# The Grammar Hewn Shop Notes · https://hewn.build/shop-notes/the-grammar 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 Grammar." Hewn Shop Notes, hewn.build/shop-notes/the-grammar. Drafted 2026-04-27; AI-drafted, not yet checked. --- Fifteen joints. One arc. If you've followed this series from the mortise to the drawbore, you've got the full vocabulary. But vocabulary without grammar is just a list. The grammar is what makes it useful. Here it is: every joint in furniture answers two questions. What loads will this connection see? And which failure mode are you protecting against? That's it. Answer those two questions and the joint selects itself. --- ## Four loads, every time Four forces act on furniture joints. Shear is gravity pulling perpendicular to the connection. A shelf in a dado is under shear. The connection wants to slide downward. The dado walls resist it. Tension is a pulling force trying to separate the joint along its axis. A chair stretcher is under tension when someone leans back. The mortise walls hold the tenon from being pulled out. Racking is lateral force trying to collapse the structure toward a parallelogram. Every table apron fights racking. It's the most common furniture failure mode, and most of what we call joinery is really a set of answers to it. Compression is force trying to close the joint. Here's the thing about compression: it almost never fails. Don't design for it. Three loads that matter. One that doesn't. --- ## The hierarchy Not all joints are equal. Some do more work. The mortise and tenon is the governing joint of furniture. It handles shear, tension, and racking simultaneously. The tenon at one-third of stock thickness is not a convention — it's the balance point where both failure modes are equally managed: thin wall on one side, thin tenon on the other. This joint has been primary for three thousand years because physics doesn't change. The dovetail is the governing joint for tension. Its mechanical interlock holds on the pull axis without glue. Everything else in this arc relies on glue for tension resistance. The dovetail does not. That's why it belongs at drawer fronts — the load on a drawer front is exactly the load this joint was designed for. The dado and the rabbet govern location. They tell a shelf where it lives. They are not tension-resistant; a shelf in a dado lifts right out. What they resist is shear, which is the load that actually acts on a shelf. That's all they need to do. Lap joints govern intersection geometry. Two pieces in the same plane. The joint's job is holding that plane relationship against racking. When someone asks why not just pocket-screw a cross-lap, the answer is that racking will find that joint in about eighteen months. The housed joints run a spectrum. Plain dado locates. Sliding dovetail adds mechanical pull-out resistance. Wedged housing locks permanently. Choose based on how much the joint needs to resist being pulled apart over time. Pocket holes govern rapid assembly in face frames and sheet goods. They're shear joints — the screw works as a clamp. They are not for structural work in solid wood across the grain. Use them for what they're good at and they perform fine. Ask them to do something else and they'll tell you, in about six months, that they weren't interested. Drawbore governs pre-stress in mortise and tenon work. Chairs need it most. Dynamic racking loads require the joint to be mechanically tight before the glue goes in. The drawbore is how you get there. --- ## The principle Every joint in this arc asked the same thing: fit to condition, not to nominal. The mortise walls need to hold under tension. The fit standard follows from that. The tenon cheeks need full long-grain contact for glue surface. The standard follows from that. The dado needs a snug sliding fit because a tight fit in wet glue becomes a fixed fit, and not the kind you wanted. The standard follows from that. Know the condition. The standard is obvious. Don't know the condition and the standard is a guess. The joint that fails is the joint that wasn't understood. --- When you pick up a plan now, you should see the joints differently. Not as techniques to execute — as arguments. The M&T at the apron is an argument about racking. The dado under the shelf is an argument about shear. The dovetail at the drawer front is an argument about tension. Every joint is making an argument. Build it like you understand the one it's making.