top of page

All dogs can talk - just not in the way you think

Wouldn’t it be amazing if dogs could talk to us? While dogs may not be able to utilize spoken language, there is evidence that dogs can use other methods to communicate desires and needs with human beings.


Researchers in Brazil at the University of São Paulo created a “keyboard” (see below) that allowed a dog named Sophia to communicate her wants and needs to her caretaker and other experimenters. After a training period, Sophia was able to use the keys to indicate when she needed or wanted the following: a walk, to be pet, a toy, water, food, to go in her crate, and when she needed to urinate. If Sophia selected the key associated with being pet, the keyboard would emit the Portuguese word for pet, and the caretaker or experimenter would give Sophia pets and attention. Sophia was reinforced for her behavior of pressing the key because she then got what she wanted! By 17 months of age, Sophia would spontaneously utilize the expression keyboard to communicate with her caretakers.


While only one dog, Sophia, was included in this study, training other dogs to communicate in a similar way would likely be possible. Humans and dogs communicate with each other every day through a series of reinforcements, punishments, and associations. You may have noticed that your dog has created ways to communicate with you in a more informal way – for example, sitting by the door when they need to go outside. Initially, your dog may sit by the door because they have associated the door with being a portal to the outside world. Once they are rewarded by access to the outside world, you have then established a reward-based link between sitting in front of the door and being given a walk.


Every day, without even realizing, we establish complex lines of communication with our dogs. How does your dog communicate with you? Comment below!


Source: Rossi, A. P. & Ades, C. (2008). A dog at the keyboard: using arbitrary signs to communicate requests. Animal Cognition, 11, 329-338. DOI 10.1007/s10071-007-0122-3

bottom of page