- Dogs that have received training in a variety of behaviors learn new tasks faster and solve problems more quickly.
- Persistence may be measured as resistance to extinction, and it may reflect how robust a learned task is.
- Highly trained dogs show more persistence and slower extinction for new tasks than untrained pet dogs.
In a classic series of studies, psychologist Harry Harlow demonstrated that in humans and animals, a trained brain thinks differently than an untrained brain. Specifically, he was able to show that solving problems (any sort of problem) makes it easier to solve new problems. He claimed, "An untrained brain is sufficient for trial-and-error, fumble-through behavior." However, solving new and novel problems and learning new tasks becomes easier once you have solved many problems and learned many tasks. He sometimes referred to this process as "learning to learn."
A simple extension of his findings to dog behavior suggests that highly trained dogs may be more successful than untrained pets in independently solving newly-presented problems, even if their previous training has nothing to do with the specific problems the animal is now confronting. A team of researchers from the University of Buenos Aires headed by Camila Cavalli looked at this phenomenon in dogs, and believe that they have discovered that one of the factors underlying this process may be persistence, a behavioral tendency or degree of motivation that comes about through high levels of training.
The Importance of Persistence
Virtually everybody who has studied the learning process knows about reinforcement, the principle that behaviors that are rewarded become stronger and more frequent. The flip side of this is extinction, which refers to the fact that behaviors that are not rewarded grow weaker and become less likely to appear. Resistance to extinction is often considered to be a measure of how well a task has been learned. That is to say, a well-learned task persists longer after rewards are no longer being provided for the behaviors.
Dogs that are highly trained, whether for service tasks, like search-and-rescue dogs, or for sports or recreational activities, such as obedience or agility competitions, often need to perform long chains of behavior with no reinforcement until the very end of the task. As the behavior sequence progresses, with step after step not being reinforced, there is the risk that extinction will set in, and the animal will simply quit before the problem is solved or before the behavior series has been completed. It is possible that highly trained dogs have learned to persist and thus have developed resistance to extinction, which would make them better problem solvers. This persistence can be learned regardless of the specific type of training that the dog experiences.
Trained Versus Untrained Dogs
In this recent study, the researchers looked at 26 adult dogs divided into two groups. The group designated as "trained dogs" all had basic obedience training; however, most had additional training of various types. There were some dogs that had advanced obedience training or participated in sports like agility or freestyle dancing, while some had recreational trick training. Two were working service dogs (an assistance dog and a cadaver detection dog).
The group designated as "pet dogs" were selected to match, as close as possible, the characteristics of the trained dogs except for the fact that they had had no formal training.
Social Learning Task and Problem-Solving
There were two tasks that the dogs were tested on. One was a social learning task where the dog had to learn to look directly into the experimenter's eyes. Every time the dog gazed at the experimenter's face, they received a treat (reinforcement). Once the dogs were performing this task, the extinction phase began, and the gazing behavior was no longer rewarded in order to see how persistent the newly learned behavior was.
The second task focused on problem-solving. It involved a flat disk that had bone-shaped depressions into which plastic bones could be fitted. The dogs had to learn to remove the plastic bones with their paws or mouth to obtain the treat hiding under each one. After the series of rewarded trials, the extinction portion of the study began, and no more reinforcements were provided for removing the bones.
For both tasks, it was expected that once the rewards were removed, the behaviors would weaken or disappear, thus showing extinction.
Trained Dogs Are More Persistent
The statistical analysis of the results that these investigators presented involves some complex mathematical modeling; however, the most practically relevant results are easily described. In the social learning task, the trained dogs gazed at the experimenters more than the pet dogs during extinction. While in the problem-solving task, the trained dogs picked up a larger number of bones during the extinction phase. In other words, for dogs who had been trained on a variety of different tasks (none of which were similar to the task that they learned in this study), their newly learned behaviors persisted longer and were more robust than the behaviors of the pet dogs who received no extensive training in any area.
The researchers summarize the results, saying, "In conclusion, trained dogs showed a greater persistence of their learned responses assessed as resistance to extinction in both a social and a non-social task. This indicates a greater ability to maintain their behaviors upon a disruption such as the absence of reinforcement."
In other words, dogs who had received high levels of training in virtually any task not only learn new tasks more quickly but are more likely to refuse to give up on those tasks that they have learned when compared to their less broadly educated cousins.
Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission.
Cavalli C., Dzik M.V., Brarda M., Bentosela M. (2022).Trained dogs do not give up. Effects of advanced training on the persistence of domestic dogs, Behavioural Processes, 104769, ISSN 0376-6357, https://doi.org/10.1016/j.beproc.2022.104769.
Harlow, H. F. (1949). The formation of learning sets. Psychological Review, 56(1), 51–65. https://doi.org/10.1037/h0062474