The visitor is announced, and Koko presses up against the glass to make her inspection. Deep black eyes take in the expectant face and opened notebook, and then Koko draws back her lips and points to one of her teeth.

"She wants to know if you have a gold tooth," Penny Patterson says.

No gold tooth, the visitor says regretfully. Koko keeps pointing. The visitor opens her mouth for confirmation. Koko peers in and swipes one large fist across her own nose.

"Koko!" Patterson looks embarrassed. "That was an insult. She said 'toilet.' "

Well, one might do worse. Koko, a 13-year-old lowland gorilla who lives principally in a large trailer 40 miles from San Francisco, is also said to address those who displease her as "dirty toilet" (curled fingers under the chin before touching the nose), or "rotten stink" (thumb brought to and pulled away from the nose, nose pinched). In a particular fit of pique, she once called Patterson "Penny toilet dirty devil," a rapid series of maneuvers that ended with fingers brandished as devil horns. She has also, according to Patterson's notes, used sign language to name her paintings ("Love," "Hate," "Alligator Bite") and lure her gorilla trailermate ("Come tickle bottom"). She is described, in Patterson's newsletters and book about her, as having discussed death:

Volunteer: Where do gorillas go when they die?

Koko: Comfortable hole bye.

Volunteer: When do gorillas die?

Koko: Trouble old.

And tigers:

Volunteer: Don't like tigers?

Koko: Frown bad red.

Volunteer: I like tigers.

Koko: Tiger nail rough.

Volunteer: What animals do you like?

Koko: Gorilla love.

The notes say that in the 12 years since Koko began learning American Sign Language as part of Francine (Penny) Patterson's graduate work at Stanford University, the gorilla has used her signing vocabulary to ask questions, lie, joke, rhyme, apologize, tease, invent and -- in a small drama that sent reporters streaming down to Woodside again this month -- grieve over the death of All Ball, the Manx kitten seen cradled in Koko's arms on the cover of January's National Geographic. All Ball was hit by a car. He had darted out one evening and was found on the road the next morning, and Patterson says she remembers precisely the wording she chose to bring the news to Koko and Koko's companion gorilla Michael.

"We've found Ball," Patterson said, signing as she spoke, the way she often does when addressing the gorillas. "He went out into the road and was hit by a car, and we won't be seeing him anymore."

She does not remember using any specific words about death, but her partner Ronald Cohn says he thinks she said "dead" or "killed" when she told him the news before turning to sign to the gorillas. Patterson left the trailer, and this, according to the volunteer who came in a few minutes later, is what Koko the gorilla did:

She was quiet for a very short time. Then she let out a high, hooting call, whoo, whoo, whoo, the cry she used to sound as a baby when Patterson left her alone at night in her cage.

"Those are the only times she gives that cry," Patterson says. "It's like 'someone's leaving me.' "

If Penny Patterson is right, there is little real precedent for the work she has done with Hanabi-Ko, which was the Japanese name given to Koko -- it means "fireworks" -- when the baby was born July 4, 1971, at the San Francisco Zoo. Other primates have learned some sign language. There are sign language experiments under way in Washington State, where five chimpanzees are reported to communicate with their trainer and one another in sign; and in Tennessee, where an orangutan named Chantek has been learning forms of American Sign Language, the gestural language used by many of the deaf. But no one else has taught signing to gorillas, no one else has worked for so long with a single animal, and if Penny Patterson is right, no one else has recorded such an astonishing variety of abstract and intelligent expression in any animal, much less the gorilla, which was once supposed to be so fearsome and nasty and thick.

If Penny Patterson is right.

"The degree to which this has scientific meaning is very much open to question," says Terry Maple, a primatologist and psychology professor who wrote the reference book "Gorilla Behavior" and now directs the Atlanta Zoo. "I personally feel that she has gone way beyond the data in her claims for that gorilla -- way beyond the data."

"The whole thing is ridiculous, of course," says Thomas A. Sebeok, an Indiana University anthropology, linguistics and semantics professor who coedited a 1980 book that attacks the techniques and conclusions of language experiments with primates. "If Penny Patterson tried to publish in a scientific atmosphere, then she would be laughed out of court."

