Kanzi seems an unlikely fellow to be at the center of raucous debate on the origin of language. A creature of simple tastes, Kanzi enjoys a good game of hide-and-seek. He is fond of dogs. And he likes M&Ms, burritos, Coke and, somewhat predictably, bananas.

Although Kanzi cannot speak, his caretakers say he can communicate with them by pointing at one or another of 250 symbols arranged on a board. Kanzi points at the symbol for "chase" and then points at the symbol for "hide" and finally points to his caretaker Liz, and Liz understands that Kanzi wants Liz to chase him and then hide with him, in that order.

This is remarkable, and controversial, because Kanzi is a chimpanzee.

Based on such communications, the psychologists and biologists who work with Kanzi reported this month that their subject can create two- and three-word sentences. Kanzi, they say, has not only learned some simple rules of grammar taught by his handlers, but has invented grammatical rules all his own. The scientists say Kanzi's language skills are roughly equal to those of a 2-year-old human child.

Over the years, there have been many claims of "talking" chimps and gorillas, who communicated with their handlers by sign language. But many researchers dismissed them as intelligent animals that simply learned to imitate their handlers to get what they wanted.

In a blistering critique a decade ago on the whole ape language field, psychologist Herbert Terrace and colleagues at Columbia University wrote an article in the journal Science entitled: "Can an Ape Create a Sentence?" Their answer, based on work with a chimpanzee they named "Nim Chimpsky" after linguist Noam Chomsky, was a definitive "no." 13,691 'Utterances' Recorded

But now a pair of researchers say the answer is "yes." Patricia Marks Greenfield, a child psychologist at the University of California at Los Angeles, and Susan Savage-Rumbaugh, a behavioral biologist at Georgia State University, believe they have shown that humans are not the only species capable of language. The duo present their studies of Kanzi in a book published this month, entitled "Language and Intelligence in Monkeys and Apes."

Kanzi "speaks" by pointing at symbols, or lexigrams, on a board that he hauls around with him on his runs through the 55-acre compound at Emory University's Yerkes Regional Primate Research Center. Over a period of five months, Kanzi's caretakers spent nine hours a day, seven days a week with the chimp and recorded 13,691 "utterances."

Kanzi appeared capable of creating several types of sentences. He could combine symbols for action and objects. To wit, Kanzi would point to the lexigrams for "hit" and "ball" or "tickle" and "Kanzi." Many of Kanzi's utterances indicated his desire for food or fun.

But Kanzi went beyond these simple sentences. Kanzi would point to the lexigrams for "chase" and "hide" and then he would gesture to one of his handlers. Translation: "You chase me and then we'll hide together." Greenfield said that she reviewed hours of videotape and found that Kanzi's caretakers had never used such grammatical combinations. In other words, Greenfield and Savage-Rumbaugh maintain that Kanzi invented his own grammar and used it consistently.

Greenfield also said she believes that Kanzi understood that a change in word order meant a change in meaning. For example, Kanzi would communicate: "Grab Matata," meaning that his mother was being grabbed. Other times, Kanzi would point to the symbols for "Matata bite," meaning Matata was doing the biting. The researchers say that their review of Kanzi's utterances show that such grammatical sentences were not simply random, but statistically significant representations of how Kanzi communicated.

The work with Kanzi has been both applauded and skewered. The question of ape language is an emotional one fought over by two warring camps.

The most influential linguist of recent times, Noam Chomsky of the Massachusetts Institute of Technology, argues that the capacity for language, with its use of symbols, syntax and grammar, belongs to humans alone. In essence, Chomsky believes that our brains, and our brains alone, are uniquely wired for language. Chimps may be clever and communicative, Chomsky says, but not clever enough to learn language.

Chomsky said last week, "It is almost unimaginable that this work has anything to do with humans." To maintain that Kanzi has language capacity, Chomsky said, is like saying a man can fly just because he can jump in the air. Chomsky dismissed the entire ape language field as gripped by "sentimental confusion."

Chomsky and his allies believe that humans underwent a period of rapid genetic change that endowed them with the biological capacity for language. Chomsky's opponents maintain instead that the capacity for language must have evolved gradually, and that chimpanzees and humans probably share a common ancestor that could handle language or something close to it. Semantics or Simian Antics

The critics are not swayed by Kanzi's "utterances."

"I myself am highly skeptical," said Mark Seidenberg of the University of Southern California. Seidenberg agrees that Kanzi is a very bright, very communicative ape. He notes that Kanzi is capable of pointing at symbols in order to get what he wants.

But there is a tremendous difference between pointing at a symbol and understanding that the symbol represents a name for a thing. Seidenberg said it is even more difficult for him to believe that Kanzi not only understands the symbols, but combines them in new and creative combinations to form sentences.

"Kanzi's problem is that there are these people who want him to use these symbols, who are stressing that these symbols are important," Seidenberg said. "Kanzi is an intelligent animal. There are things he wants. So he asks, 'How can I use these symbols to make my way through this world?' "

Other researchers have pointed out that a majority of Kanzi's "utterances" deal with requests. Terrace of Columbia University has compared animals making requests for food or affection to the users of vending machines. They are capable of pushing buttons (or pointing at symbols) to get what they want. But, according to Terrace and his colleagues, such behavior says little or nothing about language.

Chomsky and others have asked: If chimps have the capacity for language, why don't they have one? Greenfield responds that many humans have the capacity for a written language, but never develop one. She also suspects that chimpanzees -- and in particular pygmy chimpanzees like Kanzi -- may employ something like a primitive language in the wild. This assertion is sure to set off even more debate.