What Charles Rosen does to the concept of sonata form in this book is not so much deconstruction—though there is some of that, as we shall see—but rather reconstruction. The old textbook definition of sonata form has been hammered away at on many occasions; Rosen himself did a pretty thorough job of pulverization in The Classical Style of 1971. Now he builds a whole new book out of the dust. It is a paradoxical and brilliant performance.

Inasmuch as the first and often other movements of nearly every sonata, trio, quartet, quintet, and symphony by Haydn, Mozart, and Beethoven are “in sonata form,” the meaning, status, and implications of the term are obviously important to anyone interested in this music. Important in different ways, however, to listeners and critics today and to the music theorists of the early nineteenth century. The eighteenth century knew neither the term nor the definition, and Rosen traces with exemplary clarity the reasons why sonata form, when it finally got defined, got defined so poorly. The nineteenth century needed a prescriptive, didactic formula that would enable composers to emulate the classics—in particular, to emulate Beethoven. And since composers of the time cared above all about melodies, themes, and motives, the basic conflict that generates sonata form was interpreted as a conflict of themes. A pre-sonata piece by Bach—a fugue, typically—has but one theme; a sonata piece by Beethoven has two, sharply contrasted. People liked to speak of the “first theme” as masculine, the “second subject” as feminine. The fact that these adjectives were attributed (fraudulently, as has recently been shown) to Beethoven himself reflects clearly enough the classicizing, authoritarian impetus behind the original formulation of sonata form.

Any modern impetus must be historical or critical. We no longer need a pattern such as sonata form for composing in the high style; we need simply an aid to understanding Haydn and Mozart (and Beethoven). Nineteenth-century theory was never intended to provide this and so was soon found wanting, and was hammered at; later critics typically began to look below the surface level, the level inhabited by themes. In his famous “Sonata Forms” article in the eleventh edition of the Encyclopaedia Britannica, Donald Francis Tovey dealt brusquely with the thematic question by printing the first half of the Gigue from Bach’s C-major Cello Suite, which contains at least four distinct themes, next to the first-movement exposition of Haydn’s Quartet in D minor, op. 42, which contains only one. (Compare Rosen’s use of Bach’s B-flat Partita and Haydn’s Trauersymphonie.) Sonata style does indeed depend on conflict and the resolution of conflict, said Tovey, but it is not conflict of themes but a deeper-level conflict of tonality. What is fundamental is the process of modulation—Haydn’s change from the key of D minor to F major, and especially the “dramatic” way in which he makes this change perceptible.

In short, revisionists early in this century substituted keys for themes as the basic structural determinant of sonata form. This, by the way, allowed sonata-form analysis to be extended from sonatas and symphonies to concertos, opera ensembles, and even items of church music. Rosen goes further and emphasizes the function rather than the pattern of music elements, drawing on a central principle of “symmetrical resolution” which applies to tonality, themes, rhythm, texture, and nuance. Sonata form, he insists, depends on the wonderfully flexible new classical style developed by Haydn in the 1770s, a style so sensitive that almost any musical gesture that draws attention to itself asks to be confirmed, to be accounted for, and ultimately to be resolved symmetrically—by which I take Rosen to mean that gesture and resolution can be clearly perceived to be in some sort of temporal balance. Admirers of The Classical Style will find another masterly exposition of this central insight in the chapter entitled “Motif and Function,” with its fullscale analysis of the first movement of Mozart’s Prague Symphony.

Rosen is not one to overemphasize tonality, then, in the way that old writers overemphasized themes, but he does not slight it either. Some of his best writing deals with this important matter. And it is probably true that this matter more than any other eludes modern listeners as they try to deepen their appreciation of classical music. Themes, rhythm, texture, and nuance are all easier to hear than keys; perhaps as a result of the atrophy of classical tonality recounted in Rosen’s final chapter (on “Sonata Form after Beethoven”), people seem to be less sensitive to niceties of relative key quality and modulation today than they presumably were in the late eighteenth century. That is why Tovey spent so much time trying to “explain” tonality (“The curtain of hail is lifted away into blue sky, and we find ourselves in the very key in which the development started,” “Schubert in the slow movement of the Quintet produces a mysterious brightness by going from E to F sharp,” etc.). If most of the time he must have failed in this, at least he never missed a chance to assure his non-professional readers that the structural tonal elements of the sonata style existed within the sphere of their own perception and experience. Tovey strove to sophisticate his “naïve listener” by repeatedly drawing attention to aesthetic effects achieved by means of tonality.


