This book is a remarkable piece of literature, one that summarizes the major topics of cognitive biology in a comprehensible and substantial way. Frans de Waal is a professor of psychology at Emory University, and the head of the Living Links Center at the Yerkes National Primate Research Center in Atlanta. At work he is surrounded by 3,000 chimpanzees, rhesus monkeys, baboons, capuchin monkeys, orangutans, and macaques. In the early 1970s, he studied the behavior of chimpanzees in the Arnhem Zoo, and then moved to America in 1980. Having grown up in the tradition of European ethology, he brought with him a sharp eye for observation, and a flair for experimental behavioral research. Although he has studied almost exclusively animals in captivity, he has remained a classical ethologist, a scientist curious about the natural behavior of animals and animal species in their totality, including their evolutionary, physiological, ecological, and social embedding.
Among the Apes
Do animals have mental powers, and, if so, what are they? This is the broadest question of cognitive biology. The broadest of answers is surely yes. De Waal’s tactic throughout is to concentrate on concrete, largely convincing examples. It is what gives his book its flair. Meeting Kandula the elephant in the National Zoo,
scientists hung fruit high up above Kandula’s enclosure, just out of his reach. They gave the elephant several sticks and a sturdy square box. Kandula ignored the sticks but, after a while, began kicking the box with his foot. He kicked it many times in a straight line until it was right underneath the fruit. He then stood on the box with his front legs, which enabled him to reach the foot with his trunk. An elephant, it turns out, can use tools—if they are the right ones.1
Still, examples are one thing; their meaning, another. Ethology has its own research strategies, and de Waal makes refined use of them, but he is prepared to borrow from adjacent disciplines.2 A threefold approach is at work: use what you can, borrow what you need, and, from time to time, let in the philosophers and psychologists. Such an approach proved to be very fruitful in de Waal’s book Primates and Philosophers.3 The strategy has an obvious risk. Cognitive researchers too often use concepts from human psychology uncritically, either to make their own research more broad-based, or to suggest the evolutionary continuity of human beings with other species.4
Sophisticated experimental designs are the heart and soul of cognitive ethology. But experiments are often unrevealing. It is not very reasonable to ask a dog to execute a task requiring digital dexterity. Cleverness is a matter of finding a task that an animal can do. This is not always easy. In this regard, de Waal is especially adroit. He understands that well-conceived experiments require knowledge of the animal and a sympathetic appreciation of its everyday problems. “The challenge,” he notes, “is to find tests that fit an animal’s temperament, interests, anatomy, and sensory capacities.”5
Comparative studies have suffered from the dominance of primatology, and the author, too, seems to have a blind eye here. In one set of experiments, various animals were tested for their ability to copy complex tasks after seeing them performed. Chimpanzees proved able to drag food with a rake, and to open crates with pull, pressure, or rotary locks. They were also able to cooperate in order to obtain a reward. The monkeys used their hands in these experiments; the dogs, their mouths; horses, goats, and pigs used neither. Is a cognitive difference at work?
Even in birds there are significant differences in the use of the body for object manipulation. While parrots hold objects with their feet, using the beak for other manipulations, many other birds do not. Parrots thus do not find it difficult to reach a reward that has been attached to a string hanging from a branch. They draw the cord up to the branch with their beaks and then fix it to the branch with a foot. Finches, however, do not do this well.
Any conclusions drawn about their causal comprehension on the basis of how well they did in this task are clearly problematic. It is not necessary to elaborate on how much more difficult it would be to apply such studies to bats, snakes, or bees.
Skewed
Two theories have skewed our understanding of cognitive evolution. One is social intelligence theory, which, in an odd turn of phrase, has come to be known as Machiavellian intelligence theory. Complex cognition and enlarged executive brains evolved in response to challenges that are associated with social life. Would-be Machiavellians have tended to overestimate social challenge as a force driving the evolution of cognition.6 In primates, a mastery of social life does seem to have some selective value. The same is not true of the animal kingdom as a whole. Sophisticated cognitive performance is beneficial in foraging, hunting, camouflage, and navigation. There is room in nature for the crafty loner.
If Machiavellians have overestimated social life in the evolution of cognition, the impressive size of the primate brain has led many cognitive researchers to the hypothesis that in cognition, bigger is better. When thirty-six species were tested for self-control, a high correlation was found between impulse control and brain size.7 The inevitable follow-up study showed that the common raven has about as much self-control as the chimpanzee, despite a brain twenty-five times smaller in size. Jackdaws do better than capuchin monkeys.8 The fact that within birds a correlation between brain size and impulse control was found suggests that this relationship could exist within taxonomic groups, but not necessarily between them. It is possible that brain volume is less important than the number of neurons, but it is much more important to ask which constraints are dominant in the evolution of cognition. In birds, flight seems to exert tremendous pressure in favor of small neurons, high density, and, of course, weight reduction.
