The Dolphin in the Mirror

by Diana Reiss

  Mirror self-recognition: You saw in chapter 6 that Presley and Tab each definitively recognized the image in the mirror as self. And Delphi and Pan's behavior at the mirror was strongly suggestive of that ability as well. Recognition of self is, for obvious reasons, the strongest indicator that the glow of consciousness is more than a dim ember. But recall the urgency with which Presley tried to find the mark on his side, contorting himself in the cramped corner of the indoor pool after sham marking; remember his double take on first seeing his image as he casually swam by the mirror, as well as his horizontal swirl performance; think about the intensity with which he searched for a glimpse of self, looking into his own eye just inches from him in the mirror. Consider all these things, and it is not a stretch to put yourself in his mind and experience a consciousness that you recognize in yourself.

  We assume consciousness in dolphins shines less brightly than it does in our minds. But this may be just an assumption. How would we know? How could we know? There is one thing we can be certain of, though, and that is that the texture of Presley's consciousness, the texture of all dolphins' consciousness, will be different from ours, just as Wittgenstein's lion inhabits a reality that is foreign to that of humans. One major reason: While humans are primarily visual creatures, dolphins experience their world primarily through sound, through their exquisite echolocation system. If you are a dog owner, or even simply a casual observer of a dog on a walk with its owner, then you are aware that a dog's experience of a walk is dramatically different from its owner's. The dog's world is constructed from a kaleidoscope of odors—a sensory realm that is mostly invisible to us—and visual images. In the same way, dolphins can perceive their immediate environment via their exquisitely detailed natural sonar. They receive acoustic pictures built from the returning echoes of the clicks that reflect off environmental features. The upper ranges of their sonar clicks are ultrasonic to us, beyond our perception. How they integrate and interpret the returning echoes of the clicks they've produced is still a mystery. No human-made sonar comes close to this natural technology.

  Lou Herman has been a professor of psychology at the University of Hawaii for four decades, and in 1970 he founded the Kewalo Basin Marine Mammal Laboratory in Honolulu to study perception, cognition, and communication in bottlenose dolphins. A little more than two decades later, he and his colleague Adam Pack founded the Dolphin Institute, which added conservation and public education to other research interests. Lou directed what is likely the most longitudinal cognitive research program with dolphins; it began in 2004 and has resulted in many groundbreaking experiments on dolphin perception and intelligence.

  Early on, Lou Herman bravely decided to investigate dolphins' language competence. I say bravely because the investigation of animal language has a long history of contentiousness, thanks in part to John Lilly. Whereas Lilly had dreamed of humans' one day having two-way conversations with dolphins in spoken English, Herman concentrated only on dolphins' ability to comprehend, not produce, language. And he didn't use spoken words, as Lilly had wanted to. Instead, he taught one dolphin, Akeakamai, an arbitrary system of hand gestures with a simple grammar, akin to American Sign Language, as some ape-language researchers had done. A second dolphin, Phoenix, learned another arbitrary system, dolphinlike whistles created by Herman and his colleagues.

  The Kewalo dolphins quickly grasped not only the meaning of the symbols (for hoop, ball, fetch, and so on, which can be thought of as the semantics of the language) but also the meaning of the word order (the syntax). For instance, when one of the trainers gave Akeakamai the symbols for hoop-ball-fetch—in that order—she would push the ball to the hoop. But when she was given the instruction ball-hoop-fetch, she took the hoop to the ball. We all know that word order in sentences is critical to comprehension, and we respond to it without thinking. Herman put it this way: "Syntax is what tells us that a venetian blind is not a blind Venetian." Both Akeakamai and Phoenix were almost flawless in their response to the meaning of word order in the simple, and sometimes not so simple, sentences Herman gave them. For instance, they knew that "take the pipe to the basket on the right" meant that the pipe should finish up in the basket on the right, not the one of the left. Lou's dolphins proved themselves at least up to par with any chimp in their grasp of semantics and syntax.

