Science/Medicine : Chimp Called a Grammarian, of Sorts : Language: Researchers in new book claim that a 7-year-old primate named Kanzi can invent his own rules--and use them consistently.

A 7-year-old pygmy chimpanzee in a Georgia primate center may reignite a long-festering controversy over a question that is fundamental to the identification of humans as a unique product of evolution: Can other animals use language?

For the last three decades, researchers have shown that chimps, pigeons, dolphins and other animals can manipulate word symbols in a manner that some say demonstrates comprehension and correct use of language. But linguists have argued that this seeming facility simply reflects the animals’ ability to recognize subtle clues in their caregivers’ body language and respond appropriately.

In a book published last week, however, researchers from UCLA and Georgia State University report for the first time that the Georgia chimpanzee, named Kanzi, is able not only to learn and use grammatical rules--the principles that allow a small number of words to be arranged in an infinite number of sentences--but even to invent his own and use them consistently. He has done this with hand gestures and symbols on a computer keyboard.

Kanzi’s ability to use grammar is on a par with that of a 2-year-old human child, said UCLA psychologist Patricia Marks Greenfield. “He makes short, telegraphic sentences talking about relationships between actions and objects, objects and locations, and so forth,” she said.

The ability to use grammar is a “Maginot Line” that marks a sharp demarcation between humans and other primates, said psychologist Elizabeth Bates of UC San Diego, who added: “I think that that line may be crossed.” The new discovery “is potentially extremely important,” she said, but is likely to provoke heavy criticism.

A case in point is linguist Noam Chomsky of the Massachusetts Institute of Technology, who dismissed the new results out of hand. A chimpanzee using grammar “is not a logical impossibility,” he said, “but it is so outlandish that I don’t know of any biologist who has taken the possibility seriously.”

Others were not so quick to pass judgment but were demanding strong proof of the claim. “There is one non-negotiable condition for verification of the claim: an unedited wide-angle videotape or movie of the communication,” said psychologist Herbert Terrace of Columbia University. “We’ve had claims like this before and on closer scrutiny the animals were found to be responding to cuing (subtle, inadvertent gestures by caregivers) or rewards. That kind of response is not terribly pertinent to language.”

Greenfield said that she is confident her work meets such criteria.

Researchers are intensely interested in the ability of primates to understand and use language because of the insight such studies give into the roots of human communication. Such researchers are separated into two camps.

One camp, the evolutionists--who are now a minority--argues that a rudimentary ability to use language was present in the evolutionary ancestors of humans. This ability expanded slowly as the size of the human brain grew over millions of years and as the throat and voice box evolved into a shape that permitted speech.

Because humans share more than 99% of their genetic material with other primates, the evolutionists argue, modern primates should retain the rudimentary abilities of our common ancestors.

The opposite camp, the so-called discontinuity theorists, argues that language in general and grammar in particular arose as a result of a sudden mutation in the human species, perhaps as recently as 200,000 years ago. The most important characteristic of human language, they argue, is the ability to combine words spontaneously and creatively to produce new concepts and new sentences never heard before--in short, to use grammar.

Chomsky has been the leading proponent of this view.

The discontinuity theorists have been ascendant in recent years, primarily because of a phenomenon known as the “Clever Hans effect.” Clever Hans was a horse, taught by German schoolteacher Wilhelm von Osten in the early 1900s, who seemingly could solve simple addition and subtraction problems, tapping out the answers with his hoof.

But Clever Hans was not really counting. Later studies showed that the animal watched for a small, involuntary jerk of von Osten’s head when the right number was reached, at which point he stopped tapping his hoof.

Although a large number of now famous chimpanzees, such as Koko, Washoe, Lana and Nim Chimpsky, have been shown to use sign language or word symbols called lexigrams, critics have always argued that they were just demonstrating learned behavior for a reward or responding to cues given by caregivers, just as Clever Hans had before them.

