The Wrong Level of Abstraction

I was playing Love on Top by Beyoncé and stumbled upon a progression that initially seemed impossible to explain.

The song is centered around C major, yet shortly after the opening Cmaj7 it introduces a sequence of chords that appears to come from somewhere else entirely:

Cmaj7 → F#m7♭5 → Fmaj7 → F/G → Abmaj7 → G

My first reaction was to search for a functional explanation. What is F#m7♭5 doing in C major? Why does an Abmaj7 suddenly appear before G? The progression looked like a collection of unrelated harmonic events.

The more I studied it, the more I realized that the problem was not the progression. The problem was the level of abstraction at which I was observing it.

Viewed as a sequence of isolated chords, the harmony appears strange. Viewed as a system of moving voices, it becomes remarkably coherent.

The F#m7♭5 chord shares notes with the preceding Cmaj7. The following Fmaj7 differs by only a small movement in one voice. The transition through F/G creates tension that naturally seeks resolution. Even the seemingly foreign Abmaj7 functions as a chromatic approach to G, creating a brief moment of instability before release.

What initially looked arbitrary turned out to be highly structured.

The lesson extends far beyond music.

Many complex systems appear confusing when observed through the wrong lens. We focus on visible structures and struggle to explain their behavior. We see the organizational chart but not the flow of information. We see a market outcome but not the incentives that produced it. We see a person’s decisions but not the constraints under which they were made.

At one level of abstraction, the system appears irrational.

At another, it becomes almost obvious.

This is one of the recurring challenges of understanding complexity: determining what should be treated as the object of analysis.

In music, I initially treated the chords as the fundamental objects. Once I shifted my attention to voice leading, the apparent mystery largely disappeared.

The same mistake occurs elsewhere. We often become attached to the artifacts generated by a system and mistake them for the system itself.

The chord is not the music.

The organizational chart is not the organization.

The codebase is not the software system.

The observed behavior is not the process that produced it.

Understanding frequently begins when we stop asking whether a structure makes sense and start asking what dynamics could have generated it.

Sometimes the fastest path to clarity is not acquiring more information.

It is changing the level of abstraction.