  Akeakamai and Phoenix also mastered other cognitive tests. For example, a trainer would sometimes ask one of the dolphins to fetch a novel object and do something with it, and the trainer would use a symbol for the object that the dolphin had not yet learned. The dolphin had no difficulty figuring out that the symbol must refer to the only object in the pool that it hadn't seen before. Akeakamai also understood the rather abstract concept of presence and absence, and used Yes and No pedals to respond to questions. For instance, when she was asked, in the form of the symbols that meant "ball" and "question," whether there was a ball in the pool, she would press a Yes pedal if there was a ball there and a No pedal if there wasn't. Sounds simple, doesn't it? But dealing with an abstract concept, as absence is, demands more brainpower than dealing with the concrete.

  Akeakamai spontaneously and on her own took this little exercise further. Akeakamai knew that the construction Frisbee-hoop-in meant "put the hoop on the Frisbee." One day, she was told to do this but there was no Frisbee in the pool, so she got hold of the hoop, went over to the pedals, and put it on the No. Another time, when the object to be moved, the hoop, was absent, she simply went to the pedals and pressed No with her rostrum. She had found a way to respond correctly. She had worked out how to say "There's no Frisbee" and "There's no hoop." She'd devised these responses to impossible situations on her own, with no training. That's a reflection of dolphin mind: problem solving and communicating.

  For me, one particular study that Herman and Pack conducted was remarkable and extremely revealing about dolphin sensitivities—and perhaps about their own form of communication. Akeakamai not only responded to the meaning of gestures with impressive accuracy but also showed herself to be superbly sensitive to the physical form of the gestures that carried their meaning. Ordinarily, a trainer made gestures standing at the side of the pool, moving an arm from a bent position to straight out in front, moving it across the body, and so on. In this particular experiment, the goal was to determine how dolphins interpreted signals that were visually degraded. Rather than generating gestures from the normal poolside positions, the trainer gave all signals from an underwater viewing window. The first stage of degrading the signal was to have each trainer dressed in black, with his arms in white. Akeakamai would therefore see just the arms as the trainer gestured. She was able to respond just as well as she normally did. The next stage was for the trainer to wear a black body suit, with only the hands in white. Akeakamai now saw only the hands, and she still had no problem. The last stage had the trainer totally in black and holding a short black stick in each hand, a white Ping-Pong ball at the end of each stick. Akeakamai still aced it, which astonished Herman and his colleagues because, given the minimal information, they themselves had tremendous difficulty identifying which gesture was which.

  This experiment was done quite some time ago, but when I saw Lou Herman present the videos at a conference I flashed on something I had experienced when swimming with dolphins. On occasion I would swim with Circe, Terry, Pan, and Delphi. While behind them, I noticed small white areas on the leading or back edges of their pectoral fins or flukes. From the front I noticed the tips of their rostrums were lighter as well. I used these markers as directional cues so I could maintain cohesion and velocity with the group and sense which way they would move next. Birds use visual cues in similar ways in maintaining flock cohesion. Perhaps the dolphins themselves used these white-tipped cues? Recently I was swimming with dolphins during a research trip with Daisy Kaplan, a doctoral student in my lab, off Bimini, in the Caribbean, and I wondered if they were using the same visual cues that I used with them.* Perhaps you have seen films of groups of dolphins swimm
ing in unison, twisting and turning, diving and leaping, as if they were perfectly executing a piece of choreography. How do they do it? How do they avoid bumping into one another? Visual monitoring of neighbors must be a part of the system, though probably not all of it. So what are they actually monitoring? It was a wow! moment for me when I saw Lou Herman's video. Dolphins have exquisite sensitivity to minimal information, and they can accurately act on it.

