Each sonata by Beethoven has its own particular character. But is this really anything more than a platitude? Should we still be clinging to such concepts as “character” and “atmosphere”? Aren’t the musical cognoscenti interested primarily in understanding “structure,” leaving something as vague as “poetic associations” to amateurs? And haven’t the poststructuralists long since exposed “character” as a mere illusion?

Arnold Schoenberg, whom no one would accuse of being an amateur, recommended that

in composing even the smallest exercises, the student should never fail to keep in mind a special character. A poem, a story, a play or a moving picture may provide the stimulus to express definite moods. The pieces which he composes should differ widely.1

When the concept of character first began to emerge in writings about music around 1795, it was intended as a corrective to Kant’s relative deprecation of instrumental music. In his Critique of Judgment (1790), Kant had declared that music took “the lowest place among the fine arts,” because “it plays merely with sensations.”2 Music, according to Kant, was an agreeable (angenehm) rather than a fine art. Writers like Christian Gottfried Körner and Christian Friedrich Michaelis subsequently came to the aid of music and in the process consistently drew on the sonata to support their arguments.

Eighteenth-century listeners perceived the sonata as a remarkably private genre, in comparison with the Baroque suite. In place of a succession of more or less formal dances, sonatas now appeared to be “like studies of the various attitudes and passions of man.”3 Even the minuet—the only one of the suite movements to find its way routinely into the symphony, string quartet, and sonata—was capable of taking on a variety of characters, whether gracious or impetuous, solemn or humorous. This distinctive element of what could be called the humane, the personal, or the individual characterizes the sonata more aptly than the presence of any so-called “sonata-form movement.” After all, there are plenty of sonatas without even a single such movement.

From works on aesthetics written just before 1800, we know that musical character was perceived as consisting of “psychological” and “moral” components, and as mediating between these two supposedly contrasting spheres. Körner’s essay “On the Representation of Character in Music,” which appeared in 1795 in Schiller’s journal, Die Horen, goes beyond this to speak of what amounts to a masculine and feminine Ideal.4 (We might recall the report of Anton Felix Schindler, Beethoven’s secretary and biographer, that Beethoven himself spoke about the “two principles”—masculine and feminine—in music.5 ) Körner goes on to say that “the concept of character presupposes a moral life, diversity in the use of freedom, and in this diversity a unity, a rule within this arbitrariness.”6

Carl Czerny, Beethoven’s pupil, tells us that every one of Beethoven’s compositions “expresses some particular, consistently maintained mood or perspective to which the piece remains true, even in its smallest details.”7 This assertion would have found favor with Schoenberg. That Schoenberg himself adhered to this advice in his own compositions can scarcely be doubted. His String Quartet op. 7 is based on an amazingly detailed psychological program that he set down in the early stages of its composition. It lists “Rebellion; defiance; longing; rapture; depression; despair; apprehension of sinking; unfamiliar emotions of love, desire to be engrossed; solace; relief (she and he)”; and so on. And this is only part of the first movement.8 Like the titles for the movements of his Piano Concerto, this program remained unpublished during Schoenberg’s lifetime.9

In a radio broadcast, Schoenberg once remarked about his musical works: “I cannot disclose their spiritual background[s], and I would have to assert that I am inclined to be at some pains to disguise them.”10 This strangely twisted formulation reveals Schoenberg’s conflicting emotions, or, more precisely, the conflict of his emotions with a musical rationality that presupposes a work of music to be convincing without the aid of any descriptive device. Czerny’s comments on Beethoven’s Piano Sonata op. 81a—“Les Adieux”—are very much along these lines: “Moreover,” Czerny writes, “this sonata…may, and indeed should, be of interest even to those who are willing to enjoy it as pure music, without regard to the titles.”11 Yet Bee-thoven’s goal, in this sonata, had also been to set to music, as clearly as possible, feelings or sensibilities—Empfindungen—evoked by Farewell, Absence, and Return, the titles of the individual movements.

