III. Form


Brian Jarvis

Key Takeaways

  • Rondo is a form featuring a main section (referred to as either A or refrain) that returns throughout a work and is juxtaposed with contrasting sections (referred to as B, C, etc., or as episodes).
  • Common formal layouts include ABACA (five-part) and ABACABA (seven-part, likely sonata rondo).

Chapter Playlist

Conceptually, rondo is quite simple: the form consists of a recurring main section that alternates with contrasting sections. As with other forms, rondo can include a variety of auxiliary sections. In rondo form, it is tradition to call the repeating main section the refrain (or A), and the contrasting sections are referred to as episodes (and/or by letters: B, C, etc.).

The refrain material is essentially the same throughout the course of a movement (allowing for some embellishment or abbreviation) and is always heard in the tonic key. Episodes contrast with refrains tonally, and usually thematically as well. A given episode may occur multiple times in the movement or only once.

The most common manifestations of rondo form are either five-part rondo (ABACA) or sonata rondo (ABACABA), summarized in Example 1. There are two main differences between them:

  1. In sonata rondo, the C section is often akin to a sonata form’s development section, whereas in a five-part rondo, the C section is likely to be a relatively stable thematic statement like any other episode.
  2. In sonata rondo, the first ABA (i.e., ABACABA) constitutes the equivalent of a complete sonata exposition (without the repeat, and it ends in the tonic), and the second ABA (i.e., ABACABA) functions as the recapitulation, where the B section is now transposed to the tonic key.
Example 1. General outline of five-part rondo and sonata-rondo forms.

Refrains, Episodes, and Auxiliary Sections in Rondo Form

Refrains are constructed as a combination of one or more phrases and could even be an entire binary form. You can expect that they will have a clear ending punctuated with a perfect authentic cadence (PAC) and will be relatively stable thematic statements. Episodes may be structured like that as well (though with contrasting keys and melodic/motivic material), or they can include destabilizing features like modulation, chromatic harmony, and phrase expansion. Episodes may end with a clear PAC, or they may have ambiguous endings, even lacking a cadence and merging into a retransition section instead through the process of becoming.

Like other forms, rondo form can have auxiliary sections. The most common are retransitions that generate anticipation for the refrain’s return. Codas are also quite common, but introductions are not.

Analytical Challenges with Episodes: Transition ⇒ Episode or Episode ⇒ Retransition

Like the other forms, rondos may include any combination of auxiliary sections, though retransitions that dramatize the return of the refrain are particularly common. Since episodes are often relatively looser than their refrain counterparts, two sections may blend together through the process of becoming (⇒). In particular, a transition may become an episode without any clear delineation of the two formal sections. Similarly, the episode itself may not reach a clear cadential conclusion (i.e., a PAC in the key of the episode) and instead may become a retransition that prepares for the statement of another refrain. However, it is unlikely to find both of these situations occurring within a single episode. One of the important distinctions between the end of a connective section (transition/retransition) and a stable thematic statement (refrains and episodes) is that connective sections tend to emphasize their arrival on the dominant, while stable thematic statements tend to emphasize the arrival at a PAC.

Five-Part Rondo Example

Example 2. Maria Hester Park, Piano Sonata in C major, Op. 7, 3rd movement – Audio (YouTube) Click to See Score (PDF)

The final movement from Maria Hester Park’s Sonata in C major, Op. 7, from 1796 is a relatively clear example of a five-part rondo form. The A section (refrain) lasts from mm. 1–26 and consists of multiple phrases. The B section (episode 1) does not start right away at m. 27; it is instead separated from A by a transition (mm. 27–44) that modulates to the key of the dominant in m. 38 during a passage of chromaticism involving modal mixture. Unlike most transitions, this transition actually ends on the upcoming local tonic of G major (mm. 42–44) instead of the local dominant. B‘s primary thematic material lasts from mm. 45–54, and its ending is elided with a suffix that eventually becomes a retransition (around m. 71) that harmonically prepares for the return of A in m. 73. This second statement of A lasts from mm. 73–98 and is a complete repeat of the first statement of AC (episode 2) starts in m. 99, immediately after the second statement of A, and is in a contrasting key, the submediant (A minor). The C section (mm. 99–132) is longer than the B section. Its ending elides with the start of a suffix that becomes a retransition (around m. 135). The third and final statement of A starts in m. 141 and is again a complete repeat of the initial statement of A.

Sonata Rondo Example

Example 3. Ludwig van Beethoven, Piano Sonata in C minor, Op. 13, 3rd movement – Click to See Score (PDF)

The A Sections (Refrains)

The last movement from Beethoven’s Op. 13 piano sonata (subtitled Grande Sonate pathétique) is an example of a sonata-rondo form. The form is fairly complex, and while most of the sections are quite clear, the B section in particular has a few challenges. The form also includes a good amount of becoming. When approaching the form of any rondo, an efficient strategy for determining the form is to find the location of the A sections. Remember that restatements of A need to be in the tonic key and that they may contain slight variations like omitted repeats, melodic embellishments, and/or new accompaniments. In this particular form, the A sections can be found starting in mm. 1, 62, 121, and 171.

