When a paper in Science reported these experiments newspaper stories celebrated Betty’s brilliance. That Betty invented a technique for making a hook to get food does seems pretty smart. Although Abel didn’t make hooks, he had already devised the effective foraging technique of stealing from Betty. Maybe that’s just as smart. If Betty’s a crow Einstein, Abel is a crow Machiavelli. Suppose the researchers keep putting crow food in places that are harder and harder to reach. Eventually, I suppose, Betty will invent the pneumatic drill in order to obtain crow snacks embedded in concrete—and Abel will invent municipal noise ordinances, a partisan police force, and a system of fines, and will use them to obtain crow snacks from Betty. Unless Betty invents Mace first.
ALTHOUGH NO ANIMALS are clever enough to present a threat to us, there are always some people who want to emphasize the us-them divide as strongly as possible, and intelligence is often where they choose to locate the gulf. That learning is not the same thing as intelligence is shown by all the things people and animals can learn to do without understanding. Our immune systems learn to respond to pathogens without being smart.* On the other hand, an animal that knows quite a lot but can learn almost nothing does not impress us as smart.
Fans of IQ tests might like to know that Koko the gorilla is reported to score consistently between 70 and 90 on age-graded IQ tests (the human average is 100). More interesting are examples of what Koko can (and cannot) do. She can use maps. When she lost a toy outside her enclosure, and her human friends made a map of her enclosure, she was able to point out the spot on the map, just outside the enclosure, where the toy had fallen.
Despite our first-place status in the intelligence Olympics, we can’t win every single event. At the task of looking at two shapes and figuring out which is the mirror image of a third shape, a task that often appears on intelligence tests, pigeons and college students were equally accurate. But pigeons were faster. The researchers suggest that pigeons use some different, automatic process, and that they need it more than we do, because they fly around and look down on things that are oriented arbitrarily, whereas the things we look at are more consistently oriented. In other words, when we do it, it’s smart, and when they do it, it’s not. I’m not convinced that the world around me is so consistent, but I agree that navigating in three dimensions must be harder than navigating in two. So we shouldn’t feel bad about being inferior to pigeons at mental rotation. But we should avoid going on game shows where we would face teams of pigeons at mental rotation tasks, because that really would be embarrassing.
What can we say about the repeated failure of attempts to set up “Man is the only animal who…” rules? Common sense tells us that even if some animals make tools, or laugh, or invent forms of communication, we really are a very different animal. Our mistake may be in trying to find some unique quality in ourselves, when it may be that what’s different about us is quantity. We’re unusually smart, we’re really chatty, and we’ve taken the tool thing to ridiculous extremes.
Dumb brutes
It is easy to find examples of animal stupidities. Fortunately you can love animals without admiring their intellects. Lieutenant Colonel Locke described the inflexibility of tigers trying to drag away dead prey that has gotten stuck. “Once a tiger has made up his mind to do anything, he will often go on trying to do the same thing over and over again, although several simple alternatives may be available to him.” A tiger trying to drag a cow over a high fence is one example he cites, as well as several cases of tigers trying to drag an animal whose horns have caught on a raised root. The tiger keeps pulling in the same direction, when all it needs to do to disengage the horns is to pull a short distance in another direction. “Whether this obstinacy is due to lack of intelligence or to stubborn pride I cannot say.”
A peninsula? Is that like an isthmus?
Greylag geese like to nest on little islands in lakes, where they are safe from foxes. Konrad Lorenz writes that they do not understand the difference between an island and a peninsula, however. While it is doubtful that geese ever learn this distinction, they do learn that nesting at a particular site didn’t work and nest elsewhere next time.
Superstition
Like people, animals are superstitious. Spinner dolphins in an oceanarium all had “lucky corners,” which they associated with getting a fish reward. They’d do a group leap and, expecting to be rewarded for the leap, would scatter to their lucky corners. That’s where they’d be when they got fish, so the association was strengthened, when in fact they would have gotten fish anywhere.
My dogs believe that barking drives away the package delivery truck. Over the years their hysteria at its every appearance has increased. I suspect they are so vociferous because they anticipate success. The truck has come before, they barked like crazy, and it went away. Hurray! We saved us! They never do the control experiment: if they don’t bark, would the truck still go away, or would the emboldened truck driver break into our house and steal their kibble?
What is smart?
It is possible to think of intelligence as one thing, a “g factor” that can be assigned a single number, or as a collection of attributes. Richard Byrne writes, “Intelligence certainly means more than flexible learning: terms like ‘thinking clearly,’ ‘solving difficult problems,’ and ‘reasoning well’ recur in attempts to define the ability. The scope of intelligence is quite wide, including learning an unrestricted range of information; applying this information in other and perhaps novel situations; profiting from the skills of others; and thinking, reasoning, or planning novel tactics. (Which should remind us not to expect that intelligence is a single ‘thing,’ but a bag of devices and processes, endowments and aptitudes, that together produce behaviour we see as ‘intelligent.’)”
Being very good at just one thing doesn’t impress much. Many of those who exalt imitation as a crucial, rare, perhaps uniquely human, ability do not count vocal imitation by birds because it is only vocal imitation. (But don’t forget Okíchoro, the parrot who waves bye-bye.)
