Koko, Kanzi, and ape language research: Criticism of working conditions and animal care.

The Strange World of Koko, Kanzi, and the Decline of Ape Language Research

The Strange World of Koko, Kanzi, and the Decline of Ape Language Research

The state of the universe.
Aug. 20 2014 11:50 PM

What Do Talking Apes Really Tell Us?

The strange, disturbing world of Koko the gorilla and Kanzi the bonobo.

Dr. Sue Savage-Rumbaugh instructs the bonobo Kanzi to make a fire on Nov. 11, 2011 in Des Moines, Iowa.
Dr. Sue Savage-Rumbaugh instructs the bonobo Kanzi to make a fire on Nov. 11, 2011, in Des Moines, Iowa.

Photo by Laurentiu Garofeanu/Barcroft USA/Barcoft Media via Getty Images

Last week, people around the world mourned the death of beloved actor and comedian Robin Williams. According to the Gorilla Foundation in Woodside, California, we were not the only primates mourning. A press release from the foundation announced that Koko the gorilla—the main subject of its research on ape language ability, capable in sign language and a celebrity in her own right—“was quiet and looked very thoughtful” when she heard about Williams’ death, and later became “somber” as the news sank in. Williams, described in the press release as one of Koko’s “closest friends,” spent an afternoon with the gorilla in 2001. The foundation released a video showing the two laughing and tickling one another. At one point, Koko lifts up Williams’ shirt to touch his bare chest. In another scene, Koko steals Williams’ glasses and wears them around her trailer.

These clips resonated with people. In the days after Williams’ death, the video amassed more than 3 million views. Many viewers were charmed and touched to learn that a gorilla forged a bond with a celebrity in just an afternoon and, 13 years later, not only remembered him and understood the finality of his death, but grieved. The foundation hailed the relationship as a triumph over “interspecies boundaries,” and the story was covered in outlets from BuzzFeed to the New York Post to Slate.

The story is a prime example of selective interpretation, a critique that has plagued ape language research since its first experiments. Was Koko really mourning Robin Williams? How much are we projecting ourselves onto her and what are we reading into her behaviors? Animals perceive the emotions of the humans around them, and the anecdotes in the release could easily be evidence that Koko was responding to the sadness she sensed in her human caregivers. But conceding that the scientific jury is still out on whether gorillas are capable of sophisticated emotions doesn’t make headlines, and admitting the ambiguity inherent in interpreting a gorilla’s sign language doesn’t bring in millions of dollars in donations. So we get a story about Koko mourning Robin Williams: a nice, straightforward tale that warms the heart but leaves scientists and skeptics wondering how a gorilla’s emotions can be deduced so easily.


Koko is perhaps the most famous product of an ambitious field of research, one that sought from the outset to examine whether apes and humans could communicate. In dozens of studies, scientists raised apes with humans and attempted to teach them language. Dedicated researchers brought apes like Koko into their homes or turned their labs into home-like environments where people and apes could play together and try, often awkwardly, to understand each other. The researchers made these apes the center of their lives.

But the research didn’t deliver on its promise. No new studies have been launched in years, and the old ones are fizzling out. A behind-the-scenes look at what remains of this research today reveals a surprisingly dramatic world of lawsuits, mass resignations, and dysfunctional relationships between humans and apes. Employees at these famed research organizations have mostly kept quiet over the years, fearing retaliation from the organizations or lawsuits for violating nondisclosure agreements. But some are now willing to speak out, and their stories offer a troubling window onto the world of talking apes.

* * *

The first attempts to communicate with other primates began in the 1930s. Scientists knew that chimpanzees were our closest relatives and they wondered why chimps didn’t also have language.* Researchers theorized that culture could have something to do with it—perhaps if apes were raised like humans, they would pick up our language. So Indiana University psychologist Winthrop Kellogg adopted a 7½-month-old chimpanzee he named Gua. He raised Gua alongside his own human son, Donald, who was 10 months old when Gua arrived. Time magazine wrote that the experiment seemed like a “curious stunt”; others were critical of separating a baby chimp from her mother or rearing a child with a chimp. At one year of age, Gua could respond to verbal commands, but to her humans’ disappointment, she never learned to speak. The experiment was abandoned after nine months.

A human larynx.

Courtesy of Creative Commons

In the following few decades, scientists discovered that anatomical differences prevent other primates from speaking like humans. Humans have more flexibility with our tongues, and our larynx, the organ that vibrates to make the sounds we recognize as language, is lower in our throats. Both of these adaptations allow us to produce the wide variety of sounds that comprise human languages.

In a stroke of genius, researchers decided to try teaching apes an alternate, nonvocal way to communicate: sign language. Washoe, a chimpanzee, was the first research subject. Washoe was born in West Africa, then captured and brought to the United States. University of Nevada psychologists Allen and Beatrix Gardner adopted her in the 1960s. Like Gua, Washoe was raised as a child: She had her own toothbrush, books, and clothes, and the Gardners took her for rides in the family car. Over the course of her life, Washoe learned more than 250 signs, and she reportedly even coined novel words. One famous story has it that she signed “water bird” after seeing a swan. Skeptics remain unconvinced that this was evidence of spontaneous word creation, suggesting that perhaps Washoe merely signed what she saw: water and a bird.

The next decade saw an explosion of human-reared ape language research, and the same cycle of claims and criticism. Scientists named chimps as if they were human children: Sarah, Lucy, Sherman, Austin. Another was named Nim Chimpsky, a playful dig at Noam Chomsky, the linguist known for his theory that language is innate and uniquely human.