These titles provide clues toward understanding the music’s character, not only for the listener but above all for the performer. (They also made it possible for the dedicatee, Archduke Rudolf, himself a musician, to judge whether Beethoven had successfully expressed feelings associated with the titles.) Rather than deny the element of character in the music of Beethoven or Schoenberg, we should be thankful for these kinds of verbal aids. If Schoenberg conceived the first movements of his Piano Concerto under the rubric “Life was so easy,” then this will perhaps deter performers from playing the piece all too dreamily. And even if the composer covered up the “background” to his compositions and wished to leave no psychological clues, it must be granted that there is occasionally a kind of musical coherence that is above all psychologically motivated. One of the most delectable tasks of the interpreter is to sense such motivations, even if they are not to be captured in words.


Like the essays by Körner and Michaelis, Beethoven’s first piano sonata, op. 2, no. 1, in F Minor, was published in 1795. The need for “unity in variety,” as expressed by contemporary writers on aesthetics, is fulfilled in this work in a pointed manner. Compared with the plethora of ideas in the other two sonatas of op. 2, the variety in this first sonata seems to be rather tightly reined in. By contrast, the unity of deriving all the themes from the motifs stated at its beginning is clearly established in op. 2, no. 1. So, too, is the Beethovenian technique of foreshortening, a device that progressively breaks down harmony, rhythm, and melodic elements to create long spans of tension.12

Directly or indirectly, all the themes of all the movements in op. 2, no. 1, are derived from its very beginning. Beethoven did not aspire to such concentration in every sonata. A work like op. 26, for example, appears to be so loosely constructed from a thematic point of view that one can speak justifiably of a “psychological composition,” as Edwin Fischer put it, a work that reveals itself as a whole through an understanding of the relationships among the characters of all movements.13 On the other hand, the musical materials in some of the most ambitious and moving sonatas, like op. 10, no. 3, op. 31, no. 2, op. 57 (“Appassionata”), op. 106 (“Hammerklavier”), and op. 111 display a particularly strong sense of economy.

In Beethoven’s music, the motifs on which a work feeds are almost always stated at the very beginning; in keeping with this, a movement’s fundamental character can be found in its first few measures or lines. The Sonata in D Minor, op. 31, no. 2, begins with a theme whose three different motifs correspond to three different tempos. Seldom has musical material been presented to the performer and listener so clearly. The tripartite structure of this theme contains at the outset a (solemn) broken chord, at the end an (expressive) embellishment, and in the middle a layered figure on the notes A-F-E-D, which contains something like a genetic code for all the themes of all the movements, and which reappears repeatedly, either openly or in disguise.

The two other motifs are also to be found in all the themes of the D Minor Sonata. Only the opening theme of the third movement dances out of line: the embellishment is omitted here. But it soon reappears in the second theme of the movement with a persistence that is downright obstinate.

The concentration of materials in the “Hammerklavier” Sonata, op. 106, is perhaps even more astonishing. Everything in this gigantic work seems to be related to the interval of the third: the construction of the themes as well as the most important tonalities of the work are all based on this interval. Charles Rosen has pointed out how the harmonies in sections of the first movement, as well as in the development of the Adagio, progress through a series of descending thirds.14 In the Adagio, we find no fewer than sixty-five progressions by a third, along with another twenty in the Largo introduction to the fugue.

From such examples, it is easy to see that the motifs binding a work do not establish the character of that work. The character (or characters) of the “Hammerklavier” Sonata obviously cannot be derived from the interval of a third. Motifs can contribute substantially to the unity of a work, but they usually do this in an abstract sense. The impression of coherence establishes itself—even if only indirectly—when the listener’s unconscious perception is presented with recurring motifs. But musical “expression”—to which Jean-Jacques Rousseau devoted no fewer than six pages in his Dictionnaire de musique of 1768—is not dependent on motifs or constellations of motifs.15 Rather, expression makes use of them.

The fundamental character that is presented at the beginning of a movement dominates throughout. Later themes or episodes are scarcely capable of threatening its hegemony, even when they provide contrast and variety. One could perhaps point to Beethoven’s Piano Sonata in F Major, op. 54, as an exception. This little-loved, highly original work establishes two contrasting characters at its very beginning. Here is a prime example of what used to be called contrasting masculine and feminine principles; or, to put it in more contemporary terms, animus and anima. But even here, anima, which starts the piece, has the last word, while the concluding movement of this sonata synthesizes both principles.


