Chance in the wood-fired kiln

Why do potters surrender a pot's final surface to a fire they cannot control?

A potter can buy a glaze in a bucket, dip a pot, and fire it in an electric kiln that holds its schedule to within a few degrees. The result is predictable, repeatable and, in a commercial sense, safe.

Instead, some potters load a long sloping wood kiln, stoke it in shifts around the clock for several days, and then open it to find that a good share of the work is unsellable and one or two pieces are better than anything they could have designed. The obvious framing is that they are trading control for beauty. But that framing hides the thing worth examining. What exactly is being given up, and what is being obtained in exchange? "Chance" is not a substance. It cannot be added to a pot.

Start with what the fire is physically doing, because the romance obscures a fairly concrete mechanism. Burning wood leaves ash, and in a kiln with a strong draught that ash does not settle politely — it is carried through the chamber by the flame path and lands on whatever it meets. Wood ash is rich in calcium and potassium, and those are fluxes: at high temperature, somewhere in the region of 1250 to 1300 degrees Celsius, they lower the melting point of the silica and alumina already present in the clay body. The ash and the pot's own surface therefore melt into one another and form a glaze that was never applied. Where the ash falls thickly it runs and pools; where the flame licks past without much ash it leaves scorching and colour shifts, since the local atmosphere — whether that pocket of the kiln is starved of oxygen or awash in it — changes what the iron in the clay becomes.

Now notice which quantities in that account are under anyone's command. Ash landfall depends on the flame path, which depends on the draught, the stacking of the whole kiln, and the shape of the pot upwind. The local atmosphere at one shelf depends on when the last stoke went in. Neither is a knob. They are outcomes of a turbulent, several-day process with more interacting variables than a potter could measure, let alone hold steady.

So the potter is not choosing the surface of any individual pot. Then what is being chosen? Look at the decisions that actually get made: kiln design, wood species, where each piece is placed in the chamber, which face points toward the firebox, the stoking rhythm, the length of the firing. Every one of those shifts the range of outcomes — the front of the kiln reliably takes heavy ash, the back reliably takes less, a piece turned this way reliably catches more flame. What is being chosen is a distribution, and then the fire draws from it.

That reframing dissolves the paradox. The potter has not surrendered control; they have moved it up one level, from specifying a result to specifying the odds. It is the difference between naming a number and choosing which die to roll.

But why would anyone want the die at all? Here the shape of the payoff matters. A predictable process gives you the same pot every time, and the value of that pot is capped at what was designed. A process with real variance gives you a spread, and a spread has a tail. Most draws are ordinary, some are ruined, and a few land on combinations of ash flow and flashing that no one would have thought to specify — and could not reproduce even having seen them. If the failures can be discarded cheaply and the exceptional pieces carry disproportionate value, then variance itself is the product. The cost is real; potters routinely describe breaking or setting aside a substantial fraction of a firing.

Three clarifications keep this from sliding into mysticism.

First, the variance is not uniform, and mastery consists largely of narrowing it where it must be narrow. Nobody accepts randomness in whether the pot survives, whether it fuses to the shelf, or whether it warps. Wadding under each foot, careful stacking, a controlled climb in temperature and a managed cooling all exist to make the structural outcome as deterministic as possible, precisely so that the surface outcome can be left free. That is a deliberate allocation, not a general shrug.

Second, the odds are learned rather than given. A potter who has fired the same kiln forty times knows which shelf ashes heavily and which corner flashes orange. The knowledge is real, transmissible and slow to acquire, and it is knowledge about tendencies rather than about results — much closer to what a sailor knows about a stretch of coast than to what a machinist knows about a lathe.

Third, unrepeatability is not merely tolerated; it is part of what the object is worth. A surface that can be bought in a bucket carries no information about how it came to be. One produced by a single unrepeatable passage of flame is, in effect, a record of an event — which is why these pieces are valued in a register closer to a photograph of a moment than to a manufactured good, and why the aesthetic traditions built around them prize the accidental mark rather than apologising for it.

The misconception, then, is that the potter has abandoned intention. They have relocated it. Intention lives in the kiln's architecture, the stacking plan, the firing schedule and — crucially — in the judgement exercised at the unpacking, where deciding which pots leave the yard is as much an act of authorship as throwing them was.

• How ash chemistry differs between wood species, and how much of the resulting colour is predictable.
• Why reduction and oxidation change iron's colour in the clay so dramatically.
• Where else a craft deliberately keeps variance in — barrel-aging, natural dyeing, open fermentation.

Every term the collection defines is gathered in the glossary.