruction and find an idiosyncratic method which suits them personally.) Not only the individual shape of the hand counts but even the shape of the entire body. That is why there is no optimum position for sitting at the piano, in spite of what many pedagogues think. Glenn Gould sat close to the floor, while Artur Rubinstein was almost standing up. It may seem paradoxical that some pianists spend more time choosing a chair for a concert than an instrument; the piano technician at the Festival Hall in London told me that the late Shura Cherkassky decided on the piano he wanted in five minutes, but spent twenty minutes trying out different stools.

The height at which one sits does affect the style of performance. It is difficult, for example, when one is sitting very low, to play bursts of virtuoso octaves fortissimo, as with the following famous passage of the Tchaikovsky concerto in B-flat minor:


That is one aspect of piano technique that Glenn Gould, for example, could not deal with. (A recording engineer at CBS Records told me that when Gould recorded Liszt’s arrangement of Beethoven’s Symphony No. 5, he first recorded some of the virtuoso octaves in the right hand by using both hands and overdubbed the left hand afterward.) Nevertheless, the low seated position enabled Gould to achieve a beautiful technical control of rapid passage-work with different kinds of touch. The way one sits at the keyboard has had an influence on the music that composers write as well as on performance. Ravel also sat very low, for instance, and in his music there are no examples of parallel octaves fortissimo in unison for both hands which are the trademark of so much nineteenth-century virtuosity, particularly the school of Liszt, and which account for the main excitement in the concertos of Tchaikovsky and Rachmaninoff. This Lisztian style of octaves demands a play of the back and shoulder muscles more difficult to manage from a low position. Ravel’s Scarbo, perhaps the greatest tone poem for piano in the Liszt tradition, contains no parallel octaves of this kind, but only octaves alternating between the hands, equally difficult to play but not requiring a raised position of the arms.

These famous Lisztian octave passages bring up an important point: the performance of music is not only an art, but a form of sport, rather like tennis or fencing. This is particularly true of piano music, although the violinist who wields his bow aggressively like a sword has not been unknown to audiences since the early nineteenth century. The triumphant octave effects are not only the greatest crowd-pleasers (when Horowitz was young, members of the audience sometimes stood on their seats to watch him play the octaves in the first and last movements of the Tchaikovsky concerto); they also require special and painful training similar to the hours of exercise to which athletes must submit. Rubinstein, jealous of Horowitz’s glamorous success, remarked sardonically to him, “You have won the octave Olympics.”

It is interesting to note, however, that the most painful of all octave passages to execute are not to be found in Tchaikovsky or Rachmaninoff or even in Liszt, not even in the notorious Sixth Hungarian Rhapsody, but in the accompaniment to Schubert’s Erl-könig. This work must have given trouble even during the composer’s lifetime, when the pianos had a much lighter action, since he wrote out a simplified version of this song—simplified for the pianist, that is. It is, however, the brilliant loud octave passages that audiences wait for, just as they wait for the fouettés of the Black Swan in the second act of Swan Lake, another feat rather more athletic than artistic—although it would be a mistake to deny the dramatic interest of these displays of physical prowess both in piano music and ballet, which have an artistic importance at the very least equivalent to the high-altitude arabesques of the mad Lucia.

The true invention of this kind of octave display—or, at least, the first appearance of a long and relentlessly fortissimo page of unison octaves in both hands—is to be found in the opening movement of Beethoven’s Emperor concerto


It marks a revolution in keyboard sonority, but it is slower than the rapid virtuoso octaves of the early and late Romantics and not particularly hard to play.

It was initially with the generation of composers that followed Beethoven that technique first required the performer to experience physical pain, starting with Liszt and minor composers like Thalberg. Schumann does not use octaves like that, at least not at a speed to cause the pianist any discomfort except for a brief passage in the Humoresk and a much lighter one in the Toccata. Chopin employs them only once and only in the left hand, in the Polonaise in A-flat Major, and he was horrified when he heard a pianist perform this at an unreasonably fast tempo. These famous octaves in the middle section that are popularly thought to represent a cavalry charge are difficult at a rapid speed (at least, one pianist some years ago was rumored to have recorded this piece with her husband playing the left-hand octaves with both hands while she played the right-hand melody).

