Many scientists think that corvids — the family of birds that includes crows, ravens, rooks and jays — may be among the most intelligent animals on Earth, based on their ability to solve problems, make tools and apparently consider both possible future events and other individuals’ states of mind.
“There’s a lot of research that has been done with both ravens and crows because they are such intelligent species,” said Margaret Innes, an assistant curator at the Maryland Zoo in Baltimore.
Even in humans, defining and measuring intelligence is difficult, and it’s more complicated in other species, which have very different body shapes and have evolved for their niche in the environment. However, scientists who study cognition have defined a few measures of intelligence: recognising oneself in a mirror, solving complex problems, making tools, using analogies and symbols, and reasoning about what others are thinking.
For a long time, biologists expected most of these mental feats to be unique to primates. The great apes — chimpanzees, orangutans and gorillas — succeed at nearly all of these tasks, from making and using tools to learning large vocabularies of symbols, as well as recognising themselves in mirrors.
A select few other mammals also meet most of the accepted criteria for intelligence. Dogs and dolphins, for instance, are very good at tasks involving social intelligence, such as communication, conflict resolution and reasoning about what others are thinking. Dolphins are also capable of basic tool use — for instance, carrying sea sponges in their mouths to shield their noses from scrapes and bumps as they forage on the ocean floor.
However, the greatest intellectual rival to the brainy apes may be a noisy scavenger with a sharp beak, bright eyes and a brain about the size of a walnut: the crow and its corvid relatives.
Crows and ravens are clever problem-solvers, expert toolmakers and adept social movers, but scientists haven’t reached a consensus about how corvid minds handle abstract thinking or how closely their mental processes resemble those of humans.
Researchers from the University of Iowa and Lomonosov Moscow State University in Russia reported early this year that crows can use analogies to match pairs of objects. To reach that conclusion, the scientists trained crows to recognise whether two objects were identical or different, which the birds indicated by pressing one button when shown pictures of objects that matched and a different button when the objects didn’t match.
Once all the birds were good at matching objects, researchers showed the crows images of pairs of objects. Some images depicted matched pairs, while others depicted two mismatched objects with different shapes or colours. In response, crows could press buttons to choose between a matched pair or a mismatched pair.
The researchers wanted to see if crows could figure out the relationship between pairs of objects and then choose a pair with the same relationship: matched or mismatched. For instance, a crow looking at a mismatched pair would then select the mismatched pair from their response choices. Nearly 78 per cent of the time, the birds succeeded. According to the researchers, the birds recognised that the relationship between the two pairs of objects was the same. In other words, they were making analogies.
Other scientists contend that a type of reasoning less sophisticated than analogies could have produced the same results. For instance, the crows in the analogy test could have simply chosen images with similar characteristics, such as objects of the same colour, instead of reasoning about the relationship between the objects, to get the correct answer.
Some behaviours, such as those employed in the analogy test, could have more than one explanation, and until recently, scientists could only see what the birds did, then make inferences about the mental processes behind the behaviour.
Now, researcher John Marzluff and his colleagues at the University of Washington are using positron emission tomography, or PET, scans to study which parts of a crow’s brain are active when it performs such tasks as recognising friendly and unfriendly birds. And he says that another team of researchers, at the University of California at Davis, is preparing to use the same technique to study the brain activity of New Caledonian crows, a species that makes sophisticated tools. The team hopes to actually see the crows’ brains at work while they’re crafting tools.
Corvids seem to understand that other birds have minds like theirs, and their decisions often take into account what others might know, want or intend, according to several studies of crows, ravens and jays. Psychologists call this a theory of mind, and it’s a fairly sophisticated cognitive ability. Humans don’t develop it until late in childhood. Crows and their fellow corvids are social animals, much like primates, so theory of mind probably offers significant evolutionary advantages.
For one thing, it may help prevent food theft. Crows and ravens often hide food in caches and retrieve it later. “You can actually see them watching both the other birds that they are with and the humans, and if they sense that they have been seen, they will take that food and they’ll go and hide it somewhere else,” Innes said of the Maryland Zoo’s ravens. The birds appear to realise that watchers will know where they’ve hidden the food and might use that knowledge to steal it later.
Studies of several corvid species have documented this re-caching, as it is called. Sceptics of the birds’ advanced intelligence say simpler mental processes might prompt re-caching, such as making an association between being seen and later having a cache stolen.
