I first encountered Paul Rand’s work in design school. The IBM logo, in particular. Eight horizontal bars that read as three letters. A graphic solution so reduced that there’s nothing left to remove, and yet the identity is unmistakable.

That logo taught me something I didn’t have language for at the time. Rand’s method was decomposition taken to its limit. Strip the concept down to its irreducible unit. The thing that cannot be simplified further without losing the idea. The eight bars work because they are the minimum structure required to produce the letters I, B, and M through visual inference. Remove a bar and the legibility breaks. Add a bar and the elegance disappears. The concept exists at exactly one level of resolution.

Rand wrote about this explicitly. He argued that a logo is not a picture of what the company does. It’s a concept, reduced to its simplest structural expression, that acquires meaning through association over time. The IBM bars don’t depict computers. They depict nothing. They are a structural solution to the problem of visual identity, and decades of consistent application gave them meaning. The concept is the foundation. The meaning is earned.

I think about this whenever I design anything that needs to hold at scale. A brand mark. A glyph system. A design token architecture. The question is always the same: what’s the irreducible unit? What is the minimum structure that carries the concept without ornamentation?

The glyph system I built for a personal project follows Rand’s logic. Eight symbols. Fixed. No new ones. Each one a structural reduction of a concept: systems, narrative, intuition, output, craft, design, photography, music. The symbols don’t illustrate those concepts. They represent them at a level of abstraction that allows combination and recombination. Two glyphs in proximity create a meaning that neither carries alone. The system works because each unit is irreducible.

Rand’s method is decomposition with a stopping point. You don’t decompose infinitely. You decompose until you reach the unit that carries the concept and nothing else. Then you stop. That discipline, knowing when to stop removing, is harder than it sounds. The instinct is always to add. To explain. To make the concept more explicit. Rand’s work proves the opposite: the most explicit version of a concept is often the most reduced one.

In my evaluation work, I extract lenses from practitioners like Rand. The Rand lens asks: can this concept survive further reduction? If it can, it hasn’t reached its irreducible unit yet. If removing anything breaks the concept, you’ve arrived. That lens applies to a logo, to a page layout, to a blog post, to an SCSS variable architecture. The material changes. The question holds.

What I took from Rand, before I understood what I was taking, was that the concept is the structure. Not the surface. Not the style. The structure underneath everything else, stripped to the point where nothing decorative remains. The IBM logo taught me that in design school. Twenty years of practice proved him right.