This wiring diagram would account for the properties of a complex cell. As in the figure on page 18, we suppose that a large number of simple cells (only three are shown here) make excitatory synapses with a single complex cell. Each simple cell responds optimally to a vertically oriented edge with light to the right, and the receptive fields are scattered in overlapping fashion throughout the rectangle. An edge falling anywhere within the rectangle evokes a response from a few simple cells, and this in turn evokes a response in the complex cell. Because there is adaptation at the synapses, only a moving stimulus will keep up a steady bombardment of the complex cell.
A long, narrow slit of light evokes a response wherever it is placed within the receptive field (rectangle) of a complex cell, provided the orientation is correct (upper three records). A nonoptimal orientation gives a weaker response or none at all (lower record).
tion or inhibition to the cell. Because methods for obtaining this kind of knowledge don't yet exist, we are forced to use less direct approaches, with correspondingly higher chances of being wrong. The mechanism summarized in the diagram on this page seems to me the most likely because it is the most simple.
Complex cells represent the next step or steps in the analysis. They are the commonest cells in the striate cortex—a guess would be that they make up three-quarters of the population. The first oriented cell Wiesel and I recorded—the one that responded to the edge of the glass slide—was in retrospect almost certainly a complex cell.
Complex cells share with simple cells the quality of responding only to specifically oriented lines. Like simple cells, they respond over a limited region of the visual field; unlike simple cells, their behavior cannot be explained by a neat subdivision of the receptive field into excitatory and inhibitory regions.
Turning a small stationary spot on or off seldom produces a response, and even an appropriately oriented stationary slit or edge tends to give no response or only weak, unsustained responses of the same type everywhere—at the onset or turning off of the stimulus or both. But if the properly oriented line is swept across the receptive field, the result is a well-sustained barrage of impulses, from the instant the line enters the field until it leaves (see the cell-response diagram on page 17). By contrast, to evoke sustained responses from a simple cell, a stationary line must be critically oriented and critically positioned; a moving line evokes only a brief response at the moment it crosses a boundary from an inhibitory to an excitatory region or during the brief time it covers the excitatory region. Complex cells that do react to stationary slits, bars, or edges fire regardless of where the line is placed in the receptive field, as long as the orientation is appropriate. But over the same region, an inappropriately oriented line is ineffective, as shown in the illustration on this page.
The diagram on this page for the complex cell and the one on page 73 for the simple cell illustrate the essential difference between the two types: for a simple cell, the extremely narrow range of positions over which an optimally oriented line evokes a response; for a complex cell, the responses to a properly oriented line wherever it is placed in the receptive field. This behavior is related to the explicit on and off regions of a simple cell and to the lack of such regions in a complex cell. The complex cell generalizes the responsiveness to a line over a wider territory.
Complex cells tend to have larger receptive fields than simple cells, but not very much larger. A typical complex receptive field in the fovea of the macaque monkey would be about one-half degree by one-half degree. The optimum stimulus width is about the same for simple cells and complex cells—
in the fovea, about 2 minutes of arc. The complex cell's resolving power, or acuity, is thus the same as the simple cell's.
As in the case of the simple cell, we do not know exactly how complex cells are built up. But, again, it is easy to propose plausible schemes, and the simplest one is to imagine that the complex cell receives input from many simple cells, all of whose fields have the same orientation but are spread out in overlapping fashion over the entire field of the complex cell, as shown in the illustration on this page. If the connections from simple to complex cells are excitatory, then wherever a line falls in the field, some simple cells are activated;
the complex cell will therefore be activated. If, on the other hand, a stimulus fills the entire receptive field, none of the simple cells will be activated, and the complex cell won't be activated.
The burst of impulses from a complex cell to turning on a stationary line and not moving it is generally brief even if the light is kept on: we say that the response adapts. When we move the line through the complex cell's receptive field, the sustained response may be the result of overcoming the adaptation, by bringing in new simple cells one after the next.
You will have noticed that the schemes for building simple cells from center-surround ones, as in the illustration on page 18, and for building complex cells out of simple ones, as in the illustration on this page, both involve excitatory processes. In the two cases, however, the processes must be very different. The first scheme requires simultaneous summed inputs from centersurround cells whose field centers lie along a line. In the second scheme, activation of the complex cell by a moving stimulus requires successive activation of many simple cells. It would be interesting to know what, if any, morphological differences underlie this difference in addition properties.
Many complex cells respond better to one direction of movement than to the diametrically opposite direction. The difference in response is often so marked that one direction of movement will produce a lively response and the other direction no response at all, as shown in the diagram on this page. It turns out that about 10 to 20 percent of cells in the upper layers of the striate cortex show marked directional selectivity. The rest seem not to care: we have to pay close attention or use a computer to see any difference in the responses to the two opposite directions. There seem to be two distinct classes of cells, one strongly direction-selective, the other not selective.
Listening to a strongly direction-selective cell respond, the feeling you get is that the line moving in one direction grabs the cell and pulls it along and that the line moving in the other direction fails utterly to engage it—something like the feeling you get with a ratchet, in winding a watch.
We don't know how such directionally selective cells are wired up. One possibility is that they are built up from simple cells whose responses to opposite directions of movement are asymmetric. Such simple cells have asymmetric fields, such as the one shown in the third diagram of the illustration on page
17. A second mechanism was proposed in 1965 by Horace Barlow and William Levick to explain the directional selectivity of certain cells in the rabbit retina—
cells that apparently are not present in the monkey retina. If we apply their scheme to complex cells, we would suppose that interposed between simple and complex cells are intermediate cells, colored white in the diagram on the next page. We imagine that an intermediate cell receives excitation from one simple cell and inhibition from another (green) cell, whose receptive field is immediately adjacent and always located to one side and not the other. We
This type of wiring could produce a simple-cell receptive field. On the right, four cells are shown making excitatory synaptic connections with a cell of higher order. Each of the lower-order cells has a radially symmetric receptive field with oncenter and off-surround, illustrated by the left side of the diagram. The centers of these fields lie along a line. If we suppose that many more than four center-surround cells are connected with the simple cell, all with their field centers overlapped along this line, the receptive field of the simple cell will consist of a long, narrow excitatory region with inhibitory flanks.
Avoiding receptive-field terminology, we can say that stimulating with a small spot anywhere in this long, narrow rectangle will strongly activate one or a few of the center-surround cells and in turn excite the simple cell, although only weakly. Stimulating with a long, narrow slit will activate all the center-surround cells, producing a strong response in the simple cell.