I wondered why classical* composers use so many repeats? (I do not mean the repetition of themes, motives etc., but rather of whole sections.)

For instance, the 1st movement of Haydn Piano Sonata No. 30 is beautiful, but is the full repeat of the sections really necessary? Another example: why is the exposition in Beethoven's Symphony No. 5 repeated?

If composers think that repetition is useful, why not add a section that has slight differences to make the music more interesting?

I am not sure about this, but section repeats were less frequent during the Romantic era and later, is it true?

*meant as: composers from the "classical" era, not in the wider sense.

  • 7
    Prior to the late 19th century, almost all of the music people heard would need to be performed live. I'm not sure how similar most of the music that was played 300 years ago would comparedto the fraction that survives, but the lack of omnipresent recorded music would affect people's perception of live music, and the desired amount of repetition therein.
    – supercat
    Commented May 27 at 17:32
  • 20
    Not sure why you’re singling out classical music. Look at how much music throughout history has been structured around repetition. Today pop and rock stars use verses and choruses in music that is popular worldwide. If you’re wondering about literal repetitions maybe the answer you want is that convention often calls for performance variations for literal repeats, such as playing a different dynamic. Commented May 27 at 19:26
  • 8
    I can hardly think of a genre of music that doesn't lean heavily on repetition. As Adam Neely says, it legitimizes, after all. Commented May 27 at 22:39
  • 1
    Heh. For me it's the exact opposite. I don't like most classical music because there is too little repetition. Most of it sounds to me like it was produced by a random number generator without any rhyme or reason. Unless you've heard a piece 20 times already, there's no way to predict what will come and follow along in your mind. So, like all matters of taste - this is very subjective. 🙂
    – Vilx-
    Commented May 28 at 8:05
  • 1
    @Vilx- "Unless you've heard a piece 20 times already": or have listened to enough music of the period in question to know the conventions. I often find myself "just knowing" what the last several notes of a phrase will be in a piece I've never heard before, and probably 90% of the time I'm right or at least very close. The pieces that frustrate these expectations are that much more interesting if you have the expectation to begin with.
    – phoog
    Commented May 28 at 15:25

6 Answers 6


Another aspect that hasn't been mentioned is that performers often would play things differently the second time around. We often think of improvisation as a more modern trend, but that isn't necessarily the case.

Performers even back in the Baroque days could add ornaments and other slight alterations of the piece on the second time through a repeat. In some cases, the sheet music was used only as a guideline, around which the performers were expected to play their own embellishments.

Look up "basso continuo". This was a part given usually to the keyboard (harpsichord or organ in those days), though sometimes to the cello/bass as well, where they're only given simple chord figures and expected to improvise their own rendition of those chords.

The strict stipulation to only play the notes on the page came later, mostly given the ever-increasing size of orchestras. It's one thing to improvise when you've got a small group of fewer than 10 instruments, but when you've got a 100-piece orchestra, this is obviously more problematic.

  • 7
    I believe this is the answer that the asker needs most. In the Baroque era, and somewhat in the classical period, dynamics, articulations, and ornaments might not have even been written in. Commented May 27 at 19:27
  • 3
    Definitely. Modern folk tunes repeat repeatedly, and good players or good groups will make sure every repeat is different in some way.
    – Graham
    Commented May 27 at 20:14
  • Mmm.  It's interesting to see the development of (Western common-practice) music notation from a rough, limited aide-mémoire to a precise, complete definition — leaving progressively less room for creativity from the performers.  (Some other traditions and styles see things very differently, as Graham mentions.)
    – gidds
    Commented May 27 at 21:40
  • 1
    I disagree that performers have less room for performers. Notation is more precise, but no more enforceable than in the past. The venue has a big effect on such things as dynamics and tempo. A Texas Country Western bar may have 50 people on its 10x10 foot dance floor compared to 700 people in a large cathedral: echoes and sound absorption from the rugs, seats, clothes, tapestries, etc., have a big effect.
    – ttw
    Commented May 28 at 0:43
  • @gidds "Precise, complete definition" - Have you played in any large, conducted ensembles? My experience in such groups is that neither parts nor scores are "precise" nor "complete". It's just not possible to set everything down in such a way that nothing is left to interpretation. I'm not very experienced and if pressed I could come up with at least four different ways to play a staccato marking. From the composer/engraver side, you don't want to overwhelm or over-mark the music. You want to leave it open to interpretation or else it ends up annoying and lifeless, at best. Commented May 28 at 3:00

The repetition is part of the classical Sonata form, but it was standard long before the classical sonata became a recognized form.

