How did two scientists find out how the brain builds shapes from light?

They found it by listening to individual brain cells in a cat while it looked at lines. In 1959, David Hubel and Torsten Wiesel put a tiny electrode into the primary visual cortex of an anesthetised cat and projected patterns of light and dark on a screen. Some neurons fired rapidly for lines at one angle, while others preferred another angle. The brain, it turned out, does not receive a picture. It builds one, starting with simple pieces like edges.

The story begins with a gadget. While serving at the Walter Reed army institute, Hubel began recording from the primary visual cortex of cats and invented the modern metal microelectrode, learning basic machinist skills to build it. In 1958 he moved to Johns Hopkins and began working with Wiesel, and together they discovered orientation selectivity and the columnar organisation of the visual cortex.

They named the neurons that liked a line at one angle simple cells. Other neurons, which they called complex cells, responded best to lines of a certain angle moving in one direction. The studies showed how the visual system builds an image from simple stimuli into more complex representations. Their idea was that each cell's receptive field is built from the input of cells at a lower level.

Then they asked what happens when the input goes wrong. In kittens deprived of one eye, the columns fed by the other eye took over, and the kittens lacked areas receiving input from both eyes, which are needed for binocular vision. Their experiments showed that the balance between the two eyes, called ocular dominance, is set irreversibly early in childhood, which opened the door to understanding and treating childhood cataracts and a squint, called strabismus. In 1981, Hubel and Wiesel received the Nobel Prize.