I have dwelt on this technique, largely outmoded in composition today (the last example that I know in a really fine work is in the final movement of Elliott Carter’s piano sonata of 1947, more than half a century ago), not only because of its popularity, but also because the hours of practicing parallel octaves have been conjectured to be the reason for so many pianists’ having lost the control of the four and fifth fingers of their right hands in recent times. Béla Bartók in the Out of Doors suite made the effect even more athletic by writing parallel ninths. We have seen in our time the equivalent among pianists of the physical injuries experienced by tennis and football players as a result of their professions.

Keyboard performance as a sport as well as an art is already in evidence with the early sonatas of Domenico Scarlatti in the first half of the eighteenth century: here it is the gymnas-tic aspect of the athletic performance rather than physical endurance and strength that played the principal role, with the astonishing leaps of crossing hands and the rapid repeated notes in guitar effects that were Scarlatti’s specialty. With the arrival of the so-called first Viennese style of Haydn and Mozart, there is a loss of virtuosity: only a few concertos of Mozart and one or two piano trios of Haydn have anything remotely to compare with the virtuoso display that we find in Scarlatti and in Bach’s organ toccatas and his Goldberg Variations.

Late in the century there was more concern for writing for the amateur rather than the professional in order to sell sheet music, although Mozart was unable to please his publishers and accommodate himself satisfactorily to the demand for easy music. It was Beethoven who felt that the desires of the amateur—or even of the average professional—were not worth attending to, except when he wrote an easy piece to make a little extra money. Even then, his idea of an easy piece was likely to deter the average amateur, as with the first movement of opus 79, just as Mozart composed one of his hardest movements (D Major, K. 576) under the mistaken impression that he was producing something that could be negotiated by a beginner or an amateur.

It is important to realize that technical difficulty is often essentially expressive. Composers will frequently write in a detail that sounds difficult but is actually easy to play in order to add sentiment: this is particularly interesting when the difficulty is a mimicry of vocal difficulty—and most of the expression of Western instrumental music is an imitation of vocal technique. Perhaps the most obvious device is the imitation of a singer trying to reach a high note, always an expressive effect. In the Intermezzo in A Major by Brahms, opus 18 no. 2, the leap of a seventh from bar 1 to 2 is made to sound more difficult and therefore more expressive by Brahms through the addition of an arpeggiated tenth:


This mimics the difficulty a singer would have reaching a high note.

These considerations should be sufficient to show that music is not just sound or even significant sound. Pianists do not devote their lives to their instrument simply because they like music: that would not be enough to justify a dreary existence of stuffy airplanes, uncomfortable hotel rooms, and the hours trying to get the local piano technician to adjust the soft pedal. There has to be a genuine love simply of the mechanics and difficulties of playing, a physical need for the contact with the keyboard, a love and a need which may be connected with a love of music but are not by any means totally coincident with it. This inexplicable and almost fetishistic need for physical contact with the combination of metal, wood, and ivory (now often plastic) that make up the dinosaur that the concert piano has become is, indeed, conveyed to the audience and becomes necessarily part of the music, just as the audience imagines that the graceful and passionate gyrations of the conductor are an essential component of musical significance.

This aspect can be abused, we may think: the pianist who looks soulfully at the ceiling to indicate the more spiritual moments of lyricism is a comic figure, and so is the performer who throws his hands into the air to indicate a daredevil recklessness—outdone in unintentional comedy by the pianist who gestures wildly only with his right hand, as if afraid that his left hand will not easily find again its place on the ivories. But these are only excesses. For all of us, music is bodily gesture as well as sound, and its primitive connection with dance is never entirely distilled away.

