New Study Shows That Dogs Use Color Vision After All

Although their perception of color is limited, dogs discriminate between objects based on their hue—a finding that may change the way dogs are trained

The idea that dogs only see the world in black, white and shades of gray is a common misconception. What’s true, though, is that like most mammals, dogs only have two types of color receptors (commonly called “cones”) in their eyes, unlike humans, who have three.

Each of these cones is sensitive to a different wavelength (i.e. color) of light. By detecting different quantities of each wavelength and combining them, our three cones can transmit various signals for all the hues of the color wheel, the same way the three primary colors can be mixed in different amounts to do the same.

But because they only have two cones, dogs’ ability to see color is indeed quite limited compared to ours (a rough comparison would be the vision of humans with red-green colorblindness, since they, too, only have two cones). Whereas a human with full color vision sees red, orange, yellow, green, blue and violet along the spectrum of visible light, a dog sees grayish brown, dark yellow, light yellow, grayish yellow, light blue and dark blue, respectively—essentially, different combinations of the same two colors, yellow and blue:

Consequently, researchers have long believed that dogs seldom rely on colors to discriminate between objects, instead looking solely at items’ darkness or brightness to do so. But a new experiment indicates that this idea, too, is a misconception.

As described in a paper published yesterday in the Proceedings of the Royal Society B, a team of Russian researchers recently found that, at least among a small group of eight dogs, the animals were much more likely to recognize a piece of paper by its color than its brightness level—suggesting that your dog might be aware of some of the colors of everyday objects after all.

For the experiment, the researchers printed out four pieces of paper, colored dark yellow, light yellow, dark blue and light blue. Because they used these two colors in particular—and based the darkness levels on what dogs’ eyes are sensitive to—the animals would be able to discriminate among the papers in two different ways: whether they were dark or light, and whether they were blue or yellow.

For the initial ”training” stage of the experiment, the researchers took two papers that differed among both color and brightness—either a dark yellow and a light blue paper, or a light yellow and dark blue paper—and placed them each in front of a feedbox that contained a small piece of raw meat. Only one of the boxes was unlocked and accessible to the animal, and the same paper was put in front of that box every time. For each trial, the dog would be allowed to try opening only one of the boxes and then be stopped immediately.

After only a few trials, every dog learned how to routinely pick the correct box, indicating it had been trained to associated an unlocked box of meat with one of the specific pieces of paper. To reinforce the training, each dog went through a 10 trials per day for 9 days.

Then, the researchers switched things up. To a dog that had learned “dark yellow paper = meat,” they gave two new confusing choices: dark blue or light yellow. If the dog tried to open the box behind the dark blue paper, it’d show it had based their original training on brightness; if it went for the light yellow, it’d indicate it had actually memorized the color associated with the treat.

After ten tests, all the dogs went for the color-based choice more than 70 percent of the time, and six out of the eight dogs went for it 90 or 100 percent of the time. Clearly, they’d memorized the color associated with the raw meat, not whether it was dark or light.

It’s a small sample size, and all the dogs used were mixed breeds, so it’s possible that it couldn’t apply to particular breeds of dogs with different characteristic traits and behaviors. But if the finding holds up on a wider scale, it could have some effect on the field of dog training—trainers customarily avoid using color and strictly rely on brightness as a cue. For the average pet owner out there, this new research simply gives us a better idea of a dog’s-eye view of the world—and shows that it’s probably more colorful than we’d thought before.

Get the latest Science stories in your inbox.