In particular, it was used in baroque times when the entire notion of "recombining thematic material in interesting ways" was not an aesthetic principle; quite the opposite, the typical baroque piece is formed from one principal motif, theme or figure, not from the contrasts that characterize the classical sonata with its important development section.

Therefore, for a baroque suite, trio sonata or similar work, it made much more sense to simply repeat either part as an easy way to obtain double runtime without additional effort in composing and transcribing. For a piece of music that might be used just as sonic background for e.g. a banquet attended by people largely ignorant of music, this is perfectly fitting.

In contrast, Beethoven's academy concertos were visited by people keenly interested in the music as an event and marvelled at his genius. The contrast in intention and perception couldn't be greater. So even though the baroque suite morphed quite imperceptibly into the classical sonata over time, the end points of the development were quite different, and it's not surprising that the verbatim repeat fell out of favour as the classical sonata (and especially its big brother the symphony) became exercises in artistic mastery rather than just the 'muzak' of their times.

  • 2
    there are also repeats because sonatas &c were made to be danced to, even if song forms deriving from them while perhaps not meant for dance.
    – georg
    Commented May 27 at 19:17
  • 1
    @georg The classical sonata may have developed from the baroque suite of dance movements. But - well, in brief, no they weren't!
    – Laurence
    Commented May 27 at 20:14
  • @georg as far as baroque sonatas are concerned, I have the impression that "sonata" began as a designation for pieces meant to be played and listened to. Pieces meant to be danced to would more likely be called "ballo" or labeled with the name of the dance. But these would more typically appear collected into suites. Certainly by the end of the baroque, instrumental sonatas featured movements that were not particularly danceable by baroque standards.
    – phoog
    Commented May 28 at 15:38

Some repeats are varied. Other times repeats are used to establish the key of opening. Even with repeats, there are some variations. A simple AABB piece is not as repetitive as it looks. There are 5 different transitions: 0-A, A-A, A-B, B-B, B-0. Each transition will be different.

Lots of Baroque music and some later pieces are based on dances. These tend to be repeated as each dance part is usually only about 16 measures. Long-range key relations become more important in the classical era and there is a tendency to repeat parts in different keys or at least approach or exit an identical part using different keys.

  • 3
    I think dancing is the answer. Repetitive music seems to work better for dancing. For example, in a ceilidh you usually have to learn a sequence of dance moves as you go, and you get better and better at it as the dance goes on, if that makes sense. The repetitive nature of the music helps to give you a clue as to which part you should be doing at any given moment.
    – Flounderer
    Commented May 28 at 2:32
  • While some dancing today is done to live music, it used to be that almost all dancing was done to live music [some might have been performed to rhythmic chants, or perhaps silently, but certainly none to mechanically reproduced music].
    – supercat
    Commented May 28 at 16:31

The superficial answer is that the sonata form arose as an extension of baroque forms that traditionally featured repeats, as others have mentioned. Baroque dances typically had two repeated sections that might be eight or twelve or sixteen bars each. They might be the same length or different, and they frequently feature temporary tonicization of the dominant, the relative major or minor, or both, especially at the end of the first section and/or the beginning of the second, but often also at formal inflection points with a section (often after four bars, or eight or twelve if the section is longer). Another important form is the "bar form," a medieval German song form where the first section is repeated and the second is not, which is prominent in Luther's chorales and therefore in baroque protestant church music (and in modern hymnals).

This phrase level repetition is a feature of music more generally. You find it in popular music of the 20th and 21st centuries, for example. You might just as well ask why songwriters don't write slightly different versions of the chorus (a.k.a. refrain) after each verse, or why they don't change more than just the words from one verse to another. (Well, sometimes they do, actually, but it's typically seen as a departure from the standard form or as an extension of it.)

Hand copying of music notation, whether pen and ink or a hand-prepared printed edition, was expensive and time consuming. Even for short instances of repetition, it made sense to use signs to indicate the repeat rather than writing it all out. As others have noted, it was left to the performers to provide interest by adding different embellishments or other expressive variation where it was seen as desirable.

The main reason I'm writing, though, is to address the question

is the full repeat of the sections really necessary?

I think it is, yes, to preserve the balance of the large-scale form. Others of course disagree, as you will find recordings of pieces from the baroque and classical periods that ignore some or all of the repeats.

Skipping the repeat shortchanges the section, detracts from its importance in the formal scheme. It is akin to rhetorical repetition. Why does Lennon and McCartney's "Let It Be" have so much text repetition? Why did Martin Luther King say "I have a dream" so many times in his famous speech on the Washington mall?

