Sir Winston Churchill once made the following statement: “I am fond of pigs. Dogs look up to us. Cats look down on us. Pigs treat us as equals.”

While Sir Churchill’s perspective may be indicating the direction of a partnership, it also reminds us that we must be mindful of the biases and assets that can influence social interactions and other outcomes.

Jakob von Uexküll, a notable behavioral physiologist and ethologist who was most famous for his work surrounding an organism’s “Umwelt”, argued that humans must be sensitive to an organism’s internal and external environment when investigating its sensory abilities and the influence of those senses on behavior.  Honeybees “see” (via UV light) the world of flowers very differently than humans do, and what is perhaps even more remarkable, flowers have responded to the Umwelt of their primary pollinators. The image below shows a flower as it would look to a human who can’t see UV light and then how the flower looks to a honeybee using UV light.  The flowers reflect UV wavelengths, a characteristics that apparently evolved over time.

Dogs have long been of interest to humans. Whether being a human’s best companion (image of author’s dogs below), loyal and unconditionally loving, or the newest and most popular animal subject for all types of cognitive research, dogs continue to be integral members of our human lives.  As the former photographer, television personality, and animal welfare advocate, Roger A. Caras said, “Dogs are not our whole life, but they make our lives whole.”

In a recent review published by the Psychonomic Bulletin & Review, and the article focused on in this post, Sarah-Elzabeth Byosiere, Philippe Chouinard, Tiggani Howell, and Pauleen Bennett reviewed the scientific literature on the sensory abilities and perception of dogs in an effort to better understand the results of cognitive studies performed with dogs. The premise of their paper was, do we really understand the visual sensory and perceptual systems of dogs well enough to interpret and validate the results of visual-based cognitive testing?

Following a fairly recent review of the literature by Bensky and colleagues (2013), about 74% of common dog-cognition tasks use visual stimuli. Interestingly though, we have very little knowledge of the basic visual processing capabilities of dogs (i.e., color vision, sensitivity to low light conditions, brightness discrimination, visual acuity, depth perception, and sensitivity to computer monitors) and about their perception of visual stimuli such as visual and facial discriminations or visual illusions. Given that so many tasks are computer-based, it is particularly important that researchers understand the basic visual processing abilities of dogs.

Additionally, the size and morphology of the canine face across the 400-500 dog breeds currently registered in the American Kennel Club and the International Cynologic Federation are highly variable.  This variability may impact the perception of visual stimuli.  For example, having a flat face like a pug or a long nose like a dachshund would change what the dog can see as it tries to look around its nose as in the case of the dachshund.

Similarly, if a subject was almost 3ft/1m tall like a Great Dane, or only 6 in tall like a miniature Chihuahua, the height difference may have led to different evolutionary paths for visual processing.

Sense This. . .

Based on Byosiere and colleagues’ review of the literature, past and present, they concluded that dog vision may be superior to humans in some ways and less sensitive in other ways. Previously, dogs were thought to only perceive shades of grey. We now know that about 3% of their photoreceptors are cones (5% in humans), which are responsible for color vision.  However, dogs seem to only have cones that are most sensitive to blue and yellow. In contrast, humans have red, green, and blue cones. Thus, as indicated in the image below, dogs may be able to discriminate between red, blue and green when compared to grey and can discriminate between yellow and blue when compared to one another. Thus, if art appreciation was a goal, dogs would be unable to perceive the nuanced differences in color that make up an impressionist painting while most people could.

Surprisingly, dogs, like bees and other insects, fish, some birds, reptiles, and some other mammals, may also be able to sense UV light as they have lenses that transmit UV rays. Additionally, dogs may sense the Earth’s magnetic field orientation based on the presence of cryptochrome 1, a flavoprotein that is sensitive to blue light and to light-dependent magnetic orientation. Migrating birds, some reptiles, and fish respond to the Earth’s magnetic orientation, but the ability of dogs to possibly sense the magnetic field is unexpected. It will be interesting to see how these visual sensory findings spur future research – perhaps dogs can be trained as live compasses?

