What do frogs see?

Frogs have quite a unique visual system, particularly compared to humans, monkeys, and other higher mammals. The root of the vast differences in the visual perception of the frog lies in its unique retinal ganglion cells.

Retinal ganglion cells receive input from the photoreceptor cells of the retina. They are the first real stage of coding visual information that goes beyond simple changes in response according to light intensity. The nature of the connections of photoreceptor cells to a retinal ganglion cell define what is coded. Lettvin (1959) studied the ganglion cells of the frog and identified four types of ganglion cell fairly unique to frog physiology. The properties of these cells provide an insight into what frogs see.

The first type of ganglion cell identified was termed a “sustained contrast detector”. These cells show no response to diffuse light across the cell’s receptive field1, but they have a vigorous response to a lighted edge passing through the receptive field. The response of these cells are sometimes dependant on the direction of the edge of light. Frogs were also found to have “moving edge detectors” which respond to moving edges, regardless of their level of illumination. These types of ganglion cells provide information about the nature of the frog’s environment, identifying the edges of objects and how they move to identify the boundaries of the objects present in the world.

Net dimming detectors” give a prolonged response to the extinction or reduction of a diffuse light stimulus. This has the advantage of potentially alerting the frog to the presence of a predator looming near it. These cells also have the property of responding to any moving stimulus regardless of size, shape, or contrast, and does so in proportion to the amount of dimming the stimulus produces when passing across the ganglion cell’s receptive field. This has clear advantages with respect to detecting possible predators and other imminent dangers.

The most intriguing type of frog ganglion cell is the “net convexity detector”. These cells have the curious property of being selectively responsive to a dark spot anywhere in the cell’s receptive field. This spot must be darker than the background light intensity, but can be of any size. These cells continue to respond for as long as the spot is present within the receptive field. There is an increased response when the spot moves, more so if the spot is moving jerkily rather than smoothly. If the frog is presented with an array of spots which would normally produce a response, these cells do not respond. However, if one of the spots moves relative to the others, the cells respond nearly normally. The benefit of these cells is clear. They are an efficient mechanism for the frog to use to detect flies and other potential food sources. It requires little cognitive analysis as the input is fairly direct, and since frogs have very minimal cognitive abilities, this rather automatic “fly-detection” system works well.

What is clear is that the frog receives visual input that is very different from the sort we humans receive. They would appear to not have any sort of “photographic” representation of the world from the retina as we have. They lack the right sort of retinal output. All the visual information a frog receives is directly relevant to its immediate survival needs. Our needs as humans are very different to that of a frog, and much greater. It is very unlikely that a frog will have the same sort of “mental picture” of the world that we have. They might not even have any such picture at all, and may just respond to visual stimuli in a fairly autonomous fashion.


1 – The receptive field is the area of photoreceptor cells that can effect the firing rate of the ganglion cells.