Rosen does not feel the need to offer such assurances or make such explanations. He writes for readers who either share his ear for tonality or more often, I expect, take him willingly at his word. Yet in so many ways he resembles Tovey (both pianists, prodigies, classicists in music, formalists in criticism, anti-academic academics, stylists, connoisseurs, legendary conversationalists); and the fact that Rosen writes down to his readers less than Tovey did does not mean he believes any less that keys derive their aesthetic character from musical processes which can be followed over a span of time, such as modulations, preparations, or juxtapositions with other keys. Writers such as Rosen and Tovey adhere to what has been called the “gignetic” view of tonality, as opposed to the “ontic” view—tonality as a static system, abstracted from time. (These may not be the best terms, but they are ones that have recently been brought to the controversy, by Siegmund Levarie.)1 The quality of a key is defined contextually, by how it is approached: thus in the first movement of Beethoven’s Eroica Symphony, the F-major horn passage in the recapitulation section does not sound like an enhanced dominant of E flat but (Rosen almost says) an enhanced subdominant, because of the way Beethoven leads into it. And this carries Rosen’s discourse forward:

The contradiction of the normal significance of [the enhanced dominant, F major], its transformation into its opposite, makes this moment one in which all genuine harmonic motion appears to stop: the dynamics, piano and dolce, reflect the cessation of harmonic energy. The main theme is poised over these harmonies for a few bars. This is a crucial passage: the main theme is essentially a horn call in its character, but the horn has yet to play it solo. It tried to, a few bars before, in E flat, but was cut off brutally in mid-theme by the full orchestra. It plays the theme at last, but with a different sonority—not the natural E flat sonority of the instrument but F. The swift move to D flat major, and the delicacy of the flute solo confirm the exoticism of this section.

This is criticism building on what we actually hear in music, even on what we feel about what we hear (though to my taste Rosen says too little about that).

The distinctiveness of this kind of criticism, criticism based on experience and effect, becomes clear if we compare Rosen with a typical academic writer, or even with one who is more perceptive and flexible than most. David Epstein, for example, discusses the Eroica recapitulation in his recent analytical study,2 but all he wants to point out is how the bass traces the interval of a descending third, which as E flat-C (an extension of the interval E flat-C sharp at the beginning of the piece), leads into the F-major horn entry, and then as D flat-B flat also leads out of the D flat flute entry. This contributes one detail to an undoubtedly impressive demonstration of how a “basic shape,” in Schoenberg’s term, generates all parts of the composition in all musical “domains” and on all levels. Of the strange quality of F major in this context, the cessation of harmonic energy, the delicacy and exoticism, Epstein sees or says nothing.

Concern for musical experience and effect also helps to explain the paradoxical fact that Rosen has written a book on this particular subject. Sonata style, sonata principle—these have been the preferred terms in recent years, and The Classical Style included a strong attack on the canonical concept of sonata form, as has already been mentioned. But the way we appreciate this music, Rosen insists, is still through our experience of form: not any single, simplistic sonata form but multiple sonata forms, as different from one another as snowflakes all inspired by the same guiding principle. This reconstruction of sonata forms will probably cause as much raising of eyebrows (and clenching of fists) in music-theoretical circles as Rosen’s late entry into the field of mid-eighteenth-century music history will in musicological ones. Compare Epstein, once again—who in his discussion of the Eroica does not exactly ignore sonata form: in fact he points out (quite in Rosen’s spirit) how the descending-third bass figure articulates the major formal divisions. Yet his overriding, “ontic” concern is with how basic shapes control every gesture in the movement, large or small, irrespective of its location in the form. No follower of Arnold Schoenberg or Heinrich Schenker would ever dream of undertaking the dizzy taxonomy of multiple sonata forms that occupies the center of Rosen’s study.


There is much to learn and much to quibble about in this taxonomy. Rosen proposes First Movement Sonata Form, Slow Movement (or Aria) Sonata Form, Minuet Sonata Form, and Finale Sonata Form, and then goes into subtypes, three (or more) mid-century “stereotypes,” a miscellaneous series of eleven expository devices, and so on. Readers who have the musical training and the enthusiasm to follow all this will certainly be convinced of the wonderful multiplicity of formal solutions achieved by the classic masters within the guiding principles of sonata form; though perhaps it should also be added that such readers will probably not need convincing. Not only does Rosen give us this snowstorm of sonata forms, he also rearranges the basic crystal structure. Often the tonic establishes itself firmly long before the Recapitulation. Usually a Second Development comes directly after the recapitulation of the first theme. In Minuet Sonata Form the first half can close in the tonic and the fundamental contrast of key and/or material occur in the second. All this will excite debate; it will also enhance perception.

Rosen’s growing involvement with history can be followed in his articles and reviews for the NYR over the past ten years. The strong historical bent of Sonata Forms may still surprise those who remember the anti-historical or at least anti-musicological polemics of The Classical Style. Rosen’s strategy now is to trace the evolution of many different early eighteenth-century forms and genres in order to show how all contribute something to the classical synthesis. Since there is no single sonata form, there can be no single or simple history of it. In Rosen’s first book we were given a magnificent view of three great mountain peaks looming up out of an ahistorical fog. It is both a pleasure and a relief to find in the second that some lesser promontories such as Carl Heinrich Graun, Niccolò Jomelli, and the Bach sons, and even some valleys such as G.B. Sammartini, are now exposed and illuminated.

