Folk craft regionalism

Why do handmade crafts from neighbouring valleys settle into unmistakable local styles when nobody sets the rules?

Two valleys, a day's walk apart, the same clay and the same needs — and yet a dealer can glance at a pot and name the valley. No committee ever met to decide that one village flares its rims and the other does not. So where does the boundary come from? The tempting answer is that each place is expressing its identity, but that puts the conclusion in the premises: the style is what we are trying to explain, and calling it identity only renames it.

Start with the least romantic explanation available and see how far it gets. Materials differ: this valley has iron-rich clay that fires red, that one has a dye plant the other lacks. That is real, and it will separate two traditions on colour and body. But it badly under-determines what actually lets you tell them apart — the curve of a handle, the number of bands round the shoulder, a particular way of lapping a joint. Nothing in the geology cares about that.

So try the next thing. How does a potter learn? Not from the whole tradition, but from one master, perhaps two. That is a tiny sample. Every apprentice copies with small error, keeps the error, and passes it on to a sample of one again. Do you see what that does? The reference point drifts, because each generation is copying the previous copy, not the ancestral original. And in a small pool the drift is fast, for the same reason that a rare trait can sweep a small population by luck alone — there are not enough copies for the errors to average out.

Drift alone would give you slow wandering, not a crisp local look. What sharpens it? Learners do not copy at random; they lean toward whatever is most common around them. That conformist tendency is well documented in how people acquire skills, and it acts like a ratchet: a variant that happens to get slightly ahead locally gets copied disproportionately, and pulls the valley toward a single mode while the neighbouring valley is pulled toward its own. Notice that this needs no isolation in the absolute sense. It needs only that copying within a valley is much more frequent than copying across the ridge — which a day's walk comfortably supplies.

Then there is a reason that has nothing to do with learning at all. If the pots are traded, being recognisable is worth money. A buyer who was pleased last year wants the same source, and cannot inspect the kiln — so a shared visual mark becomes a collective reputation, and producers have an interest in staying inside it and in policing those who wander. Style becomes something close to a trademark that no one owns.

Now the uncomfortable part, which is where the question's own premise starts to wobble. "Nobody sets the rules" is often false. Guilds set rules. Dealers set rules by refusing to buy the odd ones. And from the nineteenth century onward, collectors, folklorists, national revival movements and tourist markets set them very deliberately, deciding what the authentic style of a region was and then supplying it. Scottish clan tartan is the notorious case: the elaborate clan-by-clan scheme now treated as ancient was largely assembled in the nineteenth century, resting partly on the Vestiarium Scoticum of 1842, a book whose claimed ancient source has never been produced. Harris Tweed went the whole way and got its definition written into statute.

And one more twist worth sitting with. How do we know an unlabelled pot came from the upper valley? Usually from its style. Museum collections were assembled partly by attributing objects to regions on the basis of how they look — which means the sharpness of regional style is, to an unknown degree, an artefact of the sorting. The pieces that did not fit went into the wrong drawer, or no drawer at all.

Three processes are stacked here and it is worth keeping them apart. Drift makes traditions diverge, given small learning pools. Conformity and market reputation concentrate each tradition around a mode, turning gradual difference into apparent boundary. Codification — by guilds, dealers, revivalists, states — then freezes a snapshot and makes it a rule, which is why so many "timeless" regional styles turn out to date from the moment someone wrote them down.

A caution about how confidently any of this can be asserted. That neutral drift models fit the changing frequencies of decorative motifs surprisingly well is a genuine finding, but a good fit is not proof that copying was actually random — selection can imitate drift in the data, and archaeologists argue about exactly this. Treat drift as the null hypothesis that keeps you from over-explaining, not as the demonstrated cause in any particular valley. And note that the people who tell you a regional style is ancient and unbroken — revival movements, tourist boards, producers protecting a name — have a direct interest in that being true.

• How archaeologists try to tell neutral drift from deliberate selection in the same pattern data.
• What happens to a living craft when a legal definition of it is written down.

Every term the collection defines is gathered in the glossary.