The Dolphin in the Mirror

by Diana Reiss

  ***

  First, a word about consciousness, a topic that has long enthralled and bewildered philosophers and biologists alike. To certain philosophers, consciousness is a private phenomenon, something we humans enjoy as individuals, giving one a sense of self and sometimes even an experience of transcendence. But according to this line of thought, consciousness doesn't materially contribute to physical well-being. And because it is private, consciousness (in humans and, if it exists, in animals) is therefore necessarily inaccessible to the forces of natural selection and to scientific investigation.

  To most biologists, however, consciousness is presumed to be a beneficial trait, the product of evolution by natural selection. If so, consciousness cannot be entirely a private thing, because it produces behaviors that are visible to natural selection. Individuals with well-developed consciousnesses should have an advantage over those who are less gifted. If this is true, it means that consciousness will become more sharply honed in the population over time, an evolutionary ratchet of a sort. Consciousness should, in principle, be accessible to scientific inquiry. But as you saw in the mirror self-recognition work with Delphi and Pan, and with Presley and Tab, self-awareness resists easy detection, at least in an objective, scientific manner. And is self-awareness the same thing as consciousness? I know I am conscious, because I experience "me." And I assume you are conscious too, because you are a fellow human being. But assumptions about nonhuman animals are much more tenuous, even though one might viscerally feel the presence of consciousness in a nonhuman, sentient animal.

  Nonetheless, let's consider what the utility of consciousness could be, why it would have evolved through natural selection. Or, to put it another way, How does it benefit a human or a dolphin to look into a mirror and recognize the image as one's self?

  Although consciousness is what we experience as self, most of what goes on in our minds is veiled from our conscious experience. Most of the brain's activity is concerned with maintaining the myriad physiological and motor systems, keeping the biological machine in good working order. We have no need to be conscious of the constant monitoring of activities and the requisite activation signals. Even an action that requires intense concentration to learn, such as riding a bike (which demands, among other things, constantly monitoring balance) or serving the ball in tennis (which requires careful monitoring of how the ball is thrown into the air and how to swing the racket), become automatic once one becomes proficient. What had been a conscious effort no longer is. And many a person has had the experience of being at the wheel of a car and suddenly realizing that he has not been actively aware of driving the car for the past couple of miles. The conscious mind was "elsewhere" while the unconscious mind was dealing with the practicalities of driving safely along the road.

  All of our active behaviors—getting out of bed in the morning, cooking breakfast, going to work—can be accompanied by conscious feelings and explanations concerning what's happening: wishing to stay in a warm bed for just a few more minutes; feeling hungry; feeling the need to make a living. In other words, the conscious mind has the capacity to monitor one's state at any particular moment of the day and to report it to the self. But, as with the "driving blind" experience, it is possible to imagine doing all the above activities on autopilot, without emotions or explanations coming to mind.

  Even the smartest of species with the cleverest of brains often operate on cognitive autopilot, deftly navigating the daily practicalities of subsistence and social interaction while beyond the realm of awareness. How, then, is consciousness an advantage? It's the first step in social reasoning. An individual who is aware of his own actions and emotions under particular circumstances is, in principle, able to predict the actions and emotions of another individual under those same circumstances. In the game of social interactions, as among the complex social systems of chimps, self-awareness confers a distinct advantage. An individual endowed with consciousness would be able to navigate the ever-changing web of relationships among the alliances within its community far more effectively than a behaviorist. Once the faintest spark of consciousness arose in the minds of individuals in a species, the forces of natural selection would inexorably fan it, generation after generation, so that eventually it would glow brighter and brighter. Ultimately that species would possess an aspect of what has been called the theory of mind—the ability to predict what is in the minds of others based on one's own experience.*

  In the literature of evolutionary biology, the discussions around the existence and utility of consciousness among nonhumans focused exclusively on higher primates, notably the great apes and some of the larger Old-World monkeys, such as various species of baboon. This primate-centered focus was presented by primatologists David Premack and Guy Woodruff in a landmark paper titled "Do Chimpanzees Have a Theory of Mind?"21 published in 1978. In the introduction to the paper, Premack and Woodruff said, "An individual has a theory of mind if he imputes mental states to himself and others." Which is an appropriately academic way of saying "I have an idea of what's on your mind, because I know what's on mine."† Premack and Woodruff answered their own question affirmatively: yes, chimpanzees do have a basic theory of mind.

