Charles Rosen
Charles Rosen; drawing by David Levine


The thirty-two Beethoven sonatas, composed between 1795 and 1822, are the most prestigious works in the solo piano literature. They are, as Charles Rosen says in his new book about them, “the great representative of Western culture in the upper middle-class household from 1850 almost to our day, as much a part of civilized life as entertaining guests and family dinners.”

As he says, almost to our day. Until radio and gramophone came along, while the piano was still a necessary domestic piece of furniture for a cultured family in Europe and America, middle-class children and especially girls were taught to play as part of their basic preparation for society. And for marriage: the piano became a tool of courtship, women demonstrating their refined sensibilities to suitors, who would occasionally share the keyboard for four-hand arrangements, a variety of which were written during the Victorian era specifically to involve cross-hand passages that necessitated physical contact. Fears of too much hanky-panky were behind such popular songs as “Keep Your Foot on the Soft Pedal.”

Learning the Beethoven sonatas, or at least learning the easier ones, was a particular sign of musical and moral aspiration. Attached to learning the piano was always what could be called the therapeutic ideal of cultural immersion, linked to the old notion that culture must be morally and spiritually uplifting. Arthur Loesser, in his indispensable book Men, Women, and Pianos, wrote amusingly about how emotions like lust were channeled and, in a sense, elevated through piano music into acceptable bourgeois forms of public expression; and no music was more emotionally exercising or spiritually elevating than Beethoven’s sonatas.

Their prestige assumed a certain obligation, however. I recall once seeing Rosen in a television interview half-jokingly admit that the prospect of hearing, not to mention playing, Beethoven sonatas, even for the most dedicated and austere music lover like himself, can sometimes be a little exhausting. The music, as much as any art ever made, declares itself to be serious and expects you to be serious listening to it. Beethoven could be witty, ecstatic, and raucous, even at his most sublime—as he is, for instance, in the last piano sonata, in C minor, Op. 111—but the sonatas were conceived as high art and this presumes a certain burden, however edifying, on audience as well as player.

Aside from being mentally taxing, some of the sonatas are nearly impossible physically for anybody to play properly except professionals, Beethoven having written the works mostly as private or semiprivate endeavors for amateurs but at the same time introducing technical difficulties that could be mastered only by the concert pianist, a relatively new profession in his day. The “Hammerklavier,” Op. 106, is the famous example, a piece of symphonic proportions and technical challenges that alarm even the most accomplished player. But some of the earliest sonatas, like the C Major, Op. 2, No. 3, and the E-flat Major, Op. 7, one of the longest of the thirty-two, with wide skips, exhausting chains of broken octaves, and rapid scales, are best, if not exclusively, suited to the virtuoso and proportioned for the public hall or concert salon. Beethoven wrote what he wanted in the end, and who could perform it was of secondary concern to him.

Average amateurs, however, might stumble through the score at home, then hear the music skillfully done in concert. Struggling with the technical hurdles gave them “a sense of contact, however tenuous, with the professional that one could get from almost no other set of serious works,” Rosen writes in his introduction. “They were a challenge which could be taken on, an ideal to which one could aspire, even if they could not, in the end, be fully mastered.” The sonatas were thereby a crucial link between public and private music-making until roughly World War II, when middle-class people stopped learning the piano as a matter of course and lost that connection with the performers they paid to hear or listened to on recordings.

These social changes didn’t cause professional pianists to play the Beethoven sonatas differently, of course, but you might say that the changes gave pianists more leeway to interpret the music in ways that might have surprised Beethoven. Rosen spends much of his strongly argued book wrestling with the impossible problem of original tempos, a quagmire to musicians and musicologists. It’s vain simply to declare what Beethoven meant by Allegro or Adagio, not least because Beethoven wasn’t always consistent or even certain himself, and sometimes spoke of these words as emotional terms, not necessarily indications of speed. (“What can be more absurd than ‘Allegro,’ which always means merry?” he wrote.) But Rosen makes the case for reconsidering some of what are now fairly commonplace interpretations that he believes seem to exaggerate, to one extreme or the other, the intended pace of the music.


The reasons for these exaggerations we can only theorize about. They possibly include competitive pressures in an overcrowded market, combined with decreased firsthand familiarity on the part of audiences with the scores: virtuosos today want to distinguish themselves by showing how fast they can play complex music and how gravely they can play slow passages, to suggest their spiritual intensity. And listeners, no longer steeped in the scores, cannot appreciate the difference between textual distortion and legitimate reinterpretation.

