Sensory Abilities: Vision

By | May 8, 2010

Dogs are equipped with a number of specialized sensory organs evolved to obtain biologically significant information from the environment. These various sensory systems gather and process chemical, mechanical, and physical inputs, transduce them into coded electrical impulses, and then conduct the raw sensory data to the brain. Once in the brain, the sensory data are further processed and encoded into meaningful representations about the surrounding environment. The animal is totally dependent on the reliability of this information processing for the procurement of vital biological needs and all forms of adaptive learning.

The sensory capacity of dogs can be divided into three broad categories:

1. Exteroception: Exteroceptors are sensitive to all stimulation acting on dogs from the external environment. These stimuli (including light, sound, chemical agents [taste and smell], heat, cold, and pressure) correspond to the special senses of sight, hearing, taste, smell, and touch.

2. Interoception: Interoceptors are responsive to stimulation arising from within the bodily organs, such as emotional reactions and some muscular sensations.

3. Proprioception: Proprioceptors coordinate kinesthetic sensations and reflexes of the body, including the sense of balance.

Dog’s: Vision

Much of the close social exchange that occurs between dogs and people depends on the visual recognition of subtle gestures and postural signals. This visual information provides a sensory foundation for socially significant communication and harmonious interaction. Another important function of sight is to scan the environment for biologically important changes in the dog’s surroundings not detected by the other senses.


The dog’s eye is structured so that reflected light energy can be efficiently gathered and focused upon the retina, which is composed of light-sensitive neural tissue located at the back of the eye. There are two types of photoreceptor cells located in the retina: the rods (sensitive to contrasts of light and dark) and the cones (sensitive to variations in color and detail). A dog’s retina contains many more rods than cones, with the latter comprising only 3% of all the photoreceptor cells found in the canine eye. The preponderance of rods makes the dog’s vision better suited for discriminating light and dark and detecting movement than seeing color and detail.

An important structural difference between the dog and human retina is the absence of a fovea. The fovea is a tightly concentrated area of cone and ganglion cells located at the center of the field of vision in the human eye. Nearly half of the human visual cortex is involved in representing information originating in the fovea. Although not possessing a fovea, the central portion of the dog’s retina does exhibit a higher concentration of cone cells than found in other retinal areas, with cones making up approximately 20% of all the receptors found there. Instead of possessing a fovea, the dog’s retina contains a visual streak and a concentration of ganglion cells called the central area. The visual streak is an elongated oval concentration of light-sensitive receptors and ganglion cells situated along the central portion of the retina. The visual streak and central area are believed to play important roles in enhancing visual acuity, binocular vision, and horizontal scanning. Peichl (1992), who compared the visual streaks of dogs and wolves, found that wolf retinas consistently possessed a pronounced visual streak, whereas the visual streak in dog retinas varied considerably (moderate to pronounced) among the different breeds studied. He attributed these differences to the effects of domestication and breeding, thus providing additional physiological support for what Hemmer (1990) has termed the decline of “environment appreciation” in dogs and other domestic animals due to sensory and neurological degeneration resulting from domestic breeding (see).

Clarity of vision requires that an optical image is precisely focused on the retina. This function is achieved by the cornea, lens, and the aqueous media within the eye. Like humans, many dogs are either farsighted (hyperopia) or nearsighted (myopia). In the case of farsightedness the image is focused behind the retina, whereas in nearsightedness the image is focused in front of the retina. Myopia is not a general characteristic of canine vision (as has been sometimes suggested), but its incidence is relatively more common among certain breeds. For example, Murphy and colleagues (1992) found that 64% of the Rottweilers they tested were myopic. They also determined that 53% of the German shepherds tested (clinical population) were myopic, but, interestingly, of the German shepherds participating in a guide-dog program, only 15% were affected by the condition. This finding suggests that dogs with poor focusing abilities had been excluded from the guide-dog population as the result of other behavioral shortcomings arising during their training. Certainly, dogs like the German shepherd and Rottweiler should be tested for myopia before being trained for various utilitarian tasks requiring good eyesight.