Third generation baby chimps from the original space research monkeys inside the nursery at the Coulston Research Center, New Mexico, September 1997
Third-generation baby chimps from the original space research monkeys play in the nursery at the Coulston Research Center in New Mexico in September 1997.

Photo by Paul Harris/Getty Images

Scientists tried raising other ape species as well: Chantek, an orangutan; Matata, a bonobo; Koko, a gorilla. Koko, especially, was a sensational hit with the media. Originally loaned as a 1-year-old to Stanford graduate student Francine “Penny” Patterson for her dissertation work, Koko remained with Patterson after the dissertation was complete, and Patterson founded the Gorilla Foundation in 1976 to house Koko and another gorilla, Michael.

* * *

Of the many ape savants studied over the years, two stand out as the most celebrated: Koko the gorilla and Kanzi the bonobo. Both have been profiled repeatedly in the media for their intellect and communication skills.


Koko’s résumé is more impressive than most humans’: She stars in a book called Koko’s Kitten written by Patterson and Gorilla Foundation co-director Ron Cohn, chronicling Koko’s relationship with a tail-less kitten that Koko named All Ball. The book, according to the Gorilla Foundation’s website, is “a classic of children’s literature,” and it was featured on Reading Rainbow in the 1980s. Koko has had her likeness turned into stuffed animals, and she was the guest of honor in two AOL chats, in 1998 and 2000. Koko has also had many celebrity supporters over the years: She’s met Leonardo DiCaprio and the late Mister Rogers. William Shatner says she grabbed at his genitals. Betty White is on her board of directors. Robin Williams tickled her in 2001. At the end of the day that Patterson and other colleagues told Koko about Williams’ death, the Gorilla Foundation announced, Koko sat “with her head bowed and her lip quivering.”

Another media favorite was Kanzi, a bonobo whose brilliance was discovered by accident. Kanzi was born at Yerkes Primate Center in 1980. Kanzi’s mother was a female named Lorel, but a dominant female named Matata laid claim to Kanzi and unofficially adopted him. Matata was being trained to communicate by pointing to symbols on a keyboard called lexigrams that corresponded to English words. Much to the chagrin of her human researchers, Matata showed little interest in her studies, but one day in 1982 Kanzi spontaneously began expressing himself using the lexigram board. From then on, researchers turned their focus to him instead. Sue Savage-Rumbaugh, a researcher at Yerkes at the time, had worked with chimps and other bonobos, so she oversaw Kanzi’s training. He quickly built up a lexigram vocabulary of more than 400 symbols. He’s also been said to invent new words by combining symbols, refer to past and present events, and understand others’ points of view, all of which are skills usually attributed only to humans.

Kanzi, at age 31, uses a lighter to start a fire on Nov. 11, 2011 in Des Moines, Iowa.
Kanzi, at age 31, uses a lighter to start a fire on Nov. 11, 2011, in Des Moines, Iowa.

Photo by Laurentiu Garofeanu/Barcroft USA/Barcoft Media via Getty Images

These apes are able to communicate with humans, and this alone is a testament to primate cognition. But in the past few decades there has been a spirited debate about whether apes are using language in the same way humans do. One major difference between ape and human communication appears to be their motivation for communicating. Humans spontaneously communicate about the things around them: Adults make small talk with the grocery store clerk about the weather; a toddler points out a dog on the street to her parents; readers write comments about stories on Slate.

Unlike us, however, it seems that apes don’t care to chitchat. Psychologist Susan Goldin-Meadow points out that studies with Kanzi show that only 4 percent of his signs are commentary, meaning the other 96 percent are all functional signs, asking for food or toys. Similar skepticism about Koko emerged in the 1980s, when Herb Terrace, Nim Chimpsky’s former foster parent, published a fairly scathing critique of ape language research, leading to a back-and-forth with Patterson via passive-aggressive letters to the editor of the New York Review of Books. Among other criticisms, Terrace asserted that Koko’s signs were not spontaneous but instead elicited by Patterson asking her questions. Patterson defended her research methods, then signed off from the debate, saying her time would be “much better spent conversing with the gorillas.”

Human and chimp skulls and brains (not to scale), as illustrated in Gervais' Histoire Naturelle des Mammifères.
Human and chimp skulls and brains (not to scale), as illustrated in the Histoire Naturelle des Mammifères from the 19th century.

Courtesy of Paul Gervais/Histoire Naturelle des Mammifères/Creative Commons

Critics also allege that the abilities of apes like Koko and Kanzi are overstated by their loving caregivers. Readers with pets may recognize this temptation; we can’t help but attribute intelligence to creatures we know so well. (Or to attribute complex emotions, such as grief over the death of a beloved comic and actor.) I recently wrote an article on a study suggesting that dogs experience jealousy, and the top response from dog owners was: “Anyone with a dog already knows this.” It’s hard to resist reading into animals’ actions, and it turns out, animals read into our actions, too. They carefully watch us for cues about what we want so they can get our attention or treats. In a classic case, a horse named Clever Hans was thought to understand multiplication and how to tell time but was actually just relying on the unconscious facial expressions and movements of his owner to respond correctly.

Long-term studies with human-reared apes are designed to create bonds between apes and their caregivers so that the pair feels comfortable communicating. This closeness often means that the caregiver is the only person able to “translate” for the ape, and it’s difficult to disentangle how much interpretation goes into those translations. As a result, the scientific community is often wary of taking caregivers’ assertions at face value. Some are straight-up skeptics. In a 2010 lecture, Stanford primatologist Robert Sapolsky alleged that Patterson had published “no data,” just “several heartwarming films” without “anything you could actually analyze.”