Dog With Math Skills

Is it a case of the Clever Hans Effect?
By Karen B. London PhD, August 2011

The news is full of stories about dogs with incredible abilities. As a dog lover, I adore hearing about amazing canine skills. As a scientist, I am often skeptical and wonder if the dog in question is really capable of doing what has been claimed. The story of Beau is one such case in which I have been made to wonder.

According to his guardian David Madsen, and to many witnesses, Beau can do math. For example, if Madsen tells Beau that there were six dogs at the park but three of them left, and then asks his dog how many dogs are left, Beau answers, “woof, woof, woof.” He will answer with five barks if asked what two plus three equals.

Madsen says Beau is correct about 85 percent of the time and that he has never had such a smart dog. To prove that he was not signaling the dog, Madsen has allowed others to test Beau when he (Madsen) was absent. Beau’s success when Madsen is not there proves that Madsen is not pulling a fast one on the rest of us, but it does not speak to the possibility that Beau’s skills are the result of the “Clever Hans Effect.”

Hans was a horse who lived in the late 1800s and early 1900s. He was owned by Wilhelm von Osten, who claimed to have taught him many skills, including arithmetic. Hans responded to questions, both oral and written, by tapping his foot. Many people observed Hans perform with von Osten at various shows throughout Germany.

In 1904, a panel of 13 people tested Hans to determine whether the horse actually knew the answers to the questions or if von Osten was tricking them all by secretly signaling his horse. They concluded that von Osten was not committing fraud and that the horse did indeed know the answers to the questions.

In 1907, psychologist Oskar Pfungst evaluated Clever Hans and shed new light on what the horse was able to do. In a series of tests, Pfungst investigated the horse’s success at answering questions under a variety of circumstances. He sometimes kept the horse away from spectators to make sure that the horse was not using any cues from them. He had people other than von Osten question Hans at times. He used blinders so that the horse could not always see the person asking the questions. He varied the distance between the questioner and Hans. Finally, in some cases, he used questioners who did not know the answer.

Pfungst noted that Hans got the answer to questions right even when von Osten was not the person asking the questions, which convinced him that Hans’ performance was not a fraud. He also observed that Hans answered correctly only when he could see the questioner and when the questioner knew the answer. For example, when von Osten knew the correct answer, Hans was correct almost 90 percent of the time, but when von Osten did not know the answer, the horse’s responses were correct only about 6 percent of the time. Hans’ performance suffered to a lesser degree if the questioner was far away from him.

What Pfungst noticed after observing the behavior of questioners was that as the horse tapped his leg, the person would change his expression and posture subtly as the horse approached the correct answer. He observed that when the horse had tapped the right amount of times for a correct answer, the person released that tension. That release in tension was the cue that the horse was using to know when to stop tapping.

Even being aware of this tendency to cue the horse, questioners, including Pfungst, could not stop their faces and bodies from giving information to the horse, as these cues are largely involuntary. Questioners were entirely unaware that they were communicating with the horse in this way. Pfungst showed that while Hans did not know the answers, von Osten was not a fraud. (Von Osten never accepted that Clever Hans was cuing off of people rather than actually solving the problems and continued to show his horse to appreciative crowds throughout Germany.)

The tendency of an observer to influence the behavior of a subject being studied with subtle and unintentional cues is called the “Clever Hans Effect.” Most experiments in psychology are now carefully designed to avoid it.

Hans may not have had the grasp of mathematics that von Osten claimed, but there is no doubt that this horse was a brilliant observer. His ability to cue off subtle cues in people’s posture and facial expressions was remarkable, and as such, this famous and talented horse certainly earned his nickname “Clever Hans.”

It would be interesting to test Beau, the dog who has so recently gained fame for his performances. Beau clearly possesses an extraordinary ability, but I want to know exactly what it is. Is it a great mathematical talent or a highly developed aptitude for observing and responding to people’s subtle, unintentional facial expressions and body language?

Karen B. London, PhD, is a Certified Applied Animal Behaviorist and Certified Professional Dog Trainer who specializes in working with dogs with serious behavioral problems, including aggression. She is the author of five books on canine training and behavior.

Sponsored Content