Kanzi: The Ape at the Brink of the Human Mind

by Sue Savage-Rumbaugh

  At least part of Yerkes’ instinct had been correct: An ape could be taught to make and use signs in a simple context. Typically, that context involved requesting some item of food or drink or initiating play—chimpanzees love to chase and tickle. The intuition that chimpanzees “have plenty to talk about” was, however, less substantiated by the Gardners’ experience with Washoe. True, food and play are important in chimps’ lives, particularly as youngsters. Using ASL in this context, therefore, may accurately reflect what is on the animal’s mind. But Washoe’s use of symbols did not give insight into what it is like to be a chimp, which was what many people longed for.

  Nevertheless, Washoe’s acquisition of so extensive a vocabulary of signs was a great achievement, one that seemed to indicate that a chimp could break the human language barrier. For linguists, however, the quintessence of language has become syntax, the underlying structure that orders the utterance of words and imposes overall meaning. The singular focus on syntax is the result of Noam Chomsky’s dominance in the field, and his position is simple: Without syntax, language, as we know it, cannot exist. In order genuinely to break the language barrier, chimpanzees therefore would have to demonstrate syntactical competence.

  During the early 1970s several more ape-language projects were established, inspired by the apparent success with Washoe. At the University of California, Santa Barbara, David Premack taught a female chimpanzee, Sarah, to use plastic shapes as “words.” Sarah did not communicate with these shapes, but rather used them to answer specific questions. The questions aimed to determine whether she possessed the cognitive “functional prerequisites” of language competence. For example, Premack tried to determine if she could respond to such typical linguistic structures as negation, class concept, and characterizations of change of state.

  Duane Rumbaugh, of Georgia State University in Atlanta, began the LANA (LANguage Analogue) Project at the Yerkes Primate Center. The project, which was closely tied to developing methods for teaching language to severely mentally retarded children, invented a computerized keyboard display of arbitrary signs, known as lexigrams.

  Explicitly or implicitly, each of these projects accepted the challenge to demonstrate syntactical competence as a criterion of language competence. Although the details of the results of these projects differed in many ways, some of which were mutually contradictory, there developed a strong conviction during the early 1970s that not only could apes learn symbols, but they also could use them in innovative and structured ways. For instance, the Gardners said of Washoe that she “learned a natural human language and her early utterances were highly similar to, perhaps indistinguishable from the early utterances of human children.”2

  Most of the utterances by the ASL-using apes involved single symbols, but there was a sufficient number of plausible two-and three-word combinations to encourage a sense of rudimentary language. And among these multisymbol utterances there was perceived to be a sufficient degree of structure—with appropriate order for verbs, nouns, and qualifiers, for instance—to encourage a sense of humanlike syntax. It was against this background of rising optimism that I entered Roger Fouts’s ape-language project at Oklahoma. After about six months Roger suggested I work with an adolescent female chimp called Lucy. Lucy, who was older than the chimps I’d been working with, lived with clinical psychologist Morris Temerlin and his wife, Jane, in their home not far from the institute. She was being raised in a human home, from birth, as a subject in Bill Lemmon’s cross-species rearing experiments. The purpose of these studies was to “determine whether or not maternal behavior was innate.” Lemmon, a clinical psychologist and Temerlin’s mentor, had become intrigued by the differences between the sexes. In his view, females were biologically programmed to fulfill deep-seated maternal urges. He was therefore interested in determining whether or not these urges were innate in man’s closest living relatives.

  Lemmon approached this question by placing infant female apes in the homes of his patients and former students. He then instructed the patients to rear the ape as if it were their own child. When the apes reached adulthood, he intended to artificially inseminate them and observe how they cared for their infants; he wanted to see how these apes behaved as mothers, having been reared themselves by a “mother” who loved them, but did not behave as an ape. During the rearing of these female apes, he also used the apes’ behavior as a means of “evaluating” his patients, whom he saw on a regular basis.

  Lucy was but the first of a number of such “human-reared” apes I was to meet at the Oklahoma Primate Institute. In order to obtain chimpanzees for his studies, Lemmon had developed his own breeding colony of apes. Thus there were many chimpanzees living in social groups who had minimal contact with human beings, apart from the essentials of being fed and having their quarters cleaned. Infants who were born in and remained in these groups were quickly pointed out as being “less intelligent” and “retarded in their development,” as contrasted with those reared in human homes. At the time, I did not realize that nearly all my energies for the next years would be poured into trying to understand the differences in behavior and development of these group-reared infants and their human-reared counterparts. The differences were remarkable, but “intelligence,” I was soon to learn, is probably the most elusive and detrimental concept prevalent in the conceptual toolkit of modern psychology.

  When I first met Lucy she seemed to be more attentive to social interactions with humans and to give crisper signs than the apes in the institute’s colony. When Roger introduced me, Lucy retrieved a plastic flower from her box of playthings and offered it to me, just as human children of one or two years do. Object offering passes through a preliminary stage in human children in which the object is first offered and then withdrawn before it is actually taken by the other party. A few months later, the child clearly seeks to transfer objects from his or her possession to that of another.

