Briggman et al. — mouse retinal direction selectivity

This is the cleanest early demonstration that a wiring asymmetry explains a computation. Direction-selective ganglion cells were identified functionally, then the retina was imaged by serial block-face SEM and the starburst amacrine inputs traced. The wiring was asymmetric in exactly the way the direction preference predicted.

What it is good for. The canonical worked example of a structural claim earning its functional interpretation — the standard the site’s rubrics point at.

What it does not support. Generalisation beyond this circuit. The retina is unusually tractable: a thin, layered, well-characterised structure with a known computation to explain.