Declare the Joint
The top and its two breadboard ends sat on the assembly table for a week. Not a long time. I want to be honest about that, because in my head it had become this looming thing I was dodging, and it was seven days. But I knew, the whole seven days, that I wasn’t going to glue them up. I’d walk past, look at the mating edges, think about it, and find something else in the shop that needed doing. There is always something else in the shop that needs doing.
The problem was the seams. Not the joints. The joints are breadboard ends and they’ll hold fine. The seams are the two visible lines where the breadboards meet the field of the top, and they run straight across the four foot width, which sounds easy right up until you try to clamp it. The outside edge of each breadboard is curved. Clamps want two flat surfaces facing each other, more or less parallel, and a curve gives you almost nothing to push against except near the middle. So I can pull the center of a four foot joint dead tight without thinking about it, and the last foot or so on either end is a negotiation.
The fix is jigs. Curved cauls that give the clamps something square to bear on, built for this table and useless after it. I knew that. I also knew I didn’t feel like building them, and that is a different sentence than “I couldn’t figure it out.”
So I didn’t have a gap problem in the abstract. I had a gap problem with a known address. Four of them, actually, one at each end of each seam. Little openings that would show every time I sat down at that table for the next thirty years.
Then Paul came by, looked at it, and asked whether I was going to put a V groove in to hide the gap.
He wasn’t trying to solve anything. He was curious about a table that isn’t his. He got an answer, we talked about something else, and he walked off with no idea he’d just taken a week-long problem off my bench. That’s the part I keep coming back to. Not that somebody helped me. Nobody helped me. Somebody was interested, out loud, near a thing I was stuck on.
Here’s what I’ve been chewing on since.
A V groove does not do what I would have told you it does. I’d have said it hides the seam. That’s roughly what Paul said, and it’s how everyone talks about it. It’s wrong.
It doesn’t hide anything. An eighth of an inch of chamfer on each edge, a quarter at the very most, and the line is more visible afterward, not less. You’ve made it wider. You’ve put a shadow in it.
What changes is what the seam is claiming. Two boards butted tight are pretending to be one board. That’s the whole proposition, and it’s all or nothing. Any gap anywhere along that line is the pretense coming apart, and your eye is very good at catching it, because your eye is looking for exactly one thing. Continuous surface, interrupted. A V groove stops making the claim. It says two pieces meet here, on purpose, and this is where. Once the line is supposed to be there, your eye reads the groove instead of the gap. It quits checking.
The tolerance didn’t get looser. The requirement moved.
Shadow lines work the same way. Set one surface back from another by a sixteenth and you’ve swapped a flush transition, which has to be perfect, for a deliberate step, which only has to be consistent. Consistent is a much easier thing to be than perfect.
Anyone who has built scenery knows this in their hands before they know it in words. Nobody butts two flats together and blends the seam on a set that loads in and out five nights a week. You batten it or you reveal it. The seam is going to move, and you’d rather it move inside a detail you designed than across a surface that’s lying about being one piece. I spent ten years around that, mostly on the lighting and rigging side rather than building the scenery myself, but you can’t work that close to it and not absorb how it goes together. And I did not, apparently, think to bring any of it to a dining table.
Which brings me to the part I don’t have sorted.
From ten feet away, the V groove I’m going to cut and the V groove a lazier version of me would cut are the same groove. Same eighth of an inch. Same shadow. You could not pick them out of a lineup, and neither could I, and I’m the one cutting it.
The difference is all upstream. It’s in whether I knew what I was avoiding when I decided not to build the cauls. And I’m not sure I can tell you honestly which one this is. I’ve got a reason that sounds like design. I also had a week of walking past that table, and the reason showed up after Paul did, not before.
Maybe those are the same thing. Craft is full of details that started as somebody covering for something and got kept because they looked right. Maybe every good design decision is a solved problem that stopped announcing itself as a problem.
Or maybe I just didn’t want to build the jigs.
I’m cutting them either way. An eighth, I think. Quarter at the outside if the shadow doesn’t read.