It is the interpreter’s responsibility to play the roles of different characters. Like every person, it would seem, every sonata has distinct qualities and potentialities. Each character lives and breathes as a sum of its attributes. If the interpreter goes beyond the boundaries of these attributes, a character would be falsified and ill-portrayed. Sometimes this character is marked by contradictions, through two or more souls dwelling in the same breast. (In the writings of Anton Reicha, one of the principal theorists of sonata form, the sections of a sonata-form movement bear the wonderfully dramatic designations of “exposition,” “intrigue,” and “dénouement.”16 ) In the three or four movements of a sonata—provided they do not traverse the exceptionally consistent emotional terrain of a work like Beethoven’s op. 31, no. 3—a wide range of conflicts can arise, yet the finale usually confirms the character of the first movement, albeit somewhat modified. Even in sonatas with a contrasting middle movement, like the so-called Moonlight Sonata, op. 27, no. 2 (“A flower between two chasms,” as Liszt is said to have called it), or op. 31, no. 2, nicknamed “the Tempest” (an angel between two demons, as it might be called)—in other words, in sonatas whose outer movements are entirely in the minor and whose middle movement never leaves the sphere of the major—we experience the bright major mode as the complement that reinforces all the more powerfully the darkness of the minor-key movements.

An exception to this pattern may be found in the two-movement sonatas whose movements suggest opposites. In the case of op. 111, we have turmoil and peace, or the real and mystical worlds. In the Sonata op. 90, we have the comment attributed to Beethoven himself about a “Conflict Between Head and Heart” that is followed by a lyrical “Conversation with the Beloved.”17

The idea of “character” can also be understood in a broader, less personal sense. Rather than “human attitudes and passions,” some sonatas appear to reflect elements from nature surrounding us—nature seen through a temperament. In the “Waldstein” Sonata, op. 53, the impression of space and three-dimensional depth is produced by several factors: the extended harmonic perspective, which now incorporates the mediants as a matter of course; the extraordinary ambit of the principal themes, which incorporate distinctive and unusual intervallic leaps; the broad, liberal use of repeated notes, resonances, and sequences; the new manner of handling a timbre that suggests the qualities of being near and far, high and low, clear and obscure; and finally the extended panorama of dynamics. Exceptionally, a motivic idea also contributes to this impression: the leap of twenty diatonic tones in the opening theme. The performer should present this gigantic interval as the ascent of one single voice; in this manner, our perspective as listeners will be expanded from the very beginning.

The external movements of the “Waldstein” Sonata seem to me like landscapes that unfold before the musical eye. Perceiving this, I would like to let my fantasy run free and imagine for myself how, in the first movement, the horizon lies low with a great deal of sky above it, whereas in the rondo, we find ourselves high in the mountains, listening to a mountaineer’s song, a chant montagnard. In the valleys, there is dancing: both episodes of the rondo are imbued with the character of a Russian dance, like the kind we know from the “Rasumovsky” Quartets, op. 59, and from the coda of the “Appassionata” Sonata, op. 57. The outer movements of the “Waldstein” reach out into brightness; we step outside of ourselves in these movements, whereas the Adagio turns us inward, into the darkness of our natures.

The “Waldstein” Sonata is often perceived as a bravura work. Here, I would like to suggest that a more careful reading of Beethoven’s notation can help lead us to a better understanding of the composition’s psychological elements. The “Waldstein” is the only one of Beethoven’s piano sonatas in which all of the movements begin pianissimo. And perhaps there is no other sonata prior to those of Schubert in which the cumulative span of pianissimo sections plays such an important part. Granted, Beethoven’s pp is not Schubert’s pianissimo espressivo, but rather almost always a pianissimo misterioso (to borrow formulations from Rudolf Kolisch), and only occasionally a pianissimo dolce, as we have in the opening theme of the “Waldstein” finale.

The pedaling Beethoven specifies for this rondo, which has precisely the effect that was forbidden during children’s piano lessons—a lack of harmonic clarity—is all too happily ignored by grown-ups. Without it, the rondo theme takes on a banality worthy of a music box. I am now convinced that the point of this heavily pedaled pedal-point is not primarily the audible maintenance of the bass note, but rather those delicately confused harmonic contours that create an atmosphere of tonic and dominant flowing in and out of one another. To be sure, the result of this should not be opaqueness, but transparent opalescence.

