Since music is a primitive and essential human activity its survival is not in question. By many eighteenth-century thinkers it was held to be the original form of language, the origin of speech. If there is a question of survival it is of Western art music, or serious music, what is called “classical” music. That its survival is in jeopardy is an opinion expressed largely by journalists and by a few disgruntled critics. This is, however, a view that has been surfacing regularly for the last 230 years. It was even passionately maintained more than four centuries ago. I presume that some conservative Greeks felt that way when the purity of the Doric mode was perverted by the introduction of the lascivious Ionic mode. Perhaps the best plan would be simply to take a summary look at what enabled Western art music to survive in the past, and see to what extent the same factors still apply today.

I should like to start with what may seem like an irrelevant detail from a related art. The troubadour poetry of late-twelfth-century Provence was originally written down with the music: the manuscripts contain both the words and the tunes. Toward the middle of the thirteenth century, however, collections of the poems appear without the music. It is also at this time that they begin to be accompanied sometimes by biographies of the poets, called vidas, or by critical commentaries, called razos. It would seem that when the poetry was isolated from the original musical context that gave it part of its meaning, a new context had to be invented to substitute for the lost significance. (It was sometimes true that the old melodies would have been known without being recopied, but this was not always the case.) Being read without the music, the poems now had a different social function; they were no longer performed to make up a part of a communal experience. The void caused by the disappearance of the poet who sang was made good by his biography. And the critical commentary replaced the experience of hearing together with others.

It was more or less at this time during the thirteenth century that Western music began to draw apart from poetry with the introduction of poly-phony, in which two or more strands of music sound simultaneously. The new degree of complexity, which developed over the centuries to an unprecedented and unequaled richness for music, resulted in a demand for greater and more intense awareness from both listener and performer, an awareness that was at least partially independent of the words. Words and music developed separate interests, each with an independent aesthetic that was sometimes at odds with the other. The poems now asked to be read on their own, but consequently they now required a new context to make good the disappearance of the musical context. Biography and critical commentary arose to provide a kind of significance that was previously not felt to be necessary. The poetry now had to survive without the music, and it stimulated the new setting that allowed it to flourish. Complex vocal music could even, at times, be executed by instruments alone, and the music began to stand by itself.

When we speak of survival here, obviously we do not consider the survival of the art itself, but of the specific works of art. An art can survive simply because its traditions survive, its practices continue. This kind of ritual survival is best accomplished in a society that remains static, and that is culturally homogeneous. The survival of a work in a rapidly changing society, on the other hand, depends not only on whether it is handed down to us unmutilated, but on its ability to adapt to changing conditions of reception, on its capacity, when its original social function has been destroyed or altered beyond recognition, to create or inspire new kinds of significance that allow its vitality full play. Just as troubadour poetry, when its original form of presentation was moving toward obsolescence, found a new social context from biography and critical commentary, so many works of Western music have inspired new forms of presentation from age to age that have brought them back to life again and again.

Essential to Western music is its ability to adapt to different social conditions, or—to put it a different way—to remain independent of the conditions that watched over its creation. By the middle of the fifteenth century, manuscript collections of music mixed together religious and secular music. Vocal works could be performed by instrumentalists; liturgical music—motets and sections of masses—could be arranged for the lute. Church music could be transformed into secular entertainment. As with troubadour poetry, when the art of music began to free itself from any rigid or absolute attachment to its social function in the late fifteenth and early sixteenth centuries, critical writing about composers and their works started to appear as if to justify and give meaning to the existence of an independent art. We learn, for example, that Josquin des Pres habitually withheld his compositions for several years before allowing them to circulate.


The survival of a work of Western music, before the twentieth century, depended essentially on a system of notation. Of course, we must not underestimate the power of memory. Thomas de Quincey, for example, was able to repeat a good deal of Wordsworth’s The Prelude from memory after having heard it read to him years before (he was certainly not permitted to copy any of it before the publication thirty-five years later). Most of the music in the world is not notated at all: it survives insofar as it does only through repeated performance, through the fidelity of the performers’ memory. A great deal of music is not repeated at all, but newly created on each occasion. In any case, some of the finest music does not need notation, or at any rate resists it. Folk music is particularly recalcitrant to being written down, and it is only in recent years that an attempt has been made to notate—with only partial success—some of the best examples of jazz.

