The scientific interest in studying canine cognition has led to the development of a slew of test protocols—some uniquely designed for dogs and others modified from the field of comparative psychology. Many of them employ visual tasks to test dogs’ capabilities. In order to succeed with touch screens, at discriminating fine details in tests of their abilities to follow gazes or gestures, to understand object permanence, to identify faces or facial expressions, their visual perception is part of the equation. However, most of the studies are designed based on human, rather than canine, visual perception.

Canine vision differs from humans in a number of ways. Their ability to perceive a range of color hues is not as good as people’s ability, nor is their ability to distinguish levels of brightness or their visual acuity. Dogs are sensitive to higher flicker rates than people are, which can affect any studies that use moving items on computers or on televisions. There is evidence that dog vision is even more sensitive to movement than human vision.

Since visual perception abilities are not consistently accounted for in many studies with dogs, it is hard to know whether the test protocols are accurately assessing canine cognition. The results may be affected by visual capabilities instead. Researchers recently tested the hypothesis that visual perceptual differences between dogs and people could affect the performance in visually-based tasks using a free online tool (http://dog-vision.com) that converts images to settings that match what humans or dogs can see best. They report their results in the study “Do you see what I see? The difference between dog and human visual perception may affect the outcome of experiments”.

The test subjects in the study were humans, and they were asked to decide which side was indicated by a person in a series of photos. The photos showed a woman indicating a direction (right or left) by either pointing that way with her arm extended, by turning her head or by moving the gaze of her eyes in that direction without moving her head. People were tested with photos in their original form (set for human vision) and in a form altered for canine vision.

Participants in the study could correctly choose the direction of all three sorts of cues in the unaltered (human vision setting) photos. In the photos that were altered to the dog-vision setting, they could identify the cues in the pointing with extended arm and with the head turn quite well. However, their performance dropped considerably when asked the direction indicated by the gaze of the woman’s eyes in the dog-vision setting.

The results of this study suggest that differing visual capabilities may affect performance in visual tasks. The researchers acknowledge that this study only shows that human performance is influenced when visual tasks are designed for the other species, but it is likely that dogs are similarly affected. Though many experiments that do not account for vision differences between dogs and humans have still revealed intriguing canine capabilities, future research could benefit from doing so. It is likely that researchers could increase the number of unambiguous results and also eliminate the hassle of a large drop-out rate of subjects who do not meet preliminary criteria for inclusion in the study. Potentially challenging visual presentations are a problem in canine studies, and avoiding them will help scientists conduct better research.

The information available in canine urine is astounding. From a proper sniff, dogs can learn about the sex, reproductive status, diet and stress level of dogs who have been there before. Urine is used to communicate about territories, to mask the smell of other dogs, to detect females who are likely to be reproductively receptive and to compete with other individuals. It’s no wonder that our canine friends find urine so compelling that they are irresistibly drawn to it. As anyone who has spent even a little time with dogs knows, urine sniffing is a favorite pastime.

A recent study called “Length of time domestic dogs (Canis familiaris) spend smelling urine of gonadectomised and intact conspecifics” was conducted to investigate whether gonadectomy (being spayed or neutered) affects urine-sniffing behavior. Since gonadectomy has significant impacts on body chemistry, it has long been suggested that it disrupts the flow of information available through urine that dogs have evolved to detect over many generations.

Researchers tested the affects of gonadectomy in urine sniffing by recording how long dogs sniffed urine from intact versus gonadectomized individuals. They found that dogs spent more time sniffing urine from spayed or neutered dogs than from intact ones. One possibility is that the dogs are spending a longer time sniffing such urine because they are trying to figure out the information it contains. Because it may have a combination of chemicals that is different than the range of compounds that the dogs have evolved to understand, it may be harder for them to make sense out of it.

Interestingly, this study contradicts the findings of Lisberg and Snowdon, whose 2009 paper also analyzed the investigation patterns of unfamiliar urine and found that dogs spent more time sniffing urine from intact dogs than from gonadectomized ones. One possible explanation for the difference may be that for the current paper, the dogs were tested indoors, but for the 2009 paper, the study took place outside. (Fewer distractions inside may also explain an average sniff length of nearly 13 seconds in this paper compared with just over 5 seconds in the older study.) Another difference between the results of the two studies is that the recent research found no difference in sniffing time related to what kind of dog was doing the sniffing (male or female, intact or gonadectomized) but Lisberg and Snowdon found that neutered males and intact females both spent more time sniffing urine from intact males than from neutered males.

