Friday, April 01, 2005

Bird Brain? It May Be A Compliment!

By: , , Gisela KaplanPh.D., and Lesley J. RogersD.Phil., D.Sc.

Our view of the human brain will be different depending on whether we believe it represents a unique biological system for producing higher-order cognition or, instead, is one possible system for producing such cognition. During most of the history of brain science, point out neurobiologist Lesley J. Rogers and ethologist Gisela Kaplan, the working assumption was that only the human neocortex made possible certain cognitive achievements. Now that assumption is being called into question by new research on a host of higher-order cognitive capabilities. Where is the competition in brainy behavior coming from? It’s coming from birds, the authors argue, citing new research on the surprising capabilities of some tiny brains.

We humans use our neocortex—the part of the brain that makes up most of the cerebral hemispheres —for most higher cognitive processes, including abstract thinking, problem solving, forming memories, and carrying out complex communication. The neocortex also plays a role in expressing our emotions and our personalities. Humans are not the only ones with a neocortex. All primates, indeed all mammals, have one, although none is as large and convoluted as ours. Scientists have tended to view the human neocortex as evolution’s crowning achievement, but are now discovering that evolution has led to some very smart brains with no neocortex at all. 

The tiny brains of birds do not have a neocortex, and their structures and connections are different from those of primates. This should not be surprising, because more than 200 million years ago birds branched from the line of evolution that led to mammals and on to humans via primates. What is surprising is that, despite the absence of a neocortex, birds can perform complex cognitive tasks once thought to be unique to primates and some even unique to humans. These tasks include seeing optical illusions, forming concepts, understanding the mental state of another individual, using and manufacturing tools, and communicating specific meanings to achieve specific goals. These discoveries challenge our notion of the evolution of the brain and show us that there is more than one way of wiring an intelligent brain.


In mammals, the top two-thirds of the brain —the neocortex—consists of six layers of nerve cells. In humans, the surface of this layered structure evolved to be so large compared with its thickness that it acquired folds and fissures. Bird brains, however, are primarily made of clusters of nerve cells (neurons) gathered together into discrete structures called nuclei. The traditional view was that the bird’s brain evolved and elaborated on an older structure called the paleocortex, which is still present in the mammalian brain but underneath the newly evolved, layered neocortex. The figure on the next page shows a side view of a section through the brain of a chicken and the visual pathways (black arrows) to the forebrain. The surface of the forebrain is smooth, and several nuclei are found along the visual pathways (labeled ROT, OPT, HA, and E in the figure). Note that the bird’s brain does not entirely lack layered structures, because the optic tectum (TO) is layered. It, however, is not part of the forebrain and is not used for higher cognition. 

For more than 100 years, biologists who studied comparative anatomy believed that animals without a layered cortex could not be intelligent. It was thought that their brains must be dedicated primarily to instinctual behaviors. But now researchers are observing a complex array of bird behavior in areas from visual processing to mental analysis. 


Birds are particularly good at visual processing and their abilities surpass those of humans for certain tasks. For example, even newly hatched chicks can recognize an object as a whole when it is partly hidden behind another object (a cognitive process called amodal completion). Human babies are unable to do this until they are four to seven months old, and, at first, they recognize an object only if its visible parts move in a coherent way (such as a partly hidden dog when the baby can observe the dog’s head moving and its tail wagging). This difference between the species probably developed because chicks are independent very soon after they hatch and must follow the hen and recognize her when she moves behind objects. Newborn humans, however, do not walk or move from their place independently, so they do not need this particular ability. It is delayed until the regions of their cortex that process vision are sufficiently developed. 

A study by Juan Delius, Ph.D., of Bochum University in Germany showed that some birds also perform better than humans in matching rotated objects or symbols. He tested pigeons on a task based on a question from the Eysenck IQ test for humans. The pigeons were trained to look at three keys in a row. An asymmetrical symbol was projected on the center key; the same symbol was projected on one of the side keys and a mirror-image of it on the other. To obtain a food reward, the pigeon had to peck the side key with the image that matched the one in the center. This sounds simple, but the symbols were rotated at various angles compared with the center image. Humans make consistently more errors as the angle of rotation increases, just as we have difficulty in recognizing a familiar face or object when we see it upside down. But pigeons have no trouble, regardless of the angle of rotation. This ability would serve the flying pigeon well, since birds need to recognize objects from many different angles from the air and ground.

