Click here to see the table of contents


Are Nonhuman Species Capable of Language Acquisition?


Promising results from early animal language studies were discarded after revelations of methodological problems were uncovered. This lent evidence to a biological explanation of the development of language, championed by some linguists and psychologists, who claim that linguistic skills are uniquely human. Later research, carefully designed to overcome procedural flaws has refuelled the debate, by providing evidence for the evolutionary explanation of language development, suggesting that a graduation of linguistic capabilities exists in human and nonhuman primates. Researchers and critics alike are cautious in evaluating these later studies. Many are focusing on the questions that arose from the language studies of nonhuman primates, concerning cognition and intelligence, and their relationship to linguistic ability.

Conventional biological wisdom suggested that communication in nonhuman animals was reflexive (Long, 1994) and limited to signals warning of danger, defining territorial boundaries and courtship behaviour (Nollman, 1994). Defining language as a system of spoken sounds and written symbols subject to rules of grammar, early cognitive scientists failed to uncover evidence of syntax, the rules for word order in sentences, inflection or vocabulary in animal communication and concluded that outside the human realm, language was nonexistent (Nollman, 1995).

Although animals communicate through cries, gestures and mating calls (Coon, 1992) that are understood by members of the same species, in the wild, even chimpanzees make only a few dozen distinct cries, limited to social interactions, but do not relate information about anything not immediately present (Savage-Rumbaugh, Rumbaugh and Boysen, 1980).

When the question of whether animals could be taught language skills arose, pioneers in simian linguistics actually attempted to teach a chimpanzee to talk, but six years of research resulted in a vocabulary of only four words.

In 1966, Beatrice and Allen Gardner perceived that chimpanzees' difficulty in language acquisition was not stupidity, but "inability to control lips and tongue" (Patterson, 1978). Chimpanzees also lack the vocal apparatus necessary in the production of modulated sound (Johnson, 1995) Using operant conditioning and imitation, they taught a chimpanzee called Washoe to use American Sign Language.

In the late 1970's, Washoe and Nim Chimpsky, an ape taught by Terrace, were "exposed as unintentional frauds" (Johnson, 1995) by researchers who detected that they were simply pleasing their teachers by moulding their hands into the hand signs (Craighead-George, 1985). These findings, combined with the revelation that the majority of the data obtained from project Washoe was "observational, interpretative and potentially open to cuing" (Seidenberg and Petitto, 1979, cited in Savage-Rumbaugh, Rumbaugh and Boysen, 1980), plunged simian linguistic research into disrepute.

The arena of animal language studies was resurrected by Savage-Rumbaugh and Rumbaugh, who recognized the methodological faults of the earlier trials.

The Gardners had equated language production with comprehension. Although true for normal human infants, this idea cannot be extrapolated to chimpanzees (Savage-Rumbaugh, 1979, cited in Savage-Rumbaugh, Rumbaugh and Boysen, 1980). Savage-Rumbaugh et al recognized the relative ease with which chimpanzees acquire skills, superficially resembling language while have difficulty achieving "symbolic communication ability and the reciprocal receptive behaviours that reflects the essence of language" (Savage-Rumbaugh, Rumbaugh and Boysen, 1980).

Starkly contrasting with early child-adult exchanges, most chimpanzee communication lacked "conversational turn taking" (Savage-Rumbaugh, Rumbaugh and Boysen, 1980) indicating that the simians may not have comprehended the representational significance of the symbols they learnt to use. Savage-Rumbaugh began by differentiating language production from comprehension, emphasising that the crucial precursor of language competence is the understanding of non spoken referential symbols. She focused on the meaning that the apes derived from words rather than how they produced them. Savage-Rumbaugh saw comprehension as "the route into language" (Johnson, 1995) since it is simpler to translate an idea in one's mind into a grammatical sequence of words than to decode a sentence spoken by another, whose intentions are unknown.

Savage-Rumbaugh's initial research with Sherman and Austin, demonstrated that these two male chimpanzees were able to convey information to one another using and to take turns in communicating during food sharing activities, which is rare in the wild. These findings highlighted the difference between language production and comprehension, suggesting to Savage- Rumbaugh that emphasis on the latter might facilitate further linguistic development in apes (Mitani, 1995).

Eliminating the problem of ambiguous hand signs (Johnson, 1995), and the possibility that researchers were unconsciously cuing their subjects and over interpreting data, Rumbaugh devised Yerkish, a language based on Chinese lexigrams, and programmed it into a computer, which the chimp communicated with, via a keyboard (Craighead-George, 1985).

The adroit selection of the appropriate lexigrams in correct syntax by a chimp called Lana, as well as the simian's desire for new lexigrams for items not originally programmed into the computer, led to Rumbaugh's confident conclusion (1984) that apes learn language in a meaningful way (Craighead-George, 1985).

Later work with Matata, a bonobo, inadvertently exposed the ape's young, adopted son, Kanzi to the artificial language. It was discovered that Kanzi understood and spontaneously produced a limited number of lexigrams without training. This lead to a shift away from intensive training sessions to an approach in which Kanzi was treated as a "developing human infant" (Johnson, 1995). The findings of this project, in which Kanzi learnt to appreciate word order and other syntactical cues, led Savage-Rumbaugh to conclude that bonobos possess "rudimentary syntactical ability." (Johnson, 1995) This apparent success of research with Kanzi was attributed to her exposure to language early in her life and tutoring instigated by the animal's curiosity (Johnson, 1995).

Critics responded that animals associating vocal sounds with objects was far from revolutionary (Johnson, 1995). However Savage-Rumbaugh argued that trials with words in "novel contexts" (Johnson, 1995) revealed that the chimpanzees' responses were not reflexive.

