From the NYT:
Good Dog, Smart Dog
Life as a Labradoodle may sound free and easy, but if you’re Jet, who lives in New Jersey, there is a lot of work to be done.
He is both a seizure alert dog and a psychiatric service dog whose owner has epilepsy, severe anxiety, depression, various phobias and hypoglycemia. Jet has been trained to anticipate seizures, panic attacks and plunging blood sugar and will alert his owner to these things by staring intently at her until she does something about the problem. He will drop a toy in her lap to snap her out of a dissociative state. If she has a seizure, he will position himself so that his body is under her head to cushion a fall.
Jet seems like a genius, but is he really so smart? In fact, is any of it in his brain, or is it mostly in his sniff?
The matter of what exactly goes on in the mind of a dog is a tricky one, and until recently much of the research on canine intelligence has been met with large doses of skepticism. But over the last several years a growing body of evidence, culled from small scientific studies of dogs’ abilities to do things like detect cancer or seizures, solve complex problems (complex for a dog, anyway), and learn language suggests that they may know more than we thought they did.
Something I've noticed over the years in this kind of article or television documentary about all the new tasks to which dogs are being applied is that they seldom mention what would have immediately occurred to a pre-20th Century reader. Contemporary readers are interested in the selection process for finding dogs with the best propensities for the job and the subsequent training process. But a 19th Century reader would have immediately thought of taking the dogs who are best at a particular skill and breeding them together.
Consider the Newfoundland, a giant water dog with webbed feet who doesn't dog paddle like the average dog, but uses a more powerful technique rather like the breast stroke. Moreover, Newfoundlands desperately want to rescue people from drowning. On shorelines all over the world, there are statues of heroic Newfoundlands who rescued humans from watery graves. Unfortunately, you can't really take a Newfoundland for a walk along a public beach because he might immediately splash into the water and start hauling protesting swimmers out.
Presumably, it took a lot of generations of selective breeding to come up with a great beast with these characteristics. Presumably, you could breed together dogs that are best at each new job and eventually come up with new breeds where a much higher percentage of the dogs would pass the selection process and would require less training. But modern readers don't want to hear about that because that would be eugenics. For example, here's Jonah Goldberg's 2002 National Review Online column:
Westminster Eugenics Show
Repugnant thinking that's died out for humans is thriving at the Westminster Kennel Club.
This is not to say that foresighted individuals aren't developing new breeds, just that the entire concept is usually left out of mainstream discussions.
For example, I've seen it claimed that a few dogs can sniff out cancer in people, at least melanomas on the skin. I don't know how accurate that is, but say you could develop over a few decades a breed of dog that could detect a variety of cancers by sniffing people. Think of what a boon that would be to the world's poor — instead of expensive scans, doctors in poor places could do cancer screenings for the price of dog food!
But this kind of thinking is unpopular today because the conventional wisdom is that eugenics is a "pseudoscience" — i.e., it's not just morally wrong, it's impossible.