n 1959 David Hubel and Torsten Wiesel reported that single cells in the cat's primary visual cortex respond best to lines and edges at particular orientations (Hubel & Wiesel, 1959).

what it is, and where to read it

A paper in The Journal of Physiology, free to read in its archive. The fuller account, with simple and complex cells and the columnar layout of orientation, is Hubel and Wiesel (1962).

what it says

Recording with microelectrodes from anaesthetised cats while projecting spots and slits of light onto a screen, they found that cortical cells, unlike retinal ones, barely responded to spots. They responded to a bar or edge of one orientation in one place, some only when it moved. By Hubel's later account the first clear response came from the edge of a glass slide being slid into the projector.

what's contested

Whether the cells are "detectors" of lines or filters in a more general code. Olshausen and Field (2005) estimated that the classic account explains only a modest share of what V1 neurons do with natural images.

methodological critiques

Anaesthetised animals, artificial stimuli, and cells sampled by an electrode that favours large neurons. Their later experiments on kittens raised with one eye closed, which showed a critical period, drew ethical objections that continue.

later reception

Hubel and Wiesel shared the Nobel Prize in 1981 with Sperry. The hierarchy of simple and complex cells was borrowed for Fukushima's neocognitron, a forerunner of convolutional networks, which closes a loop with the perceptron; see neuro and AI.

in the tour: the room called "The cat's cortex" has a placard called "Lines" ···

References

1. Hubel & Wiesel (1959). Receptive fields of single neurones in the cat's striate cortex. The Journal of Physiology. doi:10.1113/jphysiol.1959.sp006308
2. Hubel & Wiesel (1962). Receptive fields, binocular interaction and functional architecture in the cat's visual cortex. The Journal of Physiology. doi:10.1113/jphysiol.1962.sp006837
3. Olshausen & Field (2005). How Close Are We to Understanding V1?. Neural Computation. doi:10.1162/0899766054026639
4. Fukushima (1980). Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position. Biological Cybernetics. doi:10.1007/bf00344251
5. Rosenblatt (1958). The perceptron: A probabilistic model for information storage and organization in the brain. Psychological Review. doi:10.1037/h0042519

see also