# The Diagnosis Hewn Shop Notes · https://hewn.build/shop-notes/the-diagnosis AI-drafted, not yet checked: Drafted by Chip, the writing system Hewn runs on Claude, from Hewn's craft corpus. Jonathan Salomons has not read this note yet, and no outside reader has checked it. Read it as a shop note: check a number before you cut to it. Cite as: "The Diagnosis." Hewn Shop Notes, hewn.build/shop-notes/the-diagnosis. Drafted 2026-05-02; AI-drafted, not yet checked. --- The most expensive mistake in furniture repair is picking up a tool too soon. It's a common one. Look at a loose leg, think you know what happened, work for an hour, and end up with a piece that's worse than when it arrived. The joint that got reglued wasn't the problem. The problem was upstream, and nobody looked for it. Four questions. Answer them before anything else comes out of the drawer. --- **What failed?** Not what you see. What failed. The crack in the panel, the wobbly leg, the door that jams in August — those are symptoms. The failure is the decision upstream that made the symptom possible. A door that won't close in summer isn't a door problem. It's a movement problem. The door was sized for dry-season conditions, the frame was built without expansion clearance, or someone picked a species that runs wide on tangential movement and never checked the math. The door is doing exactly what wood does. The allowance was wrong. Write down what you observe. Not your interpretation — what you actually see. Your theory comes after the evidence. **When?** Sudden failure, or gradual? If a joint let go suddenly, look for impact or overload. If it's been working loose for years, look for glue degradation or chronic movement stress the original construction didn't accommodate. If the problem worsens every summer and recovers every winter, wood movement is almost certainly the mechanism. If it happened right after the piece was moved, check for racking — marginal joints often hold until transport gives them the shock they've been waiting for. Timing is usually the most diagnostic thing you've got. It tells you whether you're looking at a structural failure or a moisture failure. Those have different repairs. **Where did it originate?** Most visible failures start somewhere else. The crack is where the stress released, not necessarily where the stress built up. The finish peeling in one corner started at a contamination point six inches away. The racking that's now obvious in one joint began in the weakest joint, which may not be the one that's loose. Work backward. If the panel split at the glue line, ask why. Was it a glue failure, or did the panel have nowhere to go and the glue was holding the stress until it ran out of options? If it's the second one, regluing the panel is temporary. The repair is the slot or the attachment. **What are the constraints?** This one comes last, because you need the first three to answer it honestly. Does the repair need to be reversible? Museum restoration uses only reversible materials, because the piece may need to be undone by someone a hundred years from now. A dining table repair probably doesn't have that requirement. Does it need to match — finish color, texture, hardware? Is the construction sound enough to actually hold a repair, or is the underlying joinery compromised to the point where anything you do is a delay? And: what does the owner actually want? Full restoration, functional fix, or just stop the bleeding? That's not a small question. The answer shapes every decision that follows. --- **Document before you disassemble.** This is not optional. Photograph the piece as it sits. Multiple angles. Raking light to read the surface. Close-ups of every failure point. Measure all dimensions — a piece may have moved seasonally, and you want a baseline. If you're taking joints apart, photograph them in position first. Old hand-fitted furniture has a specific relationship between parts that isn't obvious from the components alone. The right rear leg fits that mortise, in that orientation, because it was fitted to it. Lose that information and reassembly becomes a puzzle you're solving blind. Test the finish before you touch it. Denatured alcohol on an inconspicuous area: if it gets tacky or dissolves, you've got shellac. Mineral spirits will tell you if there's wax over the top. Lacquer thinner softens lacquer and nitrocellulose. The finish system is the most consequential unknown in any repair — you can't add a compatible coat until you know what's there. --- **Write it down.** One page. What you found, when the failure appears to have started, what you think caused it, what the constraints are, and what you're proposing to do. This is the foundation of a repair quote. It's also what you refer back to when you're three hours in and something unexpected turns up — which it will. The diagnostic report is your map. The more careful it is, the more useful it becomes. A repair begun without a diagnosis is a guess. In woodworking, guesses have a way of expanding.