  By the 1980s, most biologists agreed that chimpanzees have the kinds of minds they have because of the highly complex social environment in which their ancestors evolved. Individuals produced more offspring if they were socially adept; social intelligence—the ability to correctly assume what was going on in the minds of others—was a powerful tool. Homo sapiens occupied the peak of this mountain, but the great apes were allowed to camp in the foothills; they were not as cognitively endowed as us, of course, but they were above the common ground of the rest of the animal world.

  Do dolphins have creative intelligence, as chimpanzees do? Absolutely. Do dolphins have a theory of mind to the extent that chimpanzees have? I have no doubt that they do. Dolphins are, in the words of Louis Herman, "cognitive cousins" of chimpanzees.22

  As it turns out, dolphins have other cognitive cousins, in addition to humans and chimpanzees. Gordon Gallup first suggested that dolphins might pass the mirror test because, like chimpanzees, dolphins have large, complex brains, lead complex social lives, and show evidence of empathy toward others. Gallup reasoned that these factors correlated with the ability for mirror self-recognition in chimpanzees, so he was not surprised when Presley and Tab passed the mirror test.

  Elephants also have large brains and show empathy, so a few years after we completed the test with dolphins, I teamed up with Emory University primatologist Frans de Waal and his graduate student Joshua Plotnik and tried the mirror test with three adult female elephants at the Bronx Zoo. Josh and I spent hot summer days atop the roof of the elephant house filming the elephants' behavior when they were in front of a jumbo and very strong eight-by-eight-foot mirror. This time, although the animals had no hands, they did have trunks. Remarkably, the elephants and dolphins showed strikingly similar behaviors and went through the same three stages as the chimps had upon their first encounter with mirrors. Happy, one of the three elephants, passed the mark test by using her trunk to pat the mark we had placed on her head.23

  We submitted the manuscript reporting the study to both Nature and Science, but neither journal even sent it out for review. Again, it was deemed of insufficient interest for the general public. The paper was published shortly thereafter in PNAS, and as with the dolphin paper, the story was widely covered in the media. My favorite media coverage was by Associated Press writer Andrew Bridges, who opened his article with "If you're Happy and you know it, pat your head." I now tell audiences that I work with "big gray animals with big gray brains," and they are all smart as hell! (See figure.)

  HUMAN

  DOLPHIN

  ELEPHANT

  Some have argued that dolphins' brains need to be as remarkably big as they are in order to facilitate the dolphins' superb echolocation abilities, which exceed even the U.S. Navy's most advanced equipment. Perh
aps this impressive biological skill requires an unusual amount of brainpower. Yet there's no evidence to support this idea. Bats, with much smaller brains, have a comparably sophisticated natural sonar system. A controversial suggestion was made more recently: perhaps much of the dolphin's brain is made up of thermogenic glial cells, which would have helped dolphins survive the decrease in ocean temperatures that occurred about forty million years ago. In this view, the dolphin's large brain is an accidental benefit of its having needed a kind of internal electric blanket. But the most widespread view is that the brains of dolphins and other cetaceans evolved to support complex cognitive abilities; this is the social competition thesis. It's certainly my view.