  When I reached out to accept the plastic flower from Lucy, thinking how nice it was of her to offer, she deftly sank her teeth into the back of my hand. Chimpanzee offers, it seemed, were not quite like human offers. I was only later to realize that rather than attempting to give me the flower, Lucy was daring me to take it, albeit in a rather deceptive manner. I would fall for that trick again many times in my ape career, as chimpanzees are of sufficient intelligence to disguise the trick in many ways.

  Why should Lucy want to bite me? Obviously I had not tried to bite or hurt her in any way. I just wanted to learn about her. Little did I understand how my presumption in entering her home with no proper greeting or explanation of my mission had been rude from her perspective. Nor did I realize that extending the object without the appropriate facial expressions and vocalizations was a way of determining how well I could judge her intentions. It reminded me of going into other cultures, or even distinct subcultures in the United States, where attempts are made to size up what you know by seeing if you fall for the oldest trick in the book. I flunked the test.

  Despite the tenor of our introduction, Lucy and I went on to develop a good relationship. That was because Roger advised me that the best way to make friends with chimps was simply to spend time alone with them. Lucy at the time weighed about seventy pounds and was rather intimidating. I wasn’t quite sure what she might do to me when we were alone, as pound for pound, apes are five times as strong as a human male in excellent physical condition. Moreover, their teeth are large and their jaws are able to exert enormous pressure. Since I had already had the end of my finger bitten off by a chimpanzee at the institute, I was a little hesitant to be left alone with Lucy. Yet Roger assured me that this was the best method and that Lucy was far less likely to harm me if I was alone with her than if someone she knew were also there. I did not understand this, as it seemed to go against all reason. Why would a chimpanzee be more likely to hurt a stranger such as myself when someone that it already knew and liked was present? Nonetheless, I accepted the advice and agreed to take Lucy out of her c
age, bring her into the main portion of the house, and begin to teach her signs.

  It was not as difficult as I had expected. Soon I was even taking Lucy for rides in my two-seater MG all over Norman, Oklahoma. Lucy would point in a certain direction and we would go that way. Sometimes, if I refused to go to a place she really wanted to see, she would take the steering wheel away from me and turn the corner herself. Of course, this could be dangerous, so every moment Lucy was in the car I was always fully prepared to stop immediately in case she decided to drive. She was not permitted to get out of the car, except when we drove to a sixty-acre plot of land outside of Norman, which was owned by the Temerlins. There Lucy could bound out and run free.

  I continued to work with Lucy for two years, while also beginning a behavioral project in which I observed four mothers and their infants. I was interested in infant development—the stages of maturation through which youngsters pass. The language project fascinated me, but I recognized that I could not understand what chimpanzees were doing with signs until I first understood chimpanzees far better. What kinds of things did they communicate spontaneously to each other and how did they do so? Were their nonverbal communications simply unconscious emotional expressions as all the literature at the time maintained? If so, why were their signs suddenly “expressions of conscious willed intent?”

  It made no sense to me that if you held up an object and taught a creature to place its hands in a certain position when you did so, that somehow this procedure led to the magic of symbolic awareness and the ability deliberately to communicate ideas and thoughts to others. I felt that capacity had to be present already in the creature. Additionally, perhaps because I had a two-year-old son of my own at the time, I was beginning to recognize that something seemed to be missing during my attempts to communicate with the signing apes. I recognized that they could successfully request objects using appropriate signs, but I began to be uneasy about how much they actually comprehended. The unease would always emerge when I tried to engage in true communication, that is, when I asked them a question for which I did not already know the answer.

  This missing component could similarly manifest itself whenever I asked the chimps to do simple things, such as to hand me a familiar object. Unsure how to respond, they would often begin to sign back rather than try to do what I asked. Similarly, if I asked them a question, such as “Where shall we go?” or “What shall we do?” they would frequently string together a variety of symbols they knew, particularly ones we had been using recently, in the apparent hope of hitting on one that sounded good to me. It was impossible to avoid the overriding impression that my son was far more deliberate in his attempts to communicate and that his understanding of simple requests and questions went considerably beyond that of the apes. The apes often seemed not to realize that they were being asked a “true” question, that is, an open-ended one that they could answer in any way they wished.

  Of course, they were queried all the time with questions like “What’s this?” “Who’s that?” “Where’s X?” but these questions always had a “correct” answer. That is, if asked the name of a person, the chimps were expected to produce the correct name. Questions such as these revealed little about what the chimp itself wanted or thought; they were simply rhetorical questions that required a signed response.

  I began to worry about what this meant, and tried to talk with Roger about it on several occasions. Roger was never unwilling to talk about such issues, but he seemed to be of the opinion that it was more or less impossible to “get inside” a chimp’s mind and therefore the only reasonable thing to do was to take the sign at face value and focus on questions that had clear answers. Scientifically, I had to agree with his stance, yet I knew that if I regularly took such an approach to my son’s language, our communication would soon cease to be a very satisfying affair for either of us. It was important that he understand the things I was trying to tell him, and that he express his thoughts, rather than just answer questions designed to test his knowledge.

