The Call of the Wild

Learning another language has never been an easy task, despite how accessibly Rosetta Stone might package it for us. Decoding an unknown language, however, is an even more daunting task – especially when we’re not even talking about the same species.

Last month, a team of marine biologists, Scientific American, and the Citizen Science Alliance launched a project called Whale.fm, which invites anyone to listen to whale song and match it against other patterns in hopes of translating the complex Orca communications (similar to those of the pilot whale). It’s a highly-covered topic that has seen a generally warmly-received because of the possibilities it inspires, if not the ease of its user interface (I found it hard to contribute to the nuances of the translation…). But, imagine being able to listen in on the most intimate conversations between Orcas! Imagine being able to talk back…

As far as we know, this is the first crowdsourced project that has asked amateur marine biologists and sea life enthusiasts to try their hand at translating for an animal species. It’s not, however, the first time that the crowd has helped decode complex materials. The Zooniverse Ancient Lives project from earlier this year, for example, enlisted the crowd in translating ancient Greek scrolls. Ancient Lives has uploaded some ancient and crumbling papyrus scrolls and shared them with the online community in hopes that the crowd will be able to match certain characters and translate the aging script.

What qualifies the crowd for such advanced translation? As the Ancient Lives director Chris Lintott says “You don’t need to know Greek… you can transcribe these texts letter by letter. It’s just pattern recognition.” This is certainly one of the defining qualities of most crowd tasks: you don’t have to be an expert to sort through a pattern and you don’t have to be a marine biologist to help them make breakthroughs in Orca whale song. In other words, it doesn’t matter if it’s all Greek to you, as long as you’re presented with a well-designed translation interface and a good translation key.

What do you think about the Whale.fm project? What other tasks involve pattern recognition and might be well-suited to the crowd?

Comments are closed.