Relative newbie here. I'm sure this discussion has taken place here before, and I've read many of the discussions that cover this more broadly (i.e. "How do you come up with your ideas?" etc.) but I'm interested specifically in various themes within a musical form.

My specific example. I decided recently that I'd like to write a cello sonata. I began mapping the piece out: the form (sonata for the first movement), rough length, what I'd like to convey, and some of the organizing and unifying structural principles. However, now it's time to construct the themes which will fit into that structure and I am absolutely stuck. There's something I'm looking for and I can sort of see it, but can't quite reach it.

So I'm curious how other composers create their thematic material. Do you mostly improvise? Do you draw out the shape of the theme first and then decide on the notes? Do you just write *something* and then revise and change it until you have a theme you want to use? 

I look forward to your responses.

You need to be a member of Composers' Forum to add comments!

Join Composers' Forum

Email me when people reply –


  • I don't have the exact quote with me, but Ravel said something to the effect of, "you'll never be able to predict the results you'll get from unintentional infidelity to a model." This has proven very helpful for me. Working with a model you will inevitably change directions and not end up following everything exactly as "the books" might say.

    So in the end, I agree, that the music usually ends up fitting into it's own form, but starting from a model is a great way to get the juices flowing. At least, it has been for me.

    I don't think inspiration is a prerequisite on the other hand. Inspiration can come before, during or after the writing process. The only prerequisite is actually sitting down to write. Sometimes I write junk, and then about halfway through that junk, something else comes along, and that takes me in directions I never planned. Other times, I feel inspired right from the start. But the key is starting. Inspiration comes with work. (I hope this doesn't spark a new debate... but I guess that's what forums are for, so bring on the friendly debate).

  • As for inspiration I often find it  in listening to and studying other musicians/composers I admire.  I recently set myself the challenge of transcribing parts the Schubert Quintet into Sibelius -- just so I could take a very close look at it.  And though it seemed like it might get tedious I found a lot of pleasure in doing it -- because it was almost like composing seeing it that closely and writing it out slowly -- second guessing what might have happened instead of what did.  Sounds dull but it wasn't at all.

    I hear bits from other music I love and sometimes play a game in my mind -- where might this have gone, what would I do with it.  And that can lead to new ideas.  Or I take a theme or a phrase of my own and think of how many different ways it could be set.  Sometimes I look at a building and try to imagine what it would be if it were music (thinking structurally!).  Or a conversation between two people -- what if they were musical themes talking. 

  • I can tell you've spent a lot of time with Caplin, Jon :P

    Now, I've done some research of my own of Sonata Form, using both Caplin and Hepokoski and Darcy.  I tend to follow Hepokoski and Darcy more primarily because while Caplin does a great job of laying out the nitty-gritty blocks of a composition, Hepokoski and Darcy push beyond and ask questions dealing more with the expressive purpose and rhetorical role of the individual "action spaces," which, to me, are of greater value to composers, especially those wanting to adapt Sonata Form to a post-tonal language. As a result, I now give my own synopsis of Hepokoski and Darcy's Sonata Theory:

    The Framing Events: MC, EEC, and ESC.

    The "action spaces" of a sonata are deployed around various types of Framing events.  These structural signposts signify both the expressive/dramatic goals of one section as well as providing the context to the next space.  The Sonata Exposition has two such goals: the Medial Caesura (MC), and the Essential Expositional Closure (EEC).  The MC divides the "1st part/theme/zone" (known in Sonata Theory as the P/TR space, standing for "Primary Theme / Transition") and the "2nd part/theme/zone" (the S space, for "Secondary Theme").  It represents an "interruption" of the flow of events (usually a half cadence in the common practice period), which sets the stage for a "restart" of activity in S space. The EEC represents the first attainment of the "goal" toward which the music has been moving (in the Common Practice Period, this is a PAC in the dominant), and all that comes after this first achievement reinforces this goal and is known as C (or Closing) space. 