The first section of a typical baroque dance (or classical sonata or symphony) might end on the dominant, with the second section starting on the dominant and winding up back on the tonic. If you skip the repeats, the formal plan becomes

I-V / V-I

while the music was designed to be heard as

I-V / I-V / V-I / V-I

You lose the opportunity to hear each section ending on one chord with the repeat of that section starting on a different chord. Or consider a form where the second section is twice as long as the first, with the second half of the second section resembling the first section. With repeats, this is

A A / BA' BA'

Without repeats it is just


Again, eliminating the repeats eliminates some of the transitions, but it also changes the balance of the piece, particularly in the relative roles of A and A'. This form is common in baroque dances, but it could also describe the sonata form, with A the exposition, B the development, and A' the recapitulation. Yes, the composer could have written four distinct sections instead of two, but the repetitiveness helps to provide balance and indeed to help the listener (and dancer) experience the form through hearing it.

I first noticed this in one of my favorite suites, J. S. Bach's French Suite number 5. I was listening to a really elegant lovely performance of it, and then in a couple of the inner movements, the performer skipped the repeats. It seemed to say "this movement isn't as important" or "I'm not so fond of this one so I don't want to spend more time on it than I have to." And it really ruined the recording for me. But for many years in my student days and well into early adulthood, I found repeats boring. In those days I also had very little appreciation for large-scale form. I did a lot of measure-counting analysis in school, but I didn't understand the point. The two are different sides of the same coin.

Perhaps a good way to illustrate it to yourself might be to take a favorite suite, sonata, or symphony and find recordings where they do and don't observe all of the repeats. Then set aside enough time to listen to one recording attentively all the way through, without distractions, to experience the overall effect of the entire piece as if you were in a concert. Then, probably better on a different day, do the same with the other recording. Focus on the large-scale effect: how satisfying or exciting or whatever is the final cadence? How intense is it? How does the arc of the piece progress over the 20 or 60 minutes you need to listen to it? I suspect you'll notice a difference. If not, try a few other recordings in case there are other aspects of the interpretation in play. With any luck this will help you understand the importance of repeats to the formal balance that was so critical in the baroque and classical eras, but in any event it should help you form your own opinion about whether "the full repeat of sections is really necessary."


If composers think that repetition is useful, why not add a section that has slight differences to make the music more interesting?

This already happens in classical style, but on a smaller scale, on the phrase level. Something typical would be a two bar statement repeated with variation, all written out. And, of course, theme and variation form is written out repetition with variation.

I am not sure about this, but section repeats were less frequent during the Romantic era and later, is it true?

Yes. And some modern performers don't actually perform classical style notated repeats. Apparently the idea was such repetitions were superficial, adding no new musical development.

Section repetition in classical style was about emphasizing binary form and the antecedent/consequent effect. The essential form, on both small and large scale, was...

||: I ... V :||: V ... I :||

...which represents the largest amount of harmonic reduction and shows how the first part was a harmonically incomplete opening, which then was followed by a harmonic closing. Repeating the two sections helps make clear the design is a deliberate two part form.

Moving into the Romantic era, that kind of purely formal music gave way to more "emotional" program music, tone poems, episodic forms, miniatures, etc. that often did not have a binary form to emphasize. Actually, when program music follows a "narrative" arc, sonata type sectional repeats would undermine the story telling design. That would be like re-reading the first part of a novel before reading the ending half.

But that doesn't mean sectional repeats were not used in the Romantic era. Small dance forms are the obvious place to look. Waltzes, mazurkas, etc. will use plenty of sectional repeats.

Whether it is Classical or Romantic style music don't overlook the obvious. Repeating sections is a way to make the music last longer. When the repeat is long enough to be a section, no variation is necessary, because monotony will not set in for most listeners.


I mostly agree with Darrel Hoffman's answer that points out the historical basis for repeats often involved variation/improvisation in those repeats. We have plenty of historical examples of composers writing out such repetitions with variation, and composers like C.P.E. Bach who specifically wrote collections of simpler pieces called "Sonatas with Varied Reprises" designed to teach beginning performers who to introduce variation and ornamentation when playing repeats.

Originally with soloists and small instrumental ensembles especially prior to ~1800, a better comparison might be with a jazz ensemble today, where a "repetition" might follow the lead sheet or arrangement but introduce some variation. And of course repetition is very common to most popular styles of music.

That said, I wanted to highlight another critical aspect that surprisingly seems missing from other answers so far: repetition teaches the music to the listener.