Hanabi-Ko was 1 year old, separated from her mother and recently recovered from malnutrition and an intestinal infection, when San Francisco Zoo officials approved Patterson's proposal to begin teaching the baby gorilla sign language. Patterson's field was developmental psychology -- she has a PhD from Stanford -- and she had been inspired by the work of Allen and Beatrix Gardner, two Nevada psychologists who had caused an uproar among linguists and psychologists by reporting that chimpanzees raised in a human-like environment had learned to communicate in American Sign Language gestures.

"When I entered the nursery of the Children's Zoo, Koko left the arms of her caretaker, Debbie Lee, for mine," Patterson writes in her book, "The Education of Koko," which was cowritten with Eugene Linden. "She pushed her soft face close to mine, smelling me and looking me over. Then Debbie put the 20-pound gorilla, all black save for a white rump patch, onto the nursery floor and I signed Hello (a gesture somewhat like a salute). Koko put her hand on her head and patted it and then promptly pulled my hair as I sat down."

Patterson had already decided to teach signs the way the Gardners had -- by "molding," or forming Koko's hands into a sign while showing her what it meant. She brought food into the cage for the "food" sign; she gave a bottle as a reward for the correct signing of "drink"; she divided the bottle contents and parceled them out to teach Koko "more." Within two months, according to Patterson's notes, Koko was signing in word combinations -- "drink there," "more food."

She says she remembers the day Koko asked her first question. On other days they had played a game together; Patterson would exhale on the window, and Koko would draw on the fogged glass. Now Koko was pointing, first to the window, then to her own mouth, then to Patterson's mouth, and then to the window again. Patterson could not imagine what the gorilla wanted, until it dawned on her that Koko was asking for the game. "That nearly knocked me over."

Originally envisioned as a short-term project that would end when Koko rejoined the other zoo gorillas, Patterson's study slowly began to look rather like permanent adoption. She moved Koko away from the zoo, first to a special trailer at Stanford, and later to the wooded six acres where a small white house overlooks the gorillas' trailers and fenced play area. The house is Patterson's, although gorilla memorabilia appears to be advancing on the human turf inside: gorilla portraits line the walls, a computer and two television sets fill the corners of the living room, and stacks of videotapes crowd up around the armchairs. Koko is no longer allowed inside Patterson's own quarters; some years ago, in a moment of gorilla exuberance, she smashed the frame of Patterson's bed, and that was the end of that.

A small corps of volunteers now attends Koko and Michael during the day. Along with Patterson, who also edits and writes much of her Gorilla Foundation's newsletter, the volunteers converse in sign language with the gorillas, teach them new signs and supervise daily routines mostly reminiscent of a child's preschool -- games, cleanup, lunch. Koko is partial to corn on the cob and tomatoes. She dislikes grapefruit, and Patterson says she invented a multilayer signing pun to describe that: Koko executes the "drink" sign in the position of the fruit sign, and simultaneously rolls down her lower lip to indicate "frown."

Her working vocabulary, signs Patterson says she uses appropriately and spontaneously, is about 500 signs, although Patterson says she has seen Koko use nearly a thousand different signs so far. Michael's vocabulary is reported to be a little less than half that size. He has never shown as much inclination to learn signing as Patterson says Koko has, although his acrylic abstract paintings are more interesting and attracted a good bit of attention when they were shown last year at an exhibition in Honolulu. Some of the paintings at that exhibit sold for several hundred dollars apiece but the gorillas were accustomed to celebrity by then, since Koko and Patterson have become something of a fixture in local newspapers and television programs here. Indeed, when a major central California earthquake two years ago was felt as far west as Woodside, San Francisco Chronicle columnist Herb Caen reported Koko's reaction to the tremor. Her assessment, Caen wrote, was "Darn darn floor bad bite. Trouble trouble."

In 1866, swept up in the furor over Charles Darwin's revolutionary suggestions about the origins of man, the Linguistic Society of Paris announced it would refuse any further papers on the evolution of language. That the society reportedly still banned such papers as recently as a few years ago is some indication of the passion and consternation that from the beginning have dogged the research on language among nonhuman animals. Long after most scientists had embraced evolution as a guide to explaining animal development, it was believed that only humans had language -- that the ability to speak or otherwise use language was what separated man from even the most intelligent nonhuman animals.