In odd corners and pockets of European civilizations, a few fragments of medieval and Renaissance music were perpetuated throughout the eighteenth century, half-alive, like the crumbling Gothic ruins of an earlier culture. A monastery south of Rome continued (and still continues today) to preserve Byzantine chant. The Sistine Chapel choir still had the odd piece of Palestrina in its repertoire. But the history of music came into being inspired by the antiquarian interests of the end of the 1700s, when, in literature, old ballads were printed along with the legends of Robin Hood, and the Dark Ages became romantically fashionable. The true beginning of musicology was Charles Burney’s decoding around 1798 of the puzzling notation of Josquin des Pres’s masterly Déploration sur la mort d’Ockeghem, a composition in memory of the fifteenth-century Franco-Flemish composer Johannes Ockeghem. For his General History of Music Burney set the parts into score. (In Josquin’s time, all music was copied as separate parts for the different performers; whatever full scores may have existed at that time, none have come down to us, and we have no autograph manuscripts from that period.)

The dependence of classical music on the score, or on a system of notation, has given rise to odd and eccentric philosophical speculations. To dispel some of the clouds, we may briefly examine what Burney did. Having copied the parts out into a score, he found that they made no sense. The dissonances were intolerable and absurd. The tenor part, set to the chant of the Requiem for the dead, was marked “Canon.” To any eighteenth-century musician, a canon was a piece in which different voices sang the same melody together but starting at different points, out of phase, like “Row, row, row your boat gently down the stream,” so that a single melody would make contrapuntal harmony with itself.

This could be presented as a puzzle. Musicians would be given the melody in a single voice, and challenged to find the places where the other voices entered into it to make perfect harmony. That would presume that one had to find how to add a voice singing the chant to Josquin’s Déploration but beginning later than the tenor in a way that would make proper sense. However, the dissonance of the piece was already so intolerable that adding notes could only make it worse. Burney’s discovery was that “canon” simply meant “puzzle,” and this one was resolved by transposing the tenor chant, which was in the mode of church music called Dorian (centered on D), up a step into the mode called Phrygian (centered on E); once this was done, all became clear, and a great work was revealed.

Josquin’s canon can stand for us as a symbol of the relation of notation to realization, of score to performance. For practical purposes, not every aspect of music can be written down. Notation is selective: only certain musi- cal elements, or “parameters,” are chosen. For this reason, we might consider an art heavily dependent on notation like Western art music as essentially inferior to the musics of other cultures, transmitted orally or by the imitation of practice. That is why we can say of such and such a pianist that he or she may be playing the written notes but has no idea how to interpret the music. It is almost entirely on what is not written down that different schools of pedagogy attempt to base their claims to superiority. However, it is essentially the fundamentally unsatisfactory nature of notation that has allowed the monuments of Western music to survive, to escape the ruinous erosion of time. In fact, it is the basic antagonism of score and performance, of concept and realization, that is the glory of Western music.


One metaphysical aspect of Josquin’s canon needs to be set in relief. We hear the tenor part sung in Phrygian, but in reality it is in Dorian. This reality has two facets. First, the original chant is in fact in the Dorian mode; that is the way it came down to Josquin and has come down to us. Second, the chant is written on paper in the Dorian mode: that is the way the singer saw it and still sees it (if one keeps the fifteenth-century notation). The written form acknowledges the “authentic” version of the chant, which is then, as Burney discovered, inflected and even contradicted by what one sings and hears. If the originating form is the written one that is not heard, the emotion conveyed by the music comes from its realization: to our ears, and probably to the ears of a contemporary of Josquin, Phrygian harmony has greater pathos than Dorian, and is most suited to the elegy celebrating Josquin’s greatest predecessor.


It is sometimes said that the work of music is not what is written down on paper: the score is merely a set of directions that enables one to find or realize the work. This claim, however, leads to an absurd misunderstanding of the way the music works. What we hear, the realization of the score in sound, can never be identified with the work itself because there are many different ways of realizing the score, but they are all realizations of the same work, which in fact remains invariant—remains, we might say, visible but inaudible behind all these realizations.

It is not sensible to claim that for every score there is an ideal realization. The relation between score and realization implies that various possibilities remain open. This is how Artur Schnabel could be justified in maintaining that a Beethoven sonata was greater than any realization could possibly be. The written score sets limits within which many possibilities can find a home. The critical problem remains essentially how to decide which of the many interpretations realize the work and which ones betray it. Some interpretations are, of course, out of bounds, but determining what is legitimate and what is illicit is not as straightforward or as easy as we might like to think. In fact, the most deplorable misrepresentations of a work can often come from those who believe themselves to be the legal guardians of the tradition.