“Learning occurs in all animal species and encompasses a variety of levels of adaptation, including such low levels as habituation and associative learning,…and is often highly specialized, inflexible, and limited in scope, as well as species-specific,” write Sue Taylor Parker and Patricia Poti. “Intelligence, which occurs only in a few long-lived, large-brained species, is, in contrast, quite generalized, flexible, and broad in scope, as well as species-specific.” Richard Byrne argues that we should expect to see intelligence in animals that live in uncertain conditions, that “intelligence should most benefit extreme generalists, species adapted to exploit continually changing environments since they must daily cope with novelty in order to survive.”
If learning doesn’t equal intelligence, because even dopes learn some things, but intelligence includes the ability to learn, the missing ingredient is understanding what you’ve learned. Byrne writes of understanding situations and responding appropriately. But people and animals often come to understand situations only after they have learned how to respond to them. (This may produce the “prelearning dip” described in chapter 1.) Then the question is whether we can generalize, whether similar situations in the future will be met with understanding, or whether a whole new learning process will have to take place.
Insight
Some researchers have sought intelligence as shown by understanding. Bernd Heinrich, in his work examining raven intelligence, is interested in things they don’t learn to do, but instead do instantly, through a flash of insight.
One Maine winter Heinrich put a chunk of suet out to attract birds, and many species came to hammer at the frozen fat with their bills, chipping off flecks of suet to eat. One day Heinrich went out to replenish the food and scared a wild raven away from the suet. Instead of chipping little bits off, the raven had been in the process of chiseling a groove across one corner of the suet chunk. Had Heinrich not interrupted, the bird would have been able to carry off qui
te a large chunk of suet once it had carved the groove all the way through. “This was a raven Einstein,” concluded the stunned Heinrich, photographing the suet. “Intelligence is doing the right thing under a novel situation, precisely as this bird had done.”
Scientific journals showed no enthusiasm for the suet story, however, dismissing it as a mere anecdote. Heinrich needed an experimental situation for ravens to show insight. He came across a story in Ranger Rick suggesting that kids put out bits of food for chickadees on the ends of strings, so they could watch the chickadees pull up the strings. “I could not believe that chickadees could actually do that. If they did, it seemed to me, they would have to be trained,” wrote Heinrich. Ravens, however, might just be able to grasp the pull-on-the-string concept.
Heinrich presented the string problem to some ravens he had raised in aviaries, ravens who had never seen string in their lives. They had seen hard salami, and they strongly approved. Heinrich tied pieces of salami to long strings which he tied to high perches. The strings were so long that the salami couldn’t be pulled up in one or two moves.
The ravens were interested. They jumped on the perch and pecked the top of the string. They peered at it. They lost interest. The second time Heinrich presented this setup, one raven, Matt, flew up to the perch, reached down, pulled up on the string, placed his foot on the loop he had pulled up, pulled up more of the string, placed his foot on that, and so on, until he had pulled up the salami, which he ate. Eureka!
In the five groups of ravens Heinrich tested, several birds knew how to pull up the salami right away, many eventually figured it out, and a few never did. In further tests of what the ravens learned and what they knew, Heinrich discovered that if he hung a sheep’s head from a string—and a sheep’s head is a grocery item ravens adore—they would not try to pull it up, apparently knowing that it was too heavy. Four ravens who were used to pulling up salami were presented with a pair of crossed strings, one of which led to salami and one of which led to a rock. To get the salami, a raven had to pull on the string above the rock. If it pulled on the string above the salami, it would get the rock. Matt, noting the path of the strings, pulled on the one that led to the salami. The other three, time after time, pulled on the string that led to the rock. Although they were not being rewarded for pulling that string, they pulled on it because it was above the salami, and so they knew it had to be the right string.
Heinrich also tied salami on strings to branches in the woods, but the wild ravens distrusted these and wouldn’t go near them. As he later discovered, pulling up food dangling on strings is something that quite a few birds can do, including chickadees, Darwin’s finches, and the plain titmouse. Some learn to do it, and some do it spontaneously.
Problem solving
Researchers in laboratories are always setting up bizarre and unnatural problems for animals to solve. In the wild animals sometimes encounter novel problems, but this is rare and we seldom happen to see it. Animals who live with or near people encounter weird problems more often.
Elizabeth Marshall Thomas’s dog Sheilah alertly noticed when a parrot in the household dropped a piece of cheese onto the floor of his tall cage. She tried to reach into the cage with her paw to get the cheese, but couldn’t. After surveying the situation, she took the edge of the paper lining the cage in her mouth and tugged it out, with the cheese riding on it. Clever, and like few situations wolves encounter.
The writer Philip Wylie once saw a squirrel trying to raid a bird feeder with a conical roof. It was hanging from a branch by a string too thin for the squirrel to climb down. Three times in a row the squirrel ran along the branch and leapt onto the roof of the feeder, and three times in a row the roof tipped and the squirrel slid off and fell to the ground. The squirrel then sat on the branch and stared at the feeder for a long time. Then it leaned forward and bit through the string. The feeder fell, and the squirrel ran down and ate birdseed.
