Use Predictability to Make Your Melodies More Surprising

musical swami

No composer wants to sound predictable. But unless you learn to combine predictability with nonconformity, you’ll have trouble writing vibrant, original melodies.


Musicians use the word predictable is a catchall term for music that sounds cliche –  overused, uninspired, or derivative. One of the main purposes of this blog is to show you why it’s a mistake to use the words predictable and cliché as synonyms. But for now, let’s go with it because this misnomer will help me make my case.

The website “Consequence Of” asked some music reviewers to name the music clichés they found most annoying. Here are a few choice responses:     

  • exit fake-outs at the end of a concert
  • dubstep remixes
  • late-song key change
  • rap interludes in pop and R&B songs

And a similar poll was posted on the music theory subreddit. As one might expect, these responses have more to do with the materials of music than its performance.

  • the “Axis of Awesome” progression (I-V-vi-IV)
  • 4 on the floor
  • descending chromatic bassline
  • wanting to be famous
  • the major scale
  • 4/4 time

Really? Do you expect anyone to believe that getting rid of the major scale and 4/4 time will open the way to more original-sounding music? C’mon Redditors!

As ridiculous this sounds, it turns out to be historically accurate. Starting in the early 20th century, composers stripped their music of anything that might sound “predictable.” We’re not just talking about “formulaic” (“cookie-cutter”) music. Oh no! They went much further than that! Composers avoided singable melodies, easy-to-follow beats, and all kinds of repetition. Although there are many profound, inventive works from this era, 98% of “contemporary classical music” left 99% of audiences 100% in the dark.

The following example is a case in point. If you went to music school and took a music history course, you likely encountered this piece or others like it.

“Structures I,” for two pianos, by Pierre Boulez (1951)

Structures I by Pierre Boulez

If you had trouble making sense of “Structures I,” it’s not due to any deficiency on your part. Boulez established strict mathematical formulas for every note, duration, dynamic, and register in this piece. Mathematical formulas, not musical formulas. Why?

Because Boulez composed “Structures I” shortly after World War II. The consensus at the time was that the “Romantic spirit” – once celebrated as the wild, untapped inner voice that would lead humankind to its unending greatness – was the same force that made us capable of the atrocities committed against fellow human beings. (Notice the parallel to the initial hope and tragic outcome of atomic potential.) In horror, artists like Boulez took great pains to remove all traces of human influence from their music for several decades.

All of this affected composers like me who got our training in the latter part of the 20th century. We felt a great deal of pressure to write “serious” music, as well. Even while outside of the academy walls, “popular music”—music that draws elements directly from its African roots: highly rhythmic, with repeated tunes and cyclical chord patterns—had been growing, exploding without interruption the entire time.

Case in point. Here’s a popular song from the 1930’s. One reason that it so powerfully grabs our imagination is how it’s melody uses predictability AND unpredictability.

I’ve “de-composed” the song so that it follows predictably all the way through. By that, I mean that the first three notes of each iteration—C-D-Eb—always lead to a D. That’s what we expect. But Richard Rogers’ version interrupts our expectation with a high Bb!

“My Funny Valentine,” by Rogers and Hart (sung by Frank Sinatra)


The high Bb would have none of its expressive power unless we expected something more typical. No predictability, no surprise. And this is why you need to use predictability in your melodies.


Think of a schema as a reusable plan or template. A schema carries with it some underlying strategy for how and why things work together as they do. Now there are a handful of basic schemas for melody, plus dozens of variations thereof. And a good composer can riff off of existing schemas to engineer the perfect plot twist for any occasion.

“My Funny Valentine” uses one of these common schemas. Its basic plan goes as follows.

[1] State a melodic idea.
[2] Restate that melodic idea, typically over different harmony. (This may require adjusting the melody to fit.) The new harmony casts the original idea in a new light.
[3] Begin a third restatement, don’t continue the melody as before. Instead, do something unexpected – something expressive!

Once you know this schema, you’ll notice it everywhere. Even in tunes you’ve heard thousands of times!

“Happy Birthday,” by Patty Hill

happy birthday analyzed

So schemas offer one way to use predictability to create expectation in a melody. Think of a schema as predictability on a macro level. Now I’d like to show you how to engage micro-level predictability in any melody you write.


What could be more predictable than geometry?

In geometry class, we learned that the shortest distance between two points is a straight line. The same holds true for melodic figures. Say we have two melody notes, each on a strong beat. Because they both fall on strong beats, it makes to think of the first note as a starting point and the second note as a goal. The most direct way to get from the first goal note to the next is to move by step in the same direction. That’s what option A does in the example below.

