Are Dogs Less Honest Than We Think?

Dogs can adjust their behavior and become deceptive to get what they want from people, according to a study recently published in Animal Cognition.

Deception is a cognitive behavior that benefits one individual at another individual’s expense. In the animal world, for example, subordinate chimpanzees may lead their dominant counterparts away from hidden food. Interestingly, deception may be related to a species’ social structure.

Previous studies have reported dogs’ deceptive behavior during dog—human play interactions. In other studies, dogs have demonstrated the ability to distinguish between cooperative and deceptive people.

Study Design

Training

The study’s authors set out to determine whether 27 pet dogs could deceive humans to get a treat (sausage or dog biscuits). The dogs were trained for the first 2 days of the 4-day study. On day 1, the dogs learned the roles of two unfamiliar women who would be their partners. Each woman walked to a bowl containing the treats, called the dog’s name, and either gave the dog food (“cooperative”) or withheld the food (“competitor”).

The dogs then underwent partner preference testing to confirm that they understood the different roles; partner preference testing was repeated on days 2 to 4 before other training and testing.

On day 2, the dogs learned to lead each partner to one of 2 covered boxes. The dog’s owner placed a treat in each box as the dog watched. After the owner left the area, one partner at a time commanded the dog to “Show me the food” and played her designated role after being led to a box; clear indications of a dog’s choice included sitting by a box. The partner then gave the dog back to its owner. The dog led its owner to a box and was given food if the box was not empty.

Testing

On days 3 and 4, the authors conducted testing with 3 covered boxes. With the dog watching, its owner placed one type of treat in each of 2 covered boxes, then demonstrated that the third box was empty; the partners were unaware of the food locations. Next, the dog led each partner, then its owner, to a box as was done during training.

At the end of day 4, the authors performed a food preference test to confirm each dogs’ treat preferences.

Results

The dogs chose the “preferred” food box significantly more frequently when leading the cooperative versus the competitive partner; they chose the empty food box significantly more frequently when leading the competitive versus the cooperative partner. These findings were observed on both testing days but were more notable on the second testing day.

On both testing days, a nonsignificant trend was observed for dogs choosing the “non-preferred” box more frequently when leading the competitive versus the cooperative partner.

The authors also measured whether the dogs’ choices were by chance. The chance level was 33% (1 in 3 chance to choose a given box). On both testing days with the cooperative partner, dogs chose the preferred box significantly above chance level and the empty box significantly below chance level. With the competitive partner, dogs chose the preferred box by chance only on the first testing day and significantly below chance on the second testing day.

Conclusions

The authors wrote that the dogs “seemed to choose a behavior that would give them the highest chance of obtaining their preferred food later.” Taken together, the study’s findings indicate that dogs adjust their behavior and become deceptive depending on whether a human is being cooperative or competitive.

Dr. JoAnna Pendergrass received her Doctor of Veterinary Medicine degree from the Virginia-Maryland College of Veterinary Medicine. Following veterinary school, she completed a postdoctoral fellowship at Emory University’s Yerkes National Primate Research Center. Dr. Pendergrass is the founder and owner of JPen Communications, LLC, a medical communications company.