Letter to the Editor

Dolphin Communication

David F. Siemens, Jr., Ph.D.
E. Kenwood St. Mesa,
AZ 85213-2340

From: PSCF 42 (December 1990): 271.

Dorothy J. Howell makes a curious statement in her review of John D. Lilly's work ( Perspectives >, December 1988, p. 242): "Only successful communication with dolphins, if truly possible, can refute or confirm his claims." Would not successful communication necessarily confirm his claim of the intelligence of dolphins?

A strict refutation is more difficult, for all possible phenomena would have to be shown to be negative. However, a simpler empirical test can tell us how promising the attempt to communicate with dolphins may be. We know that dolphins have a very effective sonar. We can be confident that an intelligent creature, confronted with a simple target in the usual test situation, will not inquire into its health nor extend an invitation for an evening swim. So these signals, along with the sounds of a solitary animal navigating its tank, may be taken to involve no more than sonar signals. If these types of signals are subtracted from the signals detected from pairs or groups of dolphins, the remainder represents the social component, which may then be analyzed. Even here a portion of the signal may be the localization function relevant to a group situation. This last element would tend to exaggerate the communication potential. The only factors in the experiment that might underestimate the communication potential are the possibility that solitary cetacians obsessively talk to themselves or that our analysis is too crude to differentiate ranging from language in the vocalizations. The former can be reduced by discovering the difference between vocalizations when dolphins are in the same pool and when they are physically isolated with an auditory connection. The latter can be guarded against by considering all the possibilities of signal modulation, such as rate of repetition, which are not important in human languages.

We do not need to know anything about the structure of dolphin language, if there is one, to get a reasonably accurate estimate of the range of communication between dolphins. We can depend on the complexity and variation of the signal. The complexity of the signal is a measure of how much information can potentially be transmitted. Played against the variation, we have a better estimate of the actual transmission. For example, if someone yells, "Duck!" and you do, you may be saved an injury. The sound communicates. But simply repeating the cry any additional number of times does not add to the informational content, unless there is a new danger. However, if the person continues: "Kid. Throwing rocks. Behind the wall. To the right." -more information is given with each group of words. Were the utterances in a language unknown to us, we could still tell the difference between a repetition and a non-repetitive sequence. So comparing the qualities of the specified remainder of dolphin vocalizations against vocalizations by other species, such as the restricted range of the dog or the broader variety of the chimpanzee, can give us a reasonable estimate of the dolphins "vocabulary." Indeed, someone familiar with Shannon's work can probably quantify the communication potential. Such a comparison is not determinative, for other matters also enter in. For example, the complex variations of the mockingbird song probably convey no more information than the scream of a peacock or the repetitious crowing of a rooster, despite the greater potential in the variability.

Such an investigation should result in a dissertation and either a warning to waste no more time on dolphin "language" or a new opportunity for research with unpredictable potential. Either consequence would have value.