Today, we often view "classical" music as this artistic contemplative music that we might sit around and analyze for hours, or which we might play recordings of over and over. We might even sit down with friends and compare a number of different recordings of the same piece, contemplating their subtleties.

This is a great way to listen, but it's also very ahistorical. In the 18th century, for example, when you had these increasingly long sonatas and symphonies and concertos, performance was a luxury. You might only get to hear a large ensemble piece of music once in your lifetime actually performed. And while music publishing was growing a lot in the 18th century, the heyday of keyboard arrangements to study and learn large-scale pieces like symphonies by yourself at home was more of a 19th century phenomenon.

So, given that context, now imagine you're getting to hear a new 10-minute orchestral movement for the only time you may ever hear it performed. As a listener, how the heck do you know what's going on? What do you focus on? Did you catch the main theme? Did you get distracted and forget it because the oboe player made a mistake the first time? Did you space out for a minute and miss that amazing descending bass pattern with crescendo that led into the next main theme?

If you're playing a giant sonata form movement spanning many minutes, it can help to simply hear the music again. As a listener, it helps you get oriented to the music, to catch the main themes, to understand how they connect and lead into each other.

Just to pick apart the traditional sonata form as an example, the second main section is called the "development," a term we use often because it references back to earlier themes and "develops" them. How will you as a first-time listener even catch those references to earlier themes unless you know them? Often the development is about creating drama and subverting earlier expectations too. Again, if you weren't paying attention the first time around, you may not catch the way the drama unfolds when that theme doesn't go where it's expected to (based on what it did in the exposition). So, maybe as a composer you should repeat that exposition to make sure the listener really knows what the themes are first. (Note that even as later repeats fell away in the 19th century, that first repetition of the exposition was often kept for this reason.)

Essentially, if you're in a culture without recordings, performances (particularly of larger ensemble pieces) may be rare. You want to give the listener as much information as possible to try to follow what's going on. Repetitions are simply an essential tool for doing so, particularly as the works grew longer in the Classical and early Romantic periods.

Later in the Romantic period, of course, composers often took a different perspective. They more frequently omitted repeats and instead might focus on an idea of continuously developing themes throughout a movement in a more integrated and constantly flowing way. The "drama" of a sonata form in the Classical period was to some extent created around expectations of structure. So, you needed to teach listeners that structure so that they might retain enough to sense when it deviated or was modified to deny an expectation in a dramatic fashion.

The more precise structural expectations of that earlier music gave way gradually to a more free-flowing and continuously developing form for many composers. Drama was articulated through a greater variety of timbres/instrumentation in larger orchestras, wider dynamic ranges, and overall a more "episodic" structure. Think of listening to Wagner and how recurring Leitmotifs help to give meaning to your experience and allow a listener something to latch on to while experiencing the music. That's similar to a Mozart theme in a very broad sense, but now it can just occur in different forms and be developed within a constantly changing sound environment. Exact repetition is no longer required or even desirable because the structural expectations are different. The composer gives listeners something different -- a new kind of "grammar" of musical elements -- to keep track of while experiencing a piece for the first time.

Also, as I noted above, the 19th century saw a huge increase in transcriptions and study scores sold to the public, so they had time to "digest" the music perhaps outside of a single performance. That too may have led some composers to feel more free to develop longer and less rigid structures, as they no longer had to essentially "teach listeners" the themes and the other stylistic vocabulary of the piece in real time, at least for "educated" listeners. The concept of music analysis in the modern sense really began. Whereas in earlier periods, music theory was mostly compositional and performative (in that it could aid in some aspects of performance, like improvisation, continuo-playing, etc.), not primarily analytical or about listening.

In sum, repetition is an essential concept to orient new listeners to a piece of music. It also helps to set up expectations for listeners -- e.g., that theme A will flow to theme B, or that harmony X will resolve to Y. If you look at psychological responses to music, familiarity can help to create emotional response: you come to enjoy the fact that A goes to B and X goes to Y. It's familiar and comforting. Particularly if those resolutions happen in a dramatic fashion. But then, it also gives the composer an ability to surprise you: suddenly A goes to C! Oh wow! That's shocking! Now your ears perk up -- this is new! What will happen next?

None of that kind of formal drama can happen without repetition creating expectations and then either fulfilling them or subverting them.

Such expectations happen on many levels in music -- in individual phrases or even on a chord-by-chord level. It's why many classic small forms for melodies do things like AABA song form. It's comforting to get to "know" the first phrase A, because then B can sound like a contrast, and then comforting to feel the familiar A return again. Sonata form is doing the exact same thing on a much larger scale when it repeats an exposition.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.