  Then Pancho came along. Pancho’s imminent arrival at the institute caused great excitement because he was said to be a pygmy chimpanzee (Pan paniscus), the rare and exotic cousin of the common chimp (Pan troglodytes). Pygmy chimps, or bonobos as they are known, are more humanlike than common chimps in many ways, including being more vocal and more communicative and having extremely expressive humanlike faces. They are also less aggressive than common chimps and tend to be very friendly toward human beings, whom they seem to have a remarkable ability to relate to.

  Therefore, even though Pancho was an adult male whom no one at the institute knew, I felt confident in taking Pancho for walks around the institute’s farm, and I did most of my observations of him outdoors. We even took rides around Norman, as I did with Lucy, and we stopped for root beer, hamburgers, and fries at an A & W. Unlike being with Lucy, I never worried about Pancho grabbing anyone who approached the car, nor about him taking the steering wheel. Often I even took my young son along; Pancho was such a gentleman.

  Looking back, it was a crazy thing to do—and probably illegal—but I developed a very strong sensitivity to Pancho’s skills and mine in our social and communicative interaction.

  After six to eight months we discovered that Pancho was not a bonobo after all, but a Koola-kamba, which some have suggested may be a naturally occurring chimpanzee-gorilla hybrid, or a chimpanzee-bonobo hybrid. Bonobos and gorillas have many physical similarities (small ears, raised nose, large abdomen, shorter toes, for example). The fact that Pancho had been misidentified for so long is indicative of how few primatologists are familiar with bonobos. Had I realized Pancho’s true identity, I’m sure I would have been less willing to be so free with him, thinking he might be aggressive. Fortunately, I had experienced no problems with him. The most significant aspect of our interaction, however, was this: Despite Pancho’s lack of signing ability, I could communicate with him as easily and as extensively as I could with Lucy. If learning signing is about language, I mused, and language is about communication, why couldn’t I communicate with Lucy more effectively than with Pancho?

  When I discussed these issues with Roger—about the apparent lack of comprehension and real communication—he insisted that Washoe understood what was said to her, and the younger apes would develop more and more comprehension as they got older. Okay, I responded, can you demonstrate that Washoe really understands requests that are made of her? Certainly, said Roger, as we talked sitting on the bank across from the chimp’s island one day. He turned to Washoe, looked around the island, and noticed that a long rope lay near the center of the island. Washoe, on the shore, was looking up at us. Roger turned to Washoe and signed, “Washoe, go get string there.” He gestured in the direction of the string. Washoe looked puzzled, but did begin walking in the direction that Roger had pointed. She looked at a variety of things on the island, touching them and looking back at Roger, as if trying to determine what he meant. She walked right past the string several times and each time Roger signed, “There, there, there (again pointing), there string.” Finally, as she again approached the area where the string lay on the ground, Roger began to sign “yes, yes, yes” and nod his head emphatically. As Washoe reached the spot, she picked up the piece of string and was praised fulsomely. “See,” said Roger, “she just had trouble finding the string.” I was not convinced.

  I soon became known at the institute as “the unbeliever.” It was a very friendly jest, and we all talked a great deal about my concerns; in fact, I was actually happy to be so labeled. When I compared my two-year-old’s language competence with the apes’, I still saw discrepancies that made me even more unsure about the strong claims that increasingly were being made for language competence in apes.

  First, I didn’t have to drill object-sign associations with my son, Shane. Words just popped into his vocabulary. Second, I didn’t have to stretch my imagination to understand most of the things he said. They were obvious from the context. I had bec
ome uncomfortable and suspicious of some of the rich interpretations of apes’ utterances I’d heard or read. No juxtaposition of words was deemed too strange to be interpreted as a reasoned utterance. For instance, the suggestion that the gorilla Koko was making puns and other kinds of word play, and had a concept of death, strained my credulity—even though I wanted to think that an ape was capable of such abstract conceptions. Third, Shane clearly understood more of what was said to him than the apes did. I could ask him to do simple tasks, and he would. When I asked a chimp to do a simple task, even as simple as picking up a specific object in front of her, there was often puzzlement—just as Washoe had experienced with finding the string. It was clear to me—as it is clear to any parent—that Shane’s ability to comprehend language developed ahead of his ability to produce language. This seemed not to be the case with chimps.

  I began to form the notion that comprehension, not production, was the central cognitive feature of language, particularly language acquisition. Comprehension is much more difficult to quantify than production of words, and so linguists had paid little experimental attention to it. In any case, the hegemony of syntax in linguistics ensured that production was held to be the defining characteristic of language competence. For the most part, ape-language researchers accepted what linguists said, and then strained to satisfy their criteria. Seeing this, I became discouraged at the prospects of moving ape-language research forward, and opted instead to devote the rest of my time at Oklahoma to studying infant development. I wanted to document as extensively as I could, the type of things that apes could communicate by using their accepted nonverbal system of glances, gestures, postures, and vocalizations.