in All Things Localization, Language & Culture, Translation

Fishy Conversations – Floating on a cerulean sea, you’re mindlessly soaking up the sunshine when a voice calls from below, “Care for a game of seaweed?” That sounds silly, but a research program in the Bahamas using a gadget designed to communicate with the dolphin species claims they’ve heard something similar.

Dolphin Talk

The Wild Dolphin Project has been studying dolphins’ social structure and language for the past 29 years. Now they’re living up to their “In their world, on their terms” principle. They’ve taken the dolphins’ whistles and clicks and strung them together to form words. Using a system called C.H.A.T.—Cetacean Hearing and Telemetry—developed by Thad Starner, the technical lead on Google Glass, they broadcast these words into the water and listen for responses. They finally got a long awaited reply; a younger dolphin the project works with used the researchers’ word for “sargassum,” a type of seaweed and plaything, presumably in an attempt to get a game going.

The Wild Dolphin Project’s research is unique in that it has worked on the open ocean with the same group of dolphins for over a quarter of a century. Their observations have revealed dolphins’ complex social structures—matriarchal societies, usually travelling in groups called “pods” of 2-40 animals and displaying characteristics we might call “human.” The project focuses on dolphin language and now, thanks to advances in information theory—a branch of mathematics that finds the structure within data—the first conversation with a dolphin may soon be a reality.

Armed with the analytical tools of information theory, scientists with access to the Wild Dolphin Project’s data are decoding dolphins’ speech. By feeding the recordings of the dolphins’ noises through their computers, they have separated binary code from randomness and found that the whistles and clicks are actually transmitting information.

Related:  Watch a Video Recap of CSOFT Health Sciences’ 2023 Summer Soiree!

This isn’t exactly “translation.” Researchers using the C.H.A.T. system have been attempting to use their invented language for common objects in the dolphins’ environment—e.g. a long, trilling whistle for the toy ball scientists always bring with them. These human-created noises are still very different from the dolphins’ own words. The dolphins are speaking a language, but its vocabulary and grammatical structure are a mystery.

“I’d say the prospect isn’t particularly hopeful,” said Robert Derbyshire, a machine translation specialist and Senior Manager of Product Development at CSOFT International, when asked about the likelihood of a dolphin-to-English interface. “The main prerequisite [of machine translation] is something where you tell the computer, ‘A-B-C in this language means X-Y-Z in that language,’ and you need a bunch of those, around 10,000.”

Denise Herzing, the head of the Wild Dolphin Project, hopes that through her team’s very limited interspecies interaction, they will eventually be able to uncover the foundation of the dolphins’ speech. “If you did have that sort of data,” said Derbyshire, “then feasibly you could develop an M.T. engine using audio files translating from dolphin into English.”

A true translation device is a long way off but the implications for our sense of place on the planet are profound. Perhaps establishing two-way communication with another Earthly intelligence will encourage environmental stewardship.

Maybe interspecies conversation will turn us all vegan—too uncomfortable eating animals capable of verbal objection. Maybe we’ll all dust off our bicycles when sparrows can complain about their engine exhaust-induced asthma. What will people say when the world’s innumerable inhuman inhabitants are able to talk back? My vote for first note: “Sorry about the mess.”

Related:  Beyond “Hey Siri”: Apple Moves in on Large Language Models

If you’re interested in learning more about CSOFT’s globalization and localization solutions, don’t forget to visit us at csoftintl.com

[dqr_code size="120" bgcolor="#fff"]