It has been nearly 20 years since Allen and Beatrix Gardner began teaching American Sign Language to chimpanzees as part of an experiment to study the intelligence of chimps raised like human children. Even after the Gardners reported that their first chimpanzee was correctly using at least 132 signs of ASL, the scientific debate raged: Weren't the chimps simply following human cues? Could this be accurately called "language"? Weren't word order and syntax missing from signed phrases like "listen dog" and "open eat drink"?

"Most scientists agree there's no question these animals are making some kind of gestures," says H. Lyn White-Miles, the University of Tennessee anthropology professor who has been teaching sign language to Chantek the orangutan. "The question is, what do they mean?"

And to this day, even among scientists presumably unencumbered by religious or philosophical convictions about human superiority, the arguments about that are fierce. Sebeok, coeditor of the book that criticizes ape language experiments, dismisses most of the researchers as either "intellectual crooks" or serious scientists who have fallen prey to the "Clever Hans" phenomenon. Clever Hans was a turn-of-the-century German horse who mystified everyone around him, including his own trainer, by using his hoof to tap out solutions to arithmetic problems. A psychologist finally arranged a problem to which only the horse could know the answer -- the psychologist whispered part of the sum in one ear while the trainer whispered a second number in the other -- and Clever Hans was foiled. The horse had not been adding at all, but rather following the innocent cues of questioners who would react almost imperceptibly as soon as his hoof-tapping reached the correct number.

The contemporary researchers are doing just that, Sebeok insists -- cuing their animals, either through mistakes in technique or through plain scientific sloppiness. "We found all the experiments seriously wanting," he says. "We are not saying whether this is possible. We are saying that so far the evidence is so bad, the experiments are so shabby, they just won't stand up."

One of Sebeok's better-known allies in the debate is a researcher who changed his mind about language in nonhumans only after the failure of his own efforts to teach a chimp sign language. Herbert Terrace, a Columbia University experimental psychologist, spent four years working with Nim Chimpsky (the name is a salute to linguist Noam Chomsky), a male chimpanzee who initially excited Terrace with evidence that he was learning to sign. But Terrace then reexamined his own data, studying videotapes of Nim, and concluded that most of what Nim signed was unintentionally cued by the teachers.

"Want cat?" a teacher might ask in sign language, and Nim would respond, "Cat me." The signing that appeared uncued, Terrace decided, was just Nim choosing a sort of wild-card sign, one guaranteed to elicit response, without any real understanding of what it meant.

"Chimps only demand," Terrace says. "They don't name. I don't think they understand cognitively that things have a name. If an ape doesn't understand things have names, it's crazy to worry about grammatical rules . . . I think it's a sham to refer to what Washoe the Gardners' first chimp or Koko have done as anything related to child language."

Those are fighting words, and people like Patterson and the Gardners respond by attacking Terrace's research methods, his analysis of his own data and his assumptions about language learning among humans. They say that properly conducted testing could have eliminated cuing. They say Terrace never taught Nim the way human children are taught language. They say Terrace and Sebeok reached their conclusions only by ignoring considerable properly gathered data. And it is nonsense, they say, to imagine that language is somehow removed from every other evolved animal trait. If intelligence moves gradually up the evolutionary scale to Homo sapiens, why not the capacity for language? Why this insistence that man alone can question, converse, arrange word order, name things around him and understand that they have names?

Once, observes Central Washington University psychologist Roger Fouts, tool use was thought to be the dividing line between man and the rest of the world's species. Then it was learned that animals use tools, so the dividing line became toolmaking ability. Then it was learned that apes make crude tools -- so the dividing line, as Fouts sees it, became rationality, the conviction that only man is truly rational.

"So language became the behavioral version of rationality," says Fouts, who has taken over work with the Gardners' signing chimpanzees. "It's been referred to as the last bastion . . . They wanted something big, and important, so they ended up with language."

Eerie is how it feels. A 230-pound gorilla is standing behind a wire enclosure, gazing straight into your eyes, and addressing you in what has the unmistakable look of sign language.

"Visitor -- " Koko's outstretched fingers tapping her own shoulders --

"do -- " touching index fingers together --

"knee time -- " one finger to knee, then to wrist --

"you hurry. Good come." She points, shakes a hand and beckons.

