Dreaming of a masters thesis

Update: see a more formal draft of my thesis proposal.

For my NYU masters thesis in Music Technology, I’m designing a beginner-oriented music learning app for the iPad and similar devices. It will approach music the way I wish I had been taught it, and the way I’ve been teaching it to my private students.

I’m motivated in this project by a few axiomatic beliefs:

• Everyone is born with the capacity to learn music. That capacity just needs to be activated in the right way.
• Anyone can and should participate in music actively. Like cooking or sports, music need not be totally mastered to benefit the participant, and it should definitely not be the exclusive province of specialists.
• Beginners should study music they’re familiar with, and that they like.
• Music teaching for beginners should follow an Afrocentric paradigm that relates to pop, rock and hip-hop. That means starting with rhythm, and treating melodic instruments as percussion.

So here’s what this means for music teaching.

• Beginners should learn pentatonics first, then mixolydian. Music education customarily begins with the major scale, but pentatonics and mixolydian are closer to pop, rock, hip-hop and dance common practice.
• Beginners should work modally for a long time. Being constrained to a certain unvarying group of notes frees up mental bandwidth to think melodically and rhythmically. The best mode to work in is the ambiguous major/minor tonality of the blues. Again, this reflects the majority of the American mainstream.
• Only after becoming familiar with blues should students embark on the major scale and diatonic harmony. Traditional music theory pedagogy is based on rules laid down in the eighteenth century. While these rules are of historical interest, their conflict with current music makes them tedious and alienating.

The app will start with drum programming, giving you templates for basic dance styles (hip-hop, techno, rock) and letting you customize them. Once you have some mastery of loop programming and rhythms, the app takes you into basic MIDI sequencing, first with single-note basslines, then simple pentatonics, and on to chords. For the visual aesthetic, I plan to avoid skeuomorphism entirely. The interface will consist entirely of geometric shapes in flat colors and large text. Here’s a concept image:

My main design inspiration is the first-person video game tutorial. Every game in the first-person shooter genre begins the same way: your character is in a room with a few objects in it. You can walk around, pick things up and jump, and that’s about it. Then the game lets you into the next room where there are some characters to interact with, more objects, maybe your first weapon. As you master each action and control, the game progressively opens up, adding complexity and challenge until you’re turned loose to roam the game world at will.

A similar approach works well with music. Beginners benefit from tight constraints. I have an exercise I do with beginning students called the one-note groove, where we play a single note over a repetitive drum loop. By eliminating all other musical dimensions, your attention is freed up to focus in on the beat. You can quickly achieve competency in this exercise, playing something that sounds like music (albeit very simple music). The pleasure and sense of accomplishment carries you into the next, more complex exercise: playing a single chord over the beat, or a short melodic pattern. Once you’ve mastered that, you can tackle a chord progression, a longer melody, and so on. At every step, the results should be musical — you should feel artistic satisfaction, not like you’re performing a tedious chore.

Music teaching software can easily keep the feeling of pleasure and play. You can have all of your notes quantized, so everything is on the beat. You can produce melodies simyply by clicking or tapping on the screen. The ideal music-learning app would use take advantage of these benefits. Furthermore, it would start off putting you under tight constraints, and gradually relaxing them as you gain mastery. This is the app I plan to design, and hopefully build.

The app begins with drum programming. First you’re given a kick and snare drum pattern playing a looped pattern of your choosing: rock, hip-hop, techno, etc. You can add other instruments: hi-hat, shaker, cowbell. Once you’ve produced some patterns that you like, then the snare is unlocked, and you can move that around and hear the result. Next, the kick is unlocked as well, and more drums become available. By that point, you know where to place them to get a satisfying beat, and the greater freedom is exciting rather than overwhelming.

Once you’ve sequenced a collection of drum patterns that you like, then it’s time to move on to the bass. Initially, you’re only allowed to use one note. This encourages you to think of the bass as just another rhythm instrument, keeping you in the familiar drum programming mindset. Then you’re given another note or two to work with. You can turn your one-note basslines into simple melodies. As you create more patterns, more notes become available: the pentatonic scale, the blues scale, the mixolydian mode. The blues tonality will be familiar from rock, pop and hip-hop, and frees you from having to think about chords.

Next, you get a second instrument, capable of playing the same scale as the bass, so you can experiment with more complex melodies and counterpoint.

For chords, you’re initially locked into the I7 idiomatic to pop, rock and dance. Next, you’re given the option to add the IV7 chord. The bass gets locked to the roots of the chords, and the melody line can only be placed on the chord tones. Then more chords are introduced: V7, bVII, bIII. Next the I chord becomes minor, with all other chords staying the same. Eventually the app moves into the major scale and its modes, the various minor scales, and complex chord combinations, until finally you “graduate” to full chromaticism. Now you’re free to keep using the app as a sequencing tool, or you can move on to more sophisticated software: Logic, Pro Tools, Ableton Live etc.

I don’t remotely approach having the programming chops to pull this off. I’ll be studying programming in the next few semesters, but I probably won’t become sufficiently advanced to realize everything I’ve described above. Realistically, the thesis will probably be a prototype with as much functionality as I’m able to produce, along with a detailed set of specs for the finished product. The plan then is to take it out into the marketplace, find some real programmers and designers to help me finish it, and get it into the hands of every would-be musician in the world. Wish me luck.