Wednesday, 25 January 2012

Could we breed dogs as smart as people?

I think so. A while ago I emailed an evolutionary biologist with the same question.

My first email:
How long would it take to breed a dog with intelligence comparable to that of a human? how many generations? so it could understand language, read, learn maths and so on. maybe even a few other adaptations so it could speak.
is there any research which looks at this kind of question? all i've really found on the internet is the tamed russian fox stuff.
my guess is that it would take only a couple of hundred generations.
He replied:
This is an interesting question. I suspect that it would take more than a couple of hundred generations to produce a dog with human-like linguistic abilities. That would require major restructuring of the brain, as well as the larynx, tongue, etc. Full human intelligence would probably require many 1000s of generations, and I don't think a canine, which is very different from a primate in many ways, could ever evolve the same kind of intelligence that we have, even if it could perhaps evolve to be much smarter than it is now.

On the other hand, there's some evidence that certain dog breeds have acquired an impressive ability to understand human language. Check out the link below:
I responded with this:
I guess It would be easier to select for a breed of dog which was especially good at what the border collie can do in that article. Despite the fact that dogs have much smaller brains than chimps, they're better at reading signals from humans, and know what a pointing finger means - wolves can't do this. So rather than trying to breed towards a human-brained dog, you'd have a creature which was exceptional at what modern dogs are already good at. it might not be very hard to breed a dog which was good at something only humans were previously capable of.

I think breeding dogs in this way would be interesting for another reason which you mentioned - they're canines, not primates. If we could increase their brain size to something similar to that of a human, I'm sure their brains would work in a very different way. and you wouldn't have to even get close to human brain size in the first place for some interesting things to happen.
I'm not an evolutionary biologist, but I don't think it would take very long to breed a dog which could understand human language very well. I'm essentially pulling a number out of my arse here, but I think you could do it within 100 years of breeding. I'm talking about really understanding complex sentence structure.

"Can you make sure the cat doesn't eat the parrot while I'm out?"
"Thanks buddy."

I don't see why this wouldn't be possible. The fact that dogs and wolves are so closely related is amazing enough, and this happened over a very short amount of time. Here's some info about a rapidly evolving lizard, demonstrating how fast large changes can happen under the right conditions:
In 1971, ten adult P. sicula specimens from the island of Pod Kopište were transported 3.5 km east to the island of Pod Mrčaru (both Croatian islands lie in the Adriatic Sea near Lastovo), where they founded a new bottlenecked population.[3][11] The two islands have similar size, elevation, microclimate, and a general absence of terrestrial predators[11] and the P. sicula expanded for decades without human interference, even outcompeting the (now extinct[3]) local Podarcis melisellensis population.[4]

Following the Yugoslav Wars, scientists returned to Pod Mrčaru and found that the lizards currently occupying Pod Mrčaru differ greatly from those on Pod Kopište. While mitochondrial DNA analyses have verified that P. sicula currently on Pod Mrčaru are genetically indistinguishable from the Pod Kopište source population,[3] the new Pod Mrčaru population of P. sicula was described, in August 2007, as having a larger average size, shorter hind limbs, lower maximal sprint speed and altered response to simulated predatory attacks compared to the original Pod Kopište population.[11] These population changes in morphology and behavior were attributed to "relaxed predation intensity" and greater protection from vegetation on Pod Mrčaru.[11]

In 2008, further analysis revealed that the Pod Mrčaru population of P. sicula have significantly different head morphology (longer, wider, and taller heads) and increased bite force compared to the original Pod Kopište population.[3] This change in head shape corresponded with a shift in diet: Pod Kopište P. sicula are primarily insectivorous, but those on Pod Mrčaru eat substantially more plant matter.[3] The changes in foraging style may have contributed to a greater population density and decreased territorial behavior of the Pod Mrčaru population.[3]

The most surprising[5] difference found between the two populations was the discovery, in the Pod Mrčaru lizards, of cecal valves, which slow down food passage and provide fermenting chambers, allowing commensal microorganisms to convert cellulose to nutrients digestible by the lizards.[3] Additionally, the researchers discovered that nematodes were common in the guts of Pod Mrčaru lizards, but absent from Pod Kopište P. sicula, which do not have cecal valves. The cecal valves, which occur in less than 1 percent of all known species of scaled reptiles,[5] have been described as an "evolutionary novelty, a brand new feature not present in the ancestral population and newly evolved in these lizards".[7]


  1. I'll throw this in here. I'm probably going to start a breeding process for intelligence in canines when I get time and money for it. I didn't read a lot before commenting but we would need many people on this project. "Bastards" would be the main interest as we breed the smartest individuals no matter what "race" they are.

