July 24, 2018
From Lassie to Balto, pop culture loves stories of a dog coming to a person’s rescue. Anecdotally, people experience their dogs coming to their aid every day, like when one of us found herself “trapped” by her children under a pile of pillows only to be “rescued” by her noble collie, Athos.
But is there any scientific evidence behind these sorts of tales?
Researchers know that dogs respond to human crying and will approach people – whether their owner or a total stranger – who show signs of distress. We decided to investigate whether dogs would go a step further than just approaching people: Would they take action to help a person in need?
We recruited 34 pet dogs and therapy dogs – that is, those who visit people in hospitals and nursing homes – to take part in our study. Dogs included a variety of breeds and ages, from an elderly golden retriever therapy dog to an adolescent spaniel mix.
When they got to the lab, each owner filled out a survey about the dog’s training and behaviors while we attached a heart rate monitor to the dog’s chest to measure its stress responses.
Next, we instructed the owner on how to behave during the experiment. Each owner sat in a chair behind a clear door that was magnetized shut – there as a barrier separating the dog from its owner – that the dog could easily push open. We assigned half the people to cry loudly and say “Help” in a distressed voice every 15 seconds. The other half of our volunteers we assigned to hum “Twinkle, Twinkle, Little Star” and say “Help” in a calm voice every 15 seconds. We ran the test until the dog opened the door or, if it didn’t, until five minutes elapsed.
Past research seemed to indicate that dogs would not help their human companions in distress, but it’s possible that the tasks to demonstrate “help” were too difficult for a dog to understand. So we adapted this straightforward task from previous research in rats. It seemed like dogs would be capable of nudging open a door to access their owners.
We expected to find that dogs would open the door more often if their owner was crying than if they were humming. Surprisingly, that isn’t what we found: About half the dogs opened the door, regardless of which condition they were in, which tells us that dogs in both conditions wanted to be near their owners.
When we looked at how quickly the dogs who opened the door did so, we found a stark difference: In the crying condition, dogs took an average of 23 seconds to open the door, while in the control condition, they took more than a minute and a half. The humans’ crying seemed to affect the dogs’ behaviors, taking just a quarter as long to push open the door and get to their human if they seemed distressed. We did not find any differences between therapy dogs and other pet dogs.
Other interesting results came when we looked into how the dogs were behaving in each condition. In the crying condition, we found the dogs that opened the door showed fewer signs of stress – and were reported by their owners to be less anxious – than dogs that did not open it. We also found that dogs that opened the door more quickly were less stressed than dogs that took longer to open it.
In contrast, dogs in the humming condition showed a slight tendency to open more quickly if they were reported to be more anxious. This may mean that dogs who opened in the humming condition were seeking their owners for their own comfort.
Because both humans and animals tend to be more empathetic toward individuals with whom they are more familiar or close, we thought that the strength of a dog’s bond with its owner might explain some of the differences we saw in dogs’ empathetic responses.
As soon as the test was over, we let the dog and owner reunite and cuddle for a few minutes to make sure everyone was calm before the next part of the experiment. Next, we turned to a test called the Impossible Task to learn a bit more about each dog’s emotional bond with its person.
In this task, the dog learns to tip over a jar to get to a treat; then we lock the jar onto a board with a treat inside and record whether the dog gazes at its owner or a stranger. There have been some mixed results with this test, but the idea is that a dog who spends more time looking at their owner during this task may have a stronger bond with their owner than a dog that doesn’t spend much time looking at their owner.
We found that dogs who opened the door in the crying condition did gaze at their owner more during the Impossible Task than non-openers. On the other hand, it was the dogs who didn’t open the door in the humming condition that gazed at their owners more than those who opened it. This suggests that openers in the crying condition and non-openers in the humming condition had the strongest relationships with their owners.
Taken together, we interpreted these results as evidence that dogs were behaving empathetically in response to their crying owners. To behave empathetically toward another individual, you must not only be aware of the distress of another person, but also suppress your own stress enough to help out. If you are overwhelmingly stressed, you might either be incapacitated or try to leave the situation entirely. This pattern has been seen in children, where the most empathetic kids are the ones who are skilled at regulating their own emotional states enough to give help.
It appears to be the case with these dogs as well. Dogs with weaker emotional bonds to their owners, and those that perceived their owners’ distress but were unable to suppress their own stress response, may have been too overwhelmed by the situation to provide any help.
While everyone hopes their dog would help them if they ever were in trouble, we found that many of the dogs did not. People involved in our experiment, particularly those with dogs that didn’t open the door, told us many stories of their dogs coming to their aid in the past. Our study suggests that in some cases if your dog doesn’t help you, it’s not a sign he doesn’t love you; Fido might just love you too much.