American Scientist

Intelligence of Apes and Other Rational Beings. Duane M. Rumbaugh and David A. Washburn. xvii + 326 pp. Yale University Press, 2003. $35.

How can you tell whether an animal is intelligent? Perhaps this is an impossible question to answer for species as different from us as honeybees and fish, but what about our closest relatives, the great apes? Shouldn't their cognitive abilities be easier to comprehend because of our anatomical and genetic similarities? Or does the degree of similarity cause biases in our thinking that may cloud our understanding? Ever since Darwin, biologists have been interested in the minds of these animals. But we are still far from discovering the real similarities and differences between ape and human intelligence—despite a wealth of important recent research, some of which is documented in Intelligence of Apes and Other Rational Beings, by Duane M. Rumbaugh and David A. Washburn.

There are two ways to approach the investigation of mental ability in primates. Comparative psychologists conduct laboratory tests of learning, memory and problem solving. In the first half of the 20th century, Robert Yerkes in the United States and Wolfgang Köhler in Europe were among the first to confront apes with problems whose solution required complex cognitive skills (how to traverse a maze, for example, or to obtain food that they cannot grab directly). Yerkes and Köhler wanted to determine whether apes had the mental equipment to solve such problems and to find out whether in attempting a solution they would employ the same processes or psychological mechanisms as humans.

Ethologists, on the other hand, investigate intelligence by recording the behavior of animals in their natural habitat. In the 1960s and 1970s, Jane Goodall, Dian Fossey and Biruté Galdikas observed specific groups of primates for years to ascertain their natural approach to ecological problems, such as how to fish for termites deep within mounds and out of reach. However, as important and illuminating as these studies were, they provided only a limited amount of information about the mechanisms of intelligence.

Two married couples who were, roughly, contemporaries of Yerkes and Köhler—Keith and Kathy Hayes, and Winthrop and Luella Kellogg—attempted to understand whether chimpanzees could learn to produce and use human language. These studies were doomed to failure because of their focus on testing whether chimpanzees' understanding and use of language was vocally based. Later research by Alan and Beatrix Gardner, Francine G. "Penny" Patterson and others on the use of American Sign Language in enculturated apes produced interesting results but did not get at the important aspects of human language: grammar, syntax and creativity. To do that, a new way of thinking about and testing apes' appreciation of language was needed.

Rumbaugh was one of the pioneers of this research. He invented an apparatus featuring a keyboard monitored by a computer, which primates could use to communicate by means of a system of symbolic language. By striking special keys marked with lexigrams (symbols that represent a word), chimpanzees formed sentences ordering a vending-type machine to supply them food, drink, moving pictures and music. The use of automated systems has dominated Rumbaugh's thinking about animal intelligence ever since his early training in the theory of animal learning.

The sections of the text in which Rumbaugh and Washburn give a detailed account of technological advances in the testing of primate learning and cognition are perhaps the most rewarding in the book. Here are described Rumbaugh's Herculean efforts to test different primate species with discrimination learning problems and discrimination reversal tests (so-called learning sets), which test how a subject "learns to learn." The animal is repeatedly presented with a pair of stimuli, such as a red square and a yellow triangle, and always rewarded with food for choosing one—the red square, say—and not the other. Then an entirely new pair of stimuli are presented, with only one member of the pair being rewarded. A subject who, after seeing on that first presentation which stimulus is rewarded, always chooses it on subsequent trials has learned a rule and can transfer it to new stimuli. Rumbaugh found that the ability to transfer rules correlates with brain size in primates.

Although Rumbaugh and Washburn state that they approach cognition from a comparative perspective, they offer no substantial discussion of the abilities of other animals. I was intrigued by the reference in the title to "Other Rational Beings," but after reading the book I do not know what species the authors think are rational—aside from perhaps rhesus monkeys. I found the largely primatocentric view of Rumbaugh and Washburn outdated and problematic, particularly in relation to some of the extremely convincing recent examples of complex cognition in parrots, scrub-jays, crows, dolphins, sea lions and dogs. In the light of findings in those species of concept formation, language, number, episodiclike memory, mental attribution, and tool use and manufacture, can we continue to maintain that primates actually differ from other animals in intelligence? You won't find the answer to that question in this book.

The final section, "Rational Behaviorism," presents a theoretical framework for understanding intelligent, novel, flexible animal behaviors, which cannot be explained by classical or instrumental conditioning. Briefly, Rumbaugh and Washburn propose that a new category of animal behaviors be added to the two traditional categories of respondents (innate, involuntary responses elicited by an unconditioned stimulus and shaped through classical conditioning) and operants (emitted, voluntary reactions that operate on the environment to bring about a change that leads to reward). The authors refer to members of this third category as emergents—novel ways of behaving that are not based on associative learning. Emergents include, among other things, the abilities to acquire concepts, learn complex skills through observation, and make and use tools. Such emergents develop through an integrative process, using aspects of an animal's physiology, genetics, instinctual behavior, respondents, operants and cognitive processing. I was not convinced that this all-encompassing idea makes any great leap forward in our understanding of intelligent behavior.

This book is incredibly difficult to categorize, and as such, hard to recommend. It is a confusing mix of autobiography, anecdotal observation, partial review of previous research, science journalism and perhaps most frustratingly, self-congratulation. There is little question that Rumbaugh should be congratulated for his seminal contributions to comparative psychology, particularly in the areas of ape language and animal learning, and for his work with Washburn on the automation of learning and memory tasks for rhesus monkeys. However, I didn't need to be reminded every couple of pages how great these contributions were! There is very little discussion of other groups' research, particularly studies of nonprimates and research completed in the last decade. One missed opportunity is that, strikingly, the authors fail to mention the recent finding that New Caledonian crows can invent tools to gain out-of-reach food using wire—a prime candidate for an emergent.

I have to disagree with Russ Tuttle's appraisal on the dust jacket that the book is "the best single source where one can find out what apes and monkeys know and how we know it." I suggest that readers interested in ape intelligence would be better served by directing their attention to Primate Cognition, by Michael Tomasello and Josep Call (Oxford University Press, 1997). However, if you would like a firsthand personal account of the trials and tribulations of research in this area, then this is a good place to start.—Nathan J. Emery, Animal Behaviour, University of Cambridge