Another article for the Classical Music SIG in Mensa, this one covers briefly some of my ideas about sonata form.

When I was a young piano student, I learned a number of Mozart and Beethoven sonatas from highly-edited versions which included little indicators in the score as to which section of the sonata-allegro form I was currently playing. My teacher helped to clarify the meaning of these indicators, and it was thus that I acquired the textbook definition of sonata-allegro form.

You’re probably familiar with it. In the event you aren’t, here it is in a nutshell. A sonata-allegro form consists of three primary sections, the exposition, development, and recapitulation. The exposition consists of a main theme (or themes) which are stated in the tonic key. This is followed by a transitional passage which serves to modulate into a secondary key, either the dominant (in major modes) or the relative major (in minor modes.). The modulatory passage ends with a half-cadence in secondary key, at which point there is a musical event in the tonic of the secondary key. That musical event is usually a new theme or set of themes; in the most traditional explanations this is a more lyrical theme than the primary theme. This passage is followed by more material in the secondary key, reaching a closing theme (or at least figure) that brings us to a full cadence in the secondary key. The exposition is then repeated. Here’s a little diagram:

Primary Theme (Tonic) -- Modulating Passage to Secondary Key --- Secondary Theme --- Closing Material

This is followed by the development, in which thematic materials from the exposition are recombined, varied, and (in short) developed—with the purpose being a gradual modulation back into the primary key center. Usually there is a period of retransition in which the dominant of the primary key is clearly stated, which leads us into the primary key and:

The recapitulation, which begins with the restatement of the Primary Theme in the tonic key. In the classic textbook formulation, the recapitulation is a more-or-less verbatim repeat of the exposition, the only difference being that the modulatory passage in the exposition remains in the primary key center, thus the secondary themes and closing themes are also in the primary key. At the end of the recapitulation there may or may not be a coda, or musical extension.

So that’s sonata-allegro form as most young students learn it. Stated as I have above, it almost sounds like a recipe, or a template. Fill in the blanks and come up with your own sonata-form composition with a minimum of fuss and bother.

If it sounds like a compositional template, that’s because it was created as a compositional template—i.e., a recipe for composing pieces in sonata form, and not as a description of the practices of the major composers who had written in that form. The final imprimatur to the textbook template was the work primarily of two musicians, both intimately connected with Beethoven: Carl Czerny, one of Beethoven’s best-known pupils, and Adolf Bernhard Marx, a theoretician and musical aesthetician. The sonata-form template they created was based primarily on late Mozart and early Beethoven, and was part of a series of treatises on the craft of composition—‘how-to’ books, in other words.

There’s nothing wrong with the Czerny/Marx template as a tool for composition, but as a model for understanding the compositional practices of 18th-century composers, it’s a very poor fit. An encounter with the composers of the Classical era can be quite shocking if we approach these composers expecting them to write in accordance with the textbook form—because they often don’t. The composers of the Classical era—not only Haydn and Mozart, but also Vanhal, Dittersdorf, Cannabich, Stamitz (both elder and the two youngers), Beck, Kraus, Salieri, Bonno, Monn, Wagenseil, Christian Bach, Abel, et al., had never heard of ‘sonata form’ or its components, just as they had never heard of “Classicism.”

Sonata form was a very different animal to these composers. It wasn’t exclusively a series of events happening at a particular time in a particular way. To a Classical composer, the process was a rhetoric of tension and resolution, imbalance and re-balance. The language included certain thematic events as common syntax, but was not by any means bound by these. It was the tension/resolution that really mattered.

Charles Rosen has referred to the ‘sonata principle’ which seems to me as a much better description of actual compositional practice. The sonata principle is really quite straightforward: we set up a tension between two key centers, usually the tonic and dominant in major keys or the tonic and relative major in minor keys. The secondary key is viewed as being an imbalance set against the potential re-balancing of the primary key. Any notable musical material stated in the secondary key must be restated at some point later in the primary key as a resolution.

