Clever Hans
As the story is told, sometime late in the first decade of the twentieth century, a retired German mathematics teacher named Wilhelm von Osten discovered an odd thing about his horse. He had reason to believe that this particular animal, whose name was Hans, was capable of amazing, human-like powers of intellect. Von Osten was fascinated by the scientific implications of such a feat. He began lecturing Hans in a number of disciplines, including European political history, arithmetic, and classical music. Hans could answer questions on these subjects, and others; he did so by tapping his foot an appropriate number of times to indicate his replies.
Hans was able to answer questions posed in foreign languages, questions whispered to him, even questions not asked at all, but merely thought. For a brief moment in history, Clever Hans had the German scientific community chasing its tail trying to explain the phenomenon.
Eventually, an experimental psychologist named Oskar Pfungst uncovered the fact that if the questioner did not know the answer to the question asked, Hans was also in the dark. The horse’s accuracy rate, Pfungst also noted, dropped decidedly toward the evening hours, and he was able to answer no questions at all when the questioner stood behind an opaque screen.
The intelligence Hans was displaying was, as it turned out, immense; but it was not the particular intelligence that had baffled the scientific world. Hans was using the visual cues of his questioners to know when to stop tapping. Even Pfungst, knowing this, found it nearly impossible to not cue the horse with unconscious body language or expression changes as the horse neared the correct number of taps.
Behavior or Intelligence?
The Hans incident almost single-handedly stopped the advancement of theories of animal intelligence. Not long after the Clever Hans incident, American psychologist J.B. Watson proposed a new method of experimental psychology called behaviorism to distinguish between behavior and intelligence.
Behaviorism consists of “carefully defined stimuli and the overt responses they elicit, formulating rules to predict what animals will do when presented with experimentally controlled cues.” It was meant to circumvent what was going on in the nervous system entirely — to be an objective natural science. “Its theoretical goal is the prediction and control of behavior,” wrote Watson. “Introspection forms no essential part of its method nor is the scientific value of its data dependent upon the readiness with which they lend themselves to interpretation in terms of consciousness.” In short, behaviorism assigned an animal’s response not to intelligence, but rather instinct and behavior.
Learning by Accident
E.L. Thorndike, whose Animal Intelligence (1898) is widely regarded as the beginning of modern animal intelligence studies, argued against the ability of animals to demonstrate certifiable reasoning and intelligence. His most famous experiment involved placing cats in “puzzle boxes,” testing their ability to figure out a means of escape. What he concluded from this and other experiments was that a certain (and fairly small) percentage of the animals were able to learn by accident the means of escape (which involved things like pressing a thumb latch and pushing against a door). “The great support of those who do claim for animals the ability to infer,” he wrote, “has been their wonderful performances which resemble our own.” But the animals, he concluded, did not learn as we do. “…They learn those things by accident,” he wrote. “They certainly do.”
Education or Evolution?
No one is truly capable of getting into the heads of animals, although brain imaging has made great progress. It is possible to know what areas of a brain are active, but it is not yet possible to know if animals are “thinking.” Do they, like humans, reason and manipulate objects in their minds or use cognitive maps to solve problems, or is it that their seemingly “smart” actions are really just the end result of hundreds of thousands of years of evolution — a sort of simulated “smartness” in which the animal is carrying out actions that have proven successful over the millennia? Does the absence of language in animals (a point sometimes fiercely debated) somehow prevent them from insight? Or are they capable of what Germany’s Würzburg school of psychology called an “A-ha experience” — an epiphany?
What is Intelligence?
Some define intelligence as the ability to learn, which nearly all animals — even gold fish and bumble bees — are capable of. Others, like M.I.T.’s Stephen Pinker, hold that “intelligence is an ability to figure out how things work in order to overcome obstacles.” True, many animals can learn to overcome obstacles, but it is to learn for a purpose that, according to Pinker, defines intelligence and differentiates it from behavior.
