The Hundred-Year Chair · a chapter

So Similar, Yet So Different

On the second of September I woke at five in the morning with a chair in my head, and by a quarter to eight it was a drawing, and by that night I was sitting in it.

I already had a chair I liked. The Keel Panel concentrated its construction into the long diagonal I had worked so hard to resolve. But its seat and back still needed broad pieces of plywood. MakerStock's leftovers were mostly narrow strips, in different thicknesses and finishes. I had followed a promising design away from the material I had set out to use.

The evening before, I had been looking at Piet Hein Eek's Wehkamp chair and its lounge sibling. These were oak chairs with arms, from a flat-pack collection for the Dutch retailer. I liked them and had been studying how their members met. Rietveld's crate chair was also familiar by then, not just from photographs but from having built it: boards crossing in layers, their connections visible. And I had spent weeks working with two triangles. Somewhere in the night those things came together.

What I drew that morning was ten narrow sticks, all the same width, with rounded ends. Five made each side of the chair. The arm, back, and long diagonal enclosed one triangle; the seat rail, front leg, and diagonal enclosed the other. The triangles stood beside the sitter, leaving room for small panels between the frames. I could make those panels from strips too.

The construction followed Rietveld's example. Where sticks met, I overlapped them and put a pin through the layers. I did not need to cut their ends to meet at an angle, or make a bracket to hold them there. The sticks continued a little beyond the pins, with the same rounded end on each. They looked like the sticks children use to build bridges in science class. I called it the popsicle chair, with another thought of Mr. Grove.

I worked out the geometry in Illustrator and sent the drawing to Claude. It suggested a better order for the overlapping layers. I took the suggestion and wrote, "Bam...developed design by 7:45." By the time I went to the shop, I had a drawing I could build from.

That afternoon I cut the sticks and rounded their ends on the router table against a printed guide. The seat and back rested on small cleats between the side frames. That night I sat in the chair. My report was "pretty amazing to go from idea in the morning to sitting in it tonight."

By September 4 I had made three versions. The latest had a white arm beside a darker one, black front legs, a pale birch diagonal, and seat and back panels in several finishes. It was the mosaic I had wanted at the start of the project, made from what we actually had. "I think it is lovely and interesting," I wrote. "That is all actual scrap."

I liked this chair without needing to excuse its materials. The black film, white melamine, and exposed birch were different surfaces, but every cut showed the pale layers of plywood. The rounded ends repeated around the frame. The scraps had enough in common to look as though they belonged together without pretending to be matching pieces of new stock.

Claude had expected the pale diagonal against dark members to read as a missing part. In the shop I thought the opposite. The birch had grain where the black film had none, and I liked seeing it. I had given the machine a flat color to judge. I was looking at wood.

We also disagreed about the screws. I had fastened the small panels from above, through the plywood into hardwood cleats beneath. Claude suggested hiding the heads by driving the screws up from below. That left very little plywood for the threads to grip and risked a tip breaking through the finished face. My way put the threads in the hardwood. I kept the heads on top, one size and evenly spaced. Their regularity pleased me against the changing colors of the panels. I had no wish to disguise how the chair was put together.

A few days after the first build, I found an earlier version of the idea in my notebook. About a week before that productive morning, under a heading for the thousand-year chair, I had written, "Best chair is probably simplified Krat (see Eek variants). Add triangles! Lighten up..." Beside it were sketches of narrow members and arms. I had been approaching the chair from another direction, wondering how to refine Kratstoel for a much longer life. Then I had turned to something else and forgotten the pages.

The morning's idea had felt sudden. Yet I had been looking at Eek, building Rietveld, drawing triangles, and making those earlier sketches. By September 2, I could put the pieces together quickly enough to build the result that day.

Building it also exposed a possibility I had not been pursuing. During the first few days I pulled one pin from each side of the chair, where the arm, front leg, and diagonal met. The chair folded into a standing bundle.

That joint had already annoyed me. Three sticks shared one pin, so I could not release the diagonal without disconnecting the other two. I had complained about it while drawing the chair. Now withdrawing those two pins let the assembly move. "The most annoying thing about buying deck chairs on line is that wretched assembly," I wrote. "This is two bolts."

I was getting ahead of myself. The frame could fold, but the arms could not pass the back panel. They stuck out to either side of the flattened chair.

The diagonal had been holding the side frame in shape. Once I released the connections, the sticks could turn at their pins. That was easy to see on the shop floor. Getting all the sticks and panels to lie together was harder. Their order in the layers mattered: an arm on one side of a member might swing past it, while the same arm on the other side would collide with the back.

On September 5 I tried two arrangements on paper, then wrote across the notebook page: TOO HARD! GIVE TO CLAUDE. I asked it to try the possible orders of the five sticks in three layers. It checked 243 assignments and found none that would fold with the recessed back panel in place. A fourth layer could give the parts room to pass, but it would require a spacer and make the chair wider.

