The Eye Gets Fooled

The Horizontal and Vertical of Se-sal and Jeongja-sal

To join a single door, you first divide the slats. Whether it's se-sal or jeongja-sal, it comes down to deciding how many cells to divide the frame into and setting the spacing —

and the calculation itself isn't hard. Divide the inner dimension by the number of cells and the figure comes right out. Divide it up by that figure and stand it up, and the cells look squashed flat.

Measure with a ruler, they're equal. Measure again, still equal. Set the ruler down, step back, and look — and once again the horizontal looks wide and the vertical looks short.

So you nudge the vertical a little longer. There's no fixed value for how much. This much for a big door, that much for a small one — it's set differently each time.

Even for the same se-sal door, change the size and you look at the proportion fresh from the start all over again. This is where it splits. Take one door that came out well and just scale it up, and it's accurate by the ruler but wrong to the eye. Only the numbers carry over to the drawing, and the numbers go just as wrong the moment the size changes.

A Story I Already Knew

That the human eye sees differently from a ruler — I already knew that. I was interested enough in this sort of thing that I'd even looked up material and organized it, so it's not that I missed it out of ignorance. It's a story that was already worked out a hundred-odd years ago. That two squares of the same size look wider when filled with vertical lines and taller when filled with horizontal lines. The explanation attached to it is that a filled space looks bigger than an empty one. This is exactly why a cell filled with lattice looks different from what the ruler says.

Typography has an even more vivid version of the same story. Because a horizontal stroke looks thicker than a vertical one, drawing a lowercase letter o so it looks evenly weighted actually requires drawing the vertical strokes thicker. This was old common knowledge among type designers, and only confirmed by experiment in recent years. I knew it. But knowing it and what my hands actually did were two different things.

When I only knew it from pictures, the eye being fooled was an entertaining story. Standing in front of a door panel, it's not a story anymore — it's a job.

No matter how many times you measure, the same cell still looks squashed flat no matter how many times you look at it, and knowing the answer doesn't change what the eye sees.

Since knowing doesn't fix it on its own, in the end there's nothing to do but fix it by hand.

The Ancestors Knew It Too

Measure an old door, se-sal or jeongja-sal, and the cells are never perfect squares. The vertical is always a little longer.

And that amount was never fixed at one single value. It changes as the door gets bigger, and it differs from hand to hand.

If there had been a fixed value, every door would have to be the same. That they differ means it was set fresh, in front of the door, each time.

Today we call this an optical illusion. There are experiments, there are papers. The old carpenters had no such words for it.

Having no word for it doesn't mean they didn't know. The very fact that they reset that proportion door by door is the proof that they knew.

They knew the eye sees differently from a ruler, and they even knew that how differently it sees changes from door to door.

That's what an eye for it really means.

Not memorizing which measurement to use, but being able to set it fresh, standing in front of the door.