The Sonata in D Major, op. 28 (which not coincidentally has come to be called “Pastoral”), points in two different directions. An introverted Andante in D Minor interrupts the listener’s affectionate enjoyment of country life, which in turn incorporates two thunderstorms. The three-part psychological layout of this Andante deserves closer consideration. The first and third parts (the latter of which is an embellished repeat of the first) alternate between stern composure and sighing lamentations. The middle part, by contrast, surprises us with a bucolic scene in the major mode; in this section, the quality of innocence remains wholly untouched by melancholy or fear.

It is therefore all the more disturbing that such naiveté is turned into anxiety and horror in the coda. What does Beethoven do to lead us into the daylight after the bitter resignation of this movement’s conclusion? Gradually, cautiously, and humorously, he opens our eyes. The entire Scherzo is devoted to the restoration of the major third: it persists with the note F-sharp, which had been the object of near fixation throughout forty measures of the introductory Allegro movement’s development section.

In the Sonatas op. 28 and op. 53, the performer and the listener are invited to share in the composer’s musical communion with nature. But there are also pieces that confront us like a superior power—more elemental than personal, more angelic than human. Both are to be found in the D Minor Sonata op. 31, no. 2. The outer movements are for the most part elemental, while the Adagio is angelic. Czerny, who clearly knew this sonata better than most of the others, speaks of the “tragic character” of the work but also notes “the romantic-picturesque quality of the entire tone-painting.”18 To me, Czerny’s own story about the finale having been inspired by riders galloping past Beethoven’s window seems more revealing than Schindler’s notion of its affinity with Shakespeare’s Tempest, an idea that Czerny ignores. For a performer of the piece, nothing could be more suggestive about matters of rhythm and tempo than the “continuously passionate motion”—as Czerny puts it—of horse and rider.19

Czerny’s indication of the “picturesque quality of the tone-painting,” the painterly and painted effect of op. 31, no. 2, is also useful. Of op. 31, no. 3, Czerny claimed that “this sonata is more declamatory [sprechend] than picturesque.”20 In this work, in fact, only the minuet is lyrical, in its second half even sighing a bit.

What about the first sonata of this group of three? When Beethoven gathers three sonatas under a single opus number, he always wants to show just how different sonatas written within the latest phase of his style can be. Alongside the al fresco, painterly D Minor Sonata and its eloquent, indeed wittily garrulous sister in E-flat Major, the G Major Sonata, op. 31, no. 1, strikes me as dancing, as an expression of bodily movement by dancers or mimes with a distinct tendency toward the grotesquely comic.

The connection between character and structure in this sonata is particularly significant. In and of itself, the structure might appear to be full of inexplicable shortcomings. Only when we take into account the comic character of this work does all the apparent structural nonsense make sense. If one were to grasp merely the character, however, then the formal and structural expectations that the piece violates would remain incomprehensible. Whoever denies the ability of “absolute music” to incorporate comic intentions and effects must necessarily give a poor grade to this piece (an evaluation that has in fact been made repeatedly in critical commentaries). There are plenty of indicators of a comic character in the first two movements. The unique designation of the slow movement—“Adagio grazioso,” almost a contradiction in itself—gives some indication of Beethoven’s ironic treatment of the kind of galant style found in the two early Rondos, op. 51, or in the Violin Romances, op. 40 and op. 50.

Assigning qualities like “speaking,” “painting,” or “dancing” can be helpful in understanding other sonatas as well. To these we should add a fourth possibility, “singing.” This is applicable to those pieces in which singing outweighs speaking, or in which speech is entirely absorbed by song.

Basically, the performer should welcome anything that supports our understanding of the musical characters or characteristics in a given work. One useful concept is that of the four natural elements. The finale of Beethoven’s Sonata in C-sharp Minor, op. 27, no. 2, can easily be related to fire, as can the finale of his Sonata in E-flat Major, op. 31, no. 3, marked “Allegro con fuoco.” The Rondo of op. 2, no. 2, and the second movement of op. 54 could be perceived as flowing. The first movement of the Sonata op. 109 hovers a few inches above the earth: the bass line follows the upper voices as weightlessly as possible. In contrast to this, the bass line of the second movement scrapes into the earth, as it were. The third movement, then, unites earth and air: it both hovers and rests at the same time. Every good performance of a work, incidentally, needs its breath of fresh air, so that the music does not suffocate as if under a bell jar.

