Do Dogs Feel Empathy?

If you cry and your dog nuzzles you and licks your face, is she trying to comfort you? Researchers try to figure out whether dogs feel empathy.

Jolanta Benal, CPDT-KA, CBCC-KA
5-minute read
Episode #158

Do Dogs Feel Empathy?

As most of us with dogs know, bursting into tears when you have a dog is different from bursting into tears when you don’t have a dog. If you have a dog, chances are pretty good that said dog will press herself against you, or stand in front of you with her ears pinned back, or lick your face, or all of the above. And as for you, you might pet her, or rub your teary face in her fur. You probably feel a bit better. You might even stop crying.

Meanwhile, though, what’s going through your dog’s head?

Empathy, “Personal Distress,” and Sympathy

Deborah Custance and Jennifer Mayer, of Goldsmiths College in London, designed a clever experiment to test whether dogs’ behavior around apparently distressed people is consistent with empathy. As a starting point, they explain a distinction between empathy and “personal distress.” In both cases, we enter into another person’s emotions. If we feel empathy, we may try to comfort the other person. However, if what we’re experiencing is personal distress – that is, if we just want to make ourselves feel better because the other person’s feelings have upset us – then we might be more likely to seek comfort for ourselves.

How can we tell what an animal is feeling when we can’t ask the animal to tell us in words?

Custance and Mayer also discuss sympathy. Cognitive scientists consider sympathy more complex than plain old empathy because sympathy takes a step back – you understand what the other person is feeling, but you don’t necessarily share it. There doesn’t seem to be any way to distinguish sympathy from empathy on the basis of behavior, and we can’t ask non-human animals to describe their inner experience in words. So the best we can do with dogs is to see whether their behavior looks empathetic, or just “personally distressed.”

How the Experiment Worked

Now on to the experiment! There were 18 test dogs. Professor Mayer played the role of a stranger visiting the dog and the dog’s guardian at home, while a third person stood aside and took video. Mayer didn’t interact with the dogs at all, just sat down at a distance from the guardian. Then she and the guardian talked for two minutes, after which one or the other of them would either fake crying for 20 seconds, or hum “Mary Had a Little Lamb” for 20 seconds. The researcher and the guardian took turns humming and crying, with intervals of talking, until each person had cried once and hummed once. To keep everyone’s behavior as consistent as possible, the guardians were told in advance what body language to use as they hummed or cried.

The researchers predicted that if the dogs experienced “personal distress” in response to the fake crying, they would probably approach their guardian for comfort while Professor Mayer was crying. If the dogs were mostly curious about what was going on, they would approach Professor Mayer both while she was crying and while she was humming “Mary Had a Little Lamb.” If the dogs felt empathy, the researchers predicted, they would be more likely to approach whoever was crying. Also, if the dogs felt empathy, their behavior when they approached the crying person would include subdued, appeasing elements more than playful, “neutrally calm,” or alert elements. Finally, the intervals of talking provided a baseline, because low-key human conversation is something that happens around dogs all the time and that usually isn’t of much interest to them.

Most Dogs Approached the “Crying” Person

Of the 18 dogs tested, 15 approached the crying person and 6 approached the humming person. Humming did elicit more “person-oriented” behavior than plain old talking did, and dogs showed more “person-oriented” behavior toward the crying person than to the other person, who remained silent. A finding that impressed me is that it didn’t seem to matter whether the dog’s guardian or the visiting stranger was the person crying, even though the stranger had made absolutely no attempt to befriend the dog beforehand. And, yes, the dogs’ behavior toward the crying person was, as the researchers put it, “markedly different” than the dogs’ behavior during humming or talking: “The majority of dogs … behaved in a manner that was consistent with empathic concern and comfort-offering.”

Awwwww! Dogs are empathetic! Case closed! Right?

What the Study Shows – and Doesn’t Show

Nope, sorry; hang on a sec while I fetch my little pin to burst that shiny, shiny bubble. All the dogs did was act in ways consistent with empathy. Their behavior might be explained in other ways, as well. For example, Custance and Mayer point out, pet dogs may well have a history of being rewarded for approaching distressed people. You remember my little scenario from the beginning of this article? You’re crying. Your dog comes up to you. You pet her. Maybe you say affectionate things to her. It’s a good bet that many pet dogs have learned that a crying person will offer attention and affection.

Or try this on for size: Crying by humans is sometimes accompanied by other behaviors that may upset and frighten dogs, such as yelling and throwing dishes. A dog who learns that when crying happens, yelling may not be far behind, may also learn that appeasing the upset human works better than seeking comfort elsewhere to get the upset-human sequence to stop. The dog might act just like the dogs in the study, but his inner experience would be of “personal distress,” not true empathy.

Further Research?

It’s not obvious to me how we could ever find out which of these explanations is right, or whether all of them are, or whether it depends on the individual dog. Custance and Mayer suggest several further experiments. We could test puppies as they develop, to see how and whether the “comforting” response is learned. And might toy dogs or guarding-breed dogs respond differently from the mixed-breeds and hunting breeds in Custance and Mayer’s study? The researchers had second thoughts about their choice of humming as a behavior to contrast with crying; perhaps laughing would have been better, because the sounds of laughter have about the same intensity as the sounds of crying. Would laughter elicit playful behavior? (It certainly does in my house!) Finally, in the experiment the crying was faked and came out of nowhere – it wasn’t in response to pain or fear, for example. Might dogs respond differently to genuine human distress, evoked in a realistic context?  But how could you control the level or duration of real distress? And how could you produce it without cruelty?

Custance and Mayer’s study has been widely misrepresented – one prominent blog headlined it “Dogs Are So Sweet That They Get Upset Even When Strangers Cry.” Please, no! As I’ve said many times, dogs get in terrible trouble for failing to live up to our expectations, even when those expectations are unrealistic and unfair. That’s one reason I’ve spent so many words on the limitations of Custance and Mayer’s work. Their experiment was carefully done, and it shows exactly what the researchers say it does: Dogs may feel empathy; we haven’t proved that they don’t. Right now, that’s as far as we can go. And oh, yes: when I’m upset, I take all the comfort my dog offers, no matter what’s going on in his head at the time.

You can follow me on Twitter, where I’m Dogalini. I’m The Dog Trainer on Facebook, and you can also write to me at dogtrainer@quickanddirtytips.com. I welcome your comments and suggestions, and though I can’t reply individually, I may use them as the basis for future articles. Thanks for reading.

About the Author

Jolanta Benal, CPDT-KA, CBCC-KA

Jolanta holds professional certifications in both training and behavior counseling and belongs to the Association of Professional Dog Trainers and the International Association of Animal Behavior Consultants. She also volunteered with Pet Help Partners, a program of the Humane Society of the United States that works to prevent pet relinquishment. Her approach is generally behaviorist (Pavlovian, Skinnerian and post-Skinnerian learning theory) with a big helping of ethology (animal behavior as observed in non-experimental settings).