Cognitive ethology has revealed mental faculties among animals that are not found in human beings. Dolphins and bats perceive objects by echolocation; it is a system they also use to communicate with one another. An exquisitely sensitive auditory system enables barn owls to hunt by hearing. The experimental analysis of these animals enriches the cognitive pedigree of living creatures. Animals, as de Waal properly observes, must be understood in terms of their own peculiarities, limitations, and abilities.9 Consider every species in full, classical ethology demanded.10 The phenomenal surface of animal life is often determined by everyday psychological prejudice. The righteous ethologist aims to “get under the skin.”11 Understanding the cognitive abilities of animals depends on the value of these abilities for the respective animal species. What are the special challenges that archer and cleaner fish have to master? They certainly do not have to pull ropes or roots.
From an evolutionary point of view, animals should do only what they must in order to survive and reproduce.12 This inevitably raises the question whether animals are cognitively shackled. It would seem so. A rat unaccountably able to compute prime numbers would be regarded with blank indifference by evolution. Big brains consume a great deal of energy, far more than other organs. They may be an evolutionary luxury. Functional shackling, although plausible as a Darwinian correlative to the maxim waste not, want not, may be too severe a stricture. The very essence of cognition lies in its power to overcome emerging problems.13 The faculty of cognition might demand a certain functional surplus, something in reserve.
The chief part of de Waal’s book is devoted to so-called cognitive ripples;14 it is the title of his third chapter.
The arrival of corvids on the scene illustrates how discoveries of mental life ripple across the animal kingdom, a process best summarized by what I’ll call my cognitive ripple rule: Every cognitive capacity that we discover is going to be older and more widespread than initially thought. [emphasis original]15
Do animals possess anything like language, culture, or morality, or even a soul?16 Ravens and parrots pose new challenges to our assessment of animal cognition, for well-documented experiments indicate that they possess episodic memory, self-awareness, a grasp of logic and linguistic understanding, powers of deception, and the ability to adopt the other’s perspective in matters of reconciliation and dispute. These are creatures that use and sometimes make tools.17 Ethologists may be tempted to see themselves in the eyes of a Machiavellian monkey; the temptation is less in the case of a raven. Whatever their cognitive powers, the ravens remain an alien species.
De Waal himself suffers from a primate-centric bias. This is something that de Waal recognizes: “Whereas cognition research focuses almost entirely on the tiny vertebrate minority, it is not as if the rest doesn’t move, eat, mate, fight, and cooperate.”18 His research encompasses only three percent of the animal kingdom, even though the remaining ninety-seven percent faces very similar problems and must have “a [certain] degree of cognition.”19 Even within the vertebrates, mainstream research is half-blind, because two large classes are almost invisible: the amphibians and the reptiles. These groups are indispensable in studying the development of bird or mammalian cognition. It is in the literature of comparative cognitive research that hominids undergo a demotion. As the very words might suggest, comparative cognitive research is dedicated to comparisons among species. Developmental biologists, geneticists, and neurobiologists, by way of contrast, focus on a few model organisms, such as the fruit fly or the lumpfish.
Bee-Sized Brains
Various doctrines of human cognitive superiority are made plausible by a comparison of human beings and the chimpanzees. For questions of evolutionary cognition, this focus is one-sided. Consider the evolution of cooperation in social insects, such as the Matabele ant (Megaponera analis).20 After a termite attack, these ants provide medical services. Having called for help by means of a chemical signal, injured ants are brought back to the nest. Their increased chance of recovery benefits the entire colony. Red forest ants (Myrmica rubra) have the ability to perform simple arithmetic operations and to convey the results to other ants.21
When it comes to adaptations in animals that require sophisticated neural control, evolution offers other spectacular examples. The banded archerfish (Toxotes jaculatrix) is able to spit a stream of water at its prey, compensating for refraction at the boundary between air and water. It can also track the distance of its prey, so that the jet develops its greatest force just before impact. Laboratory experiments show that the banded archerfish spits on target even when the trajectory of its prey varies.22 Spit hunting is a technique that requires the same timing used in throwing, an activity otherwise regarded as unique in the animal kingdom. In human beings, the development of throwing has led to an enormous further development of the brain.23 And the archerfish? The calculations required for its extraordinary hunting technique are based on the interplay of about six neurons. Neural mini-networks could therefore be much more widespread in the animal kingdom than previously thought.
Research on honeybees (Apis mellifera) has brought to light the cognitive capabilities of minibrains. Honeybees have no brains in the real sense. Their neuronal density, however, is among the highest in insects, with roughly 960 thousand neurons—far fewer than any vertebrate. Even if the brain size of honeybees is normalized to their body size, their relative brain size is lower than most vertebrates. Insect behavior should be less complex, less flexible, and less modifiable than vertebrate behavior. But honeybees learn quickly how to extract pollen and nectar from a large number of different flowers. They care for their young, organize the distribution of tasks, and, with the help of the waggle dance, they inform each other about the location and quality of distant food and water.