  Just like chimps, dolphins need to succeed within a complex social system. They face changing conditions that require complex cognitive abilities. Dolphin societies (also like chimps') are fission-fusion. There are shifting alliances among relatives and nonrelatives. There are even alliances of alliances in bottlenose dolphins, which is rarely seen beyond human societies. Dolphins are active social learners—they learn by observing the behavior of others, very much like us. Adults and young continue to show high levels of behavioral flexibility in the face of changing social and environmental scenarios. Dolphins learn much of their vocal and behavioral repertoires and variations within their social groups. You might say they share a culture: New vocal signals and behaviors are transmitted among members of a social group and from generation to generation. This is another trait that is shared between dolphin and primate species.

  When two different species possess a similar behavior or body form as a result of being exposed to similar environmental circumstances, biologists call it evolutionary convergence. With dolphins and chimpanzees, with self-recognition and other higher cognitive abilities, we have a remarkable example of cognitive convergence. Dolphins and great apes last shared a common ancestor thirty million years before dinosaurs became extinct. Their brains' architectures are dramatically different. Yet their abilities have evolved similarly, for similar reasons.

  The fact that we have evolutionary convergence to similar brain functions with differing brain structures is forcing cognitive biologists to rethink what underlies consciousness and the sense of self. Perhaps it isn't so important to discover the precise location in the brain of the sense of self. Perhaps it is an emergent property of any brain that has crossed some encephalization threshold or that contains the right number of neurons, has the right extent of connectivity, and includes the right organization of the connections. Whatever the answer, long-held assumptions about a revered function of the human brain have been shattered by watching dolphins in the mirror.

  8. Reflections on Dolphin Minds

  AT MARINE WORLD, in Vallejo, we were lucky to be near the Lucasfilm company (creators of Star Wars and all that). One of the model makers and creature builders there, Mark Thorpe, became very interested in our dolphin work, and he often made exotic objects for Delphi and Pan's toy collection. But as is so often the case with kids (the dolphins were just three years old at this point), they were much more interested in mundane objects than in the fanciful creations from the Star Wars folk. A boat buoy in the shape of a fish, for example. It was physically robust, made of the same tough white plastic that's used in generic boat bumpers, which is a good thing for dolphin toys.

  Delphi would grab the fish by its tail, swim directly to the bottom of the pool, and then release it. The fish would make its stately way to the surface, through sixteen feet of water, swaying from side to side as it went. The first time Delphi did this, he watched the fish's progress very carefully. He then zoomed to the surface, grabbed the fish again, took it to the bottom, and released it. This time Delphi swam up to the surface alongside the fish, swaying side to side, exactly mimicking the buoy's motion. It was like a pas de deux between a piece of plastic shaped like a fish and a marine mammal pretending to be a plastic fish. As I so often did, I gave enthusiastic encouragement from the side of the pool, shouting, "Good boy, Delph! Good boy!"

  And as he so often did, Pan closely watched what Delphi was up to. Very soon Pan swam over, and a new game ensued. From then on, the boys took turns taking the fish to the bottom of the pool and accompanying it in a dance to the surface. This was such a dolphin-like episode: Delphi observing the fish's movements, then imitating them; Pan watching them both, then taking his turn, following exactly what Delphi had done; and then the two of them taking turns cooperatively. Dolphins learn through imitation in a social context, and this was a wonderful example. But it didn't end there.

  One day not long after the fish game began, Delphi was on his way to the bottom of the pool, fish in his jaws. But before he got to the bottom, the fish slipped out of his grasp and prematurely started its return to the surface. I noticed that the fish wasn't doing its usual dance, no doubt because of its unusual orientation when it was released. But what happened next was surprising: for the first time, Delphi let out a big bubble of air that quickly morphed into a beautiful, silvery bubble ring. "Good boy, Delph," I shouted, and made a big circle movement with my finger, mimicking the shape of the bubble ring.

  I had seen other dolphins blow bubble rings in the show pool, but neither Delphi nor Pan had done it, much less in the research pool. Delphi's bubble ring was pure serendipity, I'm sure of it. He had simply let out a burst of air on being surprised at the fish's unexpected release and unusual ascent.