That last sentence gives you the 18th-century mindset in a nutshell: any important material stated in the secondary key needs to be restated somewhere down the line in the primary key to resolve the imbalance that has been set up. During the span of the second half of the 18th century, the process began to result in the ‘textbook’ three-section, multiple-theme layout—but in point of fact that textbook structure is a byproduct of thinking in sonata principle, and not the main event itself. Nor has the three-section model ever been completely accepted; many musicologists prefer a two-section model instead which is more concerned with key relationships than thematic materials.

Works of eighteenth-century composers may follow the textbook template to the last detail—after all, that’s how it managed to get itself established in the first place. However, we may well come to grief if we expect a work in “sonata form” to follow the textbook template, and in our distress begin to look for variants or explanations as to why composer X doesn’t seem to be playing by the rules. Our notion of ‘correct’ or ‘standard’ sonata form is at fault here, and not composer X’s skill. It is also a misapprehension to think that a particular musical instance which departs from our notion of the acceptable norm must represent some kind of musical wit or irony. That’s not beyond possibility, but we really do need to try to distinguish what an eighteenth-century audience considered normative, and what we here in the early twenty-first century have been trained to think of as normative.

To help illuminate the above paragraph, consider Haydn’s frequent practice of employing the same thematic material for both the ‘primary’ and ‘secondary’ theme. Put yourself in the position of a Parisian concert-goer in the 1780s, hearing Haydn’s group of “Paris” symphonies (82-87), written for the Concerts des Loges Olympiques. Perhaps if you were unfamiliar with Haydn’s work, you might well be surprised by the primary theme popping up at precisely the point where a contrasting theme were to be expected. But what about the third or fourth time you heard a symphony do that? It certainly wouldn’t seem surprising or witty any more. In fact, you would begin to recognize this as a particularly Haydnesque practice, while at the same time hearing other Haydn works which employed a contrasting second theme. So you would hear Haydn in terms of Haydn, and not in terms of some all-embracing normative thematic practice.

I’ll go through some of the textbook features of sonata form and examine them in the light of sonata-principle.

