The Dolphin in the Mirror

by Diana Reiss

  One day not long after the boys invented this new game, my colleague and friend Jill Tarter, the director of the Center for SETI Research, visited the lab. Jill spent a lot of her time looking at signals from space, and so she was happy to join me in looking over the many squiggly lines on the most recent spectrographs of the boys' whistles. She pointed to a particularly complex-looking whistle on one section on the spectrograph and said, "Look at that—doesn't that look a lot like the ring and ball whistles joined together?" I looked closely and realized she was right. I saw one long continuous whistle that began with a facsimile of the ring whistle and then connected to a facsimile of the ball whistle. This whistle was produced when the dolphins were engaged in double toy play with a ring and ball. During the following weeks, Brenda and I reviewed and analyzed the dolphins' vocal productions and discovered that they were indeed producing this novel whistle in the context of double toy play. They had invented their own new whistle with no guidance, reward, or acknowledgment from us; this was the dolphins' own doing. A reflection of these minds in the water.

  I had designed the keyboard so that when one key was hit and the appropriate sound initiated, there was a brief refractory period, about half a second, during which there could be no second sound. This meant that Pan and Delphi had never heard the ball and ring sounds combined as a single, continuous whistle. And yet this was precisely what they were producing. By placing the two sounds together, were they making a statement about the spatial relationship of the ball and ring together? Or about the game, or about something we couldn't even imagine? I suspected dolphins created combination whistles with their own sounds in their own environment when they wanted to communicate information to other individuals.

  When we wrote up our results for a paper,5 we pointed out that previous attempts in other labs to obtain vocal mimicry in dolphins through active training had required as many as a thousand trials, compared with just a handful in our case. One of the goals of the study, we said, was to answer these questions: What do dolphins learn and how do they learn it when given the freedom to interact with a self-reinforcing system?6 Pan and Delphi used that freedom to show us some of their best stuff. We had confirmed that dolphins join humans and some songbirds in the elite ranks of species that employ mimicry in vocal learning—and they do better on their own.

  4. Nonterrestrial Thinkers

  JUST A FEW WEEKS after I began working at Marine World in Redwood City, a big package was delivered to the little research building I'd established next to the dolphin research pool. My brand-new Sony video camera had arrived, and I could hardly wait to try it out. "Let's go shoot some footage in the hydrolator," I said to Bruce Silverman, my research assistant. My intention was simply to try out the new gear and get the hang of it, not to capture something I'd never seen before.

  The hydrolator was a very large, semicircular, glass-fronted elevator structure, positioned right next to one of the three large glass walls of the dolphin show pool. It was designed to glide slowly up and down, allowing people to get an underwater view of the action at different levels of the pool, which was more than twenty feet deep. But it never worked, so what we had was essentially a gigantic blue room. I had co-opted it to use as a private viewing station to observe Stormy and Schooner and the other dolphins that resided in the show pool. Bruce and I unpacked the camera equipment and did the little assembly that was required, and then I turned around to point the lens at the glass wall. There was Stormy right in front of me. Blowing bubbles. But they were no ordinary bubbles, and Stormy wasn't just idly letting air out of her blowhole. I watched as Stormy quite deliberately positioned herself horizontally almost on the floor of the pool and then moved her head upward with a small, quick jerk. At the same instant, she let out a large bubble that instantly formed itself into a perfect, lustrous bubble ring. It looked like a silver halo.1

  I was transfixed. I had heard anecdotal accounts of dolphins and whales blowing bubble rings, but I had never before seen it. It was a magical moment. Before I had a chance to savor it fully, Stormy blew another ring. She looked up and watched as it slowly expanded and rose toward the surface. And then she blew another. This one seemed to be a little smaller than the previous one, and it traveled upward faster. Within seconds the new ring met with the other ring and, in a moment of undulating creativity, they coalesced to form a much larger, hoop-size ring. Stormy had been watching very carefully too, and as soon as the two rings became one, she quickly swam straight up and passed through the hoop, shattering the large ring into a cloud of small bubbles all around her. Bruce and I looked at each other, mouths agape with astonishment. We had just witnessed this no-handed dolphin create and use a toy using only her blowhole.

