Seventh Designs

Furniture, timber and the work of the bench

Drawn plate of concentric growth rings spreading across the header.
Growth rings, read across the end of a board — third of the three header plates.

Measuring, Marking and the Story Stick

October 7, 2013 — filed in Joinery

Errors in furniture are rarely errors of measurement. A rule read carelessly is a small and obvious mistake. The expensive errors are errors of transcription: a number read correctly, remembered slightly wrong, written down in the wrong column, or applied to the wrong face. These accumulate, and they do not announce themselves until assembly.

The story stick removes the number from the process. It is a length of scrap, planed straight, on which the actual positions of every joint, shoulder, rail and opening are marked once, directly from the work or from the space the piece must fit. Afterwards every component is marked from the stick rather than from a rule. There is no number to misread because there is no number.

The stick also fixes a subtler problem, which is cumulative error. Marking five identical spaces by measuring each one from the last means five small errors added together. Marking all five from a single stick means each one carries only its own error, and the overall length stays right. Where several parts must be the same, the stick guarantees they are the same even if the stick itself is slightly out, which is usually the property that actually matters.

Marking should be done with a knife rather than a pencil wherever a cut will follow. A knife line severs the surface fibres, so a chisel or a saw dropped into it registers and stays there, and the line has no width to interpret. A pencil line is perhaps half a millimetre wide, and the question of whether to cut on it, beside it, or through it is exactly the sort of ambiguity that turns into a loose joint.

Face and edge marks are the other half of the habit. One face and one adjoining edge on every component are marked as the reference pair, and every measurement and every gauge setting is taken from those two surfaces only. This means an error in thickness or width stays where it is rather than propagating into the joints, because nothing was ever measured from the unreferenced side.

None of this is slower in practice. Marking out with a stick and a knife takes longer than reading a rule and is repaid the first time a set of rails all reach their shoulders at once.