Nobody else, as far as the visitor is able to tell, has mentioned "knee time," but Patterson explains, and the visitor performs as requested. Knee time involves resting your knee against the wire and waiting while Koko sidles up to it, gives it a fierce poke with her stubby fingers and then races the length of the enclosure, banging arms against the fence in what looks like vast amusement. Koko wants to do "elbow time" too, but Patterson says the last visitor who did that got his shirt sleeve ripped off. "Gentle," she admonishes, signing as she talks. "Nice."

A volunteer is quietly transcribing this entire exchange into a notebook, which is how Patterson and her aides have been able to pepper the book and newsletters with the more intriguing of Koko's responses. They have described the gorilla reading -- showing by sign that she recognizes printed words like "apple" and "nut." They have described her expressing opinion, like the time a volunteer concluded a session on body parts by asking Koko what was boring; Koko is reported to have signed, "Think eye ear eye nose boring." They have described her declaring correctly that a volunteer was overstaying her usual visit: "Time bye you," Koko signed, and when the volunteer signed "What?" Koko signed "Time bye goodbye."

And Patterson has recorded the signs she says Koko invents when the gorilla encounters things whose names she has not learned: "red corn drink" for pomegranate, "lettuce grass" for parsley, "my cold cup" for ice cream. Given a caramel apple on a stick when she was 6 years old, Koko is recorded as signing "lollipop food tree apple."

It is not, in the end, as mysterious as people seem to make it, Patterson says -- this is not St. Francis of Assisi, or the lone key that will somehow make audible the voices of all the world's buffalo and cockroaches and Irish terriers. One is tempted to call up the image of Helen Keller suddenly crashing out through the silence, but Patterson says nothing so cataclysmic takes place with a primate. "It's just a whole lot like talking to a young kid, and they don't know the secrets of the universe any more than we do," she says.

Such a memorable picture, the gorilla swinging from her cargo nets and the pretty blond woman smiling in through the fence.

"Entertaining," another scientist says sourly, clearly irritated that any reporter should take Patterson seriously enough even to make inquiries about her. "And you people seem to like it."

No one in the field seems to know precisely what to make of Koko and Patterson's claims about her. Patterson's scientific publishing has been minimal, her book was a well-written but popular account dismissed by one colleague as a "trade book," and even some of her milder critics say she sometimes comes off like a well-intentioned mother who sees brilliance in her child's every turn of phrase.

"I could sign to Chantek, 'What is the meaning of life?' " says White-Miles, the Tennessee scientist who has been working with an orangutan. "And since he's in the habit of conversing with us whether or not he understood it, he might make some sign like 'good go up', and I could say, 'Oh, he's saying we're all going to go to heaven . . .' Is my interpretation of his signing what he actually intended? Or a more conservative approach might be that he felt constrained to make some sign."

Patterson has heard this many times before. She knows that when a researcher spends 12 years with one animal, there may be problems of overinterpretation, and she says she regularly designs tests to check for that independently -- but she cannot work full time that way, she says. "You don't have conversation then."

And it is enormously time consuming to analyze data, particularly videotape, which must be studied in frame-by-frame detail to transcribe each sign Koko makes. Patterson says she must also go on managing the Gorilla Foundation, which receives no federal aid and relies on grant money and individual contributions. "The massive amounts of data we've accumulated are being processed slowly," she says. "All I can say is that it is in the works, and what I do not want to do is publish fast and furious and have to retract at the end."

Besides, Patterson says, she does not want to let up on the daily work with the animals. "I'd like to essentially try to define the limits -- I don't know if that's even possible -- of the gorillas' language abilities and related abilities, their intellectual development . . . and beyond that, I'm very concerned about gorillas as an endangered species, and knowing them as only language will allow us to -- being able to convey that."

Also, not to wax too unscientific about it, she loves her gorilla. "It's complicated, but I just really love the kid," Patterson says. She says All Ball's mother is pregnant again, and that Koko is to receive a new kitten from that litter. She is still touched by Koko's request at her last birthday, when Patterson offered what she thought was an appealing selection of food and toys. "This is so pathetic," Patterson says. "The poor kid . . . You know what she wanted? She wanted a diary and a pen." She got them; the pen is ball point, the diary is green and hard-bound like the logs Patterson uses, and now in the evenings Koko retires to a corner of her trailer to scribble in private. CAPTION: Picture 1, Koko making the sign for "eye while looking at a picture book. Picture 2, Peggy Patterson with Koko in 1978. Photos Copyright (c) 1978, National Geographic Society