Verified by 
A paper published today in the journal Science challenges us to consider whether every study that compares wolves and dogs can shed light on domestication.
This paper, by Miho Nagasawa and colleagues from Azabu University and the University of Tokyo, has a very technical title, but you are more likely to hear about it with a snappier heading echoing the title of a commentary Science is also publishing, by Evan MacLean and Brian Hare: “Dogs hijack the human bonding pathway.”
The paper reports two experiments. In the first, dog owners spent half an hour with their dogs in a room. Eleven wolves and their owners also participated in the same procedure. The researchers measured how often the animals and their owners looked at each other, and how often owners talked to or touched their animals. Before and after, urine samples were taken from both the people and their animals and analyzed for levels of oxytocin.
Oxytocin is a hormone that is released in the body during affectionate and intimate contact between two people.
The measures of how often the people touched or talked to their dogs or wolves produced no results, so the authors focused on how often the people and their animals looked at each other, and their levels of oxytocin. Only in the supplemental material can one see that the wolves never looked at the people. (A feature of high-profile journals like Science is that most of the details are hidden away in supplemental online material). Two thirds of the dogs barely looked at the people either.
The authors decided to split the dogs into those that looked at people a little, and those that looked a lot. For the 20 dogs that only looked at people a little, nothing happened to their owner’s oxytocin levels; but for the eight dogs that looked at their people a lot, owner oxytocin went up more than three fold. The dog’s oxytocin levels also increased, though not as dramatically (about 30%).
Nagasawa and colleagues claim that because the oxytocin of the people whose dogs looked at them a lot increased, but that of the people with wolves did not, this difference in human response to dogs and wolves must be a product of domestication - since dogs are domesticated and wolves are not. Domestication, they argue, has “hijacked” (MacLean and Hare’s term) the way that social bonds are cemented in people.
Domestication and tameness are often confused, but they are not the same thing. Domestication changes animals over generations resulting in species that more readily tolerate human proximity. Tameness happens in an individual’s lifetime so that a specific individual tolerates humans. In the 1960s experimenters raised dogs without human contact throughout their early weeks of life, and those dogs grew up to be wild animals and very fearful of people.
It is also possible to tame individual members of a non-domesticated species. I am most familiar with this at Wolf Park where people have been hand-rearing wolves for over 40 years. With careful rearing from an early age, members of most species can form intimate relationships with people.
The wolves tested in this study study chose not to look at their owners. Since I have seen wolves look into the eyes of people they care about, this leads me to suspect that the wolves here did not have strong bonds with their caretakers.
Nagasawa and colleagues want to conclude something about domestication because the wolf caretakers did not experience any oxytocin increase. But the owners whose dogs did not look at them (which was two thirds of the dogs ) didn't experience increased oxytocin levels either. Thus, there is nothing can be concluded here about the fact that dogs are domesticated and wolves are not.
The authors also attempt to draw conclusions about oxytocin and the relationship between the people and their dogs, but they cannot do this because they failed to present the dogs with any people other than their owners.
In a second experiment the researchers squirted solutions of oxytocin (or control) up the noses of dogs and then left in a room with their owners and two unfamiliar people. Here they found that female dogs that received oxytocin looked at their owners more, and owner oxytocin also increased. The authors want to conclude that they are uncovering the mechanism behind the behavior from the first experiment: that there is a loop in which oxytocin drives gazing, and gazing in turn increases oxytocin.
I like that in this experiment they had some people in the room familiar and unfamiliar to the dogs, so that the specificity of the effect to people who matter to the dogs could be established. But how much we can deduce when the effect was only present in the females, whereas in the first experiment the effect didn’t differ between the sexes of dogs? It also seems strange that, although the people’s oxytocin levels went up when their dogs looked at them, dogs’ oxytocin levels did not increase even with it squirted up their noses. We are also unable to draw conclusions regarding domestication from Experiment 2 because no undomesticated animals were tested.
Here’s a thought experiment to test whether Nagasawa and colleagues have demonstrated anything about domestication. Imagine a lion tamer: Someone who has raised a lion cub from an early age and interacts with it daily. What happens to her oxytocin when her beloved big cat looks at her? I suspect that her oxytocin goes up every bit as much as that of a dog owner with a beloved dog. If you agree with me, then Nagasawa and colleagues’ study tells us nothing about domestication.
I take it to be self-evident that domestication has changed animals. I don’t experience difficulties distinguishing dogs from wolves – neither in their shape nor behavior. I just don’t believe that this paper provides the evidence necessary to demonstrate a new insight into the mechanisms underlying those changes.