Birds are able also to understand images made from an array of moving dots, as well as optical illusions. Newborn chicks form an indelible memory (or imprint) of the first thing they see—usually their mother —and follow it faithfully. In a series of experiments by Lucia Regolin, Ph.D., and colleagues at the University of Padua, newly hatched chicks were exposed to computer-generated patterns of white dots moving on a black background. The chicks became imprinted on the particular pattern they were shown and would approach it later. Some chicks saw a dot pattern generated by points fixed on parts of the body of a hen as she walked (biological motion). When humans look at such a computer-generated pattern of moving dots, we can see the walking hen, just as we can see a human walking when we look at a pattern generated by putting dots placed on the head, body, and limbs of a person. Other chicks were shown a moving pattern of dots generated from a rotating cylinder (nonbiological motion). The researchers were trying to determine if the chicks could distinguish biological from nonbiological motion; the answer was yes. Chicks imprinted on the pattern of the moving hen walked up to and stood next to this pattern when they were shown it later, but they did not approach a pattern of dots from the rotating cylinder. The opposite was true for the chicks imprinted on the rotating cylinder dots; they approached this pattern in preference to one showing biological movement. This ability to distinguish between biological and nonbiological movement has been seen as an example of higher cognition in humans. It relies on complex processing in more than one region of the cortex, including a part of the visual cortex known as the ventral stream and a part of the temporal lobe where many sensory inputs are integrated.

In another set of experiments, Elena Clara and Lucia Regolin imprinted chicks on a cone or a cylinder. Then they tested the birds with two-dimensional images that humans see as optical illusions of a cone or a cylinder. The first illusion is composed of a flat disc onto which off-center concentric circles have been drawn. When the disc is slowly rotated, we see a cone. The illusion of the cylinder is formed by two flat discs, overlapping but slightly off-center and rotating together as one piece but with the pivot at the center of one of the discs. We call the ability to see this kind of optical illusion “stereokinesis.” The chicks that had been imprinted on a cone and then presented with a choice of these two illusions at opposite ends of a runway approached the illusion of the cone, and those imprinted on the cylinder approached the illusion of the cylinder. In other words, the chicks could see the optical illusions. This discovery is another challenge to the traditional view of the bird’s brain, because we used to think that only humans could see optical illusions. Although there are no definite answers yet, we think that our visual cortex (part of the neocortex) is the main area of the brain involved.


Following the gaze of a person or animal is another complex behavior that involves interpreting what is seen. If I notice you looking at something, I might think that you are seeing something interesting and thinking about something different than I am thinking, and so I would look in the same direction as you are looking. These cognitive steps take place so rapidly that I am barely aware of them. If my view is obstructed or I do not see anything of interest when I follow your gaze, I may ask, “What are you looking at?” 

Some recent studies have shown that chimpanzees and domestic dogs will follow a human’s gaze. Although this might be a measure of intelligence, the animal also could be simply responding to straightforward behavioral cues to orient its direction of gaze, rather than being capable of understanding that the human can see something that it cannot. Nevertheless, gaze following is certainly valuable to social life; it can be a guide to the location of predators and food and to other information important for survival. 

Until recently, no one thought that a bird could follow the direction of gaze of a human, especially those species of birds that have their eyes on the sides of their head. This is difficult to test, because many birds have no clear pupil or white of the eye that shows in what direction they are looking. Bernd Heinrich, Ph.D., at the University of Vermont and colleagues at the University of Vienna tested to see whether hand-raised ravens (Corvus corax) could follow the direction of a human’s eye gaze. To their surprise, they discovered that not only did the ravens look up when a human looked up, but when the human was looking at something hidden from the raven’s immediate view by a barrier, the raven would also come over to the barrier and peer around it. If the raven were responding only to simple behavioral cues and not using higher cognition, it would do no more than stay where it is, look at the barrier, find it uninteresting, and go on with whatever it was doing before. Instead, the raven behaved as if it were aware that something interesting was located behind the barrier. 