  Herman also taught the dolphins the symbol meaning "mimic," which told them to copy the body movements a trainer performed at the edge of the pool. The act of mimicry sounds simple, but it means that one individual must model its behavior based on that of another. And when the other is a different species with completely different anatomy, an extra cognitive step is involved: matching analogous parts of the observer's anatomy to the demonstrator's. Herman's dolphins had no problem copying a whole-body pirouette, relating whole body to whole body. No surprise there. They were also adept at copying a trainer bending backward from the waist with her hands by the sides of her head. Still not too challenging. When the trainer shook a leg, that should have presented a problem. But the dolphins shook their tails! We've already seen how accomplished dolphins are at vocal mimicry (more so than any animal other than certain birds and humans). This, combined with behavioral mimicry, makes dolphins the only species, aside from humans, that are capable of both forms of mimicry.

  Dolphins appear to be performing a dance as they swim synchronously in groups in the ocean. Who is devising the dance? And how is it communicated among the group? Lou Herman decided to dabble a little bit in this performance arena. He taught Akeakamai and Phoenix a symbol meaning "create," which told them they should go off and do something novel. (Karen Pryor had done a study earlier showing that dolphins could learn the concept of "do something novel and creative"!) He has a film segment that showed a trainer giving Akeakamai and Phoenix this command at the side of the pool. The two dolphins swam to the center of the pool and circled underwater for a few seconds. Then they dramatically broke through the surface of the water, spinning clockwise and erect, heads up, and squirted water from their mouths, all in such perfect unison it would have made Olympic synchronized swimmers proud. "None of this was trained," Lou Herman said, "and it looks to us absolutely mysterious. We don't know how they do it." These active, intelligent minds understand novel and create, and they can do it in teams.

  Lou Herman and his colleagues have produced many findings over the years that further indicate that dolphins are the cognitively complex creatures we intuitively feel they are. I've given only a glimpse of his work here, and I wish I had space for more. But I do want to add just one last finding of Lou Herman's. It's an experiment I wish I'd done.

  Dolphins' sonar systems are, quite properly, regarded with something approaching awe by technical and nontechnical people alike. They are the subject of both impressive data and questionable claims. In the first category, impressive data, the U.S. Navy has determined that a dolphin can accurately locate a three-inch sphere from a distance of four hundred yards. The second category, questionable claims, arises from the fact that the sound waves that dolphins produce are able to penetrate biological tissue. In principle, therefore, dolphins should be able to examine one another's internal organs at a glance. What they "see" and what use they make of such information is a mystery, but there are stories that imply dolphins sometimes know more than we think possible. I have heard numerous accounts of people who claim that when they swam in the wild with dolphins, injured parts of their bodies were scanned by the dolphins and manipulated. This may be the case, but it is certainly possible that one moves an injured limb differently, which could capture the attention of a dolphin and lead it to investigate. I've also heard numerous accounts of pregnant women who claim that dolphins used sonar on their abdomens. It's not surprising to think that dolphin sonar could reveal a pregnancy. After all, our less sophisticated ultrasound technology performs this task. I now conduct observations and cognitive studies with a social group of dolphins at the National Aquarium in Baltimore. The social group is composed of several generations of females and a few male offspring. We have observed an older female scan the abdominal area of her pregnant offspring. It would certainly confer an advantage to the individual and the social group to know who is expecting. Whether they are really detecting a pregnancy and understanding the ramifications, however, is far from clear.

  What isn't in doubt, however, is the insight into the dolphin mind that Lou Herman's beautiful experiment gives us about what dolphins "see" when they echolocate. Just as it is hard for vision-oriented creatures to imagine what a dog "sees" when it is creating an olfactory picture of its world, so it challenges our imagination to picture a dolphin's mind when it creates an acoustic picture of the world. But Herman and his colleagues provide compelling evidence that dolphins' sound-based perceptual systems produce something akin to visual images.

  Lou collaborated with Adam Pack and Matthias Hoffman-Kuhnt on this classic experiment with a seven-year-old female named Elele. Elele was an eager student in whatever she did, and she loved to be right, which she was most of the time. When a trainer gave her the "hooray" signal, both arms stretched up vertically, Elele would leap into the air, squeaking and clicking with delight. In this experiment, Elele had a lot to squeak about, because most of the time she was near perfect in her performance.