I looked at those versions and wrote, "I do not think four layers is a worthy solution. It is inelegant." I chose to take off the arms for shipping instead. The person receiving the chair would have something to assemble after all. I preferred that inconvenience to the extra layer on a chair that would spend most of its life standing open.

The folding experiment had changed what I might do with the chair without changing why I liked it. Narrow strips had brought me back to the scrap pile; pinned connections had given me a way to flatten the result. I still had to decide which of those possibilities deserved more parts or more work. I could make the chair easier to pack and less pleasing to look at.

Making another one was becoming simpler. The triangles were identical in size and shape, so several members shared the same spacing between their pins. They did not all have to end at the same distance from a pin: the back extended above the arm, and the legs continued to the floor. I could repeat a few critical dimensions while giving each piece the length it needed.

I made a story pole, a marked stick that transferred those dimensions to the parts. It had guides for drilling the holes and notched stops for the cut lengths. Instead of reading a number from the drawing and measuring each strip afresh, I could locate it against the guide. On the white face I wrote BACK + FRONT beside one hole spacing and KEEL, SEAT, ARM beside the other. Much of the layout work for the next chair was sitting on the table saw in a single piece of scrap.

This was a different use of repetition from the panel chair's keel. There, the notched diagonal located the other parts and stayed in the finished chair. Here, the guide stayed on the bench. Both let me put parts in the right places without separately laying out every connection. The Truss chair's diagonal could remain a plain strip with holes rather than a shaped part with notches.

Later I called the popsicle version the Keel Truss. It shared the two-triangle idea with the Keel Panel, but moving the triangles alongside the sitter had changed the chair. The arm now did the job of the low tie. With the triangles kept identical, the arrangement suited a low lounge chair; raising the seat toward dining height sent the rear feet sprawling farther into the room.

Claude initially told me this arrangement could not make a dining chair. That was too strong. I could change the lengths and abandon the identical triangles. I kept them for the lounge chair because I liked both the repeated lines and the small number of hole spacings. I did not need this version to do every job. I had another chair I liked for sitting upright.

The first Truss chair nevertheless felt surprisingly upright when I sat in it, given how laid back it looked. I began adjusting the seat's tilt. During a long discussion about reducing it from ten degrees toward six, Claude argued that a chair older guests needed to get out of should have less tilt. I replied, "for the record...I'm an 'older' user (66) and none of these are problematic :)"

Then I put shims under the feet and tried the angles. My report was "pretty subtle. Both are good." I settled on eight degrees. I had spent an afternoon considering a difference I could barely feel.

Sitting also brought the parts into a relationship the side drawing did not fully convey. My back met the panel, my arms rested on the caps, and my weight settled onto the seat. I could record an angle, but I wanted to know how I liked the whole arrangement. A cushion would change it again, raising me above the seat and moving me forward from the back. That would need its own trial.

By then I had two chairs I wanted to keep working on. On the first day of the popsicle build, I had written that Eek just kept designing, with no "one best way." I could see why. The panel chair had led me to a long diagonal that did several jobs. The Truss chair made better use of the narrow leftovers. I liked both, and neither stopped me looking at other people's work.

Back in 1999, when I was designing the Xootr scooter with my brother Nathan, I found myself studying the tubing connections on the Boston subway. I was struggling with the scooter's T-handle. Details I would otherwise have passed without a thought became worth examining. Throughout my life, being caught up in a project has changed what I notice. I walk through the world with new eyes, primed by need.

On October 4, walking past Salesforce Tower in San Francisco, I caught sight of a chair in the lobby. Something about it was familiar before I stopped to look. By then I was primed for chairs and triangles. Here were both: a long wooden diagonal running from the rear foot toward the front of the arm, with a triangle on either side. I took a photograph.

An image search led me to the Z Chair, which the Danish manufacturer dk3 credits to Poul Thorsbjerg Jensen and dates to 1957. The lobby chairs were derivatives of that design, not original Z Chairs. I did not know who had made them. In the lobby chairs I recognized the arrangement of the Truss chair, but also the fussy joinery I had been trying to avoid. Their tapered members met in shaped junctions; mine overlapped at visible pins.

I was also working on cushions for the thousand-year chair. The simple seat and back cushions on these otherwise elaborate chairs pleased me. Here was a commercial piece in a prominent lobby, looking quite finished with the sort of straightforward cushioning I was considering. I had stopped because of the triangles, but I was looking at the cushions too.

My own chairs were already in use, and I was beginning to hear what other people thought of them.

A chapter of The Hundred-Year Chair, drafted and refined in public. The same text lives in the book; the interactive tools and the fully-sourced technical notes live here. home · the book · the technical notes · hundredyearchair.com