"Long before it's in the papers"
August 14, 2015


Apes may be closer to speaking than many scientists think, study says

Aug. 14, 2015
Courtesy of the University of Wisconsin-Madison
and World Science staff

The fa­mous go­ril­la Ko­ko is best known for hav­ing learn­ed a good deal of Amer­i­can Sign Lan­guage from hu­mans, as well as for un­der­stand­ing about 2,000 spo­ken Eng­lish words.

Al­though speak­ing skills are not for her, the sim­ple sounds she does make are sur­pris­ing, a new study says—and could change a per­cep­tion that hu­mans are alone in their ca­pa­city for speech, among our close ev­o­lu­tion­ary rel­a­tives.

Koko plays with a doll. (Cour­tesy of the Go­ril­la Foun­d­a­tion)

In 2010, Mar­cus Perl­man, now a re­search­er at the Un­ivers­ity of Wisconsin-Madison, started re­search at the Red­wood ­city, Calif.-based Go­ril­la Founda­t­ion, where Ko­ko has spent more than 40 years liv­ing im­mersed with hu­mans. Ko­ko has in­ter­acted for many hours daily with psy­chol­o­gist Pen­ny Pat­ter­son and bi­ol­o­gist Ron Cohn.

“I went there with the idea of stu­dying Ko­ko’s ges­tures, but as I got in­to watch­ing vid­e­os of her, I saw her per­form­ing all these amaz­ing vo­cal be­hav­iors,” said Perl­man.

The vo­cal and breath­ing be­hav­iors Ko­ko had de­vel­oped were not nec­es­sarily sup­posed to be pos­si­ble, he said.

“Dec­ades ago, in the 1930s and ’40s, a cou­ple of husband-and-wife teams of psy­chol­o­gists tried to raise chim­panzees as much as pos­si­ble like hu­man chil­dren and teach them to speak. Their ef­forts were deemed a to­tal fail­ure,” Perl­man said. “S­ince then, there is an idea that apes are not able to vol­un­tarily con­trol their vo­cal­iz­a­tions or even their breath­ing.”

In­stead, the think­ing went, apes’ calls pop out al­most re­flex­ively in re­sponse to their en­vi­ron­ment—the ap­pear­ance of a dan­ger­ous snake, for ex­am­ple.

The vo­cal rep­er­toire of each ape spe­cies was thought to be fixed, Perl­man added. They did­n’t really have the abil­ity to learn new vo­cal and breath­ing-related be­hav­iors. These lim­its fit a the­o­ry on the ev­o­lu­tion of lan­guage, that the hu­man abil­ity to speak is un­ique among liv­ing pri­ma­tes (apes, mon­keys and their close rel­a­tives).

“This idea said there’s noth­ing that apes can do that is re­motely si­m­i­lar to speech,” Perl­man said. “And, there­fore, speech es­sen­tially evolved—com­pletely new—a­long the hu­man line since our last com­mon an­ces­tor with chim­panzees.”

But in a study pub­lished on­line in July in the jour­nal An­i­mal Cog­ni­tion, Perl­man and col­la­bo­ra­tor Na­than­iel Clark of the Un­ivers­ity of Cal­i­for­nia, San­ta Cruz, sifted 71 hours of vi­deo of Ko­ko in­ter­act­ing with Pat­ter­son and Cohn and oth­ers. 

They iden­ti­fied re­peat­ed ex­am­ples of Ko­ko per­form­ing nine dif­fer­ent, vol­un­tary be­hav­iors that re­quired con­trol over her vo­cal­iz­a­tion and breath­ing. These were learn­ed be­hav­iors, not part of the typ­i­cal go­ril­la rep­er­toire, the re­search­ers said.

Among oth­er things, Perl­man and Clark watched Ko­ko b­low a rasp­ber­ry (or b­low in­to her hand) when she wanted a treat, b­low her nose in­to a tis­sue, play wind in­stru­ments, huff mois­ture on­to a pair of glass­es be­fore wip­ing them with a cloth and mim­ic phone con­versa­t­ions by chat­ter­ing word­lessly in­to a tel­e­phone cra­dled be­tween her ear and the crook of an el­bow.

“She does­n’t pro­duce a pret­ty, per­i­od­ic sound when she per­forms these be­hav­iors, like we do when we speak,” Perl­man said. “But she can con­trol her lar­ynx enough to pro­duce a con­trolled grunt­ing sound.”

Ko­ko can al­so cough on com­mand—not par­tic­u­larly ground­break­ing hu­man be­hav­ior, but im­pres­sive for a go­ril­la be­cause it re­quires her to close off her lar­ynx.

“The mo­tiva­t­ion for the be­hav­iors varies,” Perl­man said. “She of­ten looks like she plays her wind in­stru­ments for her own amuse­ment, but she tends to do the cough at the re­quest of Pen­ny and Ron.”

These be­hav­iors are all learn­ed, Perl­man fig­ures, and the re­sult of liv­ing with hu­mans since six months after her birth in 1971.

“Pre­sum­ably, she is no more gift­ed than oth­er go­ril­las,” he said.