“It didn’t appear as if there were any strong answers to these criticisms,” Bates said. “But the work with Kanzi has posed a strong challenge to this picture. The pygmy chimpanzees appear to be learning spontaneously, actively and creatively. . . .”

Greenfield and her co-author, behavioral biologist E. Sue Savage-Rumbaugh of Georgia State, studied Kanzi, who was then 5 1/2 years old, over a five-month period at her home in the Yerkes Regional Primate Research Center outside Atlanta.

Kanzi, one of the first pygmy chimpanzees to be studied for language ability, has shown great facility in learning the meanings of lexigrams, as well as understanding spoken English. “He does seem to respond (to English) in a way I’ve never seen other chimps do,” said education professor Susan Goldin-Meadow of the University of Chicago.

Research assistants worked with the animal nine hours a day, seven days a week for five months, recording 13,691 “utterances.” Kanzi communicated with the researchers by touching lexigrams (more than 250 printed words and arbitrary geometric symbols) on a computer and by combining lexigrams with gestures. Their results are reported in the book “ ‘Language’ and Intelligence in Monkeys and Apes.”

“We focused specifically on spontaneous combinations, when Kanzi combined more than one symbol to make something like a sentence,” Greenfield said in an interview. One important discovery was that “he had the beginnings of the ability to use a change in symbol order to mark a change in meaning.”

For instance, if Kanzi were going to playfully bite his half-sister Mulika, he might sign “bite Mulika.” But if his sister bit him, he would sign “Mulika bite.” Other chimpanzees that have been studied would not be able to distinguish between such changes in symbol order, Greenfield said, but Kanzi almost always used the correct order.

Even more important, she said, was Kanzi’s ability to produce his own rules. Often, for example, he would combine a lexigram with a gesture. “When he did this, he spontaneously followed the rule to always place the lexigram first and the gesture second,” she said.

Greenfield and Savage-Rumbaugh thought that perhaps Kanzi had learned this from his caretakers, but studying videotapes, they found that the caretakers almost always used the opposite order: gesture first and lexigram second. “So he went against the model he was getting from his humans,” Greenfield said.

Another rule Kanzi devised was to put two action symbols in the same order in which he wanted to carry them out, such as “chase bite” or “chase hide.” “Essentially he was using symbol order to plan action order,” Greenfield said.

“Nobody who has tried to train chimpanzees . . . has ever reported this before,” she added. “By human standards, it seems very simple, but it sheds light on the origins of human grammar. At its beginnings, we order our actions and at some point we use that as a model for ordering our symbols.”

Perhaps even more sophisticated was Kanzi’s ability to combine two different rules of his own devising. He might sign, for example, “chase bite Mulika,” in which he used two lexigrams in the order in which he intended to carry out the actions and combined them with the gesture of pointing at Mulika, following the rule of “gesture follows lexigram.”

Kanzi’s ability to invent grammar is very important, Greenfield said, “because if you think about the evolution of human language, at some point it not only had to be learned from parents, but actually had to be invented. To show that a chimpanzee has inventive ability, particularly the ability to invent simple grammar, is important because it is taken by Chomsky to be the defining feature of human language.”

Their results, she said, thus provide a strong argument against the likelihood human language developed abruptly.

Chomsky argued, however, that if pygmy chimpanzees have had the ability to use grammar for millions of years, they surely would have used it because it is biologically advantageous. “It’s completely unknown biologically that any organism could have a capacity that would be highly valuable for survival but would not use it,” he said.

Although the authors characterize Kanzi as having the abilities of a 2-year-old human, Greenfield cautioned that there are many differences between the two. For example, it took Kanzi three years to acquire simple grammatical rules that a human child might acquire in one.

Also, he was producing a much smaller number of symbol combinations than human children do at the corresponding age. “Children would use fewer one-word utterances. Kanzi was making fewer sentences and more one-word utterances,” she said.

Nonetheless, Kanzi’s abilities are remarkably human-like and seem likely to provoke heated debate for many years.