caution warranted on dolphin name story

Mark Liberman over at the linguistics blog Language Log has some cautionary words about the story out today about dolphins having names. The actual paper itself isn't out yet; it'll be in the Proceedings of the National Academy of Sciences soon, but all of the articles so far are based on press releases. As Liberman says, "[t]he media reliably overinterpret science stories that push their buttons, and nothing pushes people's buttons like talking animals."

The study seems to moderately extend something that's been known for 15 years, that the calls of dolphins include indentifying signatures. The new work, by Vincent Janik of the University of St. Andrews, used synthetic versions of the calls instead of actual recordings of a dolphin making its identifying signature. They would play the synthetic call (analgous to a computer voice saying your name) to the relatives of a dolphin, and see whether the relatives reacted to the voice the same way as they would to the dolphin saying it's own identifying call. That's fairly interesting, in that it suggests that the identifying call is based on learned patterns of whistles and what not, like some songbirds, and not on recognition of the tone of voice in a call (as in the penguins in March of the Penguins). But they don't seem to have much evidence (aside from some somewhat rare, uncontrolled evidence) that dolphins will use the name of another dolphin, either to call it or to refer to it to a third dolphin. That would be particularly interesting, and the media articles suggest that it might be true, but Liberman wisely prefers to wait for the article itself to be published before accepting that as more than speculation.