The relation of the performance of music to sound is complex and ambiguous: this is what makes possible Mark Twain’s joke that Wagner is better than he sounds. We need to understand the peculiar nature of the production of piano sonority if we are to elucidate the history of music in Europe and America after 1750. The piano has been the principal tool of composers from that time (less than half a century after its invention) until the present. Piano music is the preeminent field of experimentation.

It has often been noted that when Beethoven struck out on a new path, he began with the piano sonata, then turned to the symphony, and consolidated his experiment with the string quartet. The innovations of the early piano sonatas were carried further in the early symphonies; it was not until he was twenty that he published the string quartets, opus 18. The new turn with the three sonatas for piano, opus 31, was followed by orchestral works: the quartets of opus 59 confirmed the new style. The piano sonatas of opus 106 to 111 mark a radical development, and were succeeded by the Ninth Symphony and the Solemn Mass: quartets were the end of this last change in style.

Most composers, in fact, have followed the same procedure. The first decade of Schumann’s composing life was devoted almost entirely to piano music—in any case that is all he saw fit to publish. Debussy’s first radical attempts at harmony are found in his piano pieces. Schoenberg’s initial move to atonality is initiated by his Three Pieces for Piano, opus 11, that were followed by the Five Pieces for Orchestra and Erwartung. The first attempt at dodecaphonic music is his gavotte from the Suite for Piano, opus 25, and the Variations for Orchestra came soon after. Most striking is the first version of Le Sacre du Printemps of Stravinsky, which is for one piano, four hands. He always intended to orchestrate it, of course, but the four-hand version astonishingly does not sound like a piano reduction of an orchestral work but like a piano piece in its own right, brilliantly conceived for the two performers. (When he later orchestrated it, he made several changes.) It is, in my opinion, Stravinsky’s finest work for piano. When he finished the piano version, he took it to Debussy, and they read it over together at the piano: at the end Debussy got up and left without a word. I wonder what the sight reading sounded like, and what Debussy actually thought at the time. In any case, in the Etude pour les agréments [“for ornaments”], he produced a clear reminiscence of the opening pages of the evocation of the Russian spring night in Le Sacre.


Composing at the piano has had a bad press. Berlioz was proud that he could not play the piano, but only the flute, guitar, and tympani: that saved him, he thought, from the terrible influence of keyboard style. The finer composer, it was felt, should be capable of elaborating the work of music solely in his head, and ought not to need the crutch of trying it out at the keyboard. This is an interesting example of the snobbish idealism that wishes to separate body and mind, with the body considered morally inferior to the less material, more ethereal mind. We have here an interesting aesthetic prejudice: the work of music should be conceived not directly in material sound, but as an abstract form. The realization in sound then oddly becomes secondary.

This prejudice against sound has determined a great part of the aesthetics of performance as it is still conceived today. What is considered primary is a set of pitches which we must imagine as independent of any instrumental color: rhythmic indications are less primary (that is, they can be inflected to some extent according to the personal taste of the performer, with rubato and expressive alterations and deformations), but they are still relatively abstract. 1 Any other indications of the composer for dynamics and phrasing may also be arbitrarily altered by the performer if he thinks he has a better idea—they are thought to have less to do with the abstract structure of the composition and more to do with the realization in sound. The directions of the composer for tempo or for the use of the pedal or for fingering are generally treated as simple suggestions that have little or no authority, although both Beethoven and Chopin, for example, indicated the structure of the phrase by fingering, and the pedal indications and metronome marks were often essential to their conceptions.

Very few pianists pay the slightest attention to Chopin’s pedal indications, and most editors have disregarded them, and still today continue to disregard them: they are regularly infringed or discounted even in the new critical edition that comes from Warsaw. Almost no pianist, however, would dream of changing the pitches of one of Beethoven’s or Chopin’s works (except, of course, when Chopin has provided us with different versions of the same work). The ideas of the composer for the actual realization in sound of his abstract pitches are oddly a secondary matter for most musicians and seem to have very little authority. The followers of the “authenticity” movement have tried to reverse this metaphysical conception, and to make, in an extremely rigid manner, the actual sound the composer would have heard, or might have heard, the primary consideration. They attack the suppleness of the Western tradition with regard to realizing musical sound. We have all forgotten the traditionally lax attitude of a good many periods of Western music with regard to the actual pitch content of a composition, which was much less fixed than we tend to believe.