This sug­gests that some of the ev­o­lu­tion­ary ground­work for the hu­man abil­ity to speak was in place at least by the time of our last com­mon an­ces­tor with go­ril­las, he added, around 10 mil­lion years ago.

“Ko­ko bridg­es a gap,” Perl­man said. “She shows the po­ten­tial un­der the right en­vi­ron­men­tal con­di­tions for apes to de­vel­op quite a bit of flex­i­ble con­trol over their vo­cal tract. It’s not as fi­ne as hu­man con­trol, but it is cer­tainly con­trol.”

Orangutans have al­so dem­on­strat­ed some im­pres­sive vo­cal and breath­ing-related be­hav­ior, ac­cord­ing to Perl­man, in­di­cat­ing the whole great ape family may share the abil­i­ties Ko­ko has learn­ed to tap.

* * *

Send us a comment on this story, or send it to a friend

Sign up for


On Home Page         


  • Stu­dy: cat com­pet­ition has rocked dog fam­ily for ages

  • “Fire­works” be­gin on com­et un­der space­craft’s eye


  • Study links global warming, war for first time—in Syria

  • Smart­er mice with a “hum­anized” gene?

  • Was black­mail essen­tial for marr­iage to evolve?

  • Plu­to has even cold­er “twin” of sim­ilar size, studies find


  • F­rog said to de­scribe its home through song

  • Even r­ats will lend a help­ing paw: study

  • D­rug may undo aging-assoc­iated brain changes in ani­mals

The famous gorilla Koko is best known for having learned a good deal of American Sign Language from humans, as well as for understanding about 2,000 spoken English words. Although speaking skills are not for her, the simple sounds she does make are surprising, a new study said—and could change a perception that humans are alone in their capacity for speech among our close evolutionary relatives. In 2010, Marcus Perlman, now a researcher at the University of Wisconsin-Madison, started investigations at the Redwood City, Calif.-based Gorilla Foundation, where Koko has spent more than 40 years living immersed with humans. Koko has interacted for many hours daily with psychologist Penny Patterson and biologist Ron Cohn. “I went there with the idea of studying Koko’s gestures, but as I got into watching videos of her, I saw her performing all these amazing vocal behaviors,” said Perlman. The vocal and breathing behaviors Koko had developed were not necessarily supposed to be possible, he said. “Decades ago, in the 1930s and ‘40s, a couple of husband-and-wife teams of psychologists tried to raise chimpanzees as much as possible like human children and teach them to speak. Their efforts were deemed a total failure,” Perlman said. “Since then, there is an idea that apes are not able to voluntarily control their vocalizations or even their breathing.” Instead, the thinking went, the calls apes make pop out almost reflexively in response to their environment—the appearance of a dangerous snake, for example. The vocal repertoire of each ape species was thought to be fixed, Perlman added. They didn’t really have the ability to learn new vocal and breathing-related behaviors. These limits fit a theory on the evolution of language, that the human ability to speak is unique among living primates (apes, monkeys and their close relatives). “This idea said there’s nothing that apes can do that is remotely similar to speech,” Perlman said. “And, therefore, speech essentially evolved—completely new—along the human line since our last common ancestor with chimpanzees.” But in a study published online in July in the journal Animal Cognition, Perlman and collaborator Nathaniel Clark of the University of California, Santa Cruz, sifted 71 hours of video of Koko interacting with Patterson and Cohn and others. They identified repeated examples of Koko performing nine different, voluntary behaviors that required control over her vocalization and breathing. These were learned behaviors, not part of the typical gorilla repertoire, the researchers said. Among other things, Perlman and Clark watched Koko blow a raspberry (or blow into her hand) when she wanted a treat, blow her nose into a tissue, play wind instruments, huff moisture onto a pair of glasses before wiping them with a cloth and mimic phone conversations by chattering wordlessly into a telephone cradled between her ear and the crook of an elbow. “She doesn’t produce a pretty, periodic sound when she performs these behaviors, like we do when we speak,” Perlman said. “But she can control her larynx enough to produce a controlled grunting sound.” Koko can also cough on command—not particularly groundbreaking human behavior, but impressive for a gorilla because it requires her to close off her larynx. “The motivation for the behaviors varies,” Perlman said. “She often looks like she plays her wind instruments for her own amusement, but she tends to do the cough at the request of Penny and Ron.” These behaviors are all learned, Perlman figures, and the result of living with humans since Koko was just six months old. “Presumably, she is no more gifted than other gorillas,” he said. “The difference is just her environmental circumstances. You obviously don’t see things like this in wild populations.” This suggests that some of the evolutionary groundwork for the human ability to speak was in place at least by the time of our last common ancestor with gorillas, he added, around 10 million years ago. “Koko bridges a gap,” Perlman said. “She shows the potential under the right environmental conditions for apes to develop quite a bit of flexible control over their vocal tract. It’s not as fine as human control, but it is certainly control.” Orangutans have also demonstrated some impressive vocal and breathing-related behavior, according to Perlman, indicating the whole great ape family may share the abilities Koko has learned to tap.