The survival of the Western musical past is due in part to the creation of an efficient system of notation. For the past 250 years, the transmission of the musical tradition has depended on the fact that in Europe and to some extent in America the proof that one belongs to a cultural elite has implied that some members of your family have learned to read this system of notation. Even members of the lower classes have been taught to read music, to sing, and to play an instrument—generally a keyboard instrument—as a proof of cultural literacy, as an aid to social mobility, to rising in the world. In particular, most women from a class of even moderate pretensions learned to play the piano just as they learned to read, to sew, and to cook. Playing the piano gave one a certain self-respect, enabled one to judge others, and allowed the amateur to arrogate the right to pass judgment even on the professional.

We think today of music as something public—performed in concerts or sold to the public in the form of recordings. In the nineteenth and early twentieth centuries, the experience of music was altogether different. Most music was heard at home, and was completely private; some of the experience was semiprivate, music played at parties or after dinner by invited professional and amateur musicians. Even symphonies were not solely a part of the public realm; in music-loving families four-hand piano arrangements of the Mozart and Beethoven symphonies were common enough. Wagner operas and Mahler symphonies were sold in four-hand piano versions for playing with a friend or another member of the family. Just as most of the audience today at a play by Shakespeare have read some of his works in school, so a good part of the audience at a symphony concert seventy-five years ago had received their first knowledge of an important part of the symphonic repertoire at home.

This private and semiprivate experience of music did more than increase one’s familiarity with the repertoire. It altered the understanding of music. Playing a symphony at the piano four hands and then hearing it in public with an orchestra made one aware of the existence of the work with a strikingly different realization of the sound. Learning to play a Beethoven sonata as a part of one’s education and then hearing a concert pianist perform it in public showed how personal the interpretation of a score could be—and was intended to be, in fact, for a great part of Western musical history. Merely the fact of displacing a work from the private to the public sphere altered its significance. Our assumption today, made unconsciously, that almost all music is basically public is a radical distortion of Western tradition. We no longer have a public that largely understands how the visual experience of a musical score is transformed into an experience of sound, and to what extent this transformation is not a simple matter but is capable of individual inflections. If the audience for classical music is not growing fast enough to make the art richly profitable—it never really was so profitable, by the way—the reason is not that there is a diminished response to the art of serious music but that there are fewer children who learn to play the piano. Learning music from records instead of playing it has radically altered our perception of the art.


Asking in what ways our musical culture can be transmitted to the future requires us to consider how our heritage has come down to us—which part of the past has been transmitted by being written down and what we have been able to gather only by a continuing tradition. Here it is worth emphasizing that notation only preserves very limited aspects of the music: many parameters are not written down at all.

In the history of European music, pitch—the quality of a sound that locates its position in the scale—is primary. Traditionally what has been notated is largely the pitch of the notes (although the notation of pitch is not as precise as we might think) and, more roughly and grossly, the rhythm. It is not clear how exactly our earliest examples of written music define the pitch: the scores seem to indicate whether the voice went up or down without allowing us to decide exactly what the intervals were. Greater precision was introduced later on and was certainly demanded by polyphonic music.

Nevertheless, what was written down was not always exactly what was sung or played. Singers could alter the written pitches up or down by the introduction of sharps and flats that did not originally appear in the text. These accidentals are called musica ficta, fictive music. We have been warned by Professor Margaret Bent of All Souls, Oxford, formerly of Princeton—and we must be grateful for the warning—not to identify the notes in a medieval or Renaissance manuscript with the white notes on the piano. Much of the ficta added to the text may have been originally understood as implicit in the notation.

However, the conception of what the pitch ought to be did not remain static over the years, and musicians developed a taste for chromatic harmony which could inflect music written as long as half a century before. To understand the development of European musical history, we should note the curious fact that the notes that were actually sung and heard are called fictive, and this implies that the written notes are somehow more real. What should be emphasized is that the musical text remained invariant while the performances could vary even as regards the absolutely primary element of pitch. Until the eighteenth century—and even later—the ornamentation of the written musical line was left to the performer. Couperin and Bach encroached on the performer’s freedom, and wrote out many of the ornaments. It is interesting to note that sometimes in Bach the ornaments infringe the academic rules of counterpoint: ornaments were not subject to the same rules as the underlying text.

