Strict adherence to this injunction has meant that scientists in general and biologists in particular have always preferred to explain the behavior of animals in terms of reflexes and genetics rather than entertain the idea that animals have a mind. This approach has been enormously successful for most of this century. It has explained very neatly how dancing bees can tell their hive-mates the direction of a rich source of food. It has given a perfectly acceptable explanation for the courtship behavior of sticklebacks and the way in which a gosling, recognizes its parents. It can even resolve the question of why some animals mate with only one partner, and others with several. Onlv quite recently have biologists begun to observe and think about behaviors which are difficult to understand without assuming that animals have minds.

How, for instance, do you account for the sort of behavior observed by primatologist Frans de Waal and one of his students in their work on a group of captive chimpanzees at Arnhem Zoo? One of the chimps in their group, a male named Yeroen, hurt his hand in a fight with another male, Nikkie. It seemed to be troubling him, because he was limping. Frans De Waal takes up the story:

The next day a student, Dirk Fokkema, reports that in his opinion Yeroen limps only when Nikkie is in the vicinity. I know that Dirk is a keen observer, but this time I find it hard to believe that he is correct. We go to watch and it turns out that he is indeed right: Yeroen walks past the sitting Nikkie from a point in front of him to a point behind him and the whole time Yeroeii is in Nikkie's field of vision he hobbles pitifully, but once he has passed Nikkie his behavior changes and he walks normally again.

Whatever is going on here, it is difficult to see how it can be explained convincingly by reflexes or genetics. It is anecdotes like this that are forcing biologists to re-evaluate the animal mind. But there is now a huge body of successful work in biology that has been accumulated thanks largely to teachers insisting that their students do not treat animals as if they had minds like humans. Among these people anthropomorphism - attributing human qualities to animals is a very dirty word. Persuading anyone from this school of thought that 'there is not the least doubt' that animals like chimpanzees and dolphins do indeed have minds is not an easy task. They would probably argue (though perhaps without much enthusiasm) that Yeroen must have learned through previous experience that when he limped he avoided trouble with Nikkie - a simple conditioned reflex.

For many of the newer generation of biologists who are steeped in the detailed behavioral studies of chimpanzees, gorillas and the like, the product of thousands upon thousands of hours of painstaking field observations over the last thirty years or more, the results of the dolphin studies come as much less of a surprise. Dolphins, after all, are intensely social animals, and it is among the social animals, especially monkeys and apes, that intelligence has evolved on land. Many researchers are now suggesting that it may in fact be social pressure that encourages the developmerit of intelligence in general and a sense of self in particular. The abilltv to imagine what another animal will do in a given situation probably allows chimpanzees to compete successfully for mates, to avoid fights and so on. And imagining what an animal is thinking and feeling inevitably must involve a mental process along the lines of 'if I were her...' The same is equally true for dolphins.

All this leaves one major question unanswered. If dolphins are intelligent, self-aware, and can understand a language similar to ours, do they have a language of their own? If the answer turns out to be yes, then many more questions automatically follow. Will we one day be able to communicate with dolphins in the way that John Lilly thought we might? Are the dolphins, as they swim through the sea, composing dolphin sagas for each other? Do they have their own history of interactions with humans, to be passed from generation to generation in some whistle-language story or sound-picture film that we cannot decipher?

One attempt to explore how dolphins might use language between themselves is a study being carried out by Diana Reiss and her co-workers at Marine World Africa in California. They built an underwater keyboard with abstract symbols which the dolphins could press with their beaks. Each key produces a computer-generated whistle which the dolphins can hear underwater. At the same time something happens the dolphins press the ball symbol, a ball is thrown into the pool. The rub symbol produces a pat for the dolphin.

What this experiment has shown so far is not wildly exciting. The dolphins learn to copy the artificial whistles, and they associate these whistles with the objects subsequently placed in the pool. When playing with a ball, for example, the dolphins reproduce the ball whistle. What is interesting about this approach, however, is that it does not involve training the dolphins to perform particular human-devised tasks, but allows them to develop their own patterns of behavior - to play with the equipment in their own way. It is the sort of development that might in future allow researchers to investigate the nature of the dolphins' own vocal repertoire.