The book is less useful, I fear, about the nineteenth century. In his introduction Rosen gives clear warning to those who, he says, are likely to pick up his book in the hope of learning the history of sonata form—its origin, development, and fate:

These questions seem reasonable enough, on the face of it; but as they are generally put, they are doomed to remain unanswered because they make untenable assumptions…. They assume that a form has a history—in other words, that it is subject to change: but if a form “changes,” it is not clear when it would be useful to consider it the same form, although changed, and when we must think of it as a new form altogether.

Rosen himself has changed enough since The Classical Style to see in the prehistory of sonata forms a fascinating record of individual stylistic elements grasped here and there and sometimes combined, sometimes not, in ways that anticipate the classic solution. But for Rosen sonata form depends on the classical style; he does not see how the form can be squared with the evolving style or styles of musical romanticism. One by one Schubert, Schumann (in a lengthy discussion of the F sharp-minor Sonata), Chopin, Mendelssohn, Brahms, and Bruckner bear melancholy witness to the decline. Bartók’s Fifth Quartet “is not so much a new version of sonata form as a brilliant twentieth-century metaphor for sonata form.”

And Beethoven? The nineteenth-century prescriptive stereotype of a single sonata form based on themes, against which Rosen so tirelessly battles, was imposed by theorists closely connected with this composer. Reicha, Czerny, and A.B. Marx molded their stereotype to fit certain Beethoven works, the stifling prestige of which tempted nineteenth-century composers into endless ill-considered bouts of imitation. According to Rosen’s little deconstruction Beethoven himself must be seen leading the decline, however unwittingly and indirectly, and Rosen returns icily to this fact in a late chapter entitled “Beethoven and Schubert.” There is much material about Beethoven in this book, all of it illuminating, but not much in this chapter, in which Rosen reminds his readers that he has “treated Beethoven throughout as if he were a late eighteenth-century composer” and seems mainly concerned to restore the composer to a state of pre-Marxist purity.

Above all, Beethoven retains the classical sense of the resolution of large-scale dissonance by the reestablishment of a symmetrical equilibrium…. The emotional climate of his music, of course, is that of the Napoleonic and post-Napoleonic eras, and so is its ideological content. Jens Peter Larsen insists that a stylistic cut must be made between Beethoven and his predecessors because Beethoven’s music is so much louder than theirs.

—but Rosen disagrees. This is droll but possibly unfair to Larsen. In the Eroica, once again, Rosen remarks on the largescale dissonance produced by the E-minor development theme, and on its resolution in the coda, where this theme

is heard first in F minor, the relative minor of the subdominant, and then in the tonic minor (E flat) as an introduction to an incredibly long passage that does nothing but repeat a V–I cadence over and over again in E flat major.

Something appears to have gone wrong with symmetrical equilibrium here. What is so “incredible” about this musical passage, so unlike Haydn or Mozart? Not just the loudness, surely. Not just the length. Rosen makes a rare and rather revealing slip when he says the passage does nothing but repeat a cadence; it does nothing less than resolve the main theme, played at last by the E flat horns. But it also provides an emotional resolution or rather an apotheosis, in which Beethoven for the first time, perhaps, expresses his response to a poem that we know attracted him in his youth:

Laufet, Brüder, eure Bahn,
Freudig, wie ein Held zum Siegen!

The ideological content skews the form, makes it incredible. In a symphony called “heroic” the emotional climate cannot, I believe, be separated off from form and style as clearly as Rosen would like.3

Shortly after the Eroica Beothoven wrote another equally incredible passage in the Appassionata Sonata, the new Presto theme in the coda of the last movement. Whether symmetrical or not, the resolution it provides can only be conceived in terms of the total work—whose three movements Beethoven has indeed linked and related in an unusually close way, as is well known. Probably the most serious lacuna in Sonata Forms is any consideration of the nineteenth century’s extension of sonata principles over total works, not just movements. One can share Rosen’s distaste for the end-product of this tendency in the so-called cyclic form of Mahler and Tchaikovsky, but still admire it in Beethoven. To judge from the way Rosen plays Beethoven, so does he. He has written about it, too, very well, in his big analysis of the Hammerklavier Sonata in The Classical Style.

Another lacuna—almost bizarre in a seemingly methodical book of 350 pages on sonata forms—is any sustained discussion of the sonata-form coda. Only one short routine paragraph is to be found on this necessary topic. It is of course in his codas that Beethoven makes his musical apotheoses, so different in spirit and form from the symmetrical resolutions of Haydn and Mozart.

But if Rosen gives us a picture of Beethoven seen through a filter, and a Brahms sadly underexposed, his Haydn and Mozart are wonderfully textured. Nobody writes better than Rosen about the music of these composers. Nobody writes better about music. One may think at first that he is just extremely ingenious with examples, finding exactly the right Haydn symphony of the 1770s or Mozart aria from Zaïde to make his points brilliantly. Then comes the Prague Symphony and the Harmoniemesse and the Emperor Concerto, and one realizes that to familiar and unfamiliar music alike Rosen brings not only an uncommonly refined ear and sensibility but also, again and again, unerring insight into just the features that make the music special and fine. That is why his writing can be so economical. He is helped by a plenitude of music examples that is without precedent in book publishing, as far as I know; more than half of this book consists of music.

This Issue

October 23, 1980