Communication with Unmotivated Dolphins: Intelligence Part 4
During my research on the last article in this series, I read about a project to develop a shared language between humans and dolphins. SETI (Search for Extra Terrestrial Intelligence) was very interested in this project, as the difficulties we face when trying to communicate with dolphins are likely to be similar to those involved in communicating with an alien race. . We can’t communicate vocally with dolphins as we do with birds (they don’t have vocal cords capable of making human speech) or with sign language as we do with apes (they don’t have hands). What dolphins do have is a very large brain and a highly complex social structure, both of which indicate they would have quite a lot to say to us if we could get around the practical difficulties.
The effort to develop a language together with another species is also a good answer to the persistent criticism in studies of animal intelligence that we are really just measuring their ability to act like a human..
The Dolphin communication project is being conducted by a group called The Wild Dolphin Project. As the name indicates, they study a group of dolphins in their natural environment. One of their key principles is to disturb the dolphins as little as possible. They never feed them, and don’t interrupt the dolphin group when they are involved with tasks such as hunting or defending themselves against sharks. Adhering to these principles can make conducting experiments quite a bit more difficult, as we will discuss later.
Previous Research on Dolphin Communication
Dolphins communicate primarily through sound. They are able to make a wide range of clicks and whistles, some of which are at a pitch too high for humans to hear. They also use some body language, such as pointing to objects with their head. Each dolphin has a unique ‘signature’ whistle, which can function like a name. Dolphins have an excellent memory for the signature whistles of other dolphins they have met, and can mimic the whistles of others when they want to communicate with them. They have shown the ability to understand some complex language constructions, such as carrying out instructions in accordance with word order in a sentence 
The Wild Dolphin Project created a device for the dolphins and humans to use in communicating with each other. If you were a dolphin what kind of keyboard could you use? One that looks basically like this:
Dolphins are not good typists, their only method of hitting the keys is by smashing their face into it. Actually, it turns out they don’t like touching foreign objects at all, so they indicate they want to press a key by pointing their head at it. Therefore, the keys need to be very large and far apart so that we can be certain which one the dolphin wants to press. Each of the symbols on the keyboard represented an object the dolphins like to play with. One represented a piece of seaweed, and another a rope. In addition to the key, each play object was also represented by a 'whistle' sound. The sound was not one the dolphins used when communicating with each other, but was one they had no problem reproducing.
The dolphins involved in the study had been observed playing with the objects in the past, sometimes allowing the humans to join in their game. When trying to get the dolphins to engage in communication, one human would hold the keyboard, while several others would model basic communication using the keyboard. For example, the four objects would be placed in the water inside jars. One human would get the attention of another human, then point to one of the symbols or play the appropriate whistle sound. The second human would then go and get the object represented by the symbol. In this way, the dolphins could learn the representations for each object, and each species could relay basic messages to the other such as ‘I want to play with the seaweed.’ From this basic foundation, greater complexity could be added until the point where we could add ‘humandolphin’ to the list of world languages. Things did not progress quite as planned however. Let’s have a look at the results of the study.
There were several occasions when the human researchers were able to request the dolphins to bring them an object. After pushing the appropriate symbol or playing the whistle sound, the dolphins went to the sea floor to retrieve the object and dropped it in front of the humans at the surface. However, communication did not progress too far beyond this.
Experiment versus Observational Study
As The Wild Dolphin Project is well aware, this study does not meet the control requirements of a true experiment. Instead it can be classified as an observational study. Observational studies can have a lot of value. In animal studies two examples of common benefits are helping to generate hypotheses for experimental studies, and providing information about how the animal is likely to respond in different circumstances.
Denise Herzing, the director of The Wild Dolphin Project, has a Ted talk where she compares the dolphins to young children, talking about the importance of holding their attention. The technique of rewarding animals with food in order to elicit a certain behavior is as old as the domestication of dogs, dating back tens of thousands of years.
The Wild Dolphin Project principal of not disrupting the wild environment by feeding the dolphins is admirable, and has allowed the project to contribute a lot of information about how dolphins behave naturally. However it does make measuring the abilities of the dolphins very difficult: as reported in the study most of the time the dolphins were not particularly interested in the keyboard, and many times attempts to communicate with the dolphins ended when the dolphins simply got bored and swam away.
How can we make the humans a little more interesting?
Communication through the keyboard was too slow and clumsy to keep the dolphin's attention. The study recognized that the humans needed a quicker and more familiar way to respond to the dolphins.
The dolphins need technologically and acoustically advanced tools both for their initiations and interactions, and humans need a real time acoustic interface to respond quickly.
Dolphins can’t pronounce human words, and we are equally bad at forming the sounds of their language. Machines may provide the answer. The ability of AI-powered computers to reproduce human speech has been rapidly improving, and some of the same technology can be applied to dolphin vocalizations.
Anyone who has watched a few Science Fiction movies will have seen a universal translator- a machine which can instantaneously translate your sentence into the language of the alien you are communicating with. Artificial intelligence pattern recognition algorithms are currently working on correlating dolphin vocalizations with behavior. Perhaps SETI is correct in its interest in the wild dolphin project, and the distant ancestor of a true future Universal Translator is currently learning how to whistle and click like a dolphin.
Herzing, D. (1996). Vocalisations and associated underwater behaviour of free-ranging Atlantic spotted dolphins, Stenella frontalis and bottlenose dolphins, Tursiops truncatus. Aquatic Mammals 22.2, 61-79.
Herman, L. Richards, D., Wolz, J. (1984). Comprehension of sentences by bottlenosed dolphins. Cognition 16, 2. 129-219. https://doi.org/10.1016/0010-0277(84)90003-9
Herzing, D., Delfour, F., Pack, A (2012). Responses of Human-Habituated Wild Atlantic Spotted Dolphins to Play Behaviors Using a Two-Way Human/Dolphin Interface. International Journal of Comparative Psychology, 25(2).