• Three sections: exposition, development, recapitulation. Haydn probably never thought of the second half of the form (development-recapitulation) as consisting of two separate sections. In general he considered the whole thing a developmental process to the very end. To greater or lesser extents other composers tend to match this mindset—Joseph Martin Kraus very much so, for example.
• Modulatory passage in the exposition. This can be omitted altogether in favor of what is often called the “bifocal close”. (I wrote about that for a previous issue.) It’s perfectly possible to jump right into the secondary key and then work out the implications of changing key without a long preparation. A few good examples: Haydn Symphony #1, first and second movements, Mozart Overture to “Marriage of Figaro.”
• Musical event at arrival in secondary key center. Sometimes we glide into the secondary key center and aren’t always aware precisely when we’ve arrived. Haydn is particularly fond of this (examples abound; take a look at symphonies numbered in the 40s and 50s in particular). Sometimes the ‘event’ may be no more than a seemingly-innocuous figure. Sometimes it might be unmistakable.
• Different or ‘lyrical’ second theme. This may be the single biggest misconception about sonata form, and it dates ‘way back. In 1793 Koch requires a “cantabler Satz” (singing theme) for the arrival in the secondary key center; in 1796 Galeazzi prescribed a lyric “passo caratteristico” at this point. In real practice (as compared to prescriptive theory) the secondary theme can just as readily be a restatement of the primary theme, or some relatively innocuous melodic/harmonic gesture, as it can be a lyrical theme. Haydn was in particular fond of using the same thematic material throughout a composition, or at least material derived from a small set of motivic cell figures and then subjected to what is usually called “developing variation.” Try the “Oxford” symphony (#92) for a good example of the main theme as ‘second’ theme, or Symphony #88 for a good example of a nearly-invisible second theme that is derived from motivic structures. On the other hand, Symphony #95 offers a textbook-perfect ‘cute’ secondary theme in the relative major contrasting a harsh primary theme in C minor. The finale of Mozart’s Symphony #39 is another example of ‘monothematic’ form.
• Closing theme. Oh, there’s usually something at the cadential point. However, it can just as easily be another statement of the main theme (Mannheim composers like Johann Stamitz did that a lot) or just a cadential figure.
• Development as using thematic material from exposition. Most composers follow this practice to be sure. But there’s no real rule about it. Johann Christian Bach, for example, frequently begins his developments with new thematic material. That’s in contrast with the extremely common practice of beginning developments with the main theme played in the secondary key. Another practice is a trio section in the development, sort of like a little slow movement plopped in there, seemingly out of nowhere. It’s more common in opera overtures than in symphonies—there’s an itty-bitty one in Mozart’s Magic Flute Overture—but you can find it in the symphonic repertoire as well, for example the first movement of Haydn #45 (Farewell).
• Retransition section emphasizing the dominant. It can certainly happen. Just as common is a sudden jump into the recapitulation, typically from the submediant (vi) or its dominant (V/vi). Haydn is in particular fond of the latter approach—check out the first movements of #s 45 (Farewell) and 85 (La Reine), as well as the Kyrie of the wonderful Harmoniemesse of 1803. I’m particularly fond of the dazzling dribble-feint of “La Reine”. Mozart has some fun with this sort of thing as well—check out the first movement of Symphony 39 in E-flat.
• Musical event at the beginning of the recapitulation. The textbooks tell us there will be a restatement of the main theme in the tonic at this point. However, it’s just as likely that something else will happen; in fact, there may be no distinct event at all. There is no structural need to restate the main theme anywhere in a sonata form after its initial statement. It has been stated in the tonic key, after all, and therefore is not in a state of imbalance. As a result, it’s perfectly possible to avoid restating the main theme altogether (the so-called “binary sonata form”) or to restate it somewhere other than the beginning of the recapitulation. Mozart was in particular fond of restating the primary theme at the end of the recapitulation, which is sometimes called a “mirror recapitulation”—the position of primary and secondary themes being reversed. This mutability of the location of the primary theme gives rise to what is erroneously called the “false recapitulation”, which is a statement of the main theme, in the tonic key, somewhere during the developmental section. I consider the label erroneous because a false recapitulation is neither false nor a recapitulation. It’s just a statement of the main theme in the tonic key—which might provide a foothold of harmonic or melodic stability amidst the typically sand-shifting nature of a development section. What is “recapitulated” in a recapitulation is the material in the secondary key, not necessarily anything in the primary key. The idea of “false recapitulation” is a nifty little indicator that people have been thinking in terms of sonata form as a recipe or template. It can only be ‘false’ if there is something that is ‘true’ or ‘proper’.
• A return to the tonic key at the beginning of the recapitulation. Interestingly enough, sometimes the recap begins with the primary theme, but in the subdominant key instead of the tonic. This was usually a somewhat later practice compared to the mainline Viennese composers; it occurs in Schubert (Symphony #5) and Schumann (“Spring” Symphony) for example. The tonic key is not needed until we begin restating material that was originally stated in the secondary key, so rebalance is perfectly possible without it for the primary thematic material. Haydn had a nifty practice of beginning a recapitulation in the tonic minor instead of the major, in major-key works of course.
• Recapitulation as restatement of exposition in the tonic. Quite a few recaps will in fact work this way. But a lot of them continue on developing right up to the last moment. Both Haydn and Mozart were exemplars here. The idea that the recapitulation must repeat almost all of the material from the exposition has given rise to some odd conceits and fancies. My own particular favorite bugaboo is the “secondary development”, which is a developmental passage that happens during the recapitulation. Sort of a second-class-citizen development, poor thing. If we think of the recapitulation as an extension of development—during which material from the secondary key will be restated in the tonic key as a rebalancing of the structure, then ideas like “secondary development” really cease to have much meaning. The same holds true for ‘variants’ like the “mirror recapitulation” or “binary sonata form”, all of which are reactions to some imagined ideal of a sectionalized, event-driven sonata form.

To conclude: I don’t want to give the impression that textbook sonata-form is an utterly useless or misleading concept. It is definitely a good starting point—but only a starting point. As our limited knowledge of eighteenth-century practice has increased and broadened, we have learned that the compositional process was a great deal more dynamic, fluid, and rhetorical than was thought during the nineteenth century, when many of these formal models were developed. As much as we can, it’s wisest to approach eighteenth-century music without expectations grafted on by a later culture. To be sure this isn’t easy or even always possible, but it’s well worth the effort. The music comes much more vividly to life as we remove it from obsolete theoretical straitjackets.