I was staring at a Vignelli subway map and a Draplin logo on the same screen, and I realized I was evaluating them with completely different criteria. The Vignelli map: I was checking for clarity, reduction, whether every element served the system. The Draplin logo: I was checking for personality, warmth, whether it felt like a human made it. Same skill (evaluation), same medium (graphic design), completely different frameworks running underneath.

That was the moment the idea of extraction clicked. These practitioners weren’t just making good work. They were operating from implicit evaluation systems. Consistent internal frameworks that determined what stayed, what got cut, what counted as good enough, and what counted as finished. And those frameworks could be pulled out, studied, and turned into operational tools.

Design history stops being academic the second you start treating it as a source of evaluation tooling.

Take Vignelli. His body of work is enormous, but the principles that produced it are remarkably narrow. A handful of typefaces. The grid as structural foundation. Color used for information, not mood. Reduction as the primary move: if it can be removed without losing the communication, remove it. Study enough of his work and you can extract a set of questions that function as an evaluation lens. Can this be reduced further? Is the typography serving the content or performing? Is complexity here structural or decorative? Would this still communicate if you removed the element you’re questioning?

Those questions are a tool. Run them against any piece of design and they’ll surface specific issues. The tool doesn’t require you to agree with Vignelli about everything. It requires you to temporarily adopt his evaluation framework and see what it catches.

Now take Draplin. Different practitioner, different values, different lens entirely. Draplin’s work radiates personality and craft. Thick strokes. Bold color. A sense that someone enjoyed making this thing. The extracted questions sound nothing like Vignelli’s. Does this feel like a person made it? Is there joy in the execution? Would this look good on a hat? (Seriously. Draplin’s hat test is a real evaluation criterion, and it catches things that more sophisticated frameworks miss.) Is the craft visible? Does the personality overwhelm the communication or serve it?

Run Draplin’s lens against the same work you just evaluated with Vignelli’s, and you’ll get different results. A piece that passes Vignelli’s reduction test might fail Draplin’s personality test. A piece that radiates personality might be carrying unnecessary complexity. The disagreement between lenses is where your actual design judgment has to show up.

Then there’s Rams. Ten principles, extensively documented, applied across decades of industrial design for Braun and later codified in a form that designers have been borrowing from ever since. The extraction is easier here because Rams made his framework explicit. But the questions still need to be derived, not just repeated. “Good design is as little design as possible” becomes: is there anything here that exists because the designer wanted to add something, rather than because the user needs it? “Good design is honest” becomes: does the form accurately communicate the function, or is the form creating expectations the function can’t meet?

The extraction protocol is the same regardless of the practitioner.

First, study the output. Not one or two famous pieces. Ten or twenty, across different contexts. The pattern you’re looking for is the recurring constraint. The thing they always do. The thing they never do. Vignelli always uses the grid. Draplin always includes personality. Rams always removes anything that doesn’t serve function. The recurrence reveals the framework.

Second, study the rejections. What do they refuse to do? What do they consider beneath the standard? Vignelli refused decorative typefaces. Rams refused ornament. Draplin refuses work that feels corporate and bloodless. The rejections are at least as informative as the commitments, because they define the boundaries of the lens.

Third, translate to questions. This is the extraction itself. You’re taking observed patterns and rejections and restating them as testable criteria. A question you can run against any piece of work and get a specific, actionable answer. “Can the number of elements be reduced?” is testable. “Is this good design?” is not.

Fourth, validate. Run the extracted lens against work the practitioner made. Do the questions correctly identify why the work succeeds on the practitioner’s own terms? If the lens doesn’t explain the practitioner’s output, the extraction missed something. Go back and look again.

The value compounds when you hold multiple lenses. I run six regularly. Each one was extracted from a different practitioner with different values and different blind spots. When three lenses agree that something is a problem, the confidence is high. When two lenses disagree, I have to decide which framework applies to this context, this audience, this project. That decision is where skill lives.

I learned more about evaluation from studying practitioners than from any design textbook. The history is full of operational frameworks if you read it looking for them.