Color Vision

Until recently, many dog authorities believed that dogs lacked color vision or, at best, it was considered a very weak aspect of canine vision. This opinion was based on early color-vision studies carried out by Smith (1912) and by Orbelli (1909) before him. Smith performed a series of color-brightness discrimination experiments (primitive by contemporary standards) with dogs and concluded that, although dogs appeared to exhibit a rudimentary ability to discriminate color, this ability was “highly unstable and cannot be supposed to play any part in the animals normal existence” (Smith, 1912). Pavlov (1927) also reported disappointing results following a number of color-vision experiments performed by his laboratory associate Orbelli, whose early findings were consistent with Smiths results. During a series of color discrimination studies, Orbelli was not able to demonstrate a differential response to color, although he was able to achieve some apparent color recognition in one dog — a feat that was accomplished only after great effort and difficulty. Pavlov reported,

The results obtained by other investigators, both Russian and foreign, lead to the conclusion that colour vision in dogs, if present, is only of a very rudimentary form, and that in most dogs it cannot be detected at all. (1927:132-133)

After several frustrated experimental efforts, Orbelli concluded that dogs did not differentiate between colors but rather responded to changes of brightness in the samples presented to them. However, other researchers during this same period — ostensibly implementing controls for brightness — reported conflicting results regarding color vision in dogs. A significant procedural difference between these experiments and the ones performed by Orbelli was the use of Pavlov’s salivary method versus instrumental methods in which a dog is required to make a voluntary response indicating a choice between color samples. Experimenters using instrumental discrimination methods involving a voluntary response found that dogs did, in fact, possess some significant color vision. Stone (1921) criticized these early efforts to establish the existence of color vision in dogs, arguing that they had failed to control adequately for differences of brightness associated with the color samples presented. He suggested that positive studies indicating the presence of color vision in dogs were confounded by uncontrolled brightness factors, and concluded along with Smith and Orbelli that “the dog possesses only very rudimentary sensitivity to colors and depends very little, or not at all, on color distinctions in daily life” (1921:415).

The question of color vision in dogs has remained controversial ever since. However, highly controlled vision studies carried out by Neitz and colleagues (1989) and Jacobs and coworkers (1993) have demonstrated that dogs do possess significant abilities to perceive and use color. Neitz and coworkers, for example, determined through a series of color discrimination experiments (e.g., sample-matching discriminations) that dogs can differentiate dichromatic colors having spectral absorption peaks at 429 nm (blue-violet range) and 555 nm (yellow-green range). Spectral neutrality (colorlessness) was found to occur at 480 nm (i.e., the greenish blue range). The dog’s dichromatic color vision enables a dog to discriminate bluish objects from yellow ones, but dogs are unable to differentiate between many other colors that are vivid to humans, for example, red, orange, and green — colors that dogs probably perceive as tints and shades of yellow or blue. The various colors that dogs see are affected by a composite of perceptual inputs other than saturated hue, for example, value (lightness/darkness) and intensity (brightness/dullness). Miller and Murphy (1995) noted that dogs are unable to differentiate between greenish blue and gray. This observation is based on findings by Neitz and colleagues (1989) that the range between 475 and 485 nm (greenish blue to humans) is spectrally neutral (i.e., colorless) to dogs. The dog’s inability to discriminate between greenish blue and gray occurs, on the one hand, because a dog’s dichromatic vision cannot perceive the greenish blue hue but, also, because the normal value of greenish blue is in the gray range. These current findings conflict with an earlier study performed by Rosengren (1969), in which dogs (three female cocker spaniels) were ostensibly trained to discriminate between red, blue, green, and yellow hues. In addition, she found that the dogs could distinguish these various colors from gray samples of different values.

In contrast to earlier reports indicating the existence of only minimal (if any) color vision in dogs, Neitz and colleagues (1989) found that dichromatic color discriminations were rapidly mastered by the dogs they studied (two Italian greyhounds and a toy poodle), noting that color discrimination was evident after only a single day of training. They concluded that “color vision for the dog is not simply a laboratory curiosity, but rather may provide a useful source of environmental information” (1989:124).

Jacobs and colleagues (1993) confirmed the findings of Neitz et al. by means of sophisticated optometric instruments for measuring relative absorption rates of the photopigments contained in the dog’s cone receptor cells (electroretinogram flicker photometry). They found that dogs, like the foxes, possess two different photopigmented cones that reach spectral absorption peaks at 430 to 435 nm and 555 nm, respectively. In the case of trichromatic vision, blue-sensitive cones reach absorption peaks at 420 nm, green-sensitive ones at 534 nm, and red-sensitive pigments at 563 nm. In general, mammalian photosensitivity is limited to a narrow electromagnetic wavelength range between ultraviolet and infrared, that is, approximately 380 to 760 nm.

