Page 72 - Encyclopedia of Philosophy of Language

Language and Mind
munication with greater frequency than does the com- mon chimpanzee. They also tend to be more affiliative. Both the bonobo and common chimpanzee maintain strong social, though somewhat unique, bonds within their group.
Matata, a wild-bora female that dwelled in the for- est until an estimated age of six years, was the first bonobo introduced to the lexigram system. Methods that had proved successful with both Sherman and Austin in the majority failed with Matata. At best, Matata learned only eight lexigrams and, then, only how to use them to request that things be given to her. Even after years of effort, Matata failed to use even these few lexigrams reliably and gave no evidence that any of them functioned as symbolic representations of their individual referents. Nevertheless, Matata otherwise appeared to be a very bright chimpanzee.
During her lexigram training, Matata had her adopted son, Kanzi, with her. No systematic effort was made to teach Kanzi lexigrams while work with Matata was underway. He was always present, however, whenever Matata was worked with. Conse- quently, he alwayshad the unintended opportunity to observeher training.
When Kanzi was two and a half years old, he was separated from Matata when she was taken to another site to be bred. Only then did it become obvious that something unexpected had happened. Kanzi had learned the symbols that experimenters had been attempting to teach Matata. He needed no special training to use symbols to request that things be given to him, to name things, or even to announce what he was about to do.
As a consequence, a significant change was made in research tactics. Structured training protocols were terminated, and Kanzi was introduced to a life in which throughout his waking hours he was part of a social scene. People talked in English to Kanzi and touched the appropriate lexigrams on his keyboard whenever they coincided with words of spoken utter- ances. Through continued observation of others' use of language Kanzi also learned language and its func- tions in the real world. Kanzi was encouraged to listen to speech and to observe others; however, he was never denied objects or participation in activities if he did not use his keyboard. Caregivers used the key- board to comment on events (present, future, and past), to communicate with each other concerning their intentions or needs, and, in particular, to talk about any thing that appeared to be of interest to Kanzi.
Kanzi quickly learned to ask to travel to several of the named sites throughout 55 acres of forested land that surrounds the laboratory. He learned to ask to play a number of games, to visit other chimps, to get and/or prepare and even cook any number of specific foods, and to watch television. Kanzi's lexigram vocabulary increased to 149 words by the time he was
five and a half years old. It was only a short time before it was also noted that, in contrast to Lana, Sherman, Austin, and Matata, Kanzi appeared to be comprehending human speech—not just single words, but sentences as well.
As a consequence, an experiment was undertaken to assess Kanzi's speech comprehension and to com- pare it with that of a human child, Alia (Savage- Rumbaugh, et al. 1993). Under controlled conditions, designed to preclude inadvertent cueing as to what should be done, both Kanzi and Alia were given 660 novel sentences which requested them to do a variety of unusual things. For example, they were asked to take a specific object to a stated location or person ('Take the gorilla [doll] to the bedroom'; 'Give Rose a carrot'), to do something to an object ('Hammer the snake!'), to do something with a specific object relative to another object ('Put a rubber band on your ball'), to go somewhere and retrieve a specific object ('Get the telephone that's outdoors'), and so on.
An everchanging and wide variety of objects was present, and the subjects were asked to act upon mem- bers of each array in a variety of ways. This practice ensured that 'compliance with a request' was not sim- ply a result of a given subject doing whatever was obvious on a given trial. Thus, when asked to 'Get the melon that's in the potty,' a second melon was on the floor—evenin Kanzi's path to the potty; and when asked to 'Get the lettuce that's in the microwave (oven),' Kanzi found not only lettuce, but a variety of other things in the oven as well.
At the time of the study, Alia was two and a half years old and Kanzi was about nine years old. Inter- estingly, Kanzi's comprehension of novel spoken sen- tences was quite comparable to Alia's. Both subjects were about 70 percent correct in carrying out the sentences of novel requests on their first presentation.
By the time Kanzi was eight years old, his pro- ductive competence of lexigrams with gestures was comparable to an 18-month-old child. His lexigram vocabulary consisted of well over 250; his com- prehension of spoken English was commensurate with that of a two and a half-year-old child.
Kanzi was the first ape that had acquired language skills without formal training programs. He developed language by observing and by living. The pattern wherebyhedidsoparalleledthatofthenormalhuman child who, first, comes to comprehend the speech of others and then subsequently talks.
Was Kanzi an exceptional bonobo? Or, could his skills of comprehension be replicated with other chim- panzees?
To answer this question, Kanzi's younger sister, Mulika, was exposed to the same kind of linguistic environment that Kanzi had, but was introduced to it at a much earlier age. She, too, first developed com- prehension of spoken English and lexigrams as used by others and then began to use them productively.
50