  The infants of many nonhuman animals engage in play behavior with one another, of course. Puppies and kittens are at their cutest in rough-and-tumble play with their littermates or going over and around their ever-patient mothers. Most of us have seen video footage of chimp and gorilla infants frolicking. But this was different. This dolphin was actually creating an object of play and seemed to be quite aware of what she was doing. Stormy's behavior had every appearance of being premeditated and skillfully executed based on experience and practice, with nothing casual about it.*

  Although a Rosetta stone to decode their whistles still eludes us, there are other ways we can glimpse the richness of dolphin intelligence. Some of their behavior (like ours) is genetically programmed, so there's no need for conscious thought. But much of the dolphins' behavior, like that of other social mammals, is learned. A great deal of what they do is quite flexible and variable, based on their social and environmental circumstances. In this chapter we will consider three interactions of dolphins with their environment that give the impression of active minds at work and that demonstrate a degree of self-awareness that is rare in nature. The first is bubble-ring play and object manipulation; the second involves using the physical environment in innovative ways to forage for food; and the last is the sort of behavior that in humans would be called deception.

  ***

  Dolphins share an ancestry with even-toed ungulates, hoofed creatures such as bighorn sheep, goats, and cows, among others, although their last common ancestor dates back some forty-five million years. Odd as it may seem, given that length of time and their very different physical environments, dolphins and bighorn sheep share some body language. Both arch their backs as a sign of aggression; when surprised, curious, or frustrated, they both let out big breaths of air—this is called displacement behavior. When a sheep does this, it's not much to see. With a dolphin, of course, it produces a big, obvious bubble.

  One of the first times I saw this in dolphins was in the early stages of the keyboard work, before I started any experiments. I needed to know how good the dolphins' visual discrimination was, both underwater and in air. Would flat two-dimensional symbols be best? Or should they be three-dimensional? I started off with flat symbols, and these seemed to work pretty well. But I thought I should test three-dimensional symbols too. The first time I did this, Circe took one look and let out a huge bubble, as if she were startled! These involuntary exhalations produce bubble bursts, which look exactly as you'd imagine.

  Bubble caps are smaller, more sedate, and seem to be produced voluntarily. Bubble caps look like silvery mushroom caps that slowly rise through the water, swaying slightly and expanding as they ascend. Presley and Tab, two male dolphins that I worked with at a later period at the New York Aquarium, seemed to relish producing bubble caps, and they did it most days at a particular time. Dolphins normally feed underwater, of course. But Presley and Tab, like many dolphins in aquariums, were fed at the water surface, getting their fish as they held their heads above the water. After feeding, Presley went to his preferred area of the pool, and Tab went to his. The two young dolphins would then proceed, apparently quite deliberately, to release pockets of air from their mouths, producing a series of bubble caps, which they watched float shimmering to the surface. I spec
ulated that they had taken in air along with the fish. It was like watching two gentlemen who, after dining at the club, repaired to the smoking room for brandy and cigars, and burped in appreciation of the fine meal they had just enjoyed. This was definitely Presley's and Tab's personal thing, as I haven't seen other dolphins do it.

  While Presley's and Tab's postprandial display seemed to have a degree of intention and control to it, it doesn't match what is required for blowing bubble rings. When a dolphin releases a blast of air from its blowhole, the air bubble doesn't automatically form itself into a bubble ring; cigarette smokers cannot form smoke rings without a lot of practice. Although the physics of smoke being expelled into air is not exactly the same as air being expelled into water, a dolphin needs to exert the same degree of control over exhalation and tension of the blowhole to routinely blow perfect bubble rings. Dolphins usually have to practice, and they get better as they do, though I did once see Delphi blow a less-than-perfect bubble ring as a result of being surprised. (He quickly and enthusiastically perfected his technique.)

