Look who's talking now

Apes, Language and the Human Mind
July 9, 1999

Sue Savage-Rumbaugh is the latest in a long line of researchers who believe that apes have been given a bad press as far as their linguistic capacities are concerned. To prove her point, she has tried to show that bonobos (in particular one Kanzi, the so-called "ape of genius") can actually understand limited samples of spoken English. Here, to help her convince a sceptical world, she has enlisted the support of a philosopher (Stuart Shanker) and a linguist (Talbot Taylor). This is just as well, because without that support most readers would find Savage-Rumbaugh alone too partisan a witness to be trusted.

Savage-Rumbaugh presents her case in a way that is frustrating on at least two levels. First, it often reminds one of elderly aunts who assure you that their pet poodle "understands every word I say". Of her bonobos, Savage-Rumbaugh declares: "I know how they feel, and they know how I feel." In more lyrical moments, this reaches peaks like: "When I observe a bonobo, it is as though I am standing at the precipice of the human soul, peering deep into some distant part of myself." She thinks she can "read" on inspection the meanings of a monkey's facial expressions; a feat that some of us have difficulty in managing for our own species. And this ability is in turn projected upon Kanzi, who "can read my facial expressions as well as, if not better than, any human being I have ever known". Although Savage-Rumbaugh regards herself as a "scientist", there is not much science in this.

Second, at a deeper and more disconcerting level, what is worrying is her tendency to describe bonobo behaviour in language that assumes what she wants to prove has indeed been proved. Thus the intended result is projected in advance into the terms of description. Whatever the monkey does is interpreted as if the animal were a surrogate human. When Kanzi presses a symbol on his keyboard, he is "talking", "saying something", "naming" objects, and so on. Even when he did nothing, he "quietly noted everything we said". How he could have done that unless blessed with an innate command of American English it is difficult to comprehend.

Savage-Rumbaugh's own theory of communication is plainly based on a crude telementational or "thought-transference" model. This she promptly ascribes to Kanzi, and attributes to him a range of telementational feats that, unsurprisingly, confirm her own assumptions. She tells us that Kanzi was "the first animal ever to learn language without training, as a child does". Even setting aside the problematic phrase "learn language", the claim is untenable on two counts. One is the implication that for children linguistic apprenticeship does not involve "training". The other is the suggestion that Kanzi's apprenticeship was just like that of a human infant. Nothing could be further from the facts of the matter. Kanzi did not learn from a natural parent already fully proficient as a mature language user. He learned initially "by accident" from members of another species trying (unsuccessfully) to teach his bonobo foster-mother. This itself is remarkable; but it does not advance our understanding to assimilate such an exceptional case to that of the average human child.

Whether Shanker and Taylor support Savage-Rumbaugh to the extent they were supposed to is another question. Shanker lays the blame for our scepticism about animals' linguistic capacities squarely on Descartes. This is a trifle unjust, not only because the idea can be traced back much earlier than Shanker traces it, but because he uses "Cartesian" as a smear term to condemn many later theorists whom Descartes would probably have disowned. The irony, which Shanker does not appear to see, is that it is not only the sceptics who are being "Cartesian": Savage-Rumbaugh's own telementational assumptions are manifestly "Cartesian" too. Both sides are committed to the same fallacy. The solution lies not in more research, but in rejecting the fallacy. But that would be hard for Savage-Rumbaugh, since Kanzi would have "proved" far less than she wanted.

Shanker is at his best when attacking the sceptics and pointing out that some of them are guilty of using double standards. But when he insists that judging Kanzi to have understood an instruction when Kanzi does what was wanted is not a matter of inference at all but of simple observation, he is begging just as many questions as Savage-Rumbaugh. He fights shy of explicitly endorsing Savage-Rumbaugh's claim that "Kanzi has shown us that we are not alone among God's creatures to have been blessed with the gift of mind". Shanker's bottom line is that "where the great importance of Savage-Rumbaugh's work with Kanzi and Panbanisha lies is simply in establishing, irrefutably, that animals are indeed capable of satisfying a vast range of the criteria that we ordinarily apply when speaking of the cognitive and linguistic abilities of young children". But even this is not quite right. For "animals" substitute "these (captive, laboratory-reared and potty-trained) monkeys (who watch television in bed at night)".

Taylor is more circumspect than Shanker and comes close to admitting that the whole current debate is a meta-linguistic dialogue of the deaf. He suggests that the question cannot be resolved until the "scientific community" comes up with a "common evaluative method" for assessing human and animal performances. But he holds out no serious hope of this until scientists abandon their "epistemological conception" of the problem; in other words, admit that the question of whether Kanzi understands a sentence (unlike the question of whether Saturn has seven moons) is not a factual question at all. Meanwhile, we are doomed to an irreconcilable clash of "rhetorics", in which different participants play different "language games". This sounds like a form of linguistic relativism in which the best result that can be hoped for is stalemate. If Taylor really believed his own arguments, what he ought to be doing is persuading Savage-Rumbaugh to give up her persistent rhetorical claims about Kanzi's mind, which merely perpetuate a sterile logomachy.

The crux of the matter is that language is made what it is by the whole context of human culture. And monkey communities, so far, have developed nothing worth calling a culture. Savage-Rumbaugh and her team have chosen to bat on the wrong wicket, where, like their opponents, they constantly risk conflating arguments about language with arguments about the biomechanical substrata of language. What Kanzi and company tell us is not that bonobos too are by nature potential language users "just like us". On the contrary, it is not until the hapless monkey's world is artificially restructured by direct human control that the monkey can be made to begin to grasp (in human judgement) some of the rather complex ways it is possible to use vocal sound to integrate other activities. As to what monkeys "think" of this form of oppression, the world is still waiting for a monkey to "say".

Roy Harris is editor, Language and Communication .

Apes, Language and the Human Mind

Author - Sue Savage-Rumbaugh, Stuart G. Shanker and Talbot J. Taylor
ISBN - 0 19 510986 4
Publisher - Oxford University Press
Price - £22.50
Pages - 244

You've reached your article limit.

Register to continue

Registration is free and only takes a moment. Once registered you can read a total of 3 articles each month, plus:

  • Sign up for the editor's highlights
  • Receive World University Rankings news first
  • Get job alerts, shortlist jobs and save job searches
  • Participate in reader discussions and post comments

Have your say

Log in or register to post comments