Another aspect of baroque performance left up to the player was the figured bass or continuo—that is, the improvised addition of harmonies to the written-out bass part by the keyboard performer at organ, harpsichord, or early pianoforte in order to fill out the texture. Bach himself wanted the rules of counterpoint to be obeyed within his own part by the improviser of the continuo, who was working only with the bass part in front of him on the music rack; but Bach allowed, and indeed was forced to allow, the rules to be broken by the existence of parallel fifths and octaves between the continuo part and the upper parts (soprano, alto, and tenor) of the ensemble; the keyboard player was not able to see these parts and could not respect the rules.

For Bach, in spite of his desire to control what might normally have been improvised and his understanding of the importance for the music of the ornamentation, there was still a hierarchy of values. Some aspects of the music were primary, and these made up the compositional structure itself; others were secondary, and basically concerned only the realization. It is significant that in his fugues for the keyboard, he largely indicated the ornaments to the main theme only when they were convenient to play. Many editors add these ornaments (sometimes in brackets) each time the theme occurs. We no longer make the distinction so fundamental to Bach between composition and realization, and our error arises to a large extent from his practice of notating some of the elements of the realization with a fullness rare at the time.

In the history of music since Bach, many of the secondary elements have one by one become primary—that is, those aspects of music once left up to the performer have gradually become incorporated into the original compositional process. It would be a mistake to think that this has occurred only because composers wished to exercise greater control over the execution of their works. What has happened is that the secondary elements became primary to the structure. Dynamics, for example, were once a way for the performer to elucidate the structure of pitch and rhythm and make them expressive and even personal.* With Haydn and, above all, Beethoven, however, the dynamics are often an integral part of the motif, which has become unthinkable and unintelligible without them.

Gradually through the nineteenth and into the twentieth century, not only dynamics but tone color are removed from the province of realization and transferred into the basic process of composition. Starting with Debussy and continuing with Boulez, indeed, tone color—balance and equilibrium of sound, pedaling, quality of staccato and legato—sometimes even outweighs the element of pitch in importance. In several pieces of Debussy, it would give the music less of a shock to play a wrong note than to play the wrong dynamics or apply the wrong touch. The composers have little by little invaded the territory of the performers. The freedom of the performer has not been completely annihilated, but it has had to be reformulated.

The distinction between primary and secondary elements has dominated the history of performance in the West, and has influenced even the way music is edited. Pitch, we agree, is determined by the composer. It would be considered immoral for a performer to change Beethoven’s pitches. If he wrote a C, most musicians feel that we ought not to play a B flat just because we might like it better. Rhythm also belongs to the composer, at least to some extent: we are allowed to inflect the rhythm, however, with rubato, with ritardandos, and with various other deformations, in order to make the expression more personal, to impose our own styles. Further, performers and editors traditionally act as if Beethoven’s dynamics are merely suggestions: it would seem to be perfectly all right to play a mezzo forte where Beethoven has prescribed a piano mezza voce if we think it more effective.

As for phrasing and pedal indications, editors and performers have ruthlessly disregarded and altered Beethoven’s indications—and those of Chopin, Schumann, Liszt, and almost anyone else—at will. It appears to be legitimate to change Beethoven’s and Chopin’s fingerings even when a different fingering will radically change the sound of the passage. Pitch and part of the rhythm are primary; all the rest is secondary—a conception that, as I have implied, is imposed by the limitations of our system of notation. At times, this has even influenced composition: Schumann’s Toccata op. 7 contains almost no indications of dynamics—i.e., degrees of volume such as piano and forte—in order not to erode the freedom of the performer. (It should be said, however, that the required dynamics are fairly obvious, and Schumann inserts those few that would not be so evident—the dynamics are, by this time in the history of musical style, incorporated almost directly into the basic musical inspiration.)

The hierarchy of musical parameters has received its blessing from musicology, mostly without musicologists being fully aware of this or of its implication. Most musical analysis deals almost exclusively with pitch. Rhythm gets a smaller amount of exam- ination. Tone color, texture, phrasing, all have to wait patiently in line for a little attention. They are, however, as important a part of our sensuous experience of music as any of the primary aspects, and as we have seen, many of them have become primary in our century. Our way of writing about music, however, is still anchored in the period around 1700. The notation of music makes the pitch and rhythm most precise: anything else is sketchily and subjectively indicated, and, after all, who has the courage to write about something so difficult to pin down? Ethnomusicologists experience this problem at its most poignant.