When it comes to one bird following the direction of gaze of another bird, one might think that the first bird is merely following the direction in which the other’s head, particularly the beak, is pointing. But the explanation is not quite that simple, because many birds have two regions of the retina specialized for detailed vision, one for looking in front and one for looking to the side. Therefore, the beak may be pointing in one direction, but the bird may be looking sideways in another direction—or it may be focusing on both places at once. But it seems that birds do know when animals (including humans) with two eyes placed frontally, facing forward, are looking at them. They might have evolved this ability because many of their predators (for example, owls and many mammals) have frontal, forward-looking eyes. Also ravens might use the ability to recognize direction of gaze to decide when other birds are observing them as they hide food.


Birds have an enormously varied and complex array of vocalizations to communicate particular messages in specific situations. They use song to advertise their territories or to attract a mate, and they use food calls to ask to be fed or to signal to other birds in their flock that they have found food. They also make alarm calls to warn others of predators. Until recently, it was believed that these vocalizations were made unintentionally. Although other birds might obtain meaning from the communication, the bird calling might have no ability to decide what call it makes and when it does so. In addition, biologists thought that the bird hearing and responding to the call might make the appropriate response, such as hiding from a predator, without being aware of why it does so. In other words, we might think of birds as acting and interacting like little automatons. 

But research on alarm calls in chickens by Peter Marler, Ph.D., and Christopher Evans, Ph.D., (then at the University of California) changed scientists’ understanding. Chickens make entirely different alarm calls to signal the approach of a predator overhead versus one on the ground. When recordings of these calls are played to a bird in the laboratory, it takes the appropriate evasive action (crouching if it hears the aerial alarm call, and standing up straight and vocalizing loudly if it hears the ground-predator alarm call, in an attempt to drive off or deter the predator). Experiments comparing the responses given by the chicken on seeing a predator when alone and when in the presence of another chicken yielded the important discovery that the chicken makes alarm calls with the intention of warning other members of its species, rather than simply doing so automatically whenever it sees a predator. The chicken made alarm calls only when another chicken was present. When alone, it could suppress its alarm call and so avoid drawing attention to itself. In other words, the bird calls only when there is another bird to protect. 

Our own current research on the Australian magpie (Gymnorhina tibicen) is looking at communication in a wild bird. Magpies have a range of different alarm calls; some signal alarm in a general sense and others announce the presence of a specific predator. For example, these magpies make a distinct call when they see an eagle circling overhead. We have recorded these calls and then played them back to groups of magpies in their natural environment. By scoring their behavior before, during, and after the playback, we are able to determine whether the birds interpret the meaning of the calls. When we play a recording of the “eagle” alarm call, the magpies look overhead to scan the sky for a flying predator—a specific response that occurs rarely when other calls are played back. In fact, the response elicited by the eagle alarm call is even more specific: the magpies look overhead with their left eye, which means that they are using the right hemisphere to process the information (input from the left eye mostly goes to this hemisphere). Previous research in our laboratory has shown that this hemisphere is specialized for detecting predators. 

Most study of communication in birds has been on their songs. Biologists know much about the complexity of bird song and how songs are learned from other birds. Some species are capable of remembering many different songs that they have heard during a sensitive period when they were young. For example, young long-billed marsh wrens (Cistothorus palustris), learn songs by imitating older members of their species when the young wrens are between 25 and 55 days old. They acquire a large repertoire through learning and develop what is called “counter-singing with matching song type.” This means that a male wren can rapidly match the song of another male from a neighboring territory, which demands an extraordinary memory, at least in the short term. Other bird species also engage in counter-singing but usually of rather simple songs so that demands on memory are much lower. The marsh wren’s song is complex and the sequence sung varies greatly; nevertheless, the singer can match the song of its neighbor in this competitive singing duel—quite a challenge.