The parts of smart
Biologist Rachel Smolker writes, “The more we have looked and conducted experiments and compared different species, the more we have been forced to rephrase the question ‘How smart are they?’ to ask ‘How are they smart?’”
We can ask what attributes go into making a creature smart or able to learn. Memory is important, since otherwise there’s no way to learn from the past. Sonja Yoerg would add that a “differentiated memory—for skills, for events, for facts—with varying expiration dates would keep things tidy.” She’d also look for a “solid grip on cause and effect…hypothesis testing…serial order and pattern recognition, categorization, and concept formation. Some numerical skills might be a bonus (‘Weren’t there three pups here a minute ago?’), and being able to read the emotions of others might lead to a happier and longer life.” Curiosity—the interest in gathering information—is another trait linked to learning, although not necessarily to intelligence.
Domains, modules, and systems
Some abilities are domain-specific. They only work in certain areas. You might have a great memory for spelling and a terrible memory for faces. As Yoerg writes in Clever as a Fox, “Birds that can remember their father’s song for months can’t remember the color of a triangle for twenty seconds…. Zebras are aces at discriminating one stripe widthover another, but not better than horses at other visual discrimination tasks.”
Birds who cache seeds for the winter have fantastic memories for hiding places. One pinyon jay might store 26,000 seeds in a season, and there wouldn’t be much point if the bird could never find them again. Russell Balda and Alan Kamil looked at whether this great spatial memory capacity spills over into other areas. Sure enough, in two-and three-dimensional maze problems calling on spatial memory, seed-storing pinyon jays did better than scrub jays, who don’t store food. On the other hand, they’re not awfully good at remembering colors.
In the field of human as well as animal cognition, many theorists like the idea of modular abilities. “It is also essential to adopt a modular view of cognition as opposed to assuming some single entity such as learning ability or intelligence that all species possess to some degree,” Sara Shettleworth writes. “Similarly, behavioral neuroscientists refer to memory systems, distinct areas of the brain that do distinct tasks or store distinct kinds of memories.”
The innate, modular aspect of vervet monkey alarm calls produces a paradox. Vervets are born knowing how to make snake and leopard alarm calls, and with a partial idea of what these refer to. Through social learning, they refine their definition of the menace and learn how to react to such calls. But vervets who know that snakes are bad news do not notice much else about snakes. They will step casually along python tracks, real or faked, and scream in terror and shock if they then encounter a python. Cheney and Seyfarth call them “excellent primatologists but poor naturalists.”
Experiments showed that it was far easier for hummingbirds to learn a task that rewarded a “win-shift” strategy than one that rewarded a “win-stay” strategy. Many laboratory tests are win-stay: an animal is supposed to find out which choice is rewarded and stay with that choice. Hummingbirds who had to pick between two fake flowers, only one of which contained syrup, took hundreds of trials to learn this. But if it was a win-shift test, in which they were supposed to go to the flower that was not rewarded last time, they learned quickly. In the wild, after all, flowers that hummingbirds drink from are depleted of nectar, and the hummingbird would be wasting its time to go back before the nectar is replenished. Win-shift is what works, and it’s difficult and unnatural (though not impossible) to learn win-stay.
Sometimes it seems clear that performance in one domain improves ability in another. In recent years biologists have marveled at the dynamic hunting strategies of Portia, a spider-eating spider. Portia visits the webs of other spiders and experiments with different patterns of web twanging, attempting to pluck a rhythm that will delude the unsuspecting web owner to come out looking for a small helpless insect dinner or a spider of its own specie
s seeking passion. Portia has a huge repertoire of these rhythms and varies them depending on the feedback it gets from its intended victim. This trial-and-error process is so scientific that scientists are filled with admiration. Sometimes Portia sneaks up on its victim by a circuitous route, and the part that impresses biologists is that Portia remembers where its victim is even when it’s out of sight.
To see if Portia’s ability to employ trial and error applies only to hunting, researchers marooned spiders on tiny islands surrounded by an “atoll”—a ring of dry land, all within a larger tray of water. From the island, Portia could see both the atoll and the edge of the tray. (Portia has great eyesight, the better to hunt spiders.) There were two ways Portia could make a break for freedom: swimming to the atoll (and then the edge) or jumping to the atoll. The researchers randomly assigned each spider to either the swimming or the jumping group. If a spider tried the method they had assigned it, they allowed it to reach the atoll. If it tried the other method, a terrible storm came up and washed it back to the island (this storm took the shape of a researcher making waves with a little plastic scoop). Spiders who were washed back to the island switched methods. Spiders who made it to the atoll didn’t switch, but used the same method to head for the edge of the tray. “In Portia, perhaps a predatory strategy that routinely demands fine control over the behaviour of dangerous prey has set the stage for the evolution of problem-solving abilities that, as a spin-off, can be readily applied to novel situations, including confinement problems,” the proud researchers write. Although lauding Portia, biologists Eytan Avital and Eva Jablonka caution, “These are probably a small minority group among spiders.” Maybe. But I’m keeping my eye on the whole bunch.
Becoming a Tiger: The Education of an Animal Child Page 38