Clever Hans

Here's a story of a German horse who was reputed to be able to perform mathematics and other tasks.  He became an animal celebrity, much like Wishbone the dog became one in more recent times.  (But Wishbone had a cuter wardrobe.)

Herr Wilhelm von Osten, a mathematics teacher, was Hans's owner.  With training, and over time, he began to think that Hans could count, solve simple mathematics problems, and solve some thought problems.  Proud of his horse, von Osten put him on exhibit.

Was this an extraordinarily gifted horse, or was it a fraud by von Osten?  Because of the commotion Hans's unexpected abilities caused and because of the the possibility that horses could, on some level, think, some investigators from the prestigious University of Berlin led by psychologist Carl Stumpf looked into why this horse, now named Clever Hans, could do so well.  As a matter of fact, he clomped out his answer with his hoof correctly about 89% of the time when quizzed by his owner/handler.

Of course, one possible thing to look into was fraud; namely, Hans's owner was giving him signals to control his counting or problem-solving performance.  The investigators tried this out by substituting someone else doing the questioning instead of Wilhelm von Osten.  However, Clever Hans did just about as well when von Osten was not visible.

Finally, a graduate student in psychology, Oskar Pfungst, came up with the real reason for Hans's exceptional performances: the human asking the questions was unwittingly providing facial or postural cues that the horse picked up.  For example, suppose the questioner asked Clever Hans how much was 3 + 6?  Hans would dutifully stamp his hoof slowly while watching the questioner.  When Hans had performed the ninth hoof stamp, the questioner might raise his head, change his facial expression, or do something else.

Interestingly, Clever Hans came up with the correct answer about 89% of the time when the questioner, whether it was von Osten or someone else knew the answer.  If the questioner did not, then Hans was right only 6% of the time.

Therefore, Oskar Pfungst concluded that Clever Hans was clever in his detection of subtle nuances of humans' behavior, although he could not do sums or solve simple thought problems.

Is this a matter of horse sense?  Non-human creatures of several species can detect meaning in behaviors emitted by a completely different species.  Dogs, of course, come to mind.

http://en.wikipedia.org/wiki/Clever_Hans