There may have been a strong moral prejudice against composing at the piano, but such a method of composition has been widely practiced. Haydn always composed at the keyboard. Mozart is traditionally supposed to have composed in his head away from the piano, but in a letter to his father he writes that he is unable to compose at the moment since there is no piano available: “I am now going off to hire a clavier, for until there is one in my room, I cannot live in it, because I have so much to compose and not a minute to be lost.”2 Shortly after Mozart’s death, his biographer, Franz Xaver Niemetschek, wrote about him that he “never touched the piano while writing. When he received the libretto for a vocal composition, he went about for some time, concentrating on it until his imagination was fired. Then he proceeded to work out his ideas at the piano; and only then did he sit down and write” (my italics).3 In this account we see the prejudice against using the piano while composing, and yet an acknowledgment of its fundamental utility.

Beethoven was the great figure who composed away from the piano: that was why his increasing deafness made so little difference to his methods of work. Nevertheless, in his case, his genius at improvising at the piano must have allowed him when young to work out many of his ideas directly at the instrument, and he had for the rest of his life a repertoire of improvised phrases and motifs that served him for decades to come. It is also clear that he hammered out many of his ideas at the instrument for a good part of his life. It was, however, his prestige that made composers after him feel guilty if they were unable to compose without the assistance of the piano. Schumann, in particular, felt ashamed of his reliance on the piano for inspiration.

The utility of the piano for composing was its theoretically neutral tone color: in theory (although not in reality) the tone quality of the bass is the same as the treble. In any case, the change in tone color over the whole range of the piano is, or should be, gradual and continuous (there are breaks, of course, when the notes go from one string in the bass to two and then to three). The monochrome piano might be used therefore just for its arrangements of pitches, and the quality of the sound could—absurdly in many cases—be considered secondary.

Keyboard instruments are the only ones capable of realizing and controlling the entire texture of polyphonic music. The use of a keyboard to work out one’s compositional inspirations dates from the increasing use of a full score, as opposed to separate parts. Exactly when composers used a full score instead of composing polyphonic vocal music in separate parts is an exceedingly complex question; but it is significant that the publication of full scores took a relatively long time to catch on. I presume that composers used scores as a working device to some extent for a considerable time before the publication of scores became widespread, or even before we have any evidence for it.

In any case, when sophisticated chromatic harmony became fashionable and increasingly sought after in the late Renaissance, the suspicion arose that composers were discovering their effects by accident when strumming a keyboard instrument—a little bit as if “the Lost Chord” of Victorian sentimental poetry were found again. The most outlandish chromatic harmonies of the late sixteenth century are in Gesualdo’s madrigals. Alfred Einstein, who claimed that these harmonies induced something like sea-sickness, thought that Gesualdo must have found the modulations at a keyboard. Einstein seemed to feel that this practice was wicked, perhaps even comparable to Gesualdo’s notorious engagement of hired assassins to kill his wife and her lover instead of doing the job honorably himself.

We can find the accusation of composing at the keyboard—which amounts, as I have indicated, to slander, above all when true—much earlier than in the work of a twentieth-century musicologist like Alfred Einstein. During his lifetime Monteverdi was attacked for the same crime; he was said to have discovered his dissonances at the keyboard: it was thought to be sinful above all to work out vocal harmony on the keyboard, principally because instrumental (above all, keyboard) tuning was different from vocal intonation, and the player could not adjust the tuning of the organ in mid-phrase as a singer could inflect the intonation of a note. In his attack on composers like Monteverdi, entitled revealingly Delle imperfezioni della moderna musica [“Of the Imperfections of Modern Music”], the conservative critic Artusi in 1600 attacked what he called the harsh dissonances of Monteverdi. They deceive the ear, he claimed:

These composers…seek only to satisfy the ear and with this aim toil night and day at their instruments to hear the effect which passages so made produce; the poor fellows do not perceive that what the instruments tell them is false and that it is one thing to search with voices and instruments for something pertaining to the harmonic faculty, another to arrive at the exact truth by means of reasons seconded by the ear.