Although color vision may not be their specialty, dogs are particularly sensitive to dim light and may take more time to recover from exposure to bright light. Dogs have a special layer of tissue in their eye that acts as a biological reflector system called the tapetum lucidum. An example of it can be seen in the image below.

This reflector system is found in other vertebrates, other than humans, and acts to enhance the sensitivity to light by reflecting the incoming light through the retina twice.  Different breeds of dogs have been observed to have different sized tapetal areas; smaller dogs tend to have smaller tapetal areas while larger dogs tend to have larger tapetal areas. Although not yet studied, these differences may affect a dog’s performance on cognitive studies depending on the lighting conditions. Similarly, if dogs take longer to recover (an hour) from bright light, as compared to humans (30 min), and visual-based cognitive studies did not consider this limitation, then the dog’s discrimination performance may have been affected.

Moreover, when compared to humans, a dog’s visual acuity is not as sharp. What humans with normal vision can see at a distance of 75 ft away, dogs could see with similar acuity at 20 ft away.

We must continue to investigate these issues to ensure that the visual stimuli being used in research on dogs are sized appropriately and positioned at correct viewing distances. Currently, work is being conducted both in the electrophysiological world through sweep visual-evoked potentials, which measure the brain’s cortical responses to stimuli that gradually change in spatial frequency, and the behavioral realm in which dogs must discriminate between different size stimuli at different distances. Behavioral measures produce lower visual acuity estimates than the electrophysiological measures, indicating that more research is needed.

Dogs perceive depth as indicated by their agility training, catching of Frisbees, stealing sandwiches from countertops, and jumps on to furniture of varying heights.

Yet, their stereopsis (i.e., ability to perceive the world in 3 dimensions, which depends on binocular overlap) may be affected by the set of their eyes and the length of their noses.  It also appears to be age-dependent as adult dogs performed better on depth perception tests than juvenile or young adults.

Perceive That. . .

Given that much of the cognitive testing is now performed with computers and television monitors that have different frame rates or flicker rates, researchers need to be sensitive to how the subject senses and perceives these variations in visual stimuli. Related to this question, can dogs discriminate between different objects based on form (2-d, 3-d, lines, shapes), size, quantity, or relative size? Do they prefer to focus on specific or local details like pigeons (the E’s that make up the H) or on the global feature like humans (the H)?  Evidently, the dogs tested preferred the global features overall and retained that preference two years later.

Ultimately, based on the review by Byosiere and colleagues, it is critical for researchers to know the physical limitations of the dog’s visual sensory system as results of cognitive studies may not be valid if the stimuli are too small or the dogs are at a viewing distance at which their visual acuity is limited.  These issues may have come into play in the research testing the sensitivity of dogs to various visual illusions and to human faces, especially when asked to discriminate between different human emotional expressions.

It’s all about your perspective. . .

In the end, while dogs are able to inter-breed, it is clear that morphology varies widely among the breeds. As Byosiere and colleagues pointed out, characteristics may have been selected because of small differences in visual perception that made them better or worse at performing tasks based on vision. Thus, breeds like pugs (a brachycephalic dog) perform better on human pointing gesture tasks than breeds like dachshunds (a dolichocephalic dog) presumably because their snout doesn’t get in the way!  It is clear that much more “basic” research is needed to ensure that the cognitive testing that is currently underway is not limited by basic physiological constraints.

Calling all dogs…

Psychonomics article focused on in this post:

Byosiere, S. E., Chouinard, P. A., Howell, T. J., & Bennett, P. C. (2017). What do dogs (Canis familiaris) see? A review of vision in dogs and implications for cognition research. Psychonomic Bulletin & Review, 1-16.  DOI: 10.3758/s13423-017-1404-7.