Innes, however, is convinced that the re-caching is a sign that ravens have a theory of mind, based on her observation of re-caching behaviour in ravens at the Maryland Zoo. “Definitely,” she said. “I think it definitely indicates that.”
Other test results are harder to dismiss as simple association. When researchers in Austria hid food behind a partition, ravens found it, apparently by noticing where the humans were looking and following their gazes to the hidden food. “You’re using the person’s gaze to infer information about something you can’t see,” Marzluff said.
Brain imaging studies could settle the question, Marzluff said, because advanced cognition in all animals uses different areas of the brain than simpler associative learning.
Corvids’ toolmaking is much more clearly the product of sophisticated cognition, according to biologists who study them.
Several animals use found objects to get food, such as otters and sea gulls that use rocks to crack shellfish, and apes that use sticks to fish termites out of nests. But deliberately crafting tools is a much more sophisticated skill.
Only four species are known to actually make tools: humans, chimpanzees, orangutans — and New Caledonian crows. Although other corvid species have learnt to make and use tools in labs, only the crows found on the Pacific island of New Caledonia have been found to actually make tools in the wild.
With their beaks, the crows sharpen forked twigs into hooks for scooping larvae and worms out of holes in wood. The crows often spend more than a minute finding the right stick and then sculpting it into the right shape. Even chimpanzees don’t craft their tools so meticulously, and some researchers say that the crows’ work is on a par with very early human tools, such as spears and sharpened digging sticks.
New Caledonian crows even take steps to avoid losing their carefully crafted tools. Biologists recently discovered that the birds sometimes stash their hooks in holes, or simply stand on them, when they aren’t in use. The crows are especially careful when the risk of losing their tools is greatest, such as when the birds forage in high branches.
The aptitude for toolmaking is probably an instinct for most corvids, as it is for humans. Corvids use found objects as tools — ravens and crows, for example, drop nuts on to flat rocks to crack the shells — and nearly all corvids seem to have a knack for solving physical problems. In one set of experiments, captive crows figured out how to bend wires into hooks to retrieve food from a tube. And captive rooks, close relatives of crows and ravens, have done the same thing.
It’s unsurprising that chimpanzees and orangutans share so many abilities with humans, because they are very closely related to us, but it’s striking that corvids share so many skills once believed to set humans apart. After all, birds and mammals have spent the last 300 million years evolving on different paths, which produced very different brain structures and bodies.
Parts of the brain that evolved earlier than 300 million years ago, such as the primitive structures in the brainstem that control basic bodily functions, look the same in most animals, including primates and corvids. But structures that developed more recently, such as those involved in cognition, are organised very differently in birds than are they are in mammals.
Mammalian brains have evolved with what is called a laminar structure, in which brain cells are organised in six layers that make up the cerebral cortex, or forebrain. The cerebral cortex handles cognitive tasks, and it’s especially well developed in humans and our fellow apes, as well as other intelligent animals, such as dolphins and dogs. In the bird brain, a structure called the nidopallium caudolaterale handles cognitive tasks, and it’s especially well developed in corvids.
“All three of those animals have very large forebrains relative to the rest of their brains, for their particular group,” Marzluff said. “Certainly the forebrain of a bird and a mammal differ, but they have the same sorts of functions — that is, you know, higher-level thought and processing of sensory information.”
In birds’ brains, cells form clusters called nuclei instead of layers. For years, biologists held that the layered cerebral cortex gave mammals some cognitive advantages, but research on corvids has cast doubt on that assumption.
That such distantly related animals with such different brains could evolve such similar abilities is surprising, but when two different species face similar evolutionary pressures, natural selection can lead to similar traits. Biologists call this convergent evolution, and it’s the same process by which birds and bats both evolved wings.
At some point, biologists say, the ancestor of today’s corvids must have found itself in an ecological niche where intelligence boosted the odds of survival, so corvid brains evolved with cognitive abilities similar to those of primates.
Convergent evolution may have led to similar wiring despite the differences in physical structure between bird and mammalian brains. The network of connections between areas of the brain looks very similar in corvids and primates, and one recently published paper compared bird and primate brains to Apple and PC computers.
“At one level of analysis, they do the same things in a similar way, but viewed from another perspective, their operating systems are indeed different,” wrote the author, Cambridge University psychologist Nicola Clayton.
In the coming years, the differences and similarities between corvids’ mental operating systems and those of mammals will be analysed with brain imaging. At that point, said Marzluff, we may have “some of those answers” to the question of how smart crows really are.