Despite his annoyingly cutesy synthetic voice, Darwin the Dolphin on the TV series SeaQuest DSV did present an intriguing possibility - what if we could create a dolphin language translator? Such a device may no longer be limited to the realm of science fiction, as two scientists are currently developing an underwater computer that they hope to someday use for two-way communications with wild dolphins.
The Cetacean Hearing and Telemetry (CHAT) project is a collaboration between Denise Herzing, who founded the Wild Dolphin Project in Florida, and Thad Starner, who studies artificial intelligence at the Georgia Institute of Technology. Over the past 12 years, Herzing has attempted two-way communications with dolphins using sounds and symbols, but has met with limited success. One-way communication has been a different story, with some captive dolphins reportedly being able to learn and respond to over 100 human-created visuals and noises.
UPGRADE TO NEW ATLAS PLUS
More than 1,200 New Atlas Plus subscribers directly support our journalism, and get access to our premium ad-free site and email newsletter. Join them for just US$19 a year.UPGRADE
Starner and some of his students are now developing a smartphone-sized computer, that will be worn across a diver's chest in a waterproof case. The device will be connected to two hydrophones, capable of picking up dolphin sounds underwater - including those beyond the range of human hearing. As it can be difficult for humans to identify the source of underwater sounds, an arrangement of LED lights within the diver's mask will indicate the direction from which the various clicks and squeals are originating.
Not only will the computer hopefully be able to decode and indicate what the dolphins are saying, but by using a handheld Twiddler (sort of a combination mouse and keyboard), the diver will be able to select and send out audible dolphin-ese responses.
Before any interspecies conversations can take place, however, the team first needs to figure out the animals' vocabulary. In order to do so, they plan on running recordings of dolphin vocalizations through a pattern detection algorithm, designed by Starner. The system analyzes data, picks out deviations from the norm, then groups similar deviations together. It is hoped that by observing dolphins, and seeing which recurring deviant sounds accompany which behaviors and situations, the researchers will be able to identify specific "fundamental units" of dolphin speech.
Herzing and Starner plan on testing the device in the middle of this year, on wild Atlantic spotted dolphins.
Source: New Scientist