In 1791 a listing was made of the several dozen musicians in the employ of the episcopal court of Bonn. “Herr Ludwig van Beethoven plays clavier concertos” is the extent of the entry on the young man who was to become Bonn’s and perhaps Germany’s most celebrated citizen. At the age of twenty his output included three concertos for piano (one arranged from a violin concerto, as was not uncommon at the time), as well as a concerto for piano with two other instruments, and possibly a fourth piano concerto, if some extant early sketches were ever completed. A young pianist-virtuoso-composer needed concertos to make his way—indeed, the encounter between the concerto’s solo instrument and orchestra can stand as a metaphor for the freelance musician and his support system of audience and patrons. Before Beethoven found himself as a symphonist, the concerto was the public genre that marked the stages of his march to success and, as the nineteenth century saw it, greatness.

Around 1801, when he published the first two of his canonical piano concertos, Beethoven knew he was becoming deaf; by 1809, when he composed the fifth of them, the so-called “Emperor,” he could no longer play in public. So in the eighteen years remaining to him he no longer wrote concertos. The story of Beethoven’s concertos is a story of the obligatory and no doubt painful relinquishment of a favorite genre. (In later years he began work on more than one concerto, only to leave them unfinished. The Eighth Symphony was first sketched as a concerto.) Yet in this arrested development Leon Plantinga can trace a grand panorama of change in the very concept of Western music and in the model of a composer. It is a transition from music as performance to music as text, from the performer-composer known to the eighteenth century to the genius-creator postulated by the nineteenth. In the sphere of the concerto, Beethoven’s deafness was both cause and symbol of this transition.

This historical process has been the subject of much discussion (as well as complaint) in recent years. It is bound up with the issue of “absolute” music, music thought to be autonomous irrespective of verbal texts, social contexts, and even performance considerations. It is the subject of an entire book by the late, much translated musicologist Carl Dahlhaus, and the subject of a full half of a book on the ontology of music by the English philosopher Lydia Goehr.1 When did European music change over from a fluid performance activity, passed on from master to student over the ages, as it was and is in other cultures, and become a written, read, and (eventually) digitally recorded artifact to be contemplated as well as—as much as, one sometimes feels—heard?

Over a period of roughly a thousand years, according to historians of the longue durée. Somewhat abruptly around the year 1800, according to Goehr, and it is easy for Plantinga in one of his argumentative footnotes to show her case to be “overdrawn.” Yet it contains “a kernel of truth”; no one doubts that the process accelerated decisively in Beethoven’s lifetime. It accelerated as a response to his extraordinary music but also as a response to Romantic musical speculation.

I am reminded here of another recent study, also overdrawn, by a sociologist who attributes Beethoven’s early success in Vienna—exactly in the years he was making his way with concertos—to a need felt by his essential patrons for a new elevated musical ideal in the face of social change.2 If the Viennese aristocracy had a historic need for Beethoven, the philosophers of Jena needed him even more. Though Beethoven’s Vienna was certainly no hotbed of Romanticism, Maynard Solomon, for one, whose classic biography of Beethoven now appears in a revised edition, has always patiently searched out Romantic traits and resonances in the composer’s thought and in his music.3 And only Beethoven could have steered the conception of music from Kant’s “pleasurable stimulation” to E.T.A. Hoffmann’s revelation of the “spirit realm.” The Romantics from Tieck and Schlegel all the way up to Adorno saw the musical masterwork as an autonomous product of the free “subject.” Beethoven’s metaphysical freedom with the musical tradition he inherited from Haydn and Mozart was as evident as the secular freedoms he assumed in daily life, already the subject of myth in his own day.

A great strength of Plantinga’s book is the light he throws on this topic from the perhaps unexpected special standpoint of Beethoven’s concertos. He draws on an enviable control of the historical record and from detailed analysis of the music itself. His scrutiny of the sources of the music—Beethoven’s complicated, overwritten autograph scores and his even more complicated musical sketches—is especially thorough. Here is a case where the musicologist’s typical preoccupation with sources results in broad historical understanding.



The earliest of the well-known concertos, the irregularly numbered Piano Concerto No. 2 in B-flat, exemplifies very well the fluidity of the concept “concerto” when Beethoven first took it up. This was his first warhorse, and every time he was called on to play it it seems he rewrote or rejected parts of the score and substituted others. The partition du jour served as a basis for improvising the piano part as well as the cadenzas.