  "Good boy, Delphi!" I repeated, and again described a big circle with my finger. Delphi was at the surface by this point, watching me closely. He immediately swam back to the bottom, stationed himself in a deliberate manner, and blew another perfect bubble ring. Fast learner. The ever-vigilant Pan was watching, and he immediately came over to where I was standing at the pool's edge, swam to the bottom, positioned himself with deliberation, and blew a perfect bubble ring too. So began a period of a couple of weeks when the boys engaged in bubble-ring behavior, often when I was standing at the pool's edge, usually taking turns. As they alternated making bubbles, they would look up at me. Was it for my benefit? I had certainly lavished them with vocal praise and lots of attention. Again, so dolphinlike: they observe; see what works; see what gets attention; and then execute it perfectly.

  I give this example of imitation and rapid learning as a lead-in to asking, Just how aware are dolphins? Every creature, no matter its cognitive endowment, is aware of itself in some sense, otherwise it would not be able to navigate the physical environment, find mates and food resources, and avoid predators. It's a matter of survival. What is it like to be in the mind of a dolphin in terms of self-awareness? How brightly does the spark of consciousness glow in these minds in the water? There is no way to answer this question directly, of course, but we can examine behaviors that reveal active minds, minds that are constantly searching, constantly exploring, constantly learning. The fish-and-bubble-ring story is one such behavior. But consider the behaviors mentioned in earlier chapters, along with some new examples:

  The time-out episode: Recall young Circe, the first dolphin I worked with, in the small marine zoo at Port Barcares, near Perpignan, in southwest France. Circe had been observing my behavior, and she used it right back at me when I gave her a food she didn't like. This episode was, for me, a visceral experience as well as a scientific observation. I felt that crunch in the stomach, the shock of recognition. This wasn't just mimicking; it was communication.

  Keyboard work and vocal mimicry: It's hard to imagine anything more alien from Delphi and Pan's natural environment than arbitrary symbols on a keyboard, and yet, without being schooled in the relationship between objects and symbols, they learned for themselves very rapidly through experience. And at the same time, they learned through experience the association of these symbols and objects with the arbitrary sounds we created.

  Delphi and Pan quickly learned to imitate the arbitrary sounds, produced their own accurate renditions of them before they hit the appropriate symbols on the keyboard, and then spo
ntaneously used them when playing with the corresponding objects. They never matched the sounds with the wrong symbol or object. Pan even seemed to use the ball whistle in a vocal exchange with Delphi, after which Delphi passed him the ball.

  The other incident around the keyboard work that speaks of active minds involved Pan and his predilection for fish. Recall that on one occasion we decided to remove the fish symbol from the keyboard since Pan repeatedly activated it because he loved getting fish so much. But when he saw the keyboard with no fish symbol on it, he apparently assessed his own situation—his desire for fish; the keyboard situation—no fish symbol; and then figured out how to communicate his wishes to us, despite the hurdle he faced. He searched around the pool, found a remnant of fish, brought it to the keyboard, and touched a blank key with it, as if to say, I want fish! Creative problem solving.

  Bubble-ring behavior: The aesthetics and easy inventiveness that take place around the bubble rings belie the cognitive abilities that bubble-ring production demands. There has to be the initial learning of how to make them, by chance (as in the case of Delphi) or by observation (in Pan's case); anticipatory planning and motivation; and an intuitive sense of the physics of what makes good rings and how they might be manipulated without destroying them. So here we have dolphins making arbitrary play objects appear out of thin air, planning and releasing the rings with exquisite timing.

  Deception: Having another individual believe, based on one's own calculated behavior, that something is true when it isn't certainly demands high cognitive abilities; one has to manipulate the environment so that another individual is somehow fooled by an otherwise innocent action. True deception implies an individual knows the rules of the game and is manipulating them. It demands an appreciation of how the self's honest or dishonest act might be interpreted by the other. As you have seen, dolphins are quite capable of deception.