  Dolphins may invent bubble play on their own and learn bubble-ring production from one another by careful observation, something we would call social learning. It's not surprising then that separate populations produce bubble rings by different methods. This variation of style from one group to the next has all the hallmarks of the expression and transmission of culture. It's the product of active minds. I've already described Stormy's particular modus: the characteristic posture near the bottom of the deep pool, a quick upward jerk of the head with a concomitant release of a bubble that immediately forms a ring. Others in Stormy's pool followed very much the same procedure, although with some variations.

  Dolphins are ardent contingency testers. On one occasion in this group, Stormy was blowing single bubble rings and watching them as they rose. She had four spectators, including Schooner, each intently monitoring the activity. Schooner suddenly peeled off, swam to the bottom of the pool, and returned with a half-eaten piece of fish. He deftly dropped the fish fragment in the center of the bubble ring, and it rose under the influence of the rising ring, spinning violently in the ring's vortex. The five dolphins watched with what to all appearances was rapt attention.

  When the ring and its fish reached the surface, Stormy again swam to the bottom of the pool, took up the I'm-going-to-blow-a-bubble-ring position, and produced another perfect shimmering ring, once more watched by the other four dolphins. Schooner again left the group, retrieved a fragment of fish, and placed it in the center of the rising ring. The five dolphins watched keenly as the ring rose and the fish spun in its vortex.

  These were active minds at work, testing contingencies: What will happen if I put a fish in the middle of the bubble ring? It gave me the surreal impression of a group of scientists doing experiments, making observations, testing the dynamics of a physical system. We weren't the only ones doing the experiments.

  Dolphins can be very inventive in their bubble-ring play. At Sea Life Park Hawaii and at SeaWorld in Orlando, where the pools are relatively shallow, the dolphins turn to one side and blow rings horizontally. Then they spin the rings like kids with hoops, using deft flicks of their rostrums to spin the rings without touching them. Bubble rings tend to rise, of course, but the SeaWorld dolphins have perfected techniques for keeping rings close to the bottom. One trick is to keep the rings small. The dolphin does this by excising a small segment from a larger ring and then splicing the ring back together in what seems to be a single, expert movement of the rostrum. In a video sequence from SeaWorld taken by a spectator, one dolphin produced a ring, spun it a little, and deftly snipped off a smaller ring, which it then guided through the water. These dolphins and others at the National Aquarium in Baltimore also produce bubble rings and water snakes, undulating silver streams of air that they chase and then sometimes stitch together to form rings.

  The main difference between Marine World and SeaWorld bubble-ring production is their orientation: vertical versus horizontal. But at Sea Life Park Hawaii, the dolphins have developed a third alternative, as Ken Marten and his colleagues described in a 1996 article in Scientific American. The dolphin creates a vortex in the water by a short, swift movement of its tail. The second step is injecting air into it. "The pressure inside a vortex is lowest in the center, or ‘eye,' of the swirl," the authors wrote. "When the dolphin exhales into the vortex, the air migrates to the region of lowest pressure and is drawn out along the core of the ring-shaped vortex."2

  One young female dolphin, Tinkerbell, was even more inventive. She released a string of bubbles while swimming in a curved path near the wall of the pool. "She then turns quickly, and as the dorsal fin on her back brushes past the bubbles, the vortex formed by the fin brings the bubbles together and coils them into a helix."3 Tinkerbell had a second, similar technique in which she swam in a curved path across the tank, producing "an invisible dorsal fin vortex." She then retraced her path and injected a stream of air into the vortex, "producing a long helix that shoots out in front of her."4 Tinkerbell was apparently a budding fluid dynamicist of some considerable skill, because none of the other dolphins at Sea Life Park matched these feats, although they did enthusiastically blow bubbles in more traditional ways, learning from one another, perfecting their techniques through practice, and always showing a lot of interest in other individuals' performances, just as we saw at Marine World.

