"I know that the data says that therapy dogs help to reduce stress in people, but as far as I'm concerned the work that we require these therapy dogs to do constitutes animal abuse." This startling statement was made to me by a clinical psychologist. She and another psychologist had been having coffee with me and we were discussing some ongoing research that looks at the effectiveness of therapy dogs in relieving the symptoms of Post-traumatic Stress Disorder (PTSD) for military veterans returning from conflict zones. "Look at it this way," she went on to explain, "the patient draws comfort and reassurance from the presence of the therapy dog and ultimately this may help him to get on with his life. But what is the dog required to do? The dog is required to stay with that person, even if the patient is not paying any attention to them, even if the patient is not socially interacting with them, no matter how sterile or boring the environment might be." 

My other colleague disagreed with her interpretation. He said, "Dogs like to be around people. They're not like kids who need to be entertained all the time. At least anecdotally, dogs follow people around just to be in their presence."

It just so happens that a report bearing on this subject was recently accepted for publication by the Journal of Applied Behavior Analysis. The study was conducted by Erica Feuerbacher from the University of Florida, and Clive Wynne from Arizona State University. In terms of the number of dogs tested, this was a small study, since there were only three pet dogs who served as test subjects. However, in terms of the number of measurements, the research was quite intensive since there were five different experimental conditions, and a total of 30 measurements were taken on each of these conditions over a period of five days.

The test sessions occurred in a fairly uninspiring and barren room that contained only a desk, a chair, and a door in one corner, which opened out to give access to the living room. The major question that had to be addressed was whether the dogs considered being in the presence of their owner to be a reward. The researchers reasoned that if it was, than the dogs would learn to make a specific response (like pressing a button) in order to gain access to their owner. All of the dogs learned such a response when their only reward was a one minute escape from the room, which resulted in one minute of interaction with their owner who was sitting on a couch in the living room. After one minute, the owner led the dog back to the room, closed the door after the dog had re-entered the room, and continued the session. Remember, the dogs are learning to perform a task when their only inducement is gaining access to their owner. The fact that they learned to do something based on that opportunity alone, and then frequently performed that task when given the chance, seems to confirm that the dogs considered being near their owner to be a reward.

Once the dogs had learned the task, the researchers set up what are called "abolishing operations," which is simply the technical term for trying to produce conditions that would reduce the likelihood that the dog will perform this newly learned task. The first two attempts involved trying to make the room more attractive and interesting. In one condition they put some toys in the room for the dog play with. Although this reduced the number of times that the dogs performed the response that opened the door to give them access to the owner, over all, the dogs still preferred to be with their owner, rather than playing with their toys, more than half of the time. Putting the dog's familiar bed in the room produced a variable response. One of the dogs simply laid down on the bed and stopped responding, while for another it made no difference and that dog continued to try to gain access to the owner at the same rate as before, while the third dog was someplace in between.

The most interesting conditions, however, were those where the owner was brought into the room. Now remember, the dogs had learned a task that could give them a chance to get out of the boring, sterile room, and that would give them access to the rest of the house. In the first condition, the owner entered the room and immediately began to interact with the dog, petting it, and talking to the dog as she moved to the chair to sit down. She continued to talk to the dog and to pet it if it were close enough throughout the session. I don't think that it is any surprise to find that in this situation the dogs stayed in the room quite happily and made very few responses to open the door and leave their owner. Obviously this kind of attention is very rewarding to dogs, and so they had no incentive to leave the room.

However, there was one final test condition that is, perhaps, the most relevant. This condition evaluated whether it was the presence of the owner alone that is rewarding — even when that person is not providing any attention or interaction with the dog. Is that owner's being nearby enough to keep their pet happily in the room without having the dog ask to leave (by performing its learned response)? For this test the owner entered the room and told the dog, "I'm going to work; be a good dog." She then gave the dog a pat, and then sat at the desk to work while ignoring the dog. If the dog performed its learned response it would be let out of the room and could move away from its owner and admission to the rest of the house for a period of one minute as in the previous conditions. Now here is the important finding, even when the dog's owner was merely present in the room, without any kind of social communication or any attention being paid to the dog, all of the dogs chose to remain in the room nearly all of the time with very few requests to leave the presence of their owner.

The researchers conclude that their data shows that simply being in the vicinity of the dog's owner is rewarding to the dog. Dogs will learn to do behaviors just to gain access to their owner, and will remain in their owner's presence although nothing exciting or interesting is going on — even when there are no social interactions between the human and the dog. Thus, my colleague's contention that it is abusive to require a therapy dog to continually stay with their owner (despite the fact that there is nothing appealing for the dog to do) is clearly not borne out by the findings of this research­. This new data suggests that dogs prefer to be near their owner. Dogs seem to find the mere presence of their human companion to be rewarding and, given free choice, dogs choose to remain nearby.

Stanley Coren is the author of many books including: Gods, Ghosts and Black Dogs; The Wisdom of Dogs; Do Dogs Dream? Born to Bark; The Modern Dog; Why Do Dogs Have Wet Noses? The Pawprints of History; How Dogs Think; How To Speak Dog; Why We Love the Dogs We Do; What Do Dogs Know? The Intelligence of Dogs; Why Does My Dog Act That Way? Understanding Dogs for Dummies; Sleep Thieves; The Left-hander Syndrome