Music composition as research

stock image of writing music at a piano

My main job is directing a program that supports undergraduate research and creative projects. Personally, I define research very broadly–simply “creating new knowledge.” But it is indeed hard to compare creative activities with research. Some folks, like composer John Croft, would argue that the two practices shouldn’t be compared at all (“Composition is not research”).

This post is not a rebuttal of Croft–he raises excellent points about the “category error” of composition as research, such as forcing composers to write grant proposals with research-like terms. But I think there are more parallels in the process than perhaps he might accept. What I will attempt to do in this post is demonstrate how a creative act like composition can be considered research. I’ll use my toy piano piece hammers as an example, and walk through its creation through the lens of a scientific paper.

Research question

Research projects generally begin with a “research question.” This is a specific, narrow, and testable question. Instead of “How can we help kids learn better” it could be something like “how many images can children ages 8-10 recall after completing this activity?” 

In my case, the question starts as simply as “What if I wrote a piece for toy piano?” Right there, I have constraints and parameters decided for me. What is the range of the instrument? How many timbres can I eke out of it? How long will this piece last? And even further: “What if I wrote a piece for toy piano that emphasized the mechanical noises of the instrument?”

Literature review

In research, a “literature review” is both a process and a section of a paper. The process is learning as much as one can about the narrow topic they’re exploring. The aim is to find an even narrower question that hasn’t been studied.

So of course, a literature review for music starts with: What music exists for toy piano?

Before composing hammers, I had performed John Cage’s Music for Amplified Toy Pianos, and I had studied the music of Phyllis Chen and Karlheinz Essl for my dissertation. I did not seek out to intimately know every piece for the instrument, but I was familiar with many of them. I was of course familiar with its use in pop music and horror soundtracks, and the tropes and associations they elicit.


Some composers have methods and processes they follow, and others are more intuitive. Either way, they need to decide how they will write. For hammers, I first started with the constraints of instrument: it is 2 octaves, has one basic timbre, and the keys are naturally noisy. It’s a small instrument, but it’s not a subtle one.

For the pitch, I constrained myself even further, deciding to use only the “white” keys. But I also explored other noises I could make with the instrument by hitting the sides or flipping the keys upward. By improvising on the toy piano, I eventually wrote three main ideas: a G mixolydian triumphant fanfare, an F lydian “moto perpetuo” idea, and a “crackling” idea where the performer plays a light tremolo on the keys.


In a scientific paper, there is usually a “results” section, which is sometimes distinct from the discussion or conclusion. If separate, “results” is usually raw data. For a composer, the result could simply be the piece. Any reflections or commentary could be a discussion or conclusion.

Here is the result of hammers: 


Reflection and discussion of one’s work is where I think we can best align with traditional research. Here are some reflections I’ve had on this piece over the years:

Reflection on the piece

  • I hoped to make a serious, substantial piece for the toy piano – succeeded
  • I hoped to eschew “cuteness” usually ascribed to the toy piano – succeeded
  • I accepted the irony of the instrument by writing the big fanfare

As creative types we often like to say that we only need to please ourselves. But if other people like your piece, you know it works on some level. What was the reaction from performers and audiences? 

Fortunately this piece has had multiple performances and recordings to compare. It has been played around the US as well as Brazil, Spain, and Russia (perhaps more–composers aren’t always informed about such things). This shows some evidence of the piece being successful (as I did not pay or coerce these performers, they chose to play it!). 

“In conclusion,” this successfully responded to the question “would a piece for toy piano be good if it emphasized mechanical noises and used an ironic fanfare gesture?” 

Leave a Reply

Your email address will not be published. Required fields are marked *