Have you ever wondered if dogs recognize that other breeds and variations of dogs are the same species as they are?  Dogs have the greatest variability of the domesticated animals when it comes to size and shape.  The colors of their fur ranges in black, yellow, white, and shades of red and orange.  They can be nearly hairless to extremely long and corded coats, textures from stiff to soft, and curly.  The shape of their heads vary from a narrow pointed muzzle of a Greyhound, to the boxy face of the Mastiff.  It is estimated that there are between 400 and 500 dog breeds, all with their own individual size, color, coat and body form registered across various international kennel clubs.
It might seem like a difficult task to expect a dog to recognize all the different varieties that make up the domestic canine species from other forms of animals.  Our casual observations of our own pet dogs is not enough to establish the point, however, a team of French researchers of the University of Paris decided to experimentally assess whether dogs could tell the difference between their own species from that of others, using vision alone.  They selected nine dogs that were pets belonging to students at the National Veterinary School of Lyon in France.  The breeds were all over the road with many being crossbreeds.
The visual stimuli was massive, involving 3000 images of dogs, and 3000 images of non-dogs consisting of purebred and mixed-bred canines selected from four major morph types:  wolf type, hound type, Mastiff or Greyhound types, including all major different features of head shape hair length and position of ears.  The non-dog set of visuals included farm animals, cats, birds, rabbits, reptiles, wild felines, humans and others.  No faces of wolves or foxes were included.  The heads in the pictures varied from full frontal to ¾ front view to an extreme profile on a background of uniform blue.  Two computer monitors were used to display the test images with a barrier separating them, making the dog clearly choose which side he was going to.  The start of training began with giving the dog the idea that choosing the dog was always the correct choice, giving the image of a dog on one screen and a blank blue screen on the other.  When the dog chose the image of the dog, a clicker was sounded and he was rewarded with a treat.  After some time the blank image was replaced with that of a non-dog image and the dog image was varied through four different images of dogs.  The correct choice was always the dog.
In the general testing session, each of the 12 trials had a new, never seen before, image of a dog, paired with a new, and never seen before, image of a non-dog animal, to determine whether dogs could discriminate between dogs and non-dog species of animals.  The dogs were tested until they reached a criterion of 10 out of 12 correct choices on two consecutive days.  All of the dogs mastered this showing they had a general concept of “dog”.  Then the experimenters reversed the choices, making the non-dog image the correct choice.  This was a difficult task for some dogs, but all of the dogs eventually learned to make the correct choice and achieved criterion.
As a result, this study found that dogs seem to have an idea of what is categorized as “dogs” and which animals are clearly not dogs.  They demonstrated this knowledge by discriminately choosing a dog or a non-dog as required by a particular task.  So what is your take on this?  Do you think the test was flawed?  Was it complete?  Personally, I think my Labs know they are dogs and not water buffalo.