Unlike Beethoven’s instrument, the modern piano, with its enhanced capacity for sustaining and projecting sound, also seems to encourage exaggerated performances, especially lugubrious tempos, as does the large modern concert hall, grand and formal compared to the circumstances of Hausmusik, or music in the home, which was the dominant condition for music-making in the Vienna of Beethoven’s day.

A peculiar situation therefore prevails. Professional musicians who would never dream of tinkering with the harmonies or melodies in the music feel free to experiment with the tempos. Sometimes the results can be excellent. A wrong tempo can be more affecting than the right one in the right hands. Sviatoslav Richter’s Schubert sonatas are famous for being almost perversely slow but they are moving and sincere. On the other hand, it is frequently just a gimmick for a musician to fool around with tempos. The easiest way for a player to make an impression on an unsophisticated audience is to play faster or slower than anybody else. Glenn Gould, his eccentric talents notwithstanding, is the obvious example.

Rosen’s book is written in clear, logical prose, suited to serious readers with basic training in music theory. It is full of his familiar intelligence and attention to detail. It is not an introduction to the Beethoven sonatas for the casual listener or record shopper. Rosen isn’t inclined toward poetic or programmatic explanations for the music, as, for example, Alfred Brendel has supplied, idiosyncratically but evocatively, for Beethoven’s Diabelli Variations. At the same time, Rosen’s book has occasional passing characterizations that will be suggestive even to readers without musical knowledge. About Beethoven’s Sonata in A-flat Major, Op. 110, he writes eloquently: “The first movement begins con amabilità, and never completely loses that quality. The scherzo is humorous, folksy, sometimes brutal, and even sardonic. The finale starts with a scena and recitative full of pain, continues with a lament that is, when it returns, literally choked with despair, goes through a condition close to death and ends with a triumphant return of life.”

He adds: “One might imagine that there is a programme connected with this work, and that is certainly possible. Nevertheless, one must distinguish a programme which renders a work of music more intelligible from a programme which is only an inspiration for the composer and is no help to an understanding of the music.” The first case is a work like Bach’s “Capriccio on the Departure of a Beloved Brother,” with its allusion to the coachman’s post horn. The second case would be the ballades by Chopin, sometimes said to be based on poems of Mieckiewicz, although Chopin refused to confirm this. A third case might be the Diabelli Variations, for which a performer like Brendel, as opposed to a composer, makes up scenarios that he finds helpful, which is not Rosen’s inclination. “There is a scenario to op. 110,” Rosen observes, and this thought applies to his approach to other sonatas, “but whether it refers to any real event or literary inspiration we do not know, and it would not help us either to play the piece or listen to it if we could find out.”

So despite evocative passages that the general reader can appreciate, this is a book by a pianist for other pianists, musicologists, and serious amateurs, about the tasks facing the player, and inevitably it casts light on Rosen’s own playing. Artists, when they talk about other artists, are talking about their own work to some degree. And in Rosen’s case, by writing about Beethoven he not only explains why he plays Beethoven as he does but, at least implicitly, justifies a general approach toward the instrument—for him, a combination of scholarship and muscularity, sometimes in interesting competition. It’s rare to find a concert pianist who can explain, or has even bothered to try to explain, in other than vague or fuzzily poetic terms what is going through his mind when he sits down at the keyboard to perform a piece of music. Rosen’s book, at least indirectly, is about one musician describing his artistry through another’s.

His authority lends assurance to his arguments, but the book is not, Rosen stresses repeatedly, an attempt to dictate how the sonatas must be played by other people. “I have emphasized, and emphasize here again,” he states on the first page of the preface, “the freedom necessary for interpretation, a freedom that Beethoven himself expected…. There are frequently good reasons for disregarding a composer’s intentions. But there is no reason for not trying to find out what they were.”


True enough. This book’s value has to do with the general common sense of its arguments combined with the sensibility that a performer, and perhaps only a performer, brings to the challenge of interpretation. Every player tries to make the best case for an interpretation, knowing that it is never the only way to play the music. True music-making is not about being correct, it is about being simultaneously true to the score and to oneself, which are not automatically the same thing.