Combinations of natural elements are entirely possible in music. In the fugue theme of the “Hammerklavier” Sonata, we can discover other elements alongside the flaming, darting, flickering: the impetus of the stream, the stormy air, and the rootedness of the bass. In a sense, this finale is the most elemental movement that Beethoven ever wrote. It unleashes the elements and yet holds them under strict structural control. This is just one example of how structure and character need not necessarily conform but can rather relate to each other as contrasting features. Not only the fugue, but also the first two movements of the “Hammerklavier” Sonata are manic, the great Adagio by contrast singularly depressive, with ecstatic outbursts.

In attempting to grasp at least a few expressions of a work’s characteristics, its mood, its atmosphere, it helps to be alert to the nuances of language. A great deal—including perhaps the most essential—must remain unspoken. And yet language can stimulate perception and support psychological memory. An awareness of contrasting pairs can clarify a good deal. Opposites like calm/agitated, stiff/flexible, opaque/translucent, active/passive, real/unreal, public/private, resisting/conceding, pious/witty, or sublime/profane help to sharpen our discrimination. In the Sonata op. 110, for example, the sublime third movement grows out of the profane second. It is always the goal of such exercises in consciousness to go beyond similarities and analogies and perceive the peculiar and unusual elements of any given work.

In this regard, let us return to Schoenberg once again. In the chapter “Character and Mood” in his Fundamentals of Musical Composition, we read the following: “It is fallacious to think that the tempo indications determine character. In classical music, at least, this is not true. There is not one adagio character, but hundreds; not one scherzo character, but thousands.”21

This is a highly judicious criticism of what Schoenberg’s friend, son-in-law, and interpreter Rudolf Kolisch tried to demonstrate twenty-four years earlier in his controversial essay “Tempo and Character in Beethoven’s Music.” In that music, according to Kolisch, there are “typical categories of tempo, corresponding to categories of expression.”22 Kolisch, as is well known, associated Beethoven’s surviving metronome indications, along with indications of meter and tempo, with such expressive “types.” Kolisch maintained that the “Allegro ma non troppo 6/8” category, for instance, has a “sombre, passionate character. Motives in 16th-notes alternate with trochaic formations.” Among the examples he gives for this are the finale of op. 28 and the fugue of op. 110.23

To perceive these movements as “sombre and passionate,” however, would scarcely be acceptable even for Kolisch’s most unreserved admirers. I count myself among those who cherish his quartet recordings of 1929; but his comments on tempo and character must be approached with a degree of skepticism. When I visited Kolisch in his old age, he himself seemed not altogether free from doubt on this point. Already on November 16, 1943, Theodor W. Adorno had written to him: “I believe that it is not possible to construct Beethovenian ‘types’ on the basis of an isolated aspect such as tempo, and that this often brings together heterogeneous elements….”24 Kolisch’s pupil and longtime assistant David Satz would later point out that “for Kolisch, as for any other serious musician, tempo was only one aspect of performance; no element of performance was to be neglected at the expense of another.”25 Thus no matter how important tempo might be as an element of character, one cannot determine the character of a given work without carefully taking into account all the other aspects of that work. In Beethoven’s music (as, incidentally, in Schoenberg’s), the metronome indications are not infrequently to be adjusted according to these other “elements of performance.”

Structure and character relate to each other: they may work hand-in-hand, or they may have a relationship of fruitful tension. But interpreters should never assume that understanding the structure of a work might automatically give them insight into the work’s character, atmosphere, or spiritual state. The interpreter would do well to concern himself with structure and character as two functions that emanate, as it were, from different sides of the same work, in the hope that he might one day unite the two at a point where the pain of interpretation can be transformed into the relief of a satisfying experience. That the character of music (or at least of Beethoven’s music) incorporates psychological and moral components, as contemporary writers on aesthetics maintained, is an idea that confirms itself for the performer today. One can talk or argue about the psychological components; about the moral ones, it is better to remain silent. At most, one can attempt to demonstrate them.

Translated from the German by Mark Evan Bonds and the author

This Issue

November 16, 2000