Use of Human Language By Captive Great Apes

By Duane Rumbaugh and William Fields

This article appeared in the World Atlas of Great Apes and Their Conservation prior to bonobos arriving at Great Ape Trust. (2005, published in association with UNEP-WCMC by the University of California Press).

Some believe that the extinction of nonhuman great apes is preferable to preserving them forever in captivity, on the grounds that their nobility is diminished in artificial habitats. Others hold that great apes in captivity can lead happy lives, that the value of the preserved genetic material will prove to be very great, and that the human psyche would be significantly damaged by the loss of these species. This view embraces preservation strategies that create a diversity of niches for great apes that include the wild, zoos, reserves, refuges, sanctuaries, and even laboratories.

Chimpanzees and bonobos have lived in a captive research facility at Georgia State University in the USA since 1971, most notably sponsored by the work of Duane Rumbaugh and Sue Savage-Rumbaugh. This research has explored the mental abilities and cognitive character of great apes and in the process significantly changed our view of Pan and how these nonhumans might exist in human-modified landscapes. Two methods have been used to teach human languages to great apes: one uses sign language; the other, explored here, uses graphical symbols that represent words (lexigrams). The following is a brief account of the research initiatives of the Rumbaughs, the great apes that have participated in the research at the Language Research Center of Georgia State University, and the future plans for their lives in coexistence with humans.

The Corpus of Ape Language Research

LANA Project (1971-1976)

Lana is a female chimpanzee born in 1970 at the Yerkes National Primate Research Center. Her name derives from the LANguage Analogue (LANA) Project, which sought to develop a computer-based language training system in an effort to investigate the ability of chimpanzees to acquire language. Lana joined the research as a subject when she was two-and-a-half years old. The research was the first to interface a keyboard with a chimpanzee. At that time, it was believed that only humans could use symbols.

Lana demonstrated that she could discriminate between lexigrams and associate them with ideas. As she progressed, she would sequence words and use them grammatically, later starting to create novel utterances in response to unplanned events that affected her life. For example, Lana would request that the research technician refill her computer vending device when it was empty of treats, or request an item she had seen outside her room that the computer had no facility to provide to her. Lana exhibited language learning, and her experimental accomplishments were extraordinary. Equally important to her legacy is the lexigram keyboard, developed by Duane Rumbaugh, which has served as the primary communicative interface for ape language research at Decatur, Georgia, for the last several decades. This keyboard is composed of three panels with approximately 384 noniconic arbitrary symbols. When the apes depress a key, the word represented there is spoken by a digital voice and the lexigram is displayed on a video screen.

Sherman & Austin Research (1975-1980)

Sue Savage-Rumbaugh argued that the essence of language does not exist outside sociality and began working with two young male chimpanzees, Sherman, born 1973, and Austin, born 1974, using the LANA keyboard. The issue of human cuing was overcome experimentally by focusing on peer communication rather than that between experimenter and subject. The receptive component of language was featured in chimp-to-chimp communication, where they structured their interactions around statements of planned intent. Unlike Lana, Sherman and Austin could categorize, pretend, plan, comprehend, and respond to each other. Attending Sherman’s and Austin’s more complex use of language features was an increase in sociality and cooperation. Despite these achievements, Sherman and Austin did not comprehend spoken English. Austin died in 1998, but the other apes at the Language Research Center have not forgotten him; they still make reference to him using his lexigram, and they enjoy seeing videotapes of him.

Kanzi Research (1980-1993)

This was the first research initiative to use bonobos in language investigations. It began with a wild-caught female named Matata and her adopted son Kanzi, who was a nine-month-old baby playing in the lab while Savage-Rumbaugh tried to teach his mother language.

Kanzi was not a focus of the research because scientists thought him too young to learn these skills. When baby Kanzi was briefly separated by his mother, he began spontaneously to demonstrate productive competence for lexigrams and receptive competence for spoken English (something Matata had not achieved through direct training). Kanzi’s acquisition of productive and receptive competence emerged following passive observational exposure.

Later, as his language complex matured, Savage-Rumbaugh demonstrated that Kanzi’s utterances included grammar, syntax, and semanticity. It also seemed that his language skill enhanced his ability to learn other skills, such as the manufacture of Oldowan-type rock tools. Kanzi’s receptive competence for spoken English contrasted dramatically with the failure of the chimpanzees Sherman and Austin to do likewise.

What was the basis for the difference between the language skills displayed by bonobo Kanzi and the chimpanzees? Savage-Rumbaugh had clearly demonstrated in Kanzi that language could be acquired spontaneously and observationally without planned training; that comprehension precedes production and drives language acquisition; and that early exposure to language can greatly improve the level of competency attained.

Panpanzee & Panbanisha (1986-1990)

Considering the question of receptive competence for spoken English, Savage-Rumbaugh proceeded to investigate the questions of species variables in a co-rearing study of a chimpanzee and a bonobo. In this study, Savage-Rumbaugh hoped to have Kanzi’s mother Matata raise chimpanzee Panpanzee and bonobo Panbanisha in identical environments. They were born within three weeks of each other. While Matata took good care of both babies, she would only allow Panbanisha to nurse. At that point, Savage-Rumbaugh and her human colleagues assumed the rearing of both babies until they were four years old. Based upon this study, Savage-Rumbaugh determined that the failure of the chimpanzees Sherman and Austin to comprehend spoken English is not a species-specific variable, as both Panpanzee and Panbanisha developed receptive competence for English.

Panbanisha & Kanzi (1990-Present)

While at the Language Research Center at Georgia State University, Kanzi lived in a bonobo community at a facility that includes a 50 acre wooded forest within a 300 acre woodland preserve. The bonobos spent as much time outdoors traveling and communicating as they did indoors with computers and joysticks. Locations in the forest at the LRC were named with lexigrams, and the bonobos know the forest as well as humans might know their own village. The bonobos are able to plan where they will go and what they will do when they get there, and they talk about these plans on the communication boards.

Kanzi and Panbanisha continue to expand their linguistic world with music, art, writing, tool making, and tool using. Savage-Rumbaugh documented on film Kanzi’s ability to ‘rock knap’, breaking off flakes of stone to produce functional cutting tools as taught by archeologist Nicholas Toth; Panbanisha’s ability to write lexigrams on the floor with chalk; and the ability of both to participate in musical performances with musicians.