Of four different versions that can be identified, dating from Bonn in the late 1780s to Vienna in the late 1790s, the earliest survives in a single loose page from the score (just enough to show that the first movement that we know was present in some form) and also in the draft of a melody for an early finale for the piece, later replaced, and now known as the Rondo in B-flat for Piano and Orchestra. On the basis of sketches that can be dated for revisions, musicologists have distinguished two further versions in the mid-1790s, which can be matched up with concerts by the composer for which we have notices and dates but no programs. Only for a fourth version in 1798 has the score survived—and even here the piano part Beethoven “entered only in fits and starts, leaving much to his celebrated powers of improvisation,” as Plantinga puts it.

Those powers were exercised most spectacularly, of course, in cadenzas, the concerto’s moments of sanctioned carnival. Extended formal cadenzas come near the ends of many movements, while numerous other short ones, “run-ins” (Eingänge), and fermatas, or pauses, are scattered throughout. In addition, improvisation of a different kind continued throughout the score, improvisation carried out against a background of harmony, phrasing, and so on laid down by the orchestra, with the surface spun out extempore by the solo. Indeed the young Beethoven never wrote out—never really “composed”—concerto piano parts until he had to, because publication was in the offing. When the B-flat Concerto reached the printer in 1801, the orchestra part, too, was still in flux, subject to the composer’s last-minute tinkering.

At this point the long-bubbling soup of the B-flat Concerto became a text and the composer could put it behind him. Concertos were performing vehicles; they would not be published until they were no longer needed, when they had worn out their welcome with the public and a new one was ready. The autograph of the C-minor Concerto, No. 3, of 1803, about which Plantinga tells us more than we probably want to know, also started life as a “performance autograph.” That is, it contained the orchestral music with, at many places, no more than piano cues to guide Beethoven’s improvisation. The piano part was only fixed conclusively when it had to be, in this case for a performance in the next year by another pianist.

The evolution of Beethoven’s compositions can be traced in astonishing detail from his many sketchbooks and loose work sheets, none of which he seems ever to have thrown away. Just as sketchy piano parts served him as the basis for improvisation in the body of his concertos, so did actual sketches serve as aides-mémoires for his improvised cadenzas. That Beethovenian free fantasy did not leave overly much to chance is clear from the rather extensive notations for cadenza ideas preserved among his papers. (These indeed provide some of the evidence for the various versions of the early concertos.) The earliest sketch we have for Concerto No. 3, an isolated jotting from around 1796, is not for the body of the piece but marked for the cadenza, and its importance for Beethoven can be gauged from the fact that he returned to this very idea in 1809 when he wrote out cadenzas for all the piano concertos he had completed by that time.

It was at just around this time that Beethoven saw he would have to give up performing; he also, as he composed his last concerto, the “Emperor,” saw to it that all the constituent cadenzas were written out in full. By assuming control of the main sites of spontaneity the composer sharply attenuated the role of the performer—as composers have continued to do ever since. By also writing out cadenzas for the earlier concertos, was Beethoven proposing to give a full text for works that he would never himself control again? Or was he trying to pump some performative life back into frozen texts which he had not, after all, finally put behind him?

He did not publish those cadenzas—for lack of a market, no doubt, rather than for any qualms about such a project; in 1810-1811 he did publish the originally improvised Fantasy for Piano, Opus 77, and another extended improvisation at the beginning of Opus 80, the Choral Fantasy for Piano, Orchestra, and Chorus. (We do not know, of course, how closely the printed versions correspond to what he originally played; we do know that the music was worked over in sketchbooks before publication.) In any case, history’s decision about the cadenzas was never in doubt. They have now been safely absorbed into the primary texts. After they were first published in 1864 they at once found their way into performance scores; and few pianists ever since have played anything else. Some musicologists have analyzed these cadenzas, too, and claim to find in them an organic relation with the body of their hosts.


One who does not play Beethoven’s cadenzas is Robert Levin, the fortepianist and musicologist who has developed an impressive gift for improvisation in the styles of Mozart and Beethoven. His recent series of Beethoven concerto recordings with John Eliot Gardiner gives us a sense—sometimes a vivid sense—of what the concerto was like in its pretextual stage when it was dependent on personal performance. Flamboyant, powerful, and—the first time you hear them, at least—admirably unpredictable, Levin’s cadenzas make these recordings distinctive, indeed unique.