The distinction between idea and realization is built into Western cultural history: in the visual arts it is at the center of the argument that opposed drawing and color. Drawing was the abstract design that held the work together: color was merely the realization in the flesh of the idea. It is the late-nineteenth-century painters who showed how flimsy the opposition was. In a dogmatic attempt to assert the old hierarchy, Ingres had claimed that you could not judge an oil painting until you could examine the reproductive engraving made of it; only line was truly important for him, at least in theory. By the time of Cézanne the claim no longer made any sense. In aesthetics, any formal opposition will always be eventually overridden by artists who have learned to blur the contrasts or synthesize the opposing distinctions, as Cézanne learned to compose and draw through color. Composers similarly placed timbre on the same level as pitch, in order to compose with tone color.

Discs and tapes will transmit to future generations what our performances sound like. But what about the sound of music before Edison? What will future generations be able to make of most of our musical heritage? The history of the revival of medieval and Renaissance music, and even of the Baroque, can show upon what a flimsy base our idea of the past can rest.

How did the great masterpieces of the fifteenth-century Flemish school sound when they were performed? I take this body of work as a largely uncontested supreme moment in Western musical history. Performances have ranged in my own experience from executions almost on tiptoe, like musical equivalents of the much-admired contemporary Flemish miniatures, to shouting matches, as if we were witnessing a medieval equivalent of crowd songs at football games or an urform of Carl Orff’s Carmina Burana. How were the notes sung: with a mellow or an incisive tone, or—inspired by a once-fashionable theory based upon strange pinched facial grimaces of the singing angels in Van Eyck’s Adoration of the Lamb in Ghent—with a continuous nasal whine? Were the singers doubled by instruments, and if so, which ones? How fast were the tempos? How high or low was the pitch? Some of these questions can be answered only with varying degrees of probability. We do not know exactly what fifteenth-century music sounded like, although we can be relatively sure that performance style and practice varied from country to country and city to city.

Does it matter what the music sounded like? Or sounds like today? This is not a question that I have heard put, perhaps because we assume the answer is an obvious Yes. Indeed, I do think the answer is more or less Yes, but it is not a simple straightforward affirmative.

Sung loud or soft, fast or slow, freely or rigidly—or not sung at all but played by instruments, or indeed, merely imagined from the score—many of the works of Guillaume Dufay, Johannes Ockeghem, Josquin des Pres, and Pierre de la Rue will seem beautiful to most cultivated musicians and amateurs of music. Not because we have been brainwashed or imposed upon by historical reputation, but because of the way the works fit in with our previous musical experience from Bach to Gershwin and Boulez, and make sense to us after a momentary adjustment of our expectations. Not every revival of the past will meet with such assent, but these works are easily embraced by our already habitual musical concerns. That, of course, is because they were an essential part of the historical process by which the musical tradition developed, and these works can still be implicitly discerned within the later developments. To understand how this functions, however, we must turn back to the distinction between score and performance, composition and realization.

In our belief that music is what is made available to the public, we have a prejudice in favor of realization. It is hard to make people understand today that the score may be beautiful irrespective of my realization. The ideal example is, of course, Bach, whose work provided the foundation of music education for two centuries. His compositions from the Art of the Fugue and The Well-Tempered Keyboard, for example, have been played on the harpsichord, the organ, the clavichord, the piano, and by the string quartet, a full orchestra, the Moog synthesizer, a jazz band, and the Swingle Singers. You may prefer one medium to another (the original intention was two hands at the keyboard, whichever keyboard you had around the house), but all of it sounds pretty good. Which means that in some senses of the word “sound,” it does not matter what Bach sounds like. The theoretical structure of pitch and rhythm is successful enough to stand however it is realized. It is not that the realization in sound is irrelevant, but it is not quite so fundamentally relevant as the devotees of authentic sound would like to maintain.

