Mind Matters News and Analysis on Natural and Artificial Intelligence

Crows Can Be as Smart as Apes

But they have quite different brains. The intelligence doesn't seem to reside in the details of the mechanism
new Caledonian crow with stick tool/Auguste von Bayern

Studying animals’ intelligence has taught us many things. But in some ways, it has deepened the mystery of intelligence.

We might have thought that intelligence, in terms of individual learning ability, would gradually increase among animals, from invertebrates to vertebrates, from exothermic (cold-blooded) animals to endothermic (warm-blooded) animals, from reptiles to primates, culminating in man. In that case, intelligence would be associated with the increasingly complex brain structures that enable it.

Research has demonstrated the opposite. Some (though not most) birds can learn new skills as readily as primates. New Zealand crows can create compound tools “by combining two or more otherwise non-functional elements, an ability so far observed only in humans and great apes”; they can though estimate sizes and shapes of paper for a reward, though they do not use paper in the wild.

Some New Caledonian crows can use three tools in a row to reach food. As a result, some researchers have called crows “feathered primates.” New Zealand crows’ causal understanding (within limits) is said to rival that of 5-7 year old children. Or 7- to 10-year old children.

It’s not just crows, either. Pigeons can work with numbers up to nine with a success that is “indistinguishable from that displayed by monkeys.” Even the chicken’s intelligence “startles” some researchers (“communication skills on par with those of some primates”).

Do such skills “redefine intelligence,” or explain “how technology evolved,” as sometimes claimed? Well, not really. It’s the same sort of intelligence as we find in primates but exhibited in a quite different brain structure:

Corvids, such as crows, ravens and magpies, are among the most intelligent birds on the planet—the list of their cognitive achievements goes on and on—yet neuroscientists have not scrutinized their brains for one simple reason: They don’t have a neocortex. The obsession with the neocortex in neuroscience research is not unwarranted; what’s unwarranted is the notion that the neocortex alone is responsible for sophisticated cognition. Because birds lack this structure—the most recently evolved portion of the mammalian brain, crucial to human intelligence—neuroscientists have largely and unfortunately neglected the neural basis of corvid intelligence…

If neuroscientists want to know how brains produce intelligence, looking solely at the neocortex won’t cut it; they must study how corvid brains achieve the same clever behaviors that we see in ourselves and other mammals. Grigori Guitchounts, “Why Neuroscientists Need to Study the Crow” at Nautilus (2016)

So brain structure cannot be the simple explanation we might have supposed. Can we turn to evolution for an explanation? Not really. We don’t know when, how, or why some birds started making and using stick tools so it’s hard to do more than speculate about the evolution of the trait.

There’s another problem as well: The birds display these skills in a setting that humans have designed so that the birds might succeed (so that, in turn, the humans can publish in a science journal, of course). It’s fair to ask what role such focused, goal-oriented human intervention plays in increasing certain types of intelligence in the birds. The clear, systematic reward system not only exists but is designed to be apparent to the bird. Nature does not usually arrange such systems. If a bird seldom gets exactly the same test repeated in a natural state, its ability to learn may be frustrated, irrespective of its native intelligence.

Some might respond by saying that only some species of birds are as clever as primates are with tools. That’s true, but then only some species of mammals are as clever as primates are about anything.

We don’t know very much about intelligence in animals and birds but we do know that it is not a function of the mechanical organization of their brains. That raises a question: If artificial intelligence really came to exist, would it depend on the organization of the mechanism?

Next: Even Lizards can be smart

See also: Furry, feathery, and finny animals speak their minds. Listen.


Crows Can Be as Smart as Apes