June 05, 2004

Duck Dialects Mimick Human Dialects?

The Guardian reports that ducks in Cornwall have different accents from those in London. According to Victoria De Rijke, Lecturer in English at Middlesex University:

The Cornish ducks made longer and more relaxed sounds, much more chilled out. The cockney quack is like a shout and a laugh, whereas the Cornish ducks sound more like they are giggling.
Victoria de Rijke, by the way, is the leader of the Quack Project, which has made recordings of children from different linguistic backgrounds (Arabic, Tamil, Vietnamese, etc.) illustrating the way in which the sounds of common animals are mimicked