We see here the formation of the prejudice against composition arrived at pragmatically by physically testing the sound instead of mentally planning it by logic, rules, and traditional reason and only using the ear in a secondary role to ratify the results arrived.

It is easy enough to demonstrate that this opposition of body and mind is unrealistic if we consider improvisation. It may not be completely true to say that the fingers of the pianist have a reason of their own that reason knows not of, because improvisation is not exactly unconscious, but it is clear that the fingers develop a partially independent logic which is only afterward ratified by the mind. Perhaps one should add that interpretation, too, works very much like improvisation. In playing a Chopin ballade, an interpretation can be as much an instinctive muscular reaction of the body as a reasoned approach.

That is, in fact, one of the problems of interpretation: a tradition of performance is often a mechanical substitute for thought or inspiration—often a happy substitute, but it becomes a disastrous inhibition when the tradition has degenerated into a lax and unquestioned reminiscence of earlier performance. The unthinking, unplanned performance—and this is an incontrovertible fact of modern concert life—is generally far less spontaneous than one which questions the traditional point of view, in which the performer questions his own instincts. The musician who has surrendered his will to tradition has abandoned the possibility of keeping the tradition alive.

The greatest interaction between keyboard instrument and the process of composition begins with the invention of the pianoforte, the Hammerklavier, in the early eighteenth century. Perhaps the first works written with the knowledge that they would be played on the new invention are the two ricercares (or fugues) from Bach’s Musical Offering. The new invention gradually asserted its supremacy over the harpsichord for use in public halls (there had never been any question of employing the clavichord for this purpose): the organ, ideologically as well as physically tied to the Church, lost its dominance with the diminished interest in ecclesiastical music. Even today the organ is irrevocably tainted with religiosity. The importance of the piano was not, however, simply its greater sonority, or even its ability to realize dynamic nuances. It was, I think, above all, the fact that it was the only instrument that could both realize an entire musical score on its own and at the same time call into play all the muscular effort of the body of the performer. A loud note on the organ does not require any extra effort on the part of the performer, and only a minimal increase for the harpsichord (since coupling the manuals to gain more sonority makes the action slightly more resistant).

Trying for a loud sonority on the clavichord only succeeds in knocking it out of tune: it is capable of a most delicate sophistication, and can achieve a lovely vibrato denied to all the other keyboard instruments, but it calls upon very little corporal force, does not engage the muscles, the body, of the performer. With the piano, every increase of sound is felt by the whole body of the pianist, bringing into play back and shoulder muscles. The performer has to cooperate directly in every crescendo and decrescendo: playing the piano is closer to the origin of music in dance than it is with the earlier keyboards that it superceded. The danger of the piano, and its glory, is that the pianist can feel the music with his whole body without having to listen to it.


With the invention of the piano came the structural use not only of a contrast of dynamics but of a gradual transition from one dynamic level to another. This kind of transition existed, of course, before the second half of the eighteenth century, but it was expressive, not structural—within the interior of the phrase, not as a means of articulating the large form.

Only an articulated contrast of dynamic levels played an important role in structure until the 1760s. It is with the gradual crescendo over a full page or more of the score that the piano came fully into its own. Later, with the invention of the steel frame that made possible the large instruments in the nineteenth century, the athletic element of performance became a basic attraction with what might be called the exhilaration of violence. The pianist produces the greatest fortissimo with an exertion that makes him or her feel as if merged with the instrument, participating directly in the creation of the volume of sound like a string or wind player. The size of the piano, however, so much greater than violin or flute, induces the belief that one is dominating the sound from within, like a singer, as if mastering it were to become part of it; and therefore to a greater extent than any other instrumentalist the pianist enters into the full polyphonic texture of the music.

