Do dogs feel guilt? If you own a dog, chances are you have seen the look. When your dog greets you with head down, eyes averted, and crouching, you look around for the mess. You probably find garbage on the floor. But is the guilty response a reaction to a guilty conscience, or to you? A new article in The Atlantic questions the widely-held belief that dogs can harbor a guilty conscience.
Popular culture and dog shaming
Whether or not dogs actually experience guilt in the same way as humans, people have capitalized on the guilty behaviors of their canine friends. Dog shaming is a popular internet meme in which owners post images of their dogs with a sign detailing some recent misbehavior. And it’s true. Many of the dogs in these images look properly abashed. Some trainers instruct dog parents to scold their dog even when they don’t catch him in the act. Popular culture assigns the full range of human emotions and logical reasoning ability to our canine companions.
Admittedly, dogs do an impressive job of convincing us that they feel guilty or remorseful. But does science provide any evidence?
Is there scientific evidence that dogs feel guilt?
In addition to our own intuition that dogs indeed experience basic human emotions like, joy, fear, anger, and disgust, there is reasonable evidence that this is the case. There is less behavioral evidence to suggest that dogs feel more complex human emotions such as jealousy or guilt. Much research has been conducted around the ways dogs relate to human emotions, but behavioral science has not prioritized studies to demonstrate whether or not dogs experience emotions. However, the topic of dog guilt has been evaluated in some excellent and convincing studies over the past decade.
Researcher Alexandra Horowitz studied 14 dog-human pairs and investigated the context of the so called “guilty look.” In a series of trials, the dogs were shown a treat and instructed by their owners not to eat it. The owner left the room and the dog was either allowed by the experimenter to eat or not eat the treat. The owner was called into the room and told whether or not her dog ate the treat.
In two of the trials, the owners were deliberately given the wrong information. The experimenter videotaped the entire trial and the dog’s responses were analyzed. This study showed that dogs expressed the guilty look equally whether or not they had eaten the treat. However, dogs that were scolded were much more likely to express the guilty behavior. Interestingly, dogs that did not eat the treat but were scolded were more likely to look guilty. The results of this study suggest that expression of guilt was associated with perception of the dog’s guilt by the owner. It should also be noted that the three dogs whose owners routinely used physical means to scold them were three of the four dogs with the highest number of guilty behaviors.
A follow-up study with a similar design demonstrated that owners were able to correctly determine whether or not their dog ate the treat by how the dog behaved. However, the study was set up in such a way that in a baseline trial owners were allowed to see whether or not their dog ate the treat. These owners were allowed to scold the dog if she ate the treat during the baseline. It is reasonable to think that these owners were likely biased by knowing their pet ate or didn’t eat the treat at the first trial. This study also did not include experimental manipulation of owners’ knowledge about their dogs guilt.
In this study, investigators repeated key feature of the Horowitz and Hecht experiments. However, they controlled the baseline period in such a way that testing began only after all dogs learned not to take the treat while owners were in the room. Similar to the Howoritz study, owners were sometimes intentionally given the wrong verbal and visual information about whether their dog ate the treat or not. Ninety-six dog/human pairs were tested in this study. The study showed that the owners were not able to detect whether or not their dog had eaten the treat based on their dog’s behavior more than they would have by chance alone. This study also suggests that dog’s guilty behavior is more to do with our behavior than it is a guilty conscience.
These studies were not designed to determine whether or not dogs feel guilt. Rather, they were designed to examine whether the guilty look expressed by dogs is related to our human behaviors. Dogs may feel guilt. Unfortunately, there is not yet good evidence that they do so. There is good evidence that the behavior we think expresses guilt is more an indicator of our dog’s response to us. This makes good sense in light of the ways in which dogs have evolved to read and react to human emotions.
Let’s come back to the garbage on the floor. Your dog probably comes to the door with a guilty look, even before you have seen the mess. Isn’t that an indication of guilt? Perhaps. But it may also be a reaction to a memory of the last time you scolded him for garbage on the floor. He’s smart, and he has a good memory. He can also, it seems, put two and two together.
Hecht, J. et al, Behavioral assessment and owner perceptions of behaviors associated with guilt in dogs, Applied Animal Behav Science, 2012; 139: 134-142
Horowitz, Disambiguating the “guiltylook”: Salient prompts to a familiar dog behaviour, Behavioural Processes, 2009; 81: 447-452.
Ostojic et al, Are owners’ reports of their dogs “guilty look”influenced by the dogs’ action and evidence of the misdeed? Behavioural Processes, 2015; 111:97-100