    Now, of great importance in Sonata Theory is the following: what we hear in the Exposition sets the stage for what we expect to happen in the Recapitulation (see Jon's post for a description of events in the Recap).  In this sense the Expositional MC and EEC may be regarded as "structures of promises:" promises which will either be affirmed or denied (depending on the expressive aims of the composer) at the same moments in the Recapitulation.  The EEC in particular is an important event, since it promises the way in which the sonata's ultimate "goal" (the achievement of the PAC in tonic at the close of the Recapitulatory S space) will be articulated (known in the Recap as the ESC or the Essential Structural Closure).

    Now, any composer writing in a sonata form enters into "dialogue" with these expectations and in whatever language a composer chooses to write, these expectations still remain (that is, an interruption of events followed by an achievement of some goal away from "home base," and a recapitulation that contains a similar sequence of events but which achieves the goal in the "home" area). The individuality of a piece comes from how the composer chooses to affirm or deny these basic principles.  For example, a composer could choose to have a "failure," a kink in the chain somewhere in the Exposition that causes the EEC to not be articulated, thus creating a sense of tension in the listening audience as to whether the same "failure" will occur in the Recap.  Or, perhaps the composer does not want to interrupt the process on the way to the EEC, instead choosing to proceed without a recognizeable MC or S space (Haydn does this often, in an exposition subtype known as the "Continuous Exposition").  The possibilities for how to go about "playing" with the audience's expectations are numerous and are of value for any composer of any style.

    What can you take from this as a composer?  To explain simply, its that Sonata Form is not a blank concept, we expect certain events to unfold.  Because an audience expects a series of large-scale expressive/dramatic events, much of the Sonata's aesthetic impact depends on how these events are articulated and how their achievement is subverted by the composer.  Much like a good narrative, a Sonata draws its audience toward expected events and can either affirm or deny these expectations during the course of the unfolding music.

    Jon Brantingham said:


    I am with you, I like to have a clear plan when I am trying to write something a little longer. I have gotten bogged down and ended up feeling like my longer pieces were not reaching their potential. I've done a lot of research on sonata form recently so here is a cliff notes version. There is a lot more I could say than this, but hopefully this gets you started. I have a website you can check out here, where I talk a lot about composing in the classical style.

    If it helps, here is a piece I wrote using this plan (minus the coda, which I did not do). Hopefully you can generally hear the theme areas as you go.
    Piano Sonata No 1, 1st Movement by Jon Brantingham

    Sonata Form Themes

    Most themes in sonata form are built generally the same, but some are "looser" than others.

    The slow introduction tends to be more harmonically based, less melodically thematic.

    The main theme is usually a tight-knit theme, either written as a sentence, period, small-ternary (rounded binary) or small binary form. This is in tonic.

    This follows with a transition. The transition brings you to the subordinate theme (in dominant). You can either modulate, ending with an authentic cadence in dominant, leading to the subordinate theme or a half-cadence in tonic (which incidentally also ends on a dominant, but it is not tonicized).

    The subordinate theme is looser, but still usually follows the general path of the tight knit themes mentioned earlier. It just tends to have more extensions and expansions, especially in the cadential area. The theme is almost always in dominant.

    Sometimes you can have a subordinate theme group, meaning two or more subordinate themes. This gives you a lot of room for playing around with different styles and textures within the subordinate theme area.

    The development can get a little crazy, but tends to be what is called core/pre-core technique. The pre-core area tends to be "the calm before the storm," type section. This can be a new theme, or use material from a previous theme like the main theme. It is in a new development key, usually vi or iii in major or VI, III or v in minor.

    The core starts off with theme like material, usually about 8 bars long, but it can be longer, that has a lot of potential for fragmentation. It is then systematically cut down, into smaller and smaller fragments, and usually consists of a lot of sequence repetition. It normally ends with a significant standing on the dominant. This is in the dominant of the home key, so you can go back to tonic for the recapitulation.