Of course all improvisation is part prestidigitation; the musician has his formulas, as the conjuror has his tricks. If a virtuoso is like an athlete in some ways, he is like an illusionist in other ways. Levin creates the magical illusion of a pocket of music history innocent of and prior to scores—though not of course innocent of cues and aides-mémoires. His art is built on internalizing historical documents like Opus 77 and Opus 80, which capture spontaneity and preserve it like a pinned butterfly. In his case this is accomplished through recording, and his recorded cadenzas and ornaments bring us closer to the actual experience of improvisation than any scores can. (The ontology of written-down and recorded improvisations would make a good new project for a musical philosopher like Lydia Goehr. If Levin spliced his recording takes—he does not—edited recorded improvisations would be something further for her to mull over.)

It takes nothing away from Levin’s art to observe that what he does depends on the fortepiano. Virtuosos play at their instruments’ limits, and limits which Mozart and Beethoven tested when they improvised come nowhere near the limits of a modern piano. In modern-piano performances Mozart’s written-out cadenzas sound tame and should be retired in favor of modern items, such as one that Benjamin Britten wrote for Svyatoslav Richter, which can be heard in some recordings of the E-flat Concerto, K. 482. Virtuosity matters more to concertos than stylistic consistency. This is certainly what Beethoven appears to have thought when he wrote those cadenzas for the early concertos, cadenzas whose unrestrained bravura makes the rest of the music sound very meek indeed. (Though on the other hand, maybe he assumed that any competent performer would upgrade the early piano parts extempore to match the new cadenzas.)

In the main body of the concertos, to be sure, Levin departs little from the scores, less than he often does with scores by Mozart.4 In works like the “Emperor” even this programmatic improviser is somewhat daunted, it seems, by the prestige of Beethovenian textuality.


Piano Concerto No. 4 in G was the first that Beethoven published before playing it in public. Plantinga takes this precipitate publication, in 1808, as the first symptom of a new concept of the concerto. Beethoven was starting to conceive of a concerto as a “work,” like a symphony or a string quartet. Publication fixed the text and with that action canonized it.

Certainly today no Beethoven concerto is more willingly granted masterpiece status than the Fourth. One doesn’t need a fastidious nose to catch a whiff of a vehicle for virtuosity in the Violin Concerto or the “Emperor”; not so the Fourth. Its unusual, pensive opening movement, its even more unusual Andante for piano and strings alone, heavy with intimations of tragedy, its ebullient finale, witty, raucous, brilliant, lush, crisp, and expansive in turn—did Beethoven ever write a more scintillating finale?—all this, plus the intimate, subtle sense of dialogue in all the movements have made the Fourth a favorite with critics and commentators. In his discussion of the music itself, as opposed to the sources, Plantinga devotes more space to this work than to any of the others.

Yet as he also shows, two source documents for the concerto—the only two manuscript sources, in fact—each put this text in limbo. One is a set of instrumental parts for an arrangement of the piece for piano and string quintet. Concertos were often played as chamber music, in domestic settings; the only concertos Mozart was able to publish (his Opus 4, containing K. 413, 414, and 415; this was in 1784) were issued with optional wind parts to allow for such performances. Arrangements were a natural feature of the polymorphous performance culture, one that survived the change from music as act to music as text when it was discovered that publication could lead to commodification as well as canonization. Contemporary with Beethoven, Schopenhauer pondered the metaphysics of music to the accompaniment of Rossini operas arranged for two flutes.

The other source for the Fourth is a copyist’s score that served as printers’ copy and also as the basis for the quintet version. What is interesting about this score is that Beethoven has entered a wealth of cues and indications for extravagant virtuosity in the piano part. He was now writing for a piano with a higher range than sufficed for the published edition; as the piano underwent rapid technological development in the 1800s (as rapid as the development of techniques of reconstructing old instruments in the 1980s), Beethoven’s piano sonatas as well as his concertos show the composer racing to keep up. But over and above this, he was making the work at hand a good deal less pensive, intimate, and subtle.

Plantinga argues that Beethoven’s drastic notations for Concerto No. 4 must relate to its first public performance in 1808, which he suspects was also the première:

To whatever degree Beethoven had begun to dissociate the idea of the concerto from his own performances, in its creator’s hands the Fourth Concerto continued to be less the presentation of a “work” than a live activity, a unique event subject to whim, its nature flowing partly from the mood of the moment. The piece seemed to exist for Beethoven on two levels. One was for publication, probably in most cases for performance with reduced forces or even with piano alone in private; the intricacies of balance between solo and tutti, and a certain intimate and subdued quality in the piano remind us of chamber music. But Beethoven also had his public conception of this music, in which the solo part (at least at points) commands, dazzling and astonishing a rapt audience in a large hall.