Rosen argues, first of all, for just looking at the music, which sounds simpler than it is. Beethoven’s markings were sometimes hard to read and engravers over the years, even during his lifetime, made mistakes copying what he wrote. A tendency of players has been to normalize discrepancies: when one passage indicates a certain phrasing or length of note, all similar places are presumed to follow suit. Wrong, says Rosen. We must be careful. Much of his analysis of tempos, accents or stress marks, slurs, and other indicators of phrasing is a plea for doing what pianists too often ignore or discount. Obey the inconsistencies (at least generally: Rosen hedges), because Beethoven knew what he wanted.

A result of this, Rosen says, is that players and audiences will need to readjust their ears to different articulations of sound, more faithful to Beethoven’s intent: less fluid, more abrupt, less rounded, more transparent and varied. Carl Czerny, Beethoven’s pupil, once recalled that Beethoven thought the eighteenth-century style of playing, Mozart’s style, was “too choppy.” At the same time, Beethoven clearly was not anticipating the smooth, interconnected sound of Chopin’s day. Rosen is trying to situate Beethoven’s goal in between:

His music and its phrasing did much to prepare the new era while still exploiting so much of the old style in which he had been reared. If we wish to grasp his conception of phrasing, we must learn to enjoy the more detached sound that he still used at times, and the consequent sense of air and transparency that he inherited from the style of his earliest years.

This is one of Rosen’s most provocative points. He cites the end of the famous “Pathétique” Sonata, Op. 13, demonstrating the relevant passage on a compact disk he has produced to accompany the book. The disk features several dozen excerpts, and the entire Sonata in F Major, Op. 54, played on a late-nineteenth-century Bechstein at the Villa Caetani in Ninfa, Italy, a piano closer to an instrument of Beethoven’s experience than a new Steinway would be.

For the last movement of the “Pathétique,” Beethoven writes a half note at the end of the first phrase of the principal theme when it initially appears. The half note indicates a sustained, singing sound. But at the close of the movement the same phrase is followed by rests, or silence, changing its shape, making the end of it, as Rosen says, sharper-angled, more abrupt. The final note of the piece is then to be played strictly as written, he insists: a sudden attack and then nothing.

This may be true to the score but the effect on the disk is jarring—even more so with the abrupt interpretation Rosen puts forward for the scherzo of the Sonata in A-flat Major, Op. 26, which audiences today are accustomed to hearing as whimsical music. We gain from this change a more dynamic and varied articulation of notes, but modern listeners may perceive the enhanced fidelity to the score coming at the cost of tonal sheen, with “choppiness” replacing elegance, and hiccups of sound occurring where smooth phrase endings typically were. Of course, this being a scherzo, boisterousness may be more apt than refinement, and beauty is a matter of shifting taste. But the new sound, as Rosen demonstrates it on the disk, is not graceful.

“We need, in short, to work up sympathy for ways of playing—and hearing, too—which feel alien to us,” Rosen reiterates. Maybe so. Habits of listening have certainly changed over centuries, affecting what we perceive to be musically sensible and even authentic. Rosen has written pointedly about the early music movement, making a distinction between fidelity and authenticity. In Critical Entertainments, a recent collection of essays, he wrote:

The old ideal of fidelity demanded that the performer try to infer the composer’s intentions, and realize them with the least possible distortion. In a faithful interpretation, the performer’s own personality and his need for expression come into play…. Authenticity dispenses with all this guesswork…. Instead, we ascertain how [the composer] was played during his lifetime, in what style, with which instruments and how many of them were in his orchestra. This substitutes genuine research for sympathy, and it makes a study of the conditions of old performance more urgent than a study of the text.*

Rosen seems to want to stress fidelity in the case of interpreting the Beethoven sonatas, but relies also on authenticity to marshal information about how music was played in Beethoven’s day (and in Bach’s, Mozart’s, and Chopin’s), and on which instruments. He concludes, not surprisingly, that what distinguishes Beethoven from sonata composers before him (and most composers after him) was the sheer, astonishing density of his ideas: the extent to which, unlike others, he exploited his instrument by varying touch, dynamics, and phrasing to add drama and breadth to the music. This is a big reason why his sonatas are so rich and, with the grandest of them, on an entirely different level of ambition than sonatas by Haydn and Mozart. Fidelity to Beethoven’s complex markings, even at the risk of offending current received taste, acknowledges what makes the music a radical statement, Rosen says. This is really the heart of his argument.


