But, we had to find some wolves, so I worked with Christina Williams, who is pictured here at The Wolf Hollow Wolf Sanctuary in Massachusetts, and there was a beautiful population of wolves there. They had all been raised by Christina, and she worked tirelessly to help me test the wolves that had been raised as puppies by people. So they were very comfortable with people. They enjoyed interacting with Christina in particular. And she did the testing of the wolves so that they would be comfortable being tested. But it ends up that they didn't use the human social cues that she gave them. In fact, here's the data where the gray bars represent the wolves. We gave them 36 repetitions of the pointing game. We did different versions of it where we actually even tapped on the location, tap, tap, tap, and pointed to where the food was hidden. And you see that they're really just at about 50% there. Whereas the same number of dogs, we looked at seven wolves and seven dogs, and the seven dogs were all above chance with all three of the cues. And both species were below chance when there was no cue provided, so they weren't using their noses and only dogs were using the social cues. So we didn't find any evidence, at least in this case, that wolves were flexibly using human social gestures in a way that we were seeing in dogs. Really, their performance was reminiscent of what we see in great apes. And even with a lot of repetition, 36 trials is a lot of repetition, they were not showing improvement and that suggested to us that really there was something different between the two species. Now we did do one other control because you could say, well the wolves, maybe they're just not good at any of these kinds of games. They're wolves. They want to run around and do wolf things, not play your human games. We did play other games, for instance we played a memory game. And in that memory game, where all they had to do was remember where food was and Christina would hide it in one of her two hands. And then they had to remember which hand it was in. That was easy for them and they actually did as well as the dogs in that game. So it's not that they couldn't play games with Christina, they were required memory and cognition, a type of cognition, it really seemed that it was specific to this social gesture where they struggled. So, that then led us to say, well it's not, we couldn't find any evidence for the exposure hypothesis. We didn't find any evidence on the first attempt to look at this, so this ancestry hypothesis. That made us then think about the domestication hypothesis. What if Sid's dogs split from wolves? There's been some kind of evolution, that dogs have actually changed, and they have evolved to read human social gestures in a way that other species can't. Oh my goodness, that would be really exciting if that was the case, and we got really interested in thinking about the idea. And of course, the reason that would be exciting is because these are the same skills that we are interested in trying to understand how they evolved in humans. Given that we haven't seen the great apes use gestures the way that we do, but dogs seem more similar to us. So, if we could see how this happened in dogs, and if it really was the case that dogs had evolved these skills during domestication, we'd understand ourselves better. But, at the same time, anytime there's a hypothesis that seems that exciting, you have to be extra careful. The real issue with the hypothesis was, how are we gonna test it? Because how are we gonna test whether dogs during domestication had been selected and actually the dogs that had some allele and that there was a change in allele frequency over time that allowed dogs to somehow pass a heritable skill in reading human gestures, that allowed those dogs to be better able to survive and reproduce. How in the world were we gonna look at that? Without having to start all over again and domesticate a population of wolves. So, this is another great point to realize what science is and isn't. And science isn't just having an idea that you can't test. To do science, your idea has to be testable. Well thankfully for this guy, we could test the hypothesis about the domestication of dogs, and the potential that the real genius of dogs was that during domestication they somehow had gotten better at understanding humans, and as a result they could use this as tools and solve all sorts of problems that other species couldn't. And the reason this fox is gonna be so instructive is because there is a population of foxes that had been experimentally domesticated over a 50 year period, and that will be what the next lecture is about, is the story of traveling to Siberia and trying to test the domestication hypothesis. So in summary, in trying to understand the origin of dog social cognition, there are really three hypothesis that have been suggested to explain the remarkable communication found in dogs. The exposure hypothesis, the ancestry hypothesis and the domestication hypothesis. The use of gestures didn't seem to be explained by the high levels of exposure to humans throughout life. When we looked at puppies, wolves look very similar to great apes and using human gestures, dogs really seemed to be the odd man out. The domestication hypothesis seems plausible but difficult to test, and if you cant test it, well we were stuck. So, that's what the next lecture will be about, is how can we actually test the domestication hypothesis.