As far as we know today, only a rather select group of species and phylogenetic orders, including dolphins and whales (cetaceans), bats, and humans—and birds— can learn vocalizations. The neocortex was once thought to be an indispensable precondition for language and vocal learning, but complex vocal learning has been shown in parrots, hummingbirds, and songbirds, three groups of birds that are not closely related taxonomically. This suggests that vocal learning evolved independently at least three times among birds. Songbirds are capable of transcending simple auditory learning (the ability that most animals have to recognize specific sounds) and engaging in vocal learning or vocal imitation; that is, the sounds they hear can be matched by the sounds they produce (within limits imposed by anatomical and other constraints of the vocal apparatus). More important, birds learn vocal imitation by observing other birds, a process called cultural, as contrasted with genetic, transmission of behavior. 

Birds may not be as vocally prolific as humans, but their range of vocalizations may extend to thousands of sound combinations, as appears to be the case with brown thrashers and starlings, lyrebirds, Australian magpies, some smaller passerines, and parrots. The ability to learn so many different sounds may also show brain plasticity, and, at least for some species such as the canary, Australian magpie, and parrot, such plasticity is known to be maintained even into adulthood. Furthermore, according to Irene Pepperberg, Ph.D., who conducted a series of experiments over years at the University of Arizona on Alex, an African Grey parrot, the bird’s ability to learn human words may not be merely “parroting” but may be based on comprehension of the meaning of the words. That, indeed, would be another benchmark of what we regard as higher cognition. 


A bird’s ability to remember many different songs or other vocalizations is just one aspect of the complex cognitive process that is memory. If something moves out of sight, humans can still think about it and remember where it disappeared, but can birds do this? Evidence suggests they can even when they are very young. Giorgio Vallortigara, Ph.D., of the University of Trieste exposed chicks to small red balls soon after hatching, and later the chicks followed them as they would their mother. He then tested each chick in an arena with two screens placed a little apart. The chick had to stand in a small cage with transparent walls and watch the red ball as it was moved behind one of the screens. About two or three minutes after the ball had disappeared, the chick was released, and the experimenter watched which screen the chick approached. The chicks he tested chose the screen behind which the ball had disappeared, indicating that they remembered the location of the ball even though it had disappeared from sight. Scientists used to think that birds respond only to those objects and other stimuli that they could see at a particular time, but this experiment shows that they can hold and use memories of things past. Even though the time passed was very short for these young chicks, we can surmise that adult birds are able to remember hidden objects for a much longer time. No one, however, has yet tested this.

Remembering hidden red balls is an interesting laboratory experiment that shows an ability essential for survival, one that chicks must use to keep near the mother hen. Remembering the location of hidden food can also be essential to some birds’ survival. Some birds hide, or cache, their food at times of plenty and retrieve it later, sometimes after only a few days. Sometimes, in species living in harsh climates, they retrieve it months later when food is scarce. The Clarke’s nutcracker (Nucifraga columbiana), an example of the latter, has an extraordinary capacity for remembering the locations of thousands of cached seeds. It does so using cognitive spatial maps, that is, using complex geometrical concepts (for example, the central spot among several trees at some distance) rather than simply remembering the details of landmarks right next to the spot where each seed was hidden—a strategy that could be useless after snow has fallen. Humans also use cognitive maps, generated by using the hippocampus, a region of the brain beneath the cortex and also layered like the cortex. Birds have a region of the forebrain called the hippocampus, but it is not layered or structured like the mammalian hippocampus. Nevertheless, it seems to be able to perform many of the same functions, and it even increases in size in birds that cache their food when they perform this behavior. 

Study of birds caching their food in the laboratory and retrieving it after short delays allows close observation of their strategies. As Nicola Clayton, Ph.D., and Nathan Emery, Ph.D., at Cambridge University have found, western scrub jays (Aphelocoma californica) remember not only where but also what kind of food they have hidden. The researchers gave the birds two types of food, one relatively imperishable and the other perishable. The birds cached both types in notch holes drilled into branches arranged in a testing room. Then, the birds were taken to another location for several hours before they were again released into the testing room. The birds retrieved the perishable food items first and then the imperishable ones. In humans, we call this kind of memory “episodic” and say that it is an example of higher cognitive ability. 