Savage-Rumbaugh reported that her chimpanzees "demonstrate the rudimentary comprehension skills of two and a half year old children," understanding complex sentences and spontaneously using symbolic language in their communication.

However for many linguists, the hallmark of language is not comprehension but performance, the ability to generate increasingly complex sentences within the confines of grammar (Johnson, 1995).

The view that mental experiences are not comparably shared between humans and nonhuman primates is prevalent among linguists (Mitani, 1995) and has been enforced by Noam Chomsky's arguments that conscious thought is only made possible by the innate mechanism to decode the syntactical structure of language (Mitani, 1995). He adheres to the view that the grammar, present in all languages, has its origins in the neural connections, unique to human brains.

Savage-Rumbaugh, subscribing to the alternate evolutionary explanation of language, believes that other animals experience similar states to humans. Shanker observed (Johnson, 1995) that it is this belief that has set her work apart from the earlier attempts at eliciting evidence of language skills from nonhuman primates.

Shanker, critical of the linguists continual refinement of their requirements of language ability in ape studies, insists that Savage-Rumbaugh's research suggests that there is not an unbridgeable chasm between human and nonhuman primates' language ability, but a graduation of linguistic ability (Johnson, 1995).

Shanker believes (Johnson, 1995) the linguists objections disclose a naive understanding of the process of language. He points out that although Kanzi initially relied on subtle gestures and contextual cues from the speaker, language is a "social act… always embedded in a situation" (Johnson, 1995) and humans use these cues continuously.

Of all the captive simian language studies, the anecdotal evidence from research involving Koko, the gorilla, has come closest to revealing the capabilities that apes possess in language acquisition skills.

Although no methodical comparisons of linguistic research with the two species of primates has been undertaken, anecdotal evidence suggests that the "pace and clarity" (Linden, 1986) of gorilla signing is superior to that of the chimpanzee and may lead to a clearer, less ambiguous interpretations of meaning.

Linden observed (1986) that "Koko seemed… more comfortable with the act of signing" and had "integrated sign language… into her life." Researchers noted that Koko and her simian companion Michael, "regularly converse in sign language (Patterson, 1978) and Koko often engages in "private monologues" during relaxation periods (Patterson, 1978).

Unfortunately methodological problems have prevented the findings from attracting scientific endorsement. Although much anecdotal material had been collated, the lack of technological resources available to project Koko meant that much of the data that Koko provided was lost and could not be used for the basis of a scientific evaluation. The secrecy with which Patterson and Cohn conducted their research with Koko did little to elicit scientific support and fuelled rumours that Koko's abilities were an elaborate hoax.

However Patterson herself acknowledged that her results were difficult to support empirically (Patterson, 1978) but remains convinced of the linguistic abilities of the gorillas in her charge. She investigated the extent of Koko's understanding of their exchanges, and found that Koko used her vocabulary not only to label the objects and events of her immediate world, but could frame propositions, recreating past events (Patterson, 1978). Koko referred to events removed in time and space, the sophisticated characteristic of human language, displacement. Patterson also found abstractions such as imagine, understand, curious, idea, gentle and stupid within Koko's repertoire. An unexpected insight into Koko's use of language came when the gorilla learnt to lie, exploiting language to distort her keeper's perception of reality (Parker, 1995).

When questioned if she was a human or a gorilla, Koko's response "Fine animal gorilla," has prompted arguments that Koko is conscious, aware of herself, her thoughts and her use of language (Craighead-George, 1985).

Terrace however, believes that the animal language studies are misguided (Johnson, 1995). He is more concerned with how chimpanzees and gorillas think without language, since the complicated cognitive processes that are being studied in the trials reveals "a lot about the evolution of intelligence." He places these questions ahead of attempting to elicit "titbits of language" from apes, preoccupied with performing to obtain rewards.

An evolutionary question that naturally arose from the ape language studies concerns the effects that language has on a species. In the instance of chimpanzees, Rumbaugh provides the insight that the chimpanzees who had acquired language skills showed more effective and communicative behaviour as well as more insight in learning and problem solving and exhibited a less destructive demeanor (Craighead-George, 1985).

Extrapolating these observations to the role of language in human evolution, however, must be approached with caution.

The disappointments of the early ape language studies elevates the apparent successes of Savage-Rumbaugh and Patterson's work to greater heights. Although their findings have important implications for teaching linguistic skills to retarded humans as well as revealing the evolutionary and biological basis for human language, it is important to temper ideological concerns with realism and refrain from over interpreting the results. Many linguists continue to cling to a view that does not allow them to accept the results of the Rumbaughs' and Patterson's research. However, an increasing number of cognitive psychologists are applauding these reports of apes' advanced cognitive and rudimentary linguistic abilities.


Coon, D. (1992). Introduction to Psychology. St. Paul: West Publishing.

Craighead-George, J. (1985). How to talk to Your animals. London: Hodder and Stoughton.

Johnson, G. (1995). Chimp Talk Debate: Is It Really Language? The New York Times, June 6.

Linden, E. (1986). Silent Partners. New York: Times Books. pp.115-129.

Long, C. (1994). Language and Communication. The Interspecies Newsletter, Summer.

Mitani, J. (1995). Review of Savage-Rumbaugh and Lewin's "Kanzi: The Ape at the Brink of the Human Mind." Scientific American, Volume 272, Number 6.

Nollman, J. (1995). Who Communicates. The Interspecies News-letter, Winter.

Patterson, F. (1978). Conversations with a Gorilla. National Geographic, October.

Savage-Rumbaugh, S., Rumbaugh, D. M., & Boysen, S. (1980). Do Apes Use Language? American Scientist, 68, 49-61.

Vessels, J. (1985). Koko's Kitten. National Geographic, January.

please click to return to the table of contents 

Copyright © 1996 Ilanit Tof, All Rights Reserved.