This sense of physically becoming one with the instrument is the origin of the various delusions about the production of a beautiful sonority. If one leaves out for the moment the use of the sustaining pedal, there is nothing one can do with a piano except play louder and softer, faster and slower. A single note on the piano cannot be played more or less beautifully, only more or less forte or piano. In spite of the beliefs of generations of piano teachers, there is no way of pushing down a key more gracefully that will make the slightest difference to the resulting sound. Inside the piano, the elaborate arrangements of joints and springs will only make the hammer hit the strings with greater or lesser force. The graceful or dramatic movements of the arms and wrists of the performer are simply a form of choreography which has no practical effect on the mechanism of the instrument.

There are indeed different kinds of tonal beauty in piano sound, and each pianist can develop a personal sonority that makes his or her work recognizable, but it does not come from the way any individual sound is produced but from the balance of sound. This balance can be both vertical, as with a chord, and horizontal. The vertical dimension is most easily explained in terms of pure volume of sound. A chord is more or less rich in sonority according to the way one exploits the vibration of the harmonics or the overtones.4 The pianist must rely on aural experience to decide which notes in a chord to emphasize: the vibrations in equal temperament are not the same as those in a system of natural tuning. In natural tuning, for example, a minor seventh is an important component of the overtones of a note, and the major seventh a remote harmonic. In equal temperament this is reversed.

The piano is the only keyboard instrument in which one can grandly vary the effects of the harmonics of a chord at will by balancing the sound in different ways. I remember that when I was eleven years old and started to study with Moriz Rosenthal, I was astonished when I saw him play a chord several times and realized that he could bring out any individual note of that chord and that his way of doing it was invisible. Composers begin to exploit the vibrations of the overtones in keyboard music beginning with the invention of the piano in the early eighteenth century. When the piano became larger in the nineteenth century, this exploitation became more significant with Chopin, Schumann, and Liszt, and the sustaining pedal was used now not as a special effect (as we find in Haydn and still in Beethoven) but as a continuous vibration added to the sound which, with the gradual development of public concerts, helped it carry in public spaces.

Chopin and Schumann, above all, arranged the accompanying harmonies to make the notes of the melody vibrate. Debussy later created extraordinary effects by this means. A beautiful tone color also depends on an intuition of the harmonic significance, and an adjustment for the graceful resolution of the more expressive harmonies (even rhythm enters into the creation of a beautiful sound in this process). What we generally call banging is simply playing the notes of a chord all equally loud with no attempt to adjust for the individual notes within a chord and the way they resonate. Artur Schnabel’s pupils have told me that when he practiced, it was above all to balance the different notes within chords, seeking for a sound that was both singing and expressive.

The vertical beauty of sound—the balance of notes from low to high within a specific harmony—depends to some extent on the horizontal dimension, and the finest pianists make it possible for the listener to trace the expressive movement of the different individual voices within the contrapuntal texture. The glory of the piano is its ability to allow the different voices of the polyphonic structure to interpenetrate each other, shifting the levels from one line to another. The horizontal dimension requires a feeling for the expression latent within the melody and the phrase—and with the bass and inner voices as well. It goes without saying that an accent on a melodic note that sticks out like a sore thumb is immediately felt as ugly. More important is the beauty of sound that comes from recognizing the harmonic meaning within the melody and the curve of its arabesque. In tonal music—at least in what is called the triadic tonality of music from 1600 to 1910—expression is always concentrated in the dissonance. It is the dissonant note within a melody that requires at least a slight emphasis, the resolving consonance a softer release except at an emphatic cadence.

Much of the tonal beauty of the piano depends today upon the pedal, which allows the sympathetic vibrations of the whole instrument to act. Beginning with the 1830s, the almost continuous use of the pedal became the rule in piano playing (although Liszt and his school were more sparing, with a somewhat drier sound). This has had a disastrous effect on the interpretation of Haydn and Beethoven, for whom the pedal was a special effect. Beethoven, in particular, used an alternation of a heavily pedaled sonority as a contrast with dry unpedaled passages.