More research is definitely needed if we want to understand the complicated behavior of urine sniffing, which may involve many interactions between environment and individual traits of the dogs—both those who are the sources of urine and those who sniff if. Research is time intensive and can be costly, which is why I’m so impressed by this particular study. It was conducted in a single home in which the 12 dogs recruited to be sniffers all live, there was no funding source for the study and all of the urine in the study came from out of state to insure that the urine came from unfamiliar dogs. Kudos to the authors for taking the initiative to conduct a cool and clever experiment!

When Maureen and Jay Neitz adopted an adorable, fluffy black puppy in the late 1980s, they had no idea how important she would be in making new discoveries about vision.

They were just looking for a dog who was size-appropriate for the small apartment they lived in as UC Santa Barbara PhD candidates. Eventually, the teacup Poodle they named “Retina” helped the couple prove that dogs see much more than just black and white, and that dogs’ color vision is similar to that of the 8 percent of the human population who are red-green colorblind.

Ten million Americans, most of them male, are affected with red-green colorblindness, a genetic trait carried on the X chromosome. People with this condition can’t clearly see the difference between red and green. They often mistake green for white and red for brown or dark gray.

Colorblindness might not seem like a serious disability, but it causes unexpected, and sometimes tragic, problems for humans. For example, airline pilots must be able to differentiate between colors, which someone with red-green colorblindness can’t reliably do. Color vision is, of course, crucial in being able to discern if a traffic light is red or green. According to Don Peters, a consultant to the biotech industry who has red/green colorblindness, “Sodium vapor lights look a lot like red stoplights to me. It’s confusing to drive in an area with these lights, especially at night.” As a child, he had difficulties with color-related tasks: “I can still hear my teacher asking me why I colored the tree red. I couldn’t tell the difference.”

Colorblind people miss a lot of detail that people with normal color vision take for granted: they might not see the lines on a map, or lettering printed in colors that seem bright to those with normal vision but blend in for them. This can be dangerous when reading traffic signs or medication labels. Jay Neitz pointed out that children who are colorblind often have trouble in school, and can be mistakenly diagnosed with learning disabilities or ADHD; in spite of these potential problems, schools do not test students’ color vision.

The Neitzes established that dogs see shades of yellow, blue and gray. Other colors, such as red and green, appear faded or indistinct. Jay Neitz had an “aha” moment when Retina could not find her orange ball in a green lawn. “Sometimes the ball was right in front of her, but she would sniff around in the grass, trying to find it by smell. We realized that she simply couldn’t see it, even though it was obvious to us,” he said.

As UC Santa Barbara post-docs with degrees in biochemistry, molecular biology and biopsychology, the couple had access to a lab in which they could set up a testing area. “I realized that I had the opportunity to find out, once and for all, what kind of color vision dogs really have.” Jay built an apparatus that placed dogs in front of a screen with three lit panels. He trained the dogs to touch the screen with their noses when they saw a different shade. If the dog got it right, she would receive a cheese-flavored dog treat. In order to get the dog to touch the screen, Jay used peanut butter as an incentive. Once the dog mastered that part of the test, Jay no longer used the peanut butter.

Right away, Maureen and Jay discovered that, like people, dogs were good at figuring out shortcuts to getting a treat. In addition, “About 30 percent of the time, the dog made a lucky guess,” according to Maureen. The dogs’ attention spans were short, and on more than a few days, they just didn’t feel like doing the tests. “It took six months per dog to train them,” Maureen said. In addition to Retina, the Neitzes used two Italian Greyhounds; like Poodles, they are small, intelligent, easily trained dogs. “The dogs were treated very well,” Maureen said. “We had the utmost concern for their welfare.”

In 1989, Jay Neitz co-authored “Color vision in the dog,” which was published in the journal Visual Neuroscience; the research paper confirmed that dogs do, indeed see more than black and white. That led to a years-long search for a cure for colorblindness in humans.