Because scrub jays, like ravens, pilfer cached food from each other, one would expect that the bird hiding the food would want to avoid being seen by another bird. In fact, Clayton and Emery found that when a bird has been observed by another when it cached food, it will retrieve the food later when alone and hide it again in a new location. In the experiment that revealed this, the researchers allowed birds, tested one at a time, to cache food in two different trays, one of which was presented when an observing bird could see the caching and the other when the test bird was alone. Three hours later, the test bird was presented with the original two trays and a third tray. The bird retrieved and recached in the third tray more food from the tray that the observer had been able to see than from the tray in which it had cached in private. These results suggest that the bird is not only aware of being watched while it makes a cache, but also that it has some concept of the observer’s intention. If this were a human, not a bird, we would say that he knows what is on the observer’s mind. Primatologists call this having a theory of mind, something that they see as evidence of a high level of cognition. 


Forming abstract concepts is another complex cognitive process that we now know bird brains can handle well. Testing pigeons pecking at keys onto which images had been projected, Juan Delius, Ph.D., showed that they could form abstract concepts of “oddity” (pecking the odd picture out of a group) and “sphericity” (pecking the image of any rounded shape). They even formed the abstract concept of “water” and could learn to peck at any image with water in it, regardless of whether the water was in a glass, a lake, or a droplet on a leaf. 

Even young chicks have some abilities to form abstract concepts, as Giorgio Vallortigara, Ph.D., has found. The chick can learn to find food buried exactly at the center of arenas of different geometrical shapes (squares, triangles, circles). The chicks do not simply measure the distance from the walls, as was shown by testing them in arenas of different sizes as well as shapes, but they find the center using geometric cues. This ability in a bird, and a young one at that, is quite unexpected and takes us well away from the traditional view that bird brains are not only small but simple. 


The abilities to abstract, to remember the past, and to plan for the future are all considered highly complex forms of cognition, but some scientists think that an even more advanced trait is self-awareness. Humans can be tested verbally to establish whether a person is able to perceive himself as a separate entity in the world. Nonverbal tests of self-awareness often use a mirror, because we can recognize ourselves in a mirror and consciously make poses that alter the mirror image. 

Experiments with primates have shown that monkeys, with few exceptions, do not recognize themselves in a mirror. According to a study by Gordon Gallup, Ph.D., and colleagues, however, chimpanzees are able to do so. The researchers placed a red dot on the forehead of anesthetized chimpanzees and, after the chimps regained consciousness, showed them a mirror. If a chimpanzee merely pointed at the dot in the mirror, this would have been a sign that it did not recognize itself in the image. However, the chimpanzees in this test (and orangutans in a follow-up study) started wiping their own foreheads, suggesting that they were able to recognize the image in the mirror—with a strange red dot on the forehead—as their own. As the great apes are our direct primate ancestors and thus our closest relatives, these tests did little to change the perception that such self-awareness was unique to the most highly advanced primates and therefore linked to the increased size of the neocortex. But a recent experiment has made a radical change in this view. 

A study by Helmut Prior, Ph.D., Bettina Pollok, and Onur Güntürkün, Ph.D., tested whether or not a bird could recognize itself in a mirror. They chose magpies (Pica pica), hand-raised them, and then gave them a series of tests. The crucial test was placing a red dot on the bird’s throat, in a spot where the bird could not see it directly, and then watching how it behaved when facing a mirror. They found convincing evidence that the bird directed its attention to its own body and attempted to reach the spot where the red dot had been placed, rather than pecking at the reflection of the red dot in the mirror. So far, this is the only test of its kind in birds, but their research protocol suggests that in at least this one avian species something akin to self-awareness is present.


Our final example of the surprising cognitive abilities of birds is that of their making and using tools, long considered the hallmark of human superiority over other species. We know now that, although humans manufacture and use many more tools than animals, our actual ability to do so is not unique. Our close relative, the chimpanzee, selects twigs and breaks them to a length suitable for inserting into termite nests to fish out the tasty insects. This discovery, as well as other examples of apes using tools (for example, cracking nuts by using a hammer and anvil), was initially surprising to biologists. But then an even more surprising discovery was made: crows make and use tools. 