I have described as mere choreography the gestures that pianists employ in playing, but the choreography has a double practical function. There is the visual effect on the audience which tells the audience what the performer is feeling when the actual sound may be inadequate for that purpose. I do not wish to defend the more extravagant gestures, but I have found that even the most emphatic final cadence will sometimes not convince an audience that the music is finished without some kind of visual indication. Without it, the applause all performers hope for will be late in coming and more tentative than one would like.

The choreography has a purpose for the performer as well, like singing or grunting when performing, and becomes a way of conducting the music or a kind of self-encouragement. Claudio Arrau’s habit, for example, of simulating a vibrato with his hand on the more expressive long notes had no effect on the mechanism inside the instrument, but it was a psychological aid to interpretation and perhaps even convinced members of the audience that the note had extra resonance. The graceful gestures keep the performance relaxed, the way jumping up and down before serving loosens a tennis player’s muscles. In the case of the pianist, too, the gestures, as I have said, become part of the interpretation.

The traditional construction of the keyboard—its arrangements of black and white keys—has had a largely unrecognized influence on the history of harmony, not completely benign, because of most composers’ dependence on the piano for inspiration. The keyboard as it is constituted with its black and white keys was perhaps best fitted for music from 1700 to 1880 and has become more and more awkward since then. Above all in the late eighteenth century, music relied heavily on the transposition of motifs, or even whole sections of a piece, from one tonality to another. Playing a melody in C major feels very different under the hand from playing it in F-sharp major. We are physically in a different realm. Most music of the late 1700s is in tonalities with mainly white keys: as the work of composition progresses we find more and more black keys, and the hand begins to take different positions in order to realize the same phrases.

This means that the center of most large works, where the most important and most distant modulations occur, is different to the ear and to the mind, but in addition the sense of touch perceives the alterations and alienations of the original forms. In the development section of the first movement of Mozart’s Concerto for Piano in B-flat Major, K. 595, for example, the main theme appears in the spectacular series of B minor, C major, C minor, E-flat major, E-flat minor, ending in the conventional G minor. Each playing feels physically different for the hand, and we may say that the harmonic structure is immediately perceived by the muscles of the performer. This is the golden classical age of Western piano music, when conception, hearing, and touch all cooperate. The synthesis of tactile, aural, and intellectual experience would be difficult to repeat.

The keyboard instruments imposed equal temperament, which swept throughout the whole field of music, instrumental as well as vocal. It is sometimes said that Bach did not use fully equal temperament, but only some compromise between equal and natural tuning. However, he transposed his French overture from C minor to B minor (apparently, as Hans Bischoff suggested, to add an H (the German B) to the keys of A B C D E F G in the first two books of his Keyboard Exercises). No two tonalities are farther apart in sound than C and B in any tuning other than equal temperament, so either Bach was using equal temperament or he did not much care what the tuning did to his compositions (perhaps B minor sounded agreeably odd and exotic).

In any case, the different tunings had little effect on his procedures of composition. Beethoven, too, implied a system of equal temperament even in his string quartets, although the string players may have adjusted their pitches for expressive reasons, and still do so. He was certainly capable of writing an A-sharp for the cello together with a B-flat for the violin. Theoretically, the equal temperament imposed by the keyboard instruments reigned supreme.

In the end, equal temperament may be said to have destroyed one of the basic elements of classical eighteenth-century triadic tonality—the distinction between modulation in the dominant, or sharp, direction and modulation in the flat, or subdominant, direction. Modulating in the flat direction brings us from the basic C major, for example, eventually to G-flat major, and modulating in the sharp direction brings us to F-sharp major: in natural tuning, these two keys are different, but they are equated by equal temperament. In the long run, equal temperament obliterated the sense of the direction of modulation.

