where she will stand

a male great bowerbird, in north queensland, builds an avenue of upright sticks two walls thick. at each end of the avenue is an open area he calls a court — about a square metre, sometimes more. he decorates it with bones, snail shells, pebbles, glass, anything pale and small enough to carry. over a season he places hundreds of objects.

from above, the court looks chaotic. it isn’t. he has sorted everything by size: smallest near the avenue, largest toward the back. the gradient is steep enough that the geometry does work for him. from inside the avenue, where the female stands when she comes to look, the distant large objects subtend the same angle as the near small ones. the court appears more uniform than it is. whatever she selects for, that uniformity predicts mating success — males whose gradient is cleanest breed most.

what surprised me: he doesn’t build the court back-to-front. he doesn’t dump everything and sort. the construction is center-outward. he starts at the avenue mouth — exactly where she will stand — and works away from her. the first ten objects already produce the gradient. researchers have stripped his court bare; in seventy-two hours he has replaced about half the objects, and the perspective is back inside the first hour.

he never sees what she sees. the avenue that frames her view blocks his.

he builds for an eye that is not his, in coordinates anchored to where she will stand. he doesn’t know perspective theory. he doesn’t need to. the build order is the theory.