Players and musicologists will find dozens of intricate points to grapple with in Rosen’s book. I recall studying Mozart’s Sonata in G Major, K. 283, as a boy. My teacher, Seymour Bernstein, pointed out to me how, in the second movement, which is full of two-note slurs, Mozart occasionally took the trouble (in measure nine, for example) to write the slur as a sixteenth note followed by a thirty-second note followed by a thirty-second-note rest. This, my teacher suggested, was Mozart’s elaborate way of stressing that at that particular measure the player should detach the second note of the slur from the tone that follows. The inference was that other slurs—which did not include this complicated notation—were not to be played the same way: the sound was to be more connected, more mellifluous. Rosen, however, says that during the eighteenth century the general presumption was that the last notes of two-note slurs were played shorter than written. They were detached from succeeding notes; which leaves me wondering why Mozart wrote what he did in that case. Whom should the conscientious interpreter believe? Can both Rosen and my teacher be right?

Should musicians from now on follow Rosen’s recommendations closely, we will certainly hear many familiar sonatas in new ways. A dramatic example, on which he lavishes particular attention, involves the tempo of the finale of the Sonata in F Major, Op. 54. The finale is marked Allegretto. Pianists typically play it quickly. If you turn on a metronome when you listen to a recording you can determine the tempo at which each beat is played. A metronome mark of 60 equals one beat per second; 120, two beats per second. Artur Schnabel and Maurizio Pollini on their recordings play this finale with the metronome set between 126 and 144 to the beat (Czerny also suggests 144), which is very brisk, especially since the movement ends with a section marked Più Allegro, or “faster.”

Rosen pleads for a speed far, far slower than that. He arrives at that idea by the following argument (it is one of the major points in the book). Today, he writes, musicians and musicologists disagree even about the basic meaning of conventional terms like Allegro or Andante in eighteenth- and nineteenth-century music. But during the second half of the eighteenth century, tempi ordinari, or standard tempos, existed. These tempos might vary between places (between Vienna and Venice, say) but they would have been commonly understood by musicians working in the same milieu. Beethoven would have known what Mozart and Haydn assumed to be standard tempos, even if he wished to depart from those tempos, as seems his intent from a letter he wrote to the publishing firm of Bernhard Schott & Sons in 1826: “We can scarcely have tempi ordinari any longer, since one must fall into line with the ideas of unfettered genius.”

But what, precisely, were these tempi ordinari? Rosen tries to establish what he thinks the tempo ordinario of an Allegretto was to Mozart. He looks at the Allegretto finale of Mozart’s Piano Concerto in G, K. 453. This is a set of variations written as a progression in which there are more and more notes per beat, the basic tempo never changing, the progression therefore creating a kind of built-in accelerando or speeding-up. The fastest speed at which the variation in sixteenth notes can be clearly articulated by both soloist and orchestra is seventy-six to the half note, Rosen decides. This means, extrapolating backward and keeping the tempo intact, that the opening theme in quarter notes must be played at what we would today consider a moderate, gentle speed.

Likewise, Rosen demonstrates the opening measures of the Allegretto of Mozart’s Sonata in D Major, K. 576, at a metronome mark of seventy-six to the quarter, on his disk accompanying the book. It also sounds slower than it is often performed, but charming. Pianists, he declares, “can perform the last movement of K. 576 in public at the proper tempo and forgo public admiration for their technique; they can play it at a brilliant, rapid tempo and miss the expressive quality inherent in the music; or they can play another sonata instead.”

This amusing statement is perhaps more dictatorial than Rosen means to be. He has just made a new recording of various works by Mozart for Fonè, on the same Bechstein at Ninfa. (The piano, by the way, was given to the Caetani family by Liszt.) The CD includes this Mozart sonata. Rosen again begins the Allegretto at seventy-six. But now he is playing the entire movement, and immediately after the first bars he picks up the speed, during some of the showier passages, and the tempo varies throughout the performance, naturally. A tempo of seventy-six, despite what he says, clearly is meant as a starting point and guidepost, not as a straitjacket.

Dogged adherence to any tempo throughout a movement is pedantry, suppleness being the essence of true musicality, as Beethoven himself pointed out. Rosen’s method of determining a tempo seems sensible, with the understanding that some deviation is allowed. The backward-extrapolating approach he employs for the Concerto K. 453 is particularly perceptive and sometimes overlooked by pianists. I am reminded of Haydn’s remarkable F Minor Variations, for example. The opening, marked Andante, is generally played at the speed of a march, sometimes almost flippantly. But the tempo is difficult to sustain without seeming to rush the music by the variations in thirty-second notes and thirty-second-note triplets. Extrapolating backward from those thirty-seconds, a performer can arrive at a slower opening that reveals the gravity of the music.