    The recapitulation brings back the main theme, and the subordinate theme (which stays in tonic, instead of dominant). But you can have considerable differences in material, particularly because you can remove redundant material (things like repetitions of ideas). You can also add more material.

    The coda is an after-the-end type of area, where you can develop some other material from the piece that you couldn't specifically develop in the recapitulation. For instance, this could be some material from the development, which you want to further develop, or resolve. You probably want to avoid new thematic material, as this will probably make the overall feel be "unresolved."

    Constructing thematic material?
    Relative newbie here. I'm sure this discussion has taken place here before, and I've read many of the discussions that cover this more broadly (i.e.…
  • Well, ultimately Thematic ideas will have to come somewhere from inspiration.  However, you can "narrow down" what you are looking for based on your expressive aims.  For example, I'll give you the following idea I had for a Sonata: I wanted the Sonata to represent the many "failures" that occur while trying to achieve something, and the feeling of satisfaction we get when we finally achieve it.  As a result, I had the following ideas for themes:

    The P theme would begin very bold, loud, and emphatic.  However it would almost immediately "dissolve" before reaching its logical conclusion, as though it were "overzealous" and in being so emphatic caused its own destruction.  Thus the P theme would start with a clear thematic idea, probably pretty simply, but would after a bar or two morph into a building texture that would become the Transition.

    The S theme would begin the same way, but would be very timid and cautious, leading to a completion of the first part of its phrase (the antecedent/presentation/etc.), then its second phrase would become more confident and would almost achieve full closure, but would fail and dissolve again just like the P theme.

    Now in the Recap, I have to decide if I'm going to let the theme achieve closure there (as though it had "matured" throughout the development and learned from its mistakes) or if I'm going to subvert the closure again (maybe with the implication that the goal is simply out of reach, the music tragically cannot come to a close) and push the attainment of the goal into a Coda of some sort. 

    Now I still haven't had any musical ideas to put down on paper, but I do know at least what I'm looking for: I need something that has at least a two-part thematic structure (so the first part can be achieved in the S space before dissolving again).  And in its first incarnation its needs to be loud, emphatic, and simple.  From there I can narrow it down further: will it be texturally thick or thin? Will it have a large or narrow register (I'm leaning toward large to highlight its "bold" character, but we shall see)? what sort of contour should it have?  etc.  The more you know about what you're looking for, the more likely you'll recognize when you've found a theme that suits your expressive purposes.

    So yes, you will eventually need pure inspiration, but before that comes, you can still do work with narrowing down and codifying your expressive aims, and thinking about exactly what it is you want to say with this music.

    Tom Readmond said:

    I didn't mean to start a debate over whether to start with form or thematic material. I think if either approach works for someone, then that's the correct approach for that composer. What I was really hoping to hear about was how others stimulate their inspiration. Once in a while I'm fortunate enough to have something just pop in there spontaneously, but most of the time, in music, work, and in life, there are things I have to do to spark creative inspiration, particularly when it comes to the themes, which, as many of you have pointed out, is really such a prerequisite to other ingredients.

    Constructing thematic material?
    Relative newbie here. I'm sure this discussion has taken place here before, and I've read many of the discussions that cover this more broadly (i.e.…
  • Agree 100%

    Complicated and atonal work is fine, but ask the same composer to write something that is melodic and will be universally liked and memorable.

    Which is the greater challenge?

    You don't measure greatness by the number of marks on a page.

    Jon Corelis said:

    It always seems to start with a melody that just comes to me.  For me, music is melody, infused with rhythm, and optionally supported by harmony.  One would not, I think, readily infer this definition from many of the compositions posted on this web site.

    Constructing thematic material?
    Relative newbie here. I'm sure this discussion has taken place here before, and I've read many of the discussions that cover this more broadly (i.e.…
This reply was deleted.