This sense was always present as an important component of the musical system from 1700 to 1800, and it was scrupulously preserved by Beethoven: the dominant was a source of drama, of raised tension, the subdominant a potential source of lyricism. The distinction was already lost for Schumann and irrelevant to Chopin, and an increasing chromaticism based on equal temperament finally drove it out, in spite of Brahms’s successful reconstruction of some part of the procedure. The symmetrical complexity of a style both diatonic and chromatic was being eroded by the piano. For more than a half-century, a complex network based on mediant relationships (or modulations by thirds instead of fifths) and on a contrast of major and minor modes was an effective substitute. With Verdi and Wagner, the tonal synthesis of an entire long work is no longer enforced, but the unity of long sections of the operas is clearly realized (and in Meistersinger and Parsifal one may even speak of the entire opera).

Modern so-called neotonal music, however, is only a hollow simulacrum of either the eighteenth- or nineteenth-century systems. In today’s neotonal works, the hierarchical richness and complexity of the eighteenth-century structures has completely disappeared; even the major-minor contrasts of nineteenth-century thought have lost their capacity for controlling the large-scale form. Each single phrase may be tonal in today’s new conservative movement, but the tonal structure of an entire piece is either abandoned or given a simplistic form which does not recognize the emotional intensity of full triadic tonality—that required an intensity of listening which most of us are perhaps no longer willing to provide. In Mozart, for example, every harmony is related to the central key, and has a different harmonic significance according to its distance from the center, and the meaning also depends on whether the harmony was reached from the sharp or the flat direction. This was an extraordinarily grand expressive system that depended on a complex hierarchy that has disappeared.

The piano, hero and villain, which helped to confirm the full hierarchical system of tonality and to destroy it from within as well, is itself becoming obsolete. No longer does every middle-class family have a piano, on which the children can pick out tunes and discover a vocation for music.


Playing the Piano’ December 16, 1999

  1. 1

    In painting, the corresponding prejudice claims that contour is primary and color is secondary. Ingres, for example, insisted that one could not judge a picture without seeing the engraving.

  2. 2

    Quoted from Robert L. Marshall (August 1, 1781), Mozart Speaks (Schirmer Books, 1991), p. 24.

  3. 3

    Mozart Speaks, p. 24.

  4. 4

    In “natural” tuning, the notes of the scale are tuned by the harmonics of a fundamental note. The harmonics, as the OED puts it, are “the secondary or subordinate tones” produced by the “parts of a sonorous body (as a string, reed, column of air in a pipe); usually accompanying the primary or fundamental tone produced by the body as a whole.” The principal harmonics of a fundamental tone are those an octave above, plus the twelfth (a transposed fifth), the fifteenth (a transposed octave), and the seventeenth (or transposed third): these are the sounds of a perfect triad. Other higher intervals appear as well. These harmonics determine the sound of an instrument: a clarinet, a flute, and an oboe can all play the same note, but it will sound different because the harmonics have different powers and ratios in the different instruments.

    When, however, the scale is divided, as on the piano, into twelve equal semitones (the system called “equal temperament”), all of the notes differ slightly from the natural harmonics—all of the notes on the piano are, in fact, out of tune with nature. In equal temperament, if we go up a series of fifths (the dominant direction, e.g., C, G, D, A, E, B, F-sharp) or down a series of fifths (the subdominant direction, C, F, B-flat, E-flat, A-flat, D-flat, G-flat), we reach the same note, since G-flat and F-sharp are the same; not so in natural tuning using the perfect fifths of the harmonics, in which case G-flat and F-sharp will turn out to be slightly different and, indeed, incompatible notes.

    Equal temperament is largely imposed by keyboard instruments: any attempt to get the subtle gradations of natural tuning on a piano would end up with a complex keyboard or a double keyboard (double keyboards for this purpose were, in fact, suggested as early as the sixteenth century, but they never caught on). Almost all Western music after 1770, and a great deal of it before that date, is written at least theoretically with equal temperament in mind (although string and wind players can alter the equal temperament for practical expressive purposes). In natural tuning, the dominant and subdominant directions have radically different harmonic meanings and effects. The composers of the late eighteenth century preserved this difference of meaning even within equal temperament, but it was gradually obscured by the hegemony of the new tuning.