The eminent value of the score—the theoretical structure of pitch and rhythm with some of the other aspects of music indicated generally in a somewhat cursory fashion—is, I think, unique to Western music. It accounts for the reason that we think music is somehow allied to philosophy (although perhaps we have inherited this view of music from the Greeks). There are two consequences: first, the music may be admired even when no one can be quite sure what it sounds like; second, a score can be realized in many different ways, with many different kinds of sonority, as if a purely ideal structure could be made to give life to a multitude of actual forms. (In our time Boulez has made this almost into a principle of composing, with many different versions of the same structure, but in his case the working-out of a new realization is in reality only a later and more advanced stage of composition.) We must conclude that although the musical works of the past have been preserved by a system of notation, what has guaranteed the life of the past through so many centuries, or allowed forgotten music to be revived, is essentially the ambiguity of the notation. The fact that it does not tell us exactly how the music sounded has made it possible for later generations to realize it in their own way, deforming or restoring the past in a variety of forms.


The love of music is a natural human instinct, but the love of classical music is something else again: it may even reasonably be considered to be perverse. It certainly demands at least a discernible amount of experience and education to become addicted to it. For serious music to play an important role in a culture requires not only a significant number of professionals who can be hired to perform it but a dedicated body of amateurs who take an active but occasional part in its production. From the sixteenth to the seventeenth century, art music was upheld by the members of society who learned to sing as well as to listen. Beginning in the eighteenth century, learning to sing was partially replaced by learning to play an instrument, above all a keyboard instrument, and finally, above all, the piano, which ended by ousting all the other keyboard instruments from the province of the enlightened amateur. In the nineteenth and twentieth centuries, the piano at home and in the salon was used not only as a solo instrument, but also to accompany lieder; opera arias were also arranged and played in the home at the piano. The transcriptions of string quartets and symphonies for two or four hands was a common way of getting acquainted with the most important works that one would hear in the concert hall. As a small child I first came into contact, for example, with the symphony of César Franck by reading a transcription for two hands.

In the middle class, knowing how to play the piano was thought to give one a social advantage, as I have remarked, although of course one pretended that it was a purely disinterested love of music that supported the thousands of piano teachers who did their bit for European civilization. The great advantage of this kind of pretense is that in many cases it eventually turns out to be true; social progress is regularly and correctly fueled by forms of hypocrisy. The audience for a concert of serious music always contained an important group of amateurs who had experienced the art at first hand. Most of these had never had any ambition or desire to became a professional, but they combined connoisseurship and enthusiasm.

Learning to sing and learning to play the piano have been supplanted today by collecting records. This is a disquieting development that is already affecting the future. The audience for serious music has become increasingly passive, and there is no longer an important body of educated listeners experienced in the making of music that can act as a bridge between the general public and the professional.

Popular music has today an astonishingly large number of young people who are active in performing it privately, for their own pleasure and for a few friends. This has been true throughout history. But in our time, popular music reverses the classical relation between composition and realization: realization becomes all. Except as a form of nostalgic and commercial kitsch that deforms the original product, popular music is no longer the faithful rendition of a traditional tune. In the great examples of popular music in our century, those that have already attained a classic status like the great jazz improvisations of musicians from Art Tatum to Miles Davis, the original composition becomes identified with the performance. The interest of the tune at the basis of the improvisation—like Cole Porter’s “Night and Day,” for example—is minimal, except that it is useful and enlightening for connoisseurs to be aware of the tune, just as one recognizes the traditional Lutheran hymn tune in a Bach chorale prelude. Tatum does not realize Porter’s composition, he composes an entirely new work into which the tune enters as a structural component. In the early part of the twentieth century, the most advanced forms of popular music were essentially improvised events, each of which was one-off, ephemeral: it was preserved not by a score but on occasion by a recording, and was basically unrepeatable.

In improvisation, the distinction between composition and realization disappears almost completely. So it does as well in another typical phenomenon of our time, electronic music: the playing of an electronically realized composition on tape is identical with the composer’s intention, except where the reproducing equipment or the acoustical setup is defective. The role of improvisation, however, has been reduced in rock music: here, recording has taken over. A public performance of rock music is rarely a newly improvised work or a new realization of a score, but the reproduction of a recording. Most of the audience knows the work from a disc, and has come to experience communally en masse. (In rock music, too, the creative role of the recording must be taken into account.)

With classical music, the impact of the recording industry has been complex. It has enabled us to grasp how many different ways of realizing a work may exist, how many different and wildly contrasting interpretations of, say, The Passion According to Saint Matthew are possible. It has also made the general public acquainted—in a largely haphazard and disorienting fashion—with a bewildering variety of period styles and genres, and helped us to absorb so much of what seems alien at first hearing.

