Swimming in Uncharted Waters as They Learned to Train Whales

Much has been said and written in recent days about Sea World’s killer whales and the people who train them. Times of crisis are also times of excitement, and those who have even a passing familiarity with the issues have found it almost impossible not to leap into the fray.

Consequently, several distortions concerning the Sea World animal training program have surfaced in the last few weeks. While it would be folly to minimize the seriousness of recent misfortunes at the park, a look at how the training system was developed might provide a new perspective.

In 1971, I entered Sea World’s animal behavior department as did many novice trainers, working midnight to 9 a.m. cutting up fish to feed the animals. Over the course of the next 14 years, I worked my way up to director of animal behavior.

In the early days the atmosphere was loose and exciting. Although there was a certain informal discipline, the training methods would best have been described as “fly by the seat of your pants.” After all, no one had ever done, or even attempted, the things we were doing with the killer whales and other marine mammals. Humans had been working with these magnificent animals only a few years, and everything we tried was an experiment. We patterned our system on B.F. Skinner’s work and learned by trial and error.

The theories of operant conditioning were the basis of our early efforts. Unlike other forms of animal training, which employ adversive elements, marine mammal training developed using primarily positive reward. In fact, in the early days we were positive to a fault, rewarding everything the whales did and, in the process, creating a “spoiled child syndrome.”

With constant reward as the norm, the whales often had considerable difficulty coping with even minor disappointments. An upset child is one thing, but an angry killer whale is an entirely different matter. Wide-eyed and tense with frustration, the animals would open their mouths and shake their heads rapidly or swim abruptly away. Neither response inspired confidence. And so the long evolution of training technique began by teaching the animals to cope with a normal, healthy level of frustration.

Occasionally, a correctly performed behavior was not followed by the customary food reward. As predicted, “temper tantrums” were the result. But over time the whales’ angry responses lessened to the point that they were no longer a problem. After all, the animals’ food intake was not reduced. In fact, during this period the food consumption actually increased.

What changed was the way in which rewards were distributed, and the result was a steady improvement in the temperment of the whales. This technique became known by the non-scientific, but highly descriptive term: random and interrupted reinforcement.

Around the same time, the late 1970s and early 1980s, other modifications to the training system began to occur as well. If changes in reinforcement technique had a positive result, we wondered what the effect might be if other forms of variability were introduced. We began experimenting with different types of reward.

Petting and playing with the whales became a popular pastime, enjoyable for trainers and animals alike. We played games by the hour. “Block the Gate” was a special favorite. In this game, trainers would swim into position in front of a gate and attempt to block the whales’ passage. The whales always won of course, but we didn’t care. It was fun and that was the whole point.

Changes in behavior sequences, alterations in the form and direction of behavior, and unpredictable events (such as a trainer disappearing from one spot and reappearing at another), all contributed to the variety that helped make the whales’ environment constantly fresh and stimulating. Boredom was alleviated and the attention and interest level of the animals remained high.

It was during this time that we learned a fundamental lesson: if the goal is to develop a training system based on something other than hunger and food reward, building a positive relationship between trainer and animal is absolutely essential.

There were limits, of course, to what a positive relationship could produce. We finally came to realize that no single factor, but a combination of factors, was most effective in shaping and controlling the behavior of killer whales and other marine mammals. Positive relationships, variability in the environment, and correct distribution of rewards all played a part. It was this balanced approach, implemented by skilled and dedicated professionals, which created the spectacular behavior for which Sea World shows became justifiably renowned.

The importance of the trainers to this system cannot be overemphasized.

They generated many of the concepts and refined the techniques. They were the ones constantly in the water, not just on sunny summer days, but on cold and rainy winter days as well. They were the ones who, despite our best efforts, sometimes had to face frightening situations with animals 50 times their size. The system was good, but not perfect, and demanded superb judgment on the part of the trainers.

Anyone who has worked for a period of time with large powerful animals and claims never to have been frightened is either a liar or a fool. During 1983, the whales began breeding for the first time and their behavior became extremely erratic. As we worked our way carefully through that summer, one trainer was particularly unnerved by the turn of events. Floating in the pool while being slowly circled by a killer whale over whom you have no control is not an occupation for the faint-hearted. Yet he never quit. Day after day he quietly went about his duties. In the end, he triumphed over both the situation and his own fear. He was one of the bravest men I have ever known.

The professionalism he and his colleagues displayed, year in and year out, was the cornerstone upon which the Sea World animal behavior program was built. It has been a great program over the years and given us many enjoyable and exciting moments. Some people believe recent events at Sea World are the death knell of the animal training program. I do not agree. Whatever form the shows may take in the future, I am confident that the tradition of excellence will continue.