(Note the reviewer, Dr. S. Goldin-Meadow is widely known for her
research on use of sign language and gestures by young children.)
Goldin-Meadow, S. (1996). Book review of Kanzi: The ape at the brink of the human mind by Savage-Rumbaugh, S. & Lewin, R. International Journal of Primatology, 17, 145-148.
Kanzi: The Ape at the Brink of the Human Mind
By Sue Savage-Rumbaugh and Roger Lewin, John Wiley and Sons, New York, 1994, xvii +299 pp., $24.95 (hardback).
What if a chimpanzee were raised with all the comforts and artifacts of a human society? Would that chimp learn language as a human child does? At one point in the history of the teaching-language-to-chimps studies (when Washoe, Nim, and Sara, the best-known of the language-learning chimps, were young), the answer would have been a confident "yes." Then came the 1980's when criticisms from many quarters led to a reevaluation of that "yes." The critiques came from linguists who claimed that the accomplishments of the chimps fell far short of human language, from ape-language researchers who claimed that even these minimal accomplishments were more apparent than real (reflecting the way that the humans structured interactions for the chimps rather than the chimps' own communicative abilities), but also from the author of Kanzi, Sue Savage-Rumbaugh, who calmly pointed out that the evidence that the chimps were using their signs as symbols was thin. Since then, when the vehement criticisms from the first two camps had all but destroyed public support for, and interest in, teaching language to chimps, Savage-Rumbaugh has methodically gone about trying to convince herself (and, in the process, others) that the chimpanzee can use symbols to communicate.
Kanzi: The ape at the brink of the human mind is convincing, up to a point. Written for the layperson, the book is not a research report of the many excellent studies Savage-Rumbaugh has done (they are mentioned in the text and references are given) but rather a well-written story of how she has attempted to break down the wall that exists between human and chimp (on one side of the wall we have humans who are proficient language-users, on the other side chimps who are said to be language-less). In conjunction with Roger Lewin, a science writer, Savage-Rumbaugh recounts how she became interested in the chimp projects first at Oklahoma and later at Yerkes, and the inroads she has made in teaching the chimp about language. One of the most interesting chapters is the account of how Sherman and Austin, two common chimps who were originally both incapable of generalizing a "word" beyond the situation in which it was taught, subsequently learned how to spontaneously communicate requests to each other using words (actually a point at an arbitrary symbol standing for an object). The chapter not only provides evidence that these words had begun to function as symbols rather than mere associations for Sherman and Austin, but it also lays out the kind of laborious training needed to get the chimps to this point.
This effortful training stands in contrast to Kanzi's education. Kanzi is the star of the book and is a bonobo (pygmy) chimp raised from birth in an environment where spoken words and the language board (arbitrary symbols to which one can point to convey an object or action) were spontaneously used to communicate with him. In response, Kanzi began to use the language board to communicate with his human companions. Kanzi's accomplishments are impressive and, given Savage-Rumbaugh's sensitivity to criticisms raised about the earlier chimp work and caution in making claims in the past, are not likely to be figments of the human observer's imagination.
Understandably given that Savage-Rumbaugh felt herself under seige from all camps, the book focuses only on what Kanzi can do and does not dwell on what he can't (or at least, doesn't) do. Nevertheless, it is important to keep in mind that there are differences between Kanzi and a language-learning child. For example, many of the illustrations of Kanzi's communications presented in the book are comments on aspects of his world. However, what we learn from the research publications (e.g., Greenfield & Savage-Rumbaugh, 1991) is that comments account for only 4% of Kanzi's communications. In other words, almost all of his communications come about because he wants something done, not because he wants to chat. Although undoubtedly real, these few comments do not appear to be the norm for the chimp, and their strikingly low frequency stands in contrast to young human language-learners who use their language at least as often to make conversation as to make requests.
Another of the significant differences between child and chimpanzee (even a bonobo) is that, while chimps appear to require a great deal of linguistic input to develop language, human children, even if lacking a model for language altogether, will actually invent a language to communicate with those around them. Deaf children, whose hearing losses prevent them from acquiring speech and whose hearing parents have not yet exposed them to a sign language, use gestures to make requests as well as to comment on the past, present, and future and even on their own gestures. Moreover, the children's gestures are structured at both the word/gesture and sentence levels and have grammatical categories, similar to the early sentences of child language Goldin-Meadow & Mylander, 1990). This difference between child and chimp is particularly important when making an evolutionary argument, since it is the ability to invent a system (and not just learn it) that is critical for such an argument to go through.
Thus, while the wall between chimp and human language may be permeable, Kanzi is not equivalent to a human child in terms of either language-learning or language-creating abilities. Savage-Rumbaugh herself is undoubtedly well aware of the differences between child and chimp, but the book pays them little mind -- an unfortunate omission particularly since these differences take on new interest in light of the fact that there do appear to be some similarities between child and chimp language. Indeed, one is tempted to ask why chimps (particularly those provided with many of the artifacts of human culture) do not invent language given that they seem to be able to learn at least its rudimentary aspects.
One possibility is that they do and have and that we haven't looked in the right place. The book does, at times, mention the natural communications that chimps use in the wild, arguing that we may well be underestimating the structure of such systems. Another place we may be underestimating the chimp is in his use of gesture. Throughout the book, Kanzi's gestures are frequently mentioned in passing but no systematic attention is paid to them (aside from the pointing gesture he uses to indicate the desired agent in requests). Indeed, over the course of her studies (and the book), Savage-Rumbaugh begins to concentrate solely on Kanzi's ability to understand human speech and on his attempts to produce it. This focus on speech (as opposed to nonverbal symbols like gesture which can have language-like properties) is perhaps excessively narrow, particularly in the origins of language chapter, given that humans are completely equipotential when it comes to learning a spoken language vs. a gestural language such as American Sign Language. Why should speech be the most important yardstick against which the chimp's accomplishments are measured?
What then do we learn from successfully teaching a bit of language to a chimpanzee? One fascinating outcome that is mentioned in the book is what we can learn about the relationship between language and thought. For example, language-competent chimpanzees are able to learn to use a joystick on a computer through simple observation while language-naive apes must be trained to do so bit by bit. By comparing chimps who are language-users with chimps who are not, we can begin to get a sense of which tasks are facilitated by knowing symbols (and which are not).
Kanzi's story is well-told. The book captures Savage-Rumbaugh's extensive knowledge of apes, as well as her deep respect and affection for them, and it does so in a very readable way. Moreover, the book paves a path away from the fruitless question of whether chimps do or don't (have language) toward more interesting questions, such as which aspects of human language are within the chimp's capabilities (and which are not), and which artifacts of human culture must be in place before those abilities can be tapped.
Department of Psychology
5730 South Woodlawn Avenue
The University of Chicago
Chicago, IL 60637
Goldin-Meadow, S. & Mylander, C. Beyond the input given: The child's role in the acquisition of language. Language, 1990, 66(2), 323-355.
Greenfield, P. M., Savage-Rumbaugh, E. S. Imitation, grammatical development, and the invention of protogrammar by an ape. (1991). In N. A. Krasnegor, D. M. Rumbaugh, R. L. Schiefelbusch, & M. Studdert-Kennedy (eds.), Biological and behavioral determinants of language development (pp. 235-258). Hillsdale, N.J.: Lawrence Erlbaum Associates.