One dominant idea about animal intelligence has been that it is linked to the use and creation of tools. Stephen Budiansky, author of a book about animal intelligence called If Lions Could Talk, points out that “the accomplishments that we as humans prize the most are often our feats of technological mastery.” So what constitutes a tool? Stones used by otters to break open mussel shells? Tree branches used by elephants to scratch their backs? What about when wasps tamp down the soil around the entrance of their burrow with small pebbles in order to hide the burrow from invaders?
Budiansky points out that many organisms we think of as “lower” animals and insects use tools, but, as in the case of many nest-building birds, do not seem to understand, in a human sense, what they’re doing. Weaverbirds construct some of the most elaborate nests, and go about their work in a “rigidly determined sequence of motions.” If the nest is damaged, their only means of repair is to start over; this behavior is common in many animals. Even weaverbirds raised in isolation, when supplied with the right grasses, have the hardwired know-how to build these fantastic nests. When they do not have the right grasses, they attempt to make the nests out of their own feathers.
Animal Language and Communication
As Joel Wallman, author of Aping Language, explains, “…language, at least in the European intellectual tradition, is the quintessential human attribute, at once evidence and source of most that is transcendent in us, distinguishing ours from the merely mechanical nature of the beast.” And no small amount of money and resources have gone into this field in the search for a link to our animal past.
Much language research has involved teaching primates to use variations of sign language including famous studies of three gorillas — Washoe, Koko, and Nim Chimpsky. World-renown linguist Noam Chomsky — after whom Nim Chimpsky is humorously named — insists that “the systems taught to apes and other species differ from human language at the most primitive and elementary level.” David and Ann Premack, primate researchers themselves, argue that “The evidence we have makes it clear that even the brightest ape can acquire not even so much as the weak grammatical system exhibited by very young children.”
Though some of these apes learned as many as 68 signs, Wallman argues that, in the end, Chomsky and the Premacks are right; the apes just weren’t capable of human communication. After years of research in which Nim was surrounded by a large cast of volunteers who “conversed” with him, Terrace, Nim Chimpsky’s trainer, was convinced that Nim could talk. However, when he began scrutinizing video tapes of Nim interacting with the volunteers, he discovered that every time Nim seemed to be expressing thoughts, he was actually just responding to prompts by the volunteers. It was Clever Hans all over again. Terrace went a step further and studied tapes of Washoe and Koko, and to the great displeasure of Allen and Beatrice Gardner (Washoe’s trainers) and Francine Patterson (Koko’s trainer), pronounced them also incapable of language.
Apples and Oranges
Linguist Noam Chomsky once said, “If you want to find out about an organism you study what it’s good at. If you want to study humans you study language. If you want to study pigeons you study their homing instinct. Every biologist knows this.”
Perhaps, as Stephen Budiansky argues, inherent in this idea lies the truth about animal intelligence. Budiansky’s book If Lions Could Talk is a singularly superb exploration of how animals are intelligent. He describes animals not as lessor versions of humans, but as practitioners of evolved skills. Budiansky likens animal intelligence testing to giving a blind person a written IQ test. “It is not,” he writes, ” a very meaningful evaluation of his mental abilities.”
In a cross-species intelligence test based in visual discriminations, monkeys performed far superior to rats, learning in a way that insinuated a cognitive process higher than a simple stimulus-response association. This was, Budiansky writes, “a conclusion that was comfortably accepted, as it fit well with our preexisting prejudices about the distribution of general intelligence in nature.” However, in the same test based in smell discrimination, the rats performed just as well as the monkeys had.
Clever Hans was also extraordinarily astute. No human would have been as adept at reading the expressions of those around him. Budiansky’s argument is that Hans was naturally attuned to such cues because a horse’s life depended on such awareness.
Every animal, Budiansky asserts, has just such a skill. “The branching tree of evolution has not just one culmination, but millions of culminations — represented in every living species on earth today,” he writes. “Each is a brilliant success at what it does.” And it is the nature of such “brilliance” that remains to ponder.