Something at or around the same moderate speed seems surprisingly plausible for most Beethoven Allegrettos, despite the fact that these movements are often finales of sonatas and symphonies. Rosen ascribes our instinct for fast speeds in finales to evolved concert behavior. During the 1820s, all the movements of a Mozart symphony would still be applauded separately. “Conductors therefore did not yet find it necessary to whip up enthusiasm with the finale, which could retain its less brilliant character,” Rosen writes. But this tradition had weakened even a generation after Beethoven, which suggests to Rosen why Czerny, who died in the 1850s, may have given brisker metronome indications for Beethoven’s music than Beethoven expected.

And this brings us back to the Allegretto of Op. 54, for which Czerny prescribed the swift metronome mark of 144. If we ignore Czerny and follow Rosen, the movement changes from a rapid finger exercise, like a toccata, into something surprisingly lyrical and melodic. One of Rosen’s most intriguing demonstrations on the accompanying disk is of this movement. His tempo is around the majestic, flowing pace Beethoven indicated for the Allegretto of the Seventh Symphony, the same as Rosen illustrates for several other Allegrettos. Pianists who resist the temptation to show off may find performing Op. 54 at this slower pace revelatory.

Beethoven’s Piano Sonatas recently inspired me to listen again to an old Rosen recording of another Beethoven sonata. During the late 1960s and 1970, he recorded the last six sonatas. Those recordings were reissued several years ago as a set of two compact disks by Sony Classical. They are among the best recordings Rosen has made. The final sonata, Op. 111, begins with a sixteen-bar introduction marked Maestoso—majestic. It leads, via a rumbling trill deep in the bass, at first quiet then growing in volume, to the main section, an Allegro con brio ed appassionato. The Maestoso’s opening notes, a rapid downward leap of octaves, from E-flat to F-sharp in the bass, the distance of a diminished seventh, instantly announces the grave and revolutionary tenor of the work. On the recording, Rosen plays this introduction expansively, at the slow pace of an Adagio. The transitional trill then quickens the speed to the Allegro, which is tempestuously rendered at more than twice the pace of the Maestoso.

This is a conventional view of the relative tempos of the two sections. Richard Goode on a recording for Elektra Nonesuch plays the Maestoso a bit slower than Rosen and the Allegro even faster. Stephen Kovacevich, on another splendid recording, from 1973, recently re-released by Philips as part of its Great Pianists series, makes the Maestoso into a very slow Lento, then plays the Allegro not as fast as Goode but with a gradual hastening after the first measures. The tempo is elastic, in other words. It breathes.

But in his new book Rosen discounts his own playing and almost everyone else’s. Recognizing that the transitional trill changes from thirty-second notes to sixteenth notes as it crosses the bar from Maestoso to Allegro, he concludes that Beethoven must have wanted the first section to be played at precisely or almost precisely half the speed of the second: “Performances generally begin with a tempo four times as slow as the following Allegro. I used to do this myself because that is the way I had always heard it, but at least I gradually accelerated the trill into the faster speed so that a semblance of uniformity was achieved. I now believe the very slow tempo to be wrong.”

He demonstrates the quicker tempo on the disk accompanying the new book. It seems irrefutable, in principle. But the impression left is that the Maestoso, played faster, is not as moving and soul-searching as the “wrong” version on the old recording. Perhaps the sense of expansiveness in the slower opening is a matter of an acculturated misperception about slowness predetermining what sounds soul-searching.

But I also think, at heart, we respond to where an interpretation comes from. Does it stem from the player’s desire to be faithful to a particular relationship of tempos or to an overall sense of the music? If our modern ears hear the slower opening tempo as grander, and moreover, if playing more slowly inspires the pianist to perform more sensitively, is the correct tempo truer to Beethoven’s goals or not?

There is no answer in the end, of course. “For many works, we might well claim that the most significant virtues of Beethoven’s music may be better captured by the performance tradition worked out by decades of experience with the music than by a simple attempt to recover the practice of Beethoven’s contemporaries,” Rosen says.

Right. Brahms was famously reluctant to supply metronome markings for his intermezzi precisely because, he said, he was not so foolish “as to play them the same way every day.” However we ultimately decide to play and hear the Beethoven sonatas, Rosen’s excellent book at least compels us to think freshly about them.

This Issue

May 23, 2002