Vision in Subdued Light

Although the dog’s ability to recognize detail and color is limited, dogs possess significant abilities to see under conditions of subdued light or in darkness. Unlike humans, who are phylogenetically adapted to diurnal (daytime) activities, dogs are biologically adapted to a crepuscular rhythm of activity — that is, they are most active around dawn and again at dusk. Selective pressures have resulted in the evolution of structures and mechanisms facilitating vision under subdued light conditions. Night vision is made possible by a photosensitive chemical called rhodopsin contained in the rod receptors. When light energy falls on the rods, the rhodopsin is chemically altered or “bleached” out, transducing a neural signal that is relayed via bipolar cells to the retinal ganglion and, finally, to the optic nerve and brain. When the light source is removed, the photochemical gradually recovers to its original state. Unlike cones, which are linked to individual fibers in the optic nerve, numerous rods are synaptically connected in the retina to the same neural fiber. This “wiring” is a structural feature of canine vision that yields a loss of visual detail but an increase in light and movement sensitivity.

The dog’s vision under subdued light is enhanced by a special reflecting surface, called the tapetum lucidum, located behind the retina. Under conditions of low lighting, the pupil dilates, allowing as much light as possible to enter the eye. Unabsorbed light passing over the retina is concentrated on the tapetum and reflected back over the light-sensitive rod receptors, thus causing added bleaching of rhodopsin and a greater sensation of light. The reflective glow of a dog’s eyes when exposed to bright light in darkness is caused by the mirrorlike tapetum. Located below the tapetum lucidum in the lower part of the eye is a heavily pigmented tapetal structure called the tapetum nigrum, which is believed to absorb excessive light entering the eye and thereby reduce glare and scatter effects. Whereas the tapetum lucidum is adapted to accommodate light reflected from the darker earth, the tapetum nigrum is adapted to handle brighter light coming from the sky. These two tapetal structures work together to optimize the amount of illumination reaching the retina.

Binocular Vision and Depth Perception

In general, the eyes of predators are set toward the front of the head, giving them a much sharper and wider field of binocular vision than experienced by prey animals. The eyes of prey animals are usually located more toward the side of the head, giving them an ability to scan the surrounding environment widely for approaching danger. Binocular vision depends on a field of ocular overlap between left and right eyes and a network of complex retinal projections involving both sides of the visual cortex. Such visual abilities enable predators to precisely locate, focus, and track a prey’s movement. As the result of the placement of the eyes and the presence of a prominent muzzle blocking a full frontal view, the average dog exhibits only approximately 40 to 60 degrees of overlap between the right and left eyes. This gives dogs binocular capabilities that are good but inferior to human abilities. Such anatomical limitations, however, are a gain in terms of peripheral vision. Whereas human peripheral vision extends to about 180 degrees, the average dog’s peripheral range is approximately 250 degrees. Of course, the amount of binocular and peripheral vision varies considerably from breed to breed depending on how the eyes are set in the skull and various neural substrates mediating visual perception.

An important aspect of binocular vision is depth perception. Although a dog’s binocular vision is good, a dog’s ability to perceive depth is somewhat mitigated by a lack of full binocular vision. Since a dog’s binocular vision is limited to a more or less narrow frontal range, a dog’s ability to perceive depth is also restricted to a narrow field of vision located directly in front of it. Miller and Murphy (1995) have pointed out, however, that depth perception does not rely on binocular vision alone. Monocular depth perception is also possible as the result of head movements that produce an appearance that objects are moving at different speeds relative to one another, thus providing information about relative distances and depth between them. Other sources of important information concerning depth include foreground/background contrast, atmospheric or aerial perspective (clarity of contour), relative size/scale of objects, linear perspective, overlapping, and vertical location in the field of vision.

Shape and Form Discrimination

Humphrey and Warner (1934) reported an interesting study suggesting that a dog’s ability to form clear object images is limited both under close-up conditions and at distances, indicating that dogs may have a very narrow range of effective vision. They reported a study by Karn (1931) in which dogs were trained to discriminate between two triangles, one with its apex pointing up and the other pointing down. These triangles had 9-inch sides, and the dogs were permitted to approach as closely as they liked before choosing between them. Karn found that the dogs usually made choices only after they were within 20 inches of the triangles. By progressively making the triangles smaller, he was able to obtain reliable discrimination between triangles with 3-inch sides but no smaller. Humphrey and Warner note that a parallel deficiency in human vision would result in our not being able to read a book unless it had “letters three inches high.”

