Adventures in amateur furniture construction

Not long after we moved into our current apartment, something like three years ago now, I guess, I built a long, low and narrow shelf for putting potted plants on, along a long row of windows. It was designed to fit snugly over a long and low radiator at the bottom of these windows. We almost never have that radiator on because it genuinely rarely gets cold enough here that dressing warmly indoors isn't enough for comfort, so the proximity isn't a problem for the plants.

My approach to building this shelf was extremely naive, or if you want to try to make it sound principled, maybe we we could say "brutalist". I bought two pieces of wood from the hardware store, a plank maybe 30cm wide by 2.5m long, or thereabouts, maybe 1.5cm wide, and a square beam, maybe 5x5cm, and probably again about 2m or 2.5m long. At home I used a handsaw to cut the beam into six equal lengths - no scrap left over! And I stuck those six short legs at the four corners of the plank and at the midpoints, and attached each by drilling a single screw through the plank right into the centre of the leg. Two pieces of wood, six screws, no wasted material, and, of course, no finesse or style or anything. I didn't even sand or stain anything.

It's in no way a thing of beauty, and after three years of service the top of the plank is very unevenly discoloured due to irregular exposure to sunlight and spilled water, but you know what, it works just fine. The plants themselves and the long row of large windows capture your attention first and foremost, so the "ugly" shelf doesn't detract from the appearance of the space at all. I call it a success and I enjoyed the experience. I was of course dissatisfied with some aspects. Using a handsaw to make a perfectly straight and square cut through a bit of wood is harder than you might think, so attaching the legs directly to the base of the plank is no guarantee that the two are perpendicular. A single screw fastening of course leaves the legs free to "roll" about their axis. And with the screws being only 2.5cm or so from the edges of the plank, there's a split or two.

About a year or so later I tried another project, a bit more ambitious, a bench with a shelf underneath it to give us more working and storage space in the kitchen. I wanted the result to be substantially "better" than the plant shelf, but I still wanted to stay close to the principles of simplicity, unfussiness and simplicity of self-construction. The two big changes I made were: using the hardware store's (surprisingly cheap) wood cutting service to get much more precise edges to things, and coming up with a way to attach pieces quickly and easily using metal angle brackets, but gluing nice decorative bits over the top in such a way as to conceal most of the visible "seams". The construction is nothing that a genuine wood-working enthusiast would be in the slightest bit impressed by, but the appearance is a very substantial improvement over ye old brutalist plant shelf with visible splits and whatnot. I'm really happy with the balance I struck. I recently extended the under-shelf of this piece off to one side because I wanted yet more kitchen storage space, for reasons I might write about later.

While that bench was a definite success, I also learned a lot along the way that I would want to somehow bring to a third project, if there is one (and there probably will be, a coffee table). I thought a lot more about the design of the bench than I did the plant shelf, I actually had to sketch the whole thing out on paper, calculate lengths, etc. The biggest learning experience was that, in doing this, I mentally modelled pieces of cheap hardware store wood as ideal Platonic solids. It turns out, unsurprisingly in retrospect, that they are not. When you buy what purports to be a 5cm x 5cm x 2.5m square wooden beam, there is no guarantee that all four of the sides are actually 5cm long. There's probably +/-2mm tolerance, and I don't mean here that you might end up with a piece where all four sides are 4.9mm or 5.2mm, but rather where one side is 5.0 and another two are 5.2mm and the last is 4.9. By extension there is no guarantee that all four faces are parallel to one another. There is no guarantee that the beam is perfectly straight and perfectly untwisted along its entire length. I assume there are places in the world you can buy pieces of wood which do have these guarantees, but I don't wanna contemplate how much they might cost.

The consequence of all this is that you can't just assume two pieces of wood will end up being parallel because they are joined by two other pieces of wood of equal length using supposedly right-angled metal brackets, which is precisely what I did. I was for a short while really annoyed, but more so utterly baffled, when the legs on one side of the bench were visibly not perpendicular to the ground, while those on the other side were. Not so badly, mind, as to affect the stability of the bench in any way. It's perfectly functional. But if you look for it, you can see it. I double checked my *explicit* assumptions with tape measure and spirit level and they were correct, but the violation of high-school-level geometric principles was right there in front of me. I eventually realised that my *implicit* assumptions about the perfection of the wood were at fault. I'm not exactly sure yet how to carry this lesson forward to the coffee table project, but I'll have to. Well, maybe I won't, it will be smaller, after all, which reduces the impact of some of these things a bit, but still..

It's not something I really want to devote myself to perfecting at all, but I am convinced there is some kind of underlying combination of philosophy and technique here which could be aggressively optimised to interesting ends; figuring out how to best to trade-off between maximising functionality/quality of construction while simultaneously minimising ugliness, cost and waste (in the form of unusable off-cuts), subject to the constraints that all materials are easily available from generic chain hardware stores and construction work is mostly done by one or two people in an apartment with basic tools. The result of this optimisation process would probably be something quite different from both fancy high-end wooden furniture with handmade dovetail joints and the kind of heavily cost-optimised, industrial mass-production you find in IKEA furniture, which relies heavily on very non-generic metal doodads that I don't even have names for. It intrigues me greatly.