    He said something about 1000 generations? Not if man is controlling the arrow of evolution.

    I'm pretty shocked by the poor bag dogs with their complications and all. I would never get a "purebred" dog anyways. Why people haven't started breeding for intelligence (I haven't really checked but I can guess by looking at dog shows) I don't know...

    I wonder if someone can start a kickstarter for this...

    I wish I had the guts to post any contact info but I'll try to check in to see if are interested in more of my thoughts around it.

    By the way, you will desperately want to watch this clip about breeding domestic silver foxes in Russia if you haven't seen it. /watch?v=EoB0pdhxfZs

    1. I've heard about that experiment with foxes, very interesting to see that video.

      We inbreed dogs for all kinds of horrible shit which leaves them with genetic diseases requiring a lifetime of surgery and medication. I'd say breeding for intelligence is far more sensible as long as they're happy and healthy. Crossbreeding is a good thing.

      I don't know what kind of tests you'd use to select the most "intelligent". Whatever tests you use, the breed will become adapted to. For example, if you breed them to be good at counting, they'll be better than other dogs at this, but for the most part exactly the same in other areas.

  2. I think it's pretty limiting to assume that the only direction they could go in is something like human intelligence. I would expect canine intelligence, if it were to expand to full equivalency to ours, to be utterly alien to our own brand of intelligence. They would have different aspirations that there is no way we could ever predict. They would have different intellectual drives, and we might never understand why those intellectual drives are so important to them. If we ever found dogs to be our equals, it is uncertain how many of us would ever recognize it or fully appreciate it.

    1. That was mentioned in the original post actually. Their brains would probably work in a very different way.

  3. Think of the importance of tool use to primate (especially human) intelligence. Now imagine going through life with paws rather than hands. Maybe canines can develop their own kind of intelligence but probably wouldn't do well with a human-like one.

  4. Here is a list of intelligence in dogs based on breed:

    You have to be careful though, because I think that list is deceptive. Each breed in the top rankings has a different KIND of intelligence. Poodles are very emotional and intuitive and great at problem solving (and I would rank them #1 in emotional intelligence). Border collies are rock stars when it comes to learning complex, multi-step tricks and tasks. Retrievers are amazing when it comes to helping people with disabilities.

    I would LOVE to breed a dog based on incrementally increasing their intelligence, though I would also be very careful about what kinds of personality traits they have.

    For example, would you want a genius dog with a high amount of aggression and dominance with a bite strength of 300 pounds per square inch? That would be a problem, especially if the dog was abused or used for nefarious purposes. That makes me think of Dean Koontz's book Watchers (it's a book about a genetically engineered genius dog - I highly recommend it, it's one of the most touching and suspenseful books I've ever read).

    I think breeding a dog for intelligence, personality traits and longevity would be a great idea. Personally, I am partial to the poodle. No, they're NOT crazy, yappy granny dogs, they're the most 'human' dogs I've ever encountered and they're showing great promise in hunting trials as well as search and rescue. I've had several poodles in my household over the past 30 years and I've never had a dog that's such a hilarious troll (they have an amazing sense of humour), an amazing problem solver and a breed who is so utterly devoted to me. There are other breeds I love as well - my whippet was a sweet girl who simply loved being around her humans, my mom's bichon was a sweet-tempered and quiet little guy who just stole your heart and I've never met a golden lab I haven't loved. I currently have two poodles and they're absolute clowns who love everyone, so perhaps I'm a little biased. Lol!

    I guess what I'm trying to say is that intelligence isn't the only trait that should be bred for - you need to take into account what kind of personality you'd like to see that intelligence expressed through. ;) Also, I'd breed for longevity as well - it would suck to have an insanely smart dog whose lifespan was only 10-15 years. The dog would likely know about its own mortality and knowing it was only going to live to be 10-15 years old might drive it insane with the knowledge of its own mortality. Then again, who knows - maybe they experience time differently. ;)

    So many things to think about - I love dogs.