Ethology is the biological study of animal behavior that arose in continental Europe right before and after World War II. It reached the English-speaking world when one of its founders, Niko Tinbergen, moved across the Channel. A Dutch zoologist, Tinbergen started out in Leiden and accepted a position in Oxford in 1949. He described the male stickleback’s zigzag dance in great detail, explaining how it draws the female to the nest where the male fertilizes her eggs. The male then chases her off and protects the eggs, fanning and aerating them, until they hatch. I had seen it all with my own eyes in an abandoned aquarium—its luxurious algae growth was exactly what the fish needed—including the stunning transformation of silvery males into brightly red and blue show-offs. Tinbergen had noticed that males in tanks in the windowsill of his lab in Leiden would get agitated every time a red mail truck drove by in the street below. Using fish models to trigger courtship and aggression, he confirmed the critical role of a red signal.
Clearly, ethology was the direction I wanted to go in, but before pursuing this goal, I was briefly diverted by its rival discipline. I worked in the lab of a psychology professor trained in the behaviorist tradition that dominated comparative psychology for most of the last century. This school was chiefly American but evidently had reached my university in the Netherlands. I still remember this professor’s classes, in which he made fun of anyone who believed to know what animals “want,” “like,” or “feel,” carefully neutralizing such terminology with quotation marks. If your dog drops a tennis ball in front of you and looks up at you with wagging tail, do you think she wants to play? How na?ve! Who says dogs have desires and intentions? Her behavior is the product of the law of effect: she must have been rewarded for it in the past. The dog’s mind, if such a thing even exists, remains a black box.
Its focus on nothing but behavior is what gave behaviorism its name, but I had trouble with the idea that animal behavior could be reduced to a history of incentives. It presented animals as passive, whereas I view them as seeking, wanting, and striving. True, their behavior changes based on its consequences, but they never act randomly or accidentally to begin with. Let’s take the dog and her ball. Throw a ball at a puppy, and she will go after it like an eager predator. The more she learns about prey and their escape tactics—or about you and your fake throws—the better a hunter or fetcher she will become. But still, at the root of everything is her immense enthusiasm for the pursuit, which takes her through shrubs, into water, and sometimes through glass doors. This enthusiasm manifests itself before any skill development.
Now, compare this behavior with that of your pet rabbit. It doesn’t matter how many balls you throw at him, none of the same learning will take place. Absent a hunting instinct, what is there to acquire? Even if you were to offer your rabbit a juicy carrot for every retrieved ball, you’d be in for a long, tedious training program that would never generate the excitement for small moving objects known of cats and dogs. Behaviorists totally overlooked these natural proclivities, forgetting that by flapping their wings, digging holes, manipulating sticks, gnawing wood, climbing trees, and so on, every species sets up its own learning opportunities. Many animals are driven to learn the things they need to know or do, the way kid goats practice head butts or human toddlers have an insuppressible urge to stand up and walk. This holds even for animals in a sterile box. It is no accident that rats are trained to press bars with their paws, pigeons to peck keys with their beaks, and cats to rub their flanks against a latch. Operant conditioning tends to reinforce what is already there. Instead of being the omnipotent creator of behavior, it is its humble servant.
One of the first illustrations came from the work on kittiwakes by Esther Cullen, a postdoctoral student of Tinbergen. Kittiwakes are seabirds of the gull family; they differ from other gulls in that to deter predators, they nest on narrow cliffs. These birds rarely give alarm calls and do not vigorously defend their nests—they don’t need to. But what is most intriguing is that kittiwakes fail to recognize their young. Ground-nesting gulls, in which the young move around after hatching, recognize their offspring within days and do not hesitate to kick out strange ones that scientists place in their nests. Kittiwakes, on the other hand, can’t tell the difference between their own and strange young, treating the latter like their own. Not that they need to worry about this situation: fledglings normally stay put at the parental nest. This is, of course, precisely why biologists think kittiwakes lack individual recognition.1
For the behaviorist, though, such findings are thoroughly puzzling. Two similar birds differing so starkly in what they learn makes no sense, because learning is supposedly universal. Behaviorism ignores ecology and has little room for learning that is adapted to the specific needs of each organism. It has even less room for an absence of learning, as in the kittiwake, or other biological variation, such as differences between the sexes. In some species, for example, males roam a large area in search of mates, whereas females occupy smaller home ranges. Under such conditions, males are expected to have superior spatial abilities. They need to remember when and where they ran into a member of the opposite sex. Giant panda males travel far and wide through the wet bamboo forest, which is uniformly green in all directions. It is crucial for them to be at the right place at the right time given that females ovulate only once per year and are receptive for just a couple of days—which is why zoos have such trouble breeding this magnificent bear. That males have better spatial abilities than females was confirmed when Bonnie Perdue, an American psychologist, tested pandas at the Chengdu Research Base of Giant Panda Breeding in China. She did so by spreading out food boxes over an outdoor area. Panda males were much better than females at remembering which boxes had recently been baited. In contrast, when the Asian small-clawed otter, a member of the same arctoidea (bear-like) family of carnivores, was tested on a similar task, both sexes performed the same. This otter being monogamous, males and females occupy the same territory. Similarly, males of sexually promiscuous rodent species navigate mazes more easily than females, whereas monogamous rodents show no sex difference.2
If learning talents are a product of natural history and mating strategies, the whole notion of universality begins to fall apart. We can expect huge variation. Evidence for inborn learning specializations has been steadily mounting.3 There are many different types, from the way ducklings imprint on the first moving object they see—whether it is their mother or a bearded zoologist—to the song learning of birds and whales and the way primates copy one another’s tool use. The more variation we discover, the shakier gets the claim that all learning is essentially the same.4