Along the way, Jay heard from a diverse group of people interested in animal color vision. Game wardens asked him to create a color that hunters could use to spot each other but would be invisible to deer (“blaze orange” was the result) and, more recently, hundreds of people wanted to know what he thought of “The Dress,” a photo that went viral on the Internet featuring a dress that looked white and gold to some, blue and black to others.

Even the U.S. Army expressed an interest in the couple’s research. “They wanted to know if we could give dogs infrared vision so they could see lights used in locating bombs without alerting the enemy.” Jay joked that more people than he can count have sent him the “Far Side” cartoon showing a dog praying next to a bed, with a caption that reads “And please let Mom, Dad, Rex, Ginger, Tucker, me and all the rest of the family see color.”

Today, the Neitzes are both professors of ophthalmology and color vision researchers at the University of Washington. They continue their work in the field of color vision, but they don’t use dogs anymore. “Our dog study was done purely out of curiosity,” Maureen said. Jay added, “We demonstrated that color vision is much more complex than previously known, both in animals and in humans. Somewhere in our evolutionary past, humans developed the ability to see colors, which has helped us in many ways. One important advantage to having color vision is that it helps us determine whether a piece of food is ripe, and therefore good to eat. Humans were able to spot high-calorie food without putting out too much effort.”

In 2009, the Neitzes cured colorblindness in several male squirrel monkeys using gene therapy.

According to Maureen, “Male squirrel monkeys see only blue and yellow. Therefore, they make good subjects for color vision research.” The tests that the Neitzes developed during their original study with Retina and the two Greyhounds are now “vastly improved. We use a version of the Cambridge color test intended for human color vision testing, which Jay adapted so that it can be used on animals. It uses 16 colors, all of varying saturations.”

This more accurate test helps speed up color-vision research. The monkeys were given a human gene, which allowed them to see color the way we do. According to an article in Science, “The result raises questions about how the brain understands color, and it could eventually lead to gene-therapy treatments for colorblindness and other visual disorders in humans.”

As the Neitzes get closer to giving colorblind people normal color vision, it’s inspiring to think that a teacup Poodle trying to find her ball in the grass helped make this possible.

Read more about the Neitzes’ work.

I’m interviewing a new client whose dog tends to bark and charge and nip the heels and dan- gling hands of retreating strangers. Her dog is smallish and stocky, with a coarse, medium-length coat of mottled blue-gray, black, white and brown. His nose and ears are pointy. While I reassure her that his behavior actually makes sense from his doggy point of view, a little voice in my head whispers, “What did she expect? She got a Cattle Dog.” I have little difficulty discounting the client’s own plaintive claim that she’s had Cattle Dogs all her life and this is the first one who’s acted this way. “You were lucky until now,” my little voice says, assuming those dogs were somehow the exceptions. But when another client complains that his large, square-headed, short-coated, yellow dog is growly around his food bowl, I take his statement that “none of my other Labs have done this,” at face value. The current dog is clearly the exception. After all, my little voice says, “everyone knows Labs love people.”

My little voice is probably wrong.
Often, we assume that each breed car- ries its own set of hard-wired impulses, which are particularly difficult to alter, even with sound behavior-modification techniques. We even expect these presumed genetic predispositions to carry over to mixed-breed dogs who physically resemble a particular breed. Dog professionals are as prone to these biases as everyone else. We’ve learned them as part of the conventional professional wisdom, and our experiences seem to confirm them — not surprising, since current behavioral and neuroscience studies show that human brains consistently prefer data that sup- port what we already believe and disparage anything that contradicts it. To top it off, a nodding acquaintance with the burgeoning field of canine genetic research indisputably demonstrates connections between genetics and behavior. One new study even appears to have found the locations on the map of the canine genome that account for pointers pointing and herders herding.

So why not use breed as the way to choose the particular puppy or dog who’s likely to help us fulfill the dream of taking a perfectly behaved, friendly dog to cheer the lives of people in nursing homes, be endlessly tolerant with our kids or have the kind of indefatigable enthusiasm for retrieving that makes a good contraband-sniffing dog? How about using breed stereotypes to guide public policy decisions on whether some dogs are more likely than others to present a danger to people, or simply to assess whether that dog coming toward us means us good or ill?