Another route for examining dolphin language may turn out to be via the study of how the animals coordinate their behavior. Among the 'words' understood by Phoenix and Akeakanial at the Kewalo marine laboratory are 'creative' and 'tandem'. While being trained to respond to the 'creative' command, the dolphins were given a fish every time they did

something new. The concept is difficult for animals to grasp, since it seems at first as if they are just being teased - what was right yesterday is wrong today. But once they learn the rules, a spectacular sequence of jumps, fin slaps, rolls, twists, and so on follows. 'Tandem' requires that the two dolphins perform the same action in synchrony. The interesting command is 'Tandem Creative'. An acceptable response to this might be that the dolphins swim around the tank together, leap out of the water, each doing, a clockwise spin, and fall back in head first. It is the sort of display the males of a coalition might put on to impress a female they are herding. The question is, how do they do it? How do they each know what the other is planning?

One possibility is that they do not, but that one animal leads the way while the other follows so closely behind that it appears they are moving together. Dolphins are indeed incredibly good mimics, and have been seen copying the swimming patterns of seals, turtles, skates and rays, and, of course, humans. But if this sort of mimicry is really the source of the synchrony in tandem-creative performances, it is truly astounding. In a recent study, out of 467 observed tandem-creative displays, a leader could be detected by human observers in only thirty. Another possibility is that one of the dolphins can simply predict what the other is going to do as it swims around the tank at the start of its routine, but with over forty possible behavior patterns to choose from, the chances of this being feasible are remote. It is certainly conceivable that the dolphins are using sound to communicate their intentions to each other, and if so this would provide a fascinating experimental context in which to study dolphin communication.

It may be, of course, that we are looking in the wrong place altogether. Humans communicate primarily through sound, which leaves us free to look around us as we walk and talk. As speech evolved in our hominid ancestors, it may well have been important for them to concentrate their visual attention on looking out for predators. Conversely, dolphins, of course, rely largely on sound for this purpose, and to avoid being distracted might be expected to use sight as their primary means of communication.

Could it be that the messages,passing between dolphins are chiefly visual, rather like the sign language of the deaf? It is after all the dolphins' play behavior that we take to be the equivalent of primate grooming, the social glue that keeps the society together. It would be in these close-contact bouts, as the animals touch and swim alongside each other, that a visual language might be most effective. The idea seems outlandish, but it might explain why dolphins in tanks almost always swim anti-clockwise.

Language comprehension in humans is thought to be primarily a function of the brain's left hemisphere. Separate the left hemisphere from the right by cutting through the nerve fibres that join them together (a practice occasionally carried out for medical purposes) and something odd happens. Show one of these 'splitbrain' patients a shoe, and he will tell you without difficulty that he sees a shoe. Cover the right eye and show him a shoe, and he will be completely unable to name it. This is because nerves from the right eye are joined to the left-hand side of the brain, and those from the left to the right-hand side of the brain. Information from the right eye is therefore processed in the left hemispheres the one responsible for naming. In a split-brain patient the information recorded bv the right eye cannot be linked with the left hemisphere.

Dolphins face a similar problem. In a dolphin brain, the left and right hemispheres are barely connected, an evolutionary adaptation that may be associated with the way in which the animals sleep. If dolphins use some sort of visual language, and if like humans, they use the left side of their brain to comprehend and process that language, then they might only be able to do this if they receive the information through their right eye.

People working with dolphins have known for a long time that there is something special about the right eye. Dolphins almost always swim anti-clockwise in their tanks, keeping the right eye facing outwards. They always roll over on their right side when catching fish they have driven on to a beach. Whether this relates in any way to dolphin communication is still anyone's guess, but it is an intriguing idea.

Whether or not dolphins actually speak to each other in ways that we might recognize as a form of language, we have established that they are intelligent animals, probably aware of themselves as individuals. If, for a moment, we leave Descartes behind, we might consider that dolphins feel emotions similar to our own. They appear to experience grief and affection in much the same way as we humans do. They can be sly and excited, mischievous and depressed. If we accept that animals like the chimpanzee and the dolphin do possess minds that are basically similar to ours, this raises some difficult questions. For Descartes, animals were machines, little different from cuckoo clocks, capable of complex behavior but quite incapable of speech or reason. Humans were unique because they had a soul which, among other things, endowed them with the ability to speak. With this barrier between man and animals securely in place, Descartes could argue that because animals were no more than machines, they did not suffer when ill-treated and so humans were innocent 'however often they may eat or kill animals'. Take away that distinction (along with all the other fences that have been erected to separate humans frorn animals) and where are we? Back on the beach in the Gilbert Islands with Arthur Grimble, watching the dolphins being butchered for their meat and feeling a sense of guilt.