While studying the crows in New Caledonia, Gavin Hunt, Ph.D., and colleagues at Massey University observed that they not only use various sorts of probing tools to obtain insects from holes in trees, but they also manufacture them, using their beaks to cut tools from the leaves of pandanus palms. In addition to making and using these tools, the crows have been seen to store them in notch holes in trees and retrieve them to use later. This suggests that they have some notion of the function of the tools and can plan to use them in the future, a form of foresight that is another mainstay of higher cognition. 


The bird’s brain has remained small for a good reason: to allow flight. Brain tissue and bones are the heaviest parts of the body. Birds have evolved bones that are as strong as those of other species but much lighter; inside, they are a mass of connecting bony material with many small spaces filled with air. Birds’ brains are made of the same material as all brains but are connected in ways that seem to be more efficient for processing information. As an analogy, we might think of the first computers in the 1970s. They were large but had limited capacities compared with modern laptop computers, which are small and perform many more operations. Of course, a mammalian brain does not have limited capacities compared to a bird brain, but it is more cumbersome. It must be efficiency and different processing that enables birds to be as intelligent as many bigger-brained mammals. 

Birds have another exceptional capacity that helps them to have small but intelligent brains. They can generate new neurons when they need them and shed them when they are no longer necessary. Of course, young, developing mammalian brains generate new neurons, but once mammals become adults, they largely lose the ability to do this and to replace damaged neurons. Birds’ brains (or at least parts of them) retain the ability to make new neurons readily throughout the life span. For example, in those parts of the world where breeding is governed by the seasons, song birds sing only during the breeding season, and they do so using parts of the brain (nuclei) that have grown larger by recruiting new neurons that have recently been generated in other parts of the brain. Once the breeding season is over, the number of cells in these song nuclei declines because some of the cells die when not needed. 

Another example of this ability to build and dismantle parts of the brain is found in birds that cache their food. They remember where they have hidden the food using the hippocampus, which processes spatial information. The hippocampus expands when the birds cache food and contracts when they do not. Because processing spatial information is also important for navigating, it is not surprising that migration also increases the size of the bird’s hippocampus, as John Krebs, Ph.D., showed by studying the European garden warbler (Sylvia borin). 

It seems that birds make the best use of their small brains by time-sharing structures, letting them perform more than one operation well. When one part expands another may contract. More research is needed to confirm this, but scientists can say with confidence that the brain of the adult bird has a much more dynamic arrangement than the mammalian brain. This is why the bird brain has become a focus for neuroscientists trying to discover how adult human brains can be made to generate new neurons and so repair brain damage from stroke, trauma, or viral attack. If the conditions for making new neurons could be created in the adult mammalian brain, a whole new area of medical therapy would emerge. 

The differences between the bird and the mammal brain have given neuroscientists a powerful new way of understanding how brains grow, function, and repair themselves. On the other hand, the many recent discoveries about birds’ cognitive abilities have forced scientists to realize that bird and mammal brains have more parallels in their functions than previously believed, despite their long separate paths of evolution. As a result, biologists have begun to rename regions of the bird forebrain to make the similarities more obvious. For example, the forebrain of the bird has been renamed the pallium, which is another name for the neocephalon, a name that early neuroanatomists gave to the neocortex of mammals. All of this might seem like it is merely a matter of nomenclature, but the changes reflect our evolving understanding, and birds are teaching us that there is more than one way to evolve an intelligent brain. It is about time we began using the term “bird brain” as a compliment.  

About Cerebrum

Bill Glovin, editor
Carolyn Asbury, Ph.D., consultant

Scientific Advisory Board
Joseph T. Coyle, M.D., Harvard Medical School
Kay Redfield Jamison, Ph.D., The Johns Hopkins University School of Medicine
Pierre J. Magistretti, M.D., Ph.D., University of Lausanne Medical School and Hospital
Robert Malenka, M.D., Ph.D., Stanford University School of Medicine
Bruce S. McEwen, Ph.D., The Rockefeller University
Donald Price, M.D., The Johns Hopkins University School of Medicine

Do you have a comment or question about something you've read in CerebrumContact Cerebrum Now.