Turns out it’s not that simple.
First, there is the “what kind of dog is that?” question. Probably at least half of the estimated 77.5 million dogs in the U.S. are mixed breeds. It’s common practice among people working in res- cues and shelters to identify the dogs in their care as “predominantly breed X” or as an “X/Y mix.” Recently, when scientists used DNA analysis to test the accuracy of such labeling, they found that among dogs labeled by adoption workers, only one dog in four actually had the named breed confirmed as significantly — much less, predominantly — represented. This would not be a surprise to any geneticist or indeed, anyone who has ever glanced at Scott and Fuller’s venerable 1960s study of canine development and breed characteristics, which found that breeding, for example, a Basenji to a Cocker Spaniel often resulted in puppies with little or no resemblance to either parent.

And even reliable identification of the ancestry of a mixed-breed dog by itself wouldn’t help us predict an increased likelihood of known, genetically driven traits — say, the blood-clotting disorder that plagues Dobermans or the heart defects of Cavaliers. The parents of any mixed-breed dog have, by definition, waded out of the closed gene pool that makes purebred dogs such fertile ground for genetic research. The inevitable inbreeding of purebred populations, combined with a phenomenon called genetic drift, gradually decreases overall genetic diversity; more and more animals have fewer and fewer variable traits, including characteristics that aren’t deliberately selected for or against. But as researchers found with a colony of wolves in Sweden, even inbreeding so severe that it causes infertility can be reversed by the introduction of just one outsider. So, if we could demonstrate such a thing as “acting like a Beagle” or “acting like a Basenji,” there would be little reason to expect either one from the offspring of a Beagle/Basenji pairing.

But what about those purebred Basenjis and Beagles and Cattle Dogs and Afghans and Golden Retrievers? Can’t we expect them to behave consistently in ways that resemble the work at which they were once selected to excel?

Yes and no.
The case of my Annie, the lovely, fawncolored Greyhound camouflaged in a pile of pillows on my couch as I write this, may be instructive. She came into rescue directly from the breeding farm. It’s obvious why she never made it to the racetrack. When my other Greyhound, Henry, a racer successful enough to stay alive until retirement at four, barks and quivers at the living room window at the sight of a squirrel or takes off in an ecstatic (albeit futile) pursuit of a jackrabbit at the local off-leash park, Annie looks up blandly and then, with a clear “Whatever,” goes back to her interrupted sniffing or chewing or resting.

And yet, every single one of her ancestors, going back scores, perhaps even hundreds, of generations, was hyper-motivated to chase. They would not have had the opportunity to reproduce otherwise. Racing Greyhounds are bred for two things only: a keen inclination to pursue small, fast-moving furry things and the physical ability to do it at great speed. Racing industry insiders estimate that only about 70 to 80 percent of the dogs who result from this ruthless selection process are keen enough to race. Now, a 75 percent incidence of a trait sounds pretty high. You’d certainly take those odds in Vegas at the roulette wheel. But this is a trait that’s already extremely common across the species; it is, in all likelihood, the most widespread of the predation behaviors of hunting, stalking, chasing, killing, dissecting and eating first observed and described by the famous wolf ethologist, David Mech. Most dogs already do this. If you take more complex behaviors that are actually selected against in the wild, like compulsively fighting other dogs and failing to respond to the doggy body language equivalent of “crying uncle,” for example, your odds of reliably producing the behavior through artificial selection go down dramatically. This explains how so many of the so-called “game-bred” dogs from fight busts (like the ones rescued from Michael Vick’s fighting operation) have gone on to live companionably with other dogs as relative couch potatoes in normal homes.

Reliably increasing the likelihood of complex behaviors through selective breeding isn’t easy. And racing Greyhounds are one of only a handful of dog breeds where this is still even attempted. Since the advent of modern purebreds in the late 19th century and the subsequent closing of breed registries, selection criteria have focused almost exclusively on appearance. Qualities of temperament are sometimes mentioned, although not in ways that can be practically applied in the show ring, where — as biologist Ray Coppinger has pointed out — the behavior required is standing, and to a lesser degree, trotting alongside a handler. Most purebred dogs come out of this selection system.

So these days, when people look fondly at the breed they fancy or angrily at the one they fear and say to me, “They’re not like other dogs,” I remind my little voice to recite, “Well, actually, they kind of are.”