The paradox is reflected in the program of that 1808 concert, a notorious marathon at which the heating system broke down, and so did the orchestra, to Beethoven’s fury. Among the array of new works premièred on that occasion was the Fifth Symphony, which, ever since the influential essay by E.T.A. Hoffmann from 1810, has become the very model of autonomy and musical textuality. But the concert also included two improvisations by Beethoven, in works already mentioned, the Piano Fantasy and the Choral Fantasy, improvisations that Beethoven soon fixed in print.

The Choral Fantasy participates in the evolution of musical textuality in its own peculiar way, through its utter metamorphosis into another great canonic text, the Ninth Symphony. In the fantasy, melody settles the capricious to and fro of a piano cadenza. In the symphony, fifteen years later, melody (much the same melody) answers the distressed questions of an orchestral recitative. On the Levin-Gardiner discs, which include the Choral Fantasy, Levin plays Beethoven’s opening cadenza (marvelously well) and adds not one but two alternative improvisations of his own on separate CD tracks, allowing for three versions of this music. Like Beethoven, Levin does not flaunt “freedom”; both of his efforts pay homage to Beethoven’s text before spinning off into the uncharted and the fantastic. The first of them goes to the flat side and develops a lyric fragment with some passion, whereas the second goes sharp, turns pathetic, and forecasts a masterpiece from his third period.

Neither is quite as wild as the Beethoven. “Beethoven the architect turns into a genius running amok,” Alfred Brendel has written. “No composer has ever hazarded cadenzas of such provoking madness.”5

The fifth and last of the piano concertos, the “Emperor,” carries the process of textualization to a terminus. Not only was this concerto published before it was ever played, there is no evidence that Beethoven had any occasion in mind for its performance. (In the event it was not premièred until 1811, and then in Leipzig, presumably at the instigation of the publishers, Breitkopf & Härtel of that city.) That he never expected to play it himself is suggested by the fact that this work—and the concerto now really counts as a “work”—was written out completely at once. The piece is famously virtuosic; it even begins with mini-cadenzas for the piano, in between grandiose single chords in the orchestra. They are written out note for note both at the beginning and also when the passage returns, most unexpectedly, later in the movement.

Moreover Beethoven took unusual trouble with the published score. He added simplified versions for difficult passages. He filled in the piano part dutifully during the orchestral tuttis, thus contributing to a long-lasting controversy on performance practices that Plantinga discusses at length. The entire project reeked of paradox:

With this concerto Beethoven and Breitkopf & Härtel confronted a central conundrum of the nineteenth-century virtuoso’s address to the public. Glittering virtuosity in concert performance was counted on by publishers to spur sales of printed music to an admiring public; but the technical wizardry that had most excited audiences in the first place was the element least reproducible at the living-room piano.

Plantinga has little to say about the prehistory of the Fifth Concerto—he does not discuss the sketches—and I wonder about the impetus for it, and for the Fourth Concerto also, for that matter. By 1807 Beethoven’s fame had grown and he was publishing more and more. Would Plantinga go so far as to say that the Fourth was projected as a relatively intimate concerto for piano and a reduced orchestra with an eye to a market of amateurs, for informal performances outside the concert hall? In any case, I think that by this time Beethoven must have felt there was something missing in his output, a concerto in his characteristic “heroic” style, the style he had launched with the “Eroica” Symphony of 1803. He might well have felt impelled to write such a piece, entirely apart from questions of performance and publication.

For Plantinga, the concept of the heroic

is a very poor designation for Beethoven’s middle period as a whole…. There is no ignoring that this was also the period of the “Pastoral” Symphony, the Violin Concerto, the String Quartet Op. 59, No. 1,

and so on. Still, there is a kernel of truth to it, much more than a kernel, and Michael Steinberg, in his adroit new book about concertos, says just what needs to be said when he characterizes the “Emperor” as both the summit and the termination of Beethoven’s “heroic decade.”6 What we have here in the first movement is a grandiose opening, an episode of crisis in the middle (or development) section, and a transcendent coda at the end—heroic markers that occur in none of the other concertos. Past Piano Concerto No. 2, indeed, all of them actually start piano or even pianissimo: a rather remarkable circumstance.