Early research by Karl von Frisch suggested that such abilities cannot be the result of inflexible information processing and rigid behavioral programs. Honeybees learn and they remember.24 The most recent experimental research has, in confirming this conclusion, created an astonishing picture of the honeybee’s cognitive competence. Their representation of the world does not consist entirely of associative chains. It is far more complex, flexible, and integrative. Honeybees show configural conditioning, biconditional discrimination, context-dependent learning and remembering, and even some forms of concept formation. Bees are able to classify images based on such abstract features as bilateral symmetry and radial symmetry; they can comprehend landscapes in a general way, and spontaneously come to classify new images.25 They have recently been promoted to the set of species capable of social learning and tool use.26
In any case, the much smaller brain of the bee does not appear to be a fundamental limitation for comparable cognitive processes, or at least their performance. Jumping spiders and cephalopods are similarly instructive. The similarities between mammals and bees are astonishing, but they cannot be traced to homologous neurological developments. As long as the animal’s neural architecture remains unknown, we cannot determine the cause of their similarity.
Fish Food
Comparative cognitive ethology very often offers challenges to systematic evolutionary schemes. A good example is the comparison between fish and monkeys in solving the ephemeral reward task. An animal must learn that a given food source is unpredictable. Two vessels are offered, both containing the same amount of food. If the animal selects the first, the second is removed. If it chooses the second, the first is allowed to remain. Animals should learn to choose the second vessel. For many animals, this experiment represents an almost insoluble problem. Neither capuchin monkeys, nor orangutans, nor half of the chimpanzees tested, could solve the problem. After one hundred attempts, testing was stopped. And yet, the common bluestreak cleaner wrasse (Labroides dimidiatus), a recent study reports, masters this task with ease as soon as it is mature.27
Redouan Bshary and colleagues have proposed an eco-ethological explanation for this. Small cleaner fish remove parasites and dead skin from other fish in so-called cleaning stations. In doing so, they have both stationary and visiting clients. Cleaner fish benefit from being able to distinguish between temporary and permanent sources of food, the visitors from the inhabitants. The superiority of the cleaner fish to apes in performing this task is striking, and illustrates the pitfalls of species comparisons. The task is always seen in a specific ecological–ethological context to which species are differently adapted.28
When cognitive biologists claim that animals should not be tested with abstract tasks but rather with those tailored to their needs, species comparisons become more difficult. In 1988, Alan Kamil proposed a synthetic approach to the problem of species comparison. Species should be tested with a battery of tasks that vary in different noncognitive dimensions, but each should test a single cognitive ability. Questions arise almost at once. If one species performs well across the board, does this mean that it has a domain-general rather than a domain-specific intelligence?29 This is certainly the assumption behind modern IQ tests. Birds, rats, or apes that perform well on tests raise an obvious question. Shouldn’t evolution equip organisms with a single blade, rather than a Swiss Army knife?30
There is, in any case, little prospect of finding a general measure of intelligence, something that might establish unambiguously that the donkey is less intelligent than the mongoose. Every attempt, as de Waal properly notes, to establish a rank ordering of various species encounters obvious methodological weaknesses.31
Plus X
In his Traité des Animaux, Étienne de Condillac noted that, “it would be less interesting to know what animals are, if it were not a means to know what we are.”32 Men are not dogs or apes. They must be something more, some animal plus X.33 Whatever the X, de Waal demonstrates, it is not fundamentally new.34 Our most impressive cognitive categories may already be found in the animal kingdom. As de Waal notes,
Proponents of human uniqueness face the possibility that they have either grossly overestimated the complexity of what humans do or underestimated the capacities of other species. Neither possibility is a pleasant thought, because their deeper problem is evolutionary continuity.35
There are, nevertheless, significant differences in expression. Some animal species can detach themselves from the here and now, but no dog can expect its master in a week. Many animals use sounds in communication, but this is a far cry from human speech. Some animal species play social games, demonstrating empathy and cooperation. But as Philip Kitcher points out, human and animal altruism differ in expression, intensity, and extent.36 Functional similarity in cognitive performance is, by itself, not evidence of homology.
Unlike Kitcher, de Waal regards cognitive continuity as fundamental.37 In truth, he means continuity among apes, where the difference between man and animal is smallest. Continuity of a trait is usually based on common descent. In cognitive evolution, we are facing both divergence and convergence, where similarities may be due to common selective pressures or common descent. The assumption of continuity is, of course, plausible with respect to conserved structures like the cerebellum. Cognitive behavior is otherwise. When Jane Goodall watched chimpanzees fishing for termites, the evolutionary continuity with man was obvious. Over the course of eight million years of divergent evolution, chimpanzees and humans underwent many genetic changes. We cannot simply assume continuity as the cause for functional similarities.
The more we understand about how animals came to be what they are, the closer we get to figuring out how smart they are.
Translated and adapted from the German by the editors.