Recordings, nevertheless, stand on its head the traditional aesthetic on which art music has been based. For most of Western history, the composition is fixed, the performance is fragile, existing only for the moment. Recordings, however, privilege the performance over the score. The composition is held at a distance: the performance is up front. This sense of privilege is increased by the possibility of splicing. Performances are no longer flawed except by the idiocy of the interpreters or an injudicious placement of the microphones. Technical mistakes may be easily eradicated.

With recordings, one of the most profound aspects of the experience of art music is almost completely obliterated: the resistance of the work to interpretation. I have spoken of the relation of score to realization as an opposition. The work does not hand itself over easily to interpretation; it does not surrender without a fight; it puts obstacles in the way of the executant. If you have played a Beethoven sonata, then you know what the problems are of looking at a score and how it is to be turned into sound. But with a recording, what you hear is the sound without being aware of the score behind it. Both experiences are fine, but most people today no longer have the experience of knowing what it is to transform a score into sound and the kind of resistance that the work exhibits. With five or ten recordings of the Matthew Passion, we can see that they’re all different but we’re not exactly sure of the process that turns the score into the sound. This process is a physical experience absolutely essential to the Western music tradition. Listening to a record, we neither feel the physical difficulty of realizing a musical text, nor can we witness, as in a concert hall, the exciting spectacle of the torments of the performer.

Every musical culture has its own idiosyncratic physical experience. The peculiarity of the Western tradition is the difficulty of translating the score of someone else’s idea into sound. Recordings have, of course, made possible the expansion of our experience of the musical repertoire, given us an unprecedented contact with past ages. But there is a small qualitative difference between the experience of music on records and the experience of music in concert and a large qualitative difference between listening to it on records and playing it oneself. Sometimes there are technical obstacles, as in the works of Liszt, which can literally cause physical pain to the pianist. Sometimes the problem is emotional: how to realize the extraordinary passion that lies within a Chopin polonaise without falling into the sentimental clichés of performance style taught at conservatories around the world.

Often enough the obstacle is intellectual: How can one make audible for an audience the intricate interweaving of the voices in a Bach fugue without merely hammering out the theme and without obscuring the affective power of the work? For a great part of the repertoire, including the examples I have given, the difficulties are technical, emotional, and intellectual all at once. Behind all these lies the fundamental one: How is the physical pleasure of performance to be made manifest to the listener—and to oneself—without obscuring the qualities of the original composition that make everything possible?

For centuries, the tradition of art music has rested on this opposition of composition and realization, and on the sense that the composition has a value that transcends all imaginable realizations. Whether this fundamental aesthetic requirement will last into the future, I do not know. It is attacked not only by the replacement of public concerts and private performance by records. It is subtly undermined by directors of opera who think that they can invent stage business which has no relation or relevance to the music, as if the musical score can be overridden by the stage director’s imagination. It is violated by the fans of authentic period style who think that a work of music can be simply identified with the way it might have sounded to the composer. It is misunderstood by the widespread failure to grasp the different ways a work could be realized during the lifetime of the composer and, in addition, the ways in which it is altered by the generations that follow, while retaining an important part of its validity. It is ignored by so many who do not recognize the special character of the Western tradition, and who imagine that it can be treated as another variety of folk music.

The eventual survival of the tradition is ultimately not at stake. There are the documents available—the scores and the recordings and a bewildering pile of published research. If the tradition disappears, it may be revived, as we have revived the singing of Gregorian chant. Nevertheless, there is a difference between the continuous survival of a tradition and one that has to be brought back to life from death or near-death. It is not, however, what might be thought, a difference of authenticity. The continued existence of a tradition brings with it gradual changes and corruptions in the performance and the realization of its works. A Viennese performance of Mozart today is incomparably different from one during the composer’s lifetime. Recovering the past with Mozart is as difficult as with Monteverdi, but the problems are not the same. For Mozart, we have to correct and adjust—not only to go back to the earlier techniques and style, but to find ones more adequate to modern concert halls and instruments and modern sensibility than the nineteenth- century stylistic habits with which he has been saddled. For Monteverdi, on the other hand, we have to invent a style of performance, inspired by contemporary documents and accounts, in a process like digging up the ruins of Assyrian civilization. In the twenty-second century, will the monuments of Beethoven and Chopin have to be excavated or merely tactfully restored?

This Issue

December 20, 2001