  In my current research program with dolphins at the National Aquarium in Baltimore, I've observed many of these same innovations in bubble play. Bubble play seemed to begin in this social group of dolphins with the introduction of Jade, a female dolphin from the SeaWorld "culture" of bubble-ring blowers. Her young offspring Foster (a male) and Bayley (a female), born at the National Aquarium, showed incredible creativity in their bubble creations in their first years. They often blew bubble caps and watched them as they rose to the surface. They chased water snakes they'd created by blowing air and then whooshing through it with their dorsal fins. When Bayley was two years old, she produced exquisite bubble rings with her tail, blowing air out quickly from her blowhole and flicking it with her tail in a movement that was so subtle we could hardly make it out even when we watched it in slow motion on the videotape. The young dolphins and some of the others at the aquarium also produced the classic vertical and horizontal bubble rings seen in other social groups.

  Dolphins' bubble-ring play is very complex and demands at least a modest degree of skill in both producing bubble rings and playing with them. The dolphins seem to embark on the task with some planning and deliberation, they seem to know what they are doing, and they seem to have some grasp of the physics of how to produce them and how to manipulate them in creative ways. In other words, bubble-ring play appears to be the outcome of higher cognitive activities, the product of active minds. If different dolphin populations in different facilities followed the same methods in producing and playing with bubble rings or had only slight variations on a principal theme, then one could reasonably argue that it was the product of a basic dolphin behavior pattern and didn't require higher cognitive function. But this is absolutely not the case. There is tremendous variability in bubble-ring behaviors among different groups of dolphins, and within any one group, the patterns of behavior change over time. Active minds are at work.

  ***

  In the mid-1990s, a few years after I left Marine World in Vallejo and moved to New Haven, Connecticut, I had the chance to study bubble rings more systematically with Brenda McCowan.5 She had completed her PhD and decided to continue the research in Vallejo as acting research director, working with a new population of dolphins, three adult females and a batch of youngsters. Brenda noticed that the dolphins had recently taken an interest in blowing bubble rings. She watched, casually at first, as infants seemed to learn bubble-ring production by observing their mothers doing it and then trying it themselves. They weren't very successful at the outset, but as they practiced, their expertise increased. Before lon
g, the youngsters were quite proficient, and they became inventive. They enthusiastically manufactured bubble rings, developing ways to, among other things, spin the rings and flip them 180 degrees. What really caught Brenda's attention, however, was the youngsters' apparent enthusiasm for blowing double bubble rings, one right after the other, so that the two coalesced into one large ring, just as I had seen Stormy do many years before.

  With our colleagues Lori Marino and Erik Vance, we collected data, systematically recording what the dolphins did, minute by minute, around bubble-ring play, capturing it on video and later having it analyzed by two different people.

  Four youngsters, Avalon, Brisbee, Liberty, and Norman (aged between three and eight years), were participants in the study. After three months, the youngsters' enthusiasm for double-bubble-ring play began to wane, and they became interested in other activities. But in the meantime we learned several things. Although the dolphins were capable of blowing bubble rings anywhere in the water, from the bottom of the pool to near the surface, and although the quality of the bubble ring wasn't affected by where it was produced, the dolphins overwhelmingly chose to produce rings near the bottom of the pool. This strongly suggests that blowing a bubble ring is a planned event. Making the rings near the bottom of the pool allows more time and space for play, including producing a second ring to join with the first.

  We found that the quality of the first bubble ring had a very large influence over whether a second one would catch and coalesce with it. An excellent ring (we rated them as excellent, good, fair, or poor) was twenty times more likely to have a second ring catch and coalesce with it than a ring of inferior quality. This meant that if the dolphins were indeed interested in two-ring events with coalescence, they had a strong incentive to monitor the quality of the first ring so they could decide whether to bother producing a second. That's precisely what we saw. Although a dolphin didn't always blow a second ring, the odds that it would do so after an excellent first ring were six times higher than after a ring of lesser quality. And the dolphins were always quick to show their displeasure at low-quality rings, instantly biting them or dispersing them with a flick of the tail. The dolphins monitored the rings with laserlike focus, watching what the rings did when propelled this way or that. This was a spectator sport for active watchers.