What the proof tells you

Designs lie when they're in your head. The moment you build one — print it, cut it, hold it — it starts telling the truth.

I notice this every time I move from screen to material. On screen, a curve is just a curve. In oak, the same curve has a grain running through it, a weight in the hand, a sound when it's tapped. The screen can't show me any of that. The proof can.

The first oak prototype for TaBasket looked nothing like the renders. The proportions I had been so sure about — five petals around a pentagonal frame, mathematically correct — felt wrong the moment I stood it on the floor. The numbers were the same. The object disagreed with them. I rebuilt it three times before the proof stopped arguing.

This happens with two-dimensional work too. A composition that looks balanced in the file feels lopsided when printed. A typeface that looks gentle on a 27-inch monitor looks brittle on paper. None of this is information I could have predicted by staring at the screen longer. It's information the proof gave me by existing.

I've started to treat the proof as a collaborator. It has opinions. It tells me when a joint is going to fail, when a margin is too tight, when a colour is too cold. My job is to listen — not to argue with it because the screen said something different.

The hardest version of this is when the proof contradicts a decision I'm attached to. The temptation is to push through, to insist that the next iteration will fix it. Sometimes it does. More often the proof was right the first time, and I lose a week pretending otherwise.

Designers I trust seem to share a habit: they build the cheapest possible proof as early as possible, and then they actually look at it. Not at what it could become, but at what it is.

The proof is rarely the design I imagined. But it's usually the one that ends up worth keeping.