In order to understand the differences between the A sections, a more detailed look at the structure of the initial A section is required. The initial A section ends in m. 17, and there are two phrase-expansion techniques used to generate that length. The first is the one-more-time technique that starts in m. 9 (mm. 9–12 are a varied repetition of mm. 5–8). The second expansion technique is the suffix that starts in m. 12. This suffix starts with an elision and ends in m. 17. So, what could have been an 8-measure theme is now 17 measures because of these two expansion techniques.

With that more detailed analysis in mind, we now have a model we can use to compare the other A sections. While the second A section is a repeat of the first, the third and fourth A sections have been altered. The third A section is the shortest, lasting only 8 measures. It achieves this brevity by removing both of the expansion techniques used in the A section’s initial statement. Instead of having the one-more-time technique in its ninth measure, new, unstable music enters that is derived from the A section but is used as the start of a transition away from the A section’s material. The fourth and last A section ends at measure 182, making it 12 measures long. In this version, the one-more-time repetition is still included, but A‘s original suffix has been replaced by a new, much longer suffix that functions as the work’s coda (mm. 183–210).

The B sections (Episodes 1 & 3)

This movement’s episodes vary in key and melodic/motivic material, and they contain multiple auxiliary sections including transitions, retransitions, and suffixes. The first episode, B, is the most complex, and it is stated twice throughout the work. As is expected for sonata-rondo form, the first statement of B is in a contrasting key (the mediant) and its restatement in the “recapitulation” is in the tonic key of C—though instead of being in the minor mode to match the global key, it is in the parallel major key of C major (I).

While it’s very clear that the initial B section occurs somewhere between the first two A sections (mm. 18–62), the initiation of B is obscured by a number of features, resulting in an ambiguous starting point. There are four possible candidates for the beginning of B: mm. 25, 33, 37, and 44. There is certainly a transition starting in m. 18 due to the harmonic/melodic sequence and harmonic instability, but m. 25’s relatively stable presentation of melodic material is obscured by the lack of separation between the rhythmic activity leading into it and the fact the this melodic/motivic material is not new—it’s derived from the A section’s suffix (see m. 12)—obscuring a possible initiation function. The next candidate, m. 33, is marked because of the dominant arrival and because it introduces new melodic material involving a triplet figure. However, this candidate for the start of B is also obscured by lack of rhythmic separation between mm. 32 & 33 and the fact that m. 33 starts on a dominant harmony instead of the tonic (III). M. 37 is a similarly unclear starting point because it continues the melodic/motivic material introduced in m. 33 instead of introducing its own material, even though it is the first statement of that material in the local tonic of E♭ major. The last candidate is m. 44. Of all the options, this is the clearest starting point because of the textural gap that preceded it. However, it is very uncommon for episodes to start after a PAC has been sounded in the local key, which is the case here (see m. 43). Typically a suffix would begin after a PAC occurs in the local key of an episode. This is all to say that, after careful consideration, no clear starting point of B can be determined, and yet it seems very clear that the presence of an episode in a contrasting key occurs. In order to avoid making a dubious factual statement about exactly where B starts, I think the best way to capture the reality of this passage is to invoke the concept of becoming. So, while it’s clear that a transition begins in m. 18, the B section has no clear beginning; instead, the transition “becomes” the B section somewhere between mm. 25 and 44 but in no precise location.

The space between B and the second statement of A is a little clearer. The PAC in III at m. 51 marks the end of B and the simultaneous start of a suffix. As is common in many forms, this suffix turns into a connective section (in this case a retransition) without a clear division between the two sections, again achieving this fluidity through the process of becoming. The dominant arrival in the global key at m. 58 is a common marker of being in a retransition.

When the B section returns in the recapitulation, it is approached with a newly composed transition (m. 129). The return of B‘s material starts around mm. 134–135 (compare with mm. 25–26). Similar to the exposition’s B section, the exact location of the initiation of B in the recapitulation is also obscured. In the recapitulation, B‘s original suffix (m. 51) is omitted; instead, the melodic/motivic material first stated in m. 44 is recomposed in the recapitulation so that it now blends into a new retransition through the process of becoming before it leads to the final statement of A in m. 171.

The C Section (Episode 2)

In this movement, the C section is much easier to identify than the B sections were. It features a very clear beginning in m. 79 (notice also the clear separation of the end of A in the previous measure) and begins in the contrasting key of the submediant (A♭ major). The internal form of this section could be considered a rounded binary form in which the returning A section is not complete and instead becomes a retransition. The internal A section features a modulating period (mm. 79–86) with a written-out repeat to accommodate some textural changes (mm. 87–94). The internal B section is short (mm. 95–98), and the return of the internal A (m. 99) introduces a new texture, but the consequent phrase does not close; instead, it becomes a retransition around m. 104 that reaches a strong dominant arrival (m. 107), featuring a dramatic preparation for the return of A at m. 121. Though many sonata-rondo forms contain a C section akin to a sonata form’s development section, this movement does not, and C is instead a clear statement of an episode in a contrasting key that leads to a retransition.

  1. Joseph Bologne, String Quartet, Op. 1, no. 4, ii, Rondeau (.pdf, .docx). Access score and audio.

Media Attributions



Icon for the Creative Commons Attribution-ShareAlike 4.0 International License

OPEN MUSIC THEORY Copyright © 2023 by Brian Jarvis is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License, except where otherwise noted.

Share This Book