Karn and Munn’s results conflict somewhat with earlier findings by Pavlov’s associate, Shenger-Krestovnikova, who experimented with very subtle shape discriminations in dogs (Pavlov, 1927). In one of her experiments, dogs were required to discriminate between a circle and an ellipse. Over the course of several trials, the shape of the ellipse was gradually expanded in the direction of a circular shape. This was accomplished by altering the ratios of the semiaxes of the circle to the ellipse from 3:2, 4:3, 5:4 … 9:8. She found that dogs could master discriminations as fine as 9:8 but only with great difficulty. A few of the dogs studied developed striking neurotic sequelae as a result of the perceptual and emotional distress caused by the difficult visual discrimination (see). I once participated in a feasibility study involving military scout dogs that required them to perform a number of sophisticated remote tasks. These tasks were shaped through progressive preliminary training that began with a simple pattern discrimination in a Y maze. The cards (checker patterned and blank) were approximately 12 inches square and placed about 15 feet away from the decision point. The dogs showed great difficulty in mastering this simple discrimination task. After several days of frustrated effort, a flashing light stimulus was added to augment the positive card and to facilitate the discrimination required. The difficulty exhibited by the dogs (German shepherds bred at Biosensor Research for trainability and intelligence) may have been related to a perceptual factor similar to the one discovered by Karn — that is, perhaps the choice point was too far away for them to differentiate accurately between the discriminative stimuli being presented.

Despite the dog’s apparent difficulty in discriminating stationary shapes and patterns, most dogs unquestionably possess excellent abilities to discriminate between individuals at distances and in groups — a common and readily demonstrable observation. This ability may be due to an acute sensitivity to movement and subtleties of gesture. Whitney (1961) reported that dogs that had been previously addicted to morphine would copiously salivate whenever they saw him coming toward their kennel. When approached by strangers, the salivation effect was never evident unless he happened to be part of the group. Whitney claims to have observed, through field glasses, addicted dogs salivating as he approached their kennel at variable distances up to 120 feet away. Miller and Murphy (1995) reported a study performed in 1936 with 14 police dogs. In this study, dogs that could identify moving objects at 810 to 900 meters (m) could only recognize these same objects when stationary at much closer distances (585 m or less).


Occasionally, dogs lose their sight. Some common causes of blindness or loss of visual acuity include cataracts, progressive retinal atrophy, and glaucoma. Although sight is an extremely important sensory ability for dogs, blindness need not be a cause for euthanasia. Dogs appear to adjust well through compensatory reliance on other senses like hearing and smell and probably with the help of kinesthetic learning of the environment. Chester and Clark (1988) carried out a survey of 50 dog owners with blind dogs. Only 22% of those surveyed noticed a change in their dog’s temperament. The most common temperament changes reported were an increase of dependency and attention-seeking behavior. Other changes included increased fearfulness toward family members or other dogs. Of the owners, 74% reported that there was no change in their dog’s response to strangers; 12% reported that their dog failed to compensate adequately within familiar surroundings. Only two dogs were reported to experience an increase in aggressiveness — one of which was explained as the result of painful glaucoma and “resentment” about being medicated. This is a somewhat surprising finding, since many behavioral specialists regard blind and deaf dogs as being more prone to develop aggression problems.

To facilitate a blind or vision-impaired dog’s adjustment, appropriate training and management efforts must be carried out. Much of what such dogs performed effortlessly in the past may need to be laboriously relearned. Managing to climb up and down stairs, for example, often proves to be a particularly difficult challenge for blind dogs, but, with patient and gentle encouragement, most blind dogs can learn to climb steps without assistance. Blind dogs appear to form a mental map of the house and quickly learn to avoid bumping into things, provided that the owner is careful not to rearrange furniture or leave objects in the dog’s path (e.g., kitchen chairs left out from under the table). Such dogs should be fed in the same place and, for added safety, crated when the owner is absent. Also, the owner might wear a bell as an auditory means to communicate his or her whereabouts to the dog (Campbell, 1992). Olfactory cues (e.g., citronella oil) can be dabbed lightly on the corners of doorways, furniture, and other objects that may be bumped into as the dog moves through the house. The strategic placement of gates and other barriers is also useful. As in the case of deaf dogs, training blind dogs is based on the utilization of sensory modalities other than the disabled one, especially hearing and touch. Van der Westhuizen (1990) has recommended teaching dogs to respond to directional cues such as “left” and “right” to help guide their movements. He suggests that heeling lessons are facilitated by allowing the dog to make contact or lean into the handler’s leg, thus providing additional means to orient the dog’s movement while walking in close quarters. In addition to the use of gentle physical prompts, training blind dogs depends on the use of expressive verbalization, using tonal variations and inflections to promote effective communication. Both blind and, as will be seen in a moment, deaf dogs are at considerable risk of being injured by pedestrians (e.g., bicyclists and skaters) or by vehicular traffic and should be leashed whenever away from home.

Incoming search terms:

  • sensory abilities of dogs