In 1891, a horse in Germany, named "Hans", stunned the world when he demonstrated the ability to do math and to spell out words and sentences. Eventually, the whole thing was demonstrated to be a trick--but not a deliberate one.
Clever Hans, a newspaper photo photo from Wiki Commons
In the last years of the 19th century, in Berlin, a retired schoolteacher named Wilhelm Von Osten, who had an interest in animal intelligence and dabbled in phrenology (the practice of measuring IQ by measuring the shape of the head), decided to carry out an experiment: he attempted to teach three animals--a kitten, a bear cub, and his own horse "Hans", how to count. Von Osten began by holding the animal's foot, showing them a number, then tapping on the foot the appropriate number of times. The cat and the bear never showed any ability to understand this--but the horse was a different story. After a time, Von Osten later reported, Hans seemed to catch on, and Von Osten began teaching him simple arithmetic (tapping four times when asking Hans "what is 3 plus 1?"). then, Von Osten made a "letter board" with the alphabet on it, and taught Hans to spell out words and sentences by tapping a number for each letter on the board. He carried out this training for over two years.
What really astonished Von Osten, though, was that Hans seemed to be able to answer on his own. When Von Osten stopped holding the horse's hoof and tapping out the numbers he wanted, Hans continued to reply to questions--and he continued to respond correctly. If Von Osten asked the horse "What is 2 plus 3?" the horse would tap his hoof on the ground five times. If Von Osten asked "What day of the week is it?", the horse would tap his hoof to indicate each letter on the "letter board" and spell out the right answer. Von Ostend was convinced that Hans had incredible intellectual abilities far above those seen in any other animal.
In 1891 he began publicly demonstrating the intellectual abilities of "Clever Hans". The horse became a sensation, and crowds gathered everywhere in Berlin to see the amazing feats, as Hans solved math problems, identified people both present and historic, identified days on the calendar, and answered other questions that were put to him by audience members. In 1904, a story about Clever Hans even appeared in the New York Times in the United States, on the front page, under the headline: "Berlin's Wonderful Horse: He Can Do Almost Everything But Talk".
The German board of education appointed a panel (the "Hans Commission") to study Clever Hans and look for evidence of trickery or hoax. The panel ruled out the possibility that Von Osten was prompting the horse, by questioning Hans while Von Osten was not present, and finding that Hans could still answer questions. He was obviously not being prompted or given the answer by Von Osten. (And Von Osten did not have any financial motive for a fraud, since he performed all his shows for free.)
It wasn't until a doctor from the Berlin Psychological Institute, Oskar Pfungst, looked into the matter that the mystery was solved. Pfungst noticed an interesting detail that others had overlooked: Clever Hans only knew the answer to a particular question if the questioner also knew the answer. If asked something that the questioner himself did not know, the horse was not able to give a response--he just kept tapping without stopping. And, in another telling detail, Pfungst then noticed that Hans also could not answer if the questioner did know the correct response but the horse was not able to see him.
Pfungst realized that Hans was indeed being prompted in his answers by visual cues, but not by Von Osten, and not intentionally. Whenever he was asked a question, Hans would simply begin tapping, then examined the humans nearby for subtle clues in body language that indicated he had reached the "correct" answer, and stopped. He had learned to recognize these cues while watching Von Osten during his years-long initial training process. The horse had no idea what he was "saying"--he was simply responding to the nonverbal cues he was being given by the humans. The humans, including Osten, were not aware that they were prompting the horse--their cues were unconscious and unintentional, the result of tension as the horse approached the "right" answer and then a release of tension as Hans reached it. And if the humans didn't know the answer, they were unable to give out any cues and Hans was unable to answer. Pfungst, meanwhile, found that by consciously doing things like subtly relaxing his shoulder muscles or briefly holding his breath, he was now able to provide the cues that Hans was looking for, and could now prompt the horse into giving incorrect answers to simple questions. Although Clever Hans had become very skilled at detecting and responding to human nonverbal cues, he was, in reality, no better in math than any other horse. Although the trickery was unintentional on everyone's part, it was there nevertheless.
Today, the "Clever Hans Effect" is an important part of psychology, where it is known as the "Cueing Effect". Infant humans do it, and so do pet dogs. The phenomenon has also become a standard part of many professional magician's training--and also that of researchers into so-called paranormal and psychic phenomena. It plays a role in research into animal communication, in famous cases like Alex the parrot or Washoe the chimpanzee. And it has even found its way into courtrooms, as defendants in drug trials have used it as a defense against the "responses" of drug-sniffing dogs.