  The experiment was simple in concept but demanded careful execution. The question was this: Would Elele be able to visually identify an object after she had sonared it? In other words, does the sonar echo translate into a visual image in a dolphin's mind?

  Lou Herman and his colleagues had Elele examine the contents of a box that was transparent to sonar but visually opaque. Inside the box was a complexly shaped object constructed from PVC piping, measuring about ten inches in size. Elele's task was to visually identify the object she'd "seen" by way of sonar. It involved her discriminating between two very similarly shaped objects and picking the right one. Elele had never seen the objects before, so there was no element of recognizing something familiar. She had to form a clear image of a complex novel object based only on sonar and prove that she'd seen the object by choosing it over something very similar. She got it right almost 100 percent of the time. And she did it almost instantly. We still don't know exactly how the sonar-produced image is manifested in the dolphin's brain, but we can now say that it is not merely recorded in sound. Perhaps it is some kind of holograph? Only dolphins know.

  Elele showed herself to be equally skilled in the reverse direction, going from visual image to sound image: she had no difficulty identifying an object with sonar that she had seen visually just once. This was a very important study. It would be fascinating to repeat it while rotating the objects slightly in the visual-identification phase to see if dolphins can mentally rotate something. I bet they can.

  ***

  Spock and Shiloh were inseparable, always swimming together, always playing together, always resting together. They were a major item in the show pool at Marine World in Redwood City. Everybody loved this devoted pair. Spock was the male dolphin who had apparently tricked Jim Mullen into giving him multiple rewards for cleaning up the pool by bringing him multiple scraps of paper torn from one large piece. And Shiloh was the first dolphin I saw producing bubble rings, soon after I arrived at the facility in 1981. Then, quite suddenly, Spock became ill, and died. I'd known him for only a few years and hadn't worked closely with him in the way I worked with Terry and Circe and their boys, but I was stricken, as were Jim Mullen and the other trainers. And we were not alone.

  Shiloh was right by Spock's side as we lifted his inert body out of the pool. She looked bewildered and bereft. She wouldn't eat. She no longer swam around the pool with joie de vivre. She spent a lot of time simply lying on the bottom, deeply lethargic, stirring only when she had to surface to breathe. The other dolphins became very solicitous of her, swimming up to her side, apparently trying to encourage her to sw
im with them. But the grieving Shiloh would not be consoled. She continued like this for some time, and we became quite concerned for her health. Thankfully, after several days, like a human emerging slowly from mourning, Shiloh began to eat and socialize with the other dolphins. Oddly familiar. A pattern that connects.

  There are a number of words in the preceding paragraph that, strictly speaking, should be in quotation marks. Bewildered and bereft, for instance; and solicitous and grieving. These words are inferences of the states of mind of animals, not states that we know for certain. With known human behavior as our model, we thought Shiloh looked as if she were bewildered and bereft; her buddies behaved as if they were being solicitous of Shiloh, concerned about her well-being. If someone you knew behaved after the death of her mate as Shiloh behaved when Spock died, you would know she was grieving, and you would have a very good idea of how she felt, especially if you yourself had experienced such a loss. Shiloh looked as if she were grieving. Does that mean that she was experiencing something like the raw, searing, heart-rending emotions we associate with grieving? How can we know?

  Consider this. Back in the 1960s there were several reports of dolphins in aquariums displaying caregiving, or epimeletic, behavior with other dolphins. Occasionally, an individual dolphin in a pool becomes sick, loses strength, and is in danger of drowning because it is unable to swim to the surface to breathe. Often when this happens, other dolphins come to its aid; they stay at the sides of the ailing individual, holding it up so that it doesn't sink and drown. It looks as if the rescuers recognize that the individual is in trouble and care enough about it to keep it from drowning, often forgoing feeding for a long time while on the rescue mission. Is this a hard-wired response to a flailing individual, or do the dolphins really understand the situation and know what they are doing? Is their caregiving behavior evidence that they understand the plight of another?