In symphonies like the “Eroica” and the Fifth, the central crisis typically stages some kind of breakdown, from which the music heroically recoups. In concertos, where a likely hero is actually present, namely the solo instrument, heroism is projected differently. In the “Emperor” a dramatic confrontation confirms the soloist’s domination. The orchestra starts a big surge driven by a martial motif, the solo breaks in tumultuously with the same motif, they battle back and forth, and the solo emerges as winner. The effect is electrifying, not only because the fortepiano bangs out eight- and ten-finger chords at maximum volume, but also because it plays something that up to now, in the concerto’s first ten minutes, has been played only by the orchestra. Having seized the initiative (along with the motif), the piano now turns the martial gesture into a Byronic outcry which the orchestra can only mimic lamely. Next the hero veers unhesitatingly toward the sublime, appropriating another of the orchestra’s themes that it had never sampled previously.

The “Emperor” is also the first concerto whose first movement receives (or, rather, requires) a typical Beethoven coda, a transcendent resolution added to the traditional sonata-form model of dynamic statement, uncertainty, and stable restatement. A grand passage of orchestra-piano dialogue carries the main theme up and up in triumph—a familiar effect in all heroic genres except, until now, the concerto. (The passage was echoed by Brahms in his Piano Concerto No. 2.) After this the piano takes charge of the discourse by again playing an orchestral theme it had not played before.

Although the Fourth Concerto has something similar, the difference is beautifully characteristic. In the Fourth the piano steps up quietly, almost abashed to find itself more eloquent than the band whose place it has taken. In the Fifth the piano plays not the whole of the theme but waits and seizes just the end of it dramatically, loudly, heroically.


In addition to the historical and the source study outlined above, Beethoven’s Concertos includes over a hundred pages of analytical and critical matter, “close readings” of the six major concertos, the Triple Concerto, and the lesser works. Musical analysis is more or less doomed to bristle with technicalities, and its application at this length does not make for easy reading. The student rather than the general reader seems the target here, and while he or she is spared the tortures of the higher analysis, the emphasis on considerations of musical form, in particular, becomes wearying. Musicologists like to talk about form because it is one thing that texts appear to make manifest, but the plain fact is that despite the many innovations Beethoven introduced into concerto form, overall he was less innovative about form in this genre than in most others. Nor is emphasis on musical form the only academic feature of Plantinga’s discussions.

Yet his prose is alive and graceful, and the musicality it communicates is individual and true. Besides a good deal of what the critic Hans Keller once called “eminently professional tautology,” there is also much sensitive observation and insight on display here. At the same time, there are some baffling omissions. Among these are the two “heroic” episodes of the “Emperor” Concerto noted above, the confrontational passage in the development section of the first movement and the emphatic coda. To read a leisurely book on Beethoven’s concertos that glides past these salient episodes is strange indeed.

But Plantinga’s exhaustive study is full of good things, the result of much careful deliberation and very wide-ranging research. He observes that the entry about “Bethoven” in the 1791 court listing lacks the asterisk provided for all the important musicians under survey, and draws the conclusion that Beethoven’s sphere of activity was less the court of Bonn than Bonn high society—a forecast of his later experience in Vienna. He has discovered a newspaper report from 1820 of something called “To Psyche: a Melodramatic Essay after the Andante of Beethoven’s Concerto, Op. 58,” in which a poem said to be related to Goethe’s “Kennst du das Land?” was “spoken rhythmically to the music.” (What is Psyche doing on Orpheus’ turf, never mind Mignon?)

In another vein, Plantinga contrasts Piano Concertos No. 2 and 1 very effectively to pinpoint Beethoven’s growing maturity of style in the mid-1790s. His reading of certain movements is especially original and illuminating: the finale of Piano Concerto No. 4, and the first movement of No. 1 (though the development section is again skimped). He convincingly associates the Triple Concerto with Beethoven’s tropism toward France and things French, and writes toughly and justly about the music.

He also reports that Hans von Bülow—pianist, editor, and an influential nineteenth-century editor of Beethoven—boasted that he could always count on applause for his playing of the opening mini-cadenzas in the “Emperor” Concerto. Did people applaud during the orchestral music that follows directly? Did Bülow stop the orchestra? Can he really have stimulated applause by playing the notes as written in the score, which his audiences knew as well as we do, or did he improvise something even more magnificent? Beethoven and Robert Levin to the contrary notwithstanding, Bülow would have been within his rights to do so.

This Issue

June 24, 1999