“6 different ways to get from D to G,”

6 ways to get from d to g

But options B-F take less direct routes from D to G. Each route creates a unique profile when we look at the notation. The important thing, however, is not how they look, but HOW THEY FEEL WHEN SUNG OR PLAYED.

  • Option A looks, sounds, and feels the most direct, natural, and predictable.
  • Options B and C take a roundabout route to the destination, and glide into it smoothly.
  • Options D and E lead upward, as if intending to land on some note in that register, but then each suddenly plunges downward. This sort of indirect approach gives the final note (G) an accent.
  • As for option F, how would you describe it? Complicated? Zig-zaggy? Playful? Drunk?

In my research of melodic “routes” and their effects, I’ve found 24 commonly-used figures, shown on the chart below. Turns out that 1/2 of the figures have no particular inclination for where they “go” – that is, what the next note will be. I’ve fogged out such figures on the chart below. Each of the remaining figures have a “most-expected destination.”

the 24 universal melodic figures, with predictable figures highlighted


To show you what I mean by the “most-expected destinations” here are a handful of the “predictable figures” from the chart above. The audio file plays the predictable version first, followed by the version below it (“some other outcome”).

“some predictable figures compared with another possible outcome,”


Generally speaking, the most predictable destination for any figure sounds “natural,” “easy-going,” or “logical.” Whe a composer wants to create some other effect, he’ll “interrupt” or “alter” the destination note.

Now, it’s one thing to compare versions of a figure in an abstract graphic and sound file. The true test comes when we see how real musicians use predictability and disruption to make an actual song say what they want it to say. So let’s do that.


The Beach Boy’s hit song, “Wouldn’t It Be Nice” uses two of the figures illustrated in the previous example. Here is the marked-up score. Listen a few times and see what sense you might make of my annotations before reading my comments.

“Wouldn’t It Be Nice,” by The Beach Boys

Wouldn’t It Be Nice analyzed

Only about 4% of songs begin with their hook. This is one of them. And it makes perfect sense for the song’s message. Try something. Sing the words “wouldn’t it be nice,” returning to the first note on “nice” as I’ve notated in the graphic. This predictable version kills off all dreaminess, along with every ounce of zest for life!   

Next, check out the setup for the money note – that high G on “stay.” The main body of this idea is the Run that I’ve highlighted. Breaking the natural tendency of the Run (which we heard in the previous example) would be surprising enough on its own. But Wilson makes it even more surprising by letting us hear the Run continue predictably to the D the first time through (a few bars earlier on the lyric “that”). Again, I urge you to test this for yourself. Try singing this pair of phrases the same way both times. The first time, let both Runs lead to D. The second time, let both Runs leap to the high G. I think you’ll agree that we have a prime example of the title of this blog. Wilson uses predictable to make this melody more surprising. (And delightful, and expressive, and memorable, and …)

Points for review

  • Don’t fear predictable patterns. Use them to set up expectations. Then break those expectations in expressive ways. 
  • Schemas not only offer blueprints for sketching melodies, but they are also potent devices for creating expectations. Listeners anticipate certain outcomes based on prior experience. Schemas produce macro-level predictability.
  • Most of the 24 common melodic figures have a most-predictable outcome—a way of moving to the next figure that feels smoother than all other options. Predictable figures produce micro-level predictability.
  • Once you know exactly what makes music predictable, you can learn a lot from studying your favorite composers’ melodies. You’ll begin to understand how they achieved a particular effect and why. By listening and thinking in this way over time, you will strengthen and develop your musical intuitions.  

Want to go further? You CAN!

Each chapter in this 300-page eBook breaks down everything that effective composers do intuitively. Along the way, you’ll find plenty of step-by-step instructions to help you create the same effects in your own music.

One Response

Leave a Reply

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

David Fuentes

David Fuentes

Professor Fuentes is a composer, author, teacher, and clinician. He brings over 30 years of teaching experience from institutions that include Berklee College of Music, the University of Iowa, Brandeis University, and Calvin University.

Dr. Fuentes’ music has been performed all across the world. It includes music for the classical concert stage, theater and musical theater, television, art installations, popular music genres, and the church.

A published author, his writing on composition, vocation and the arts, and the place of music in human flourishing has been influential in the development of music curriculum in an ever-changing world.

You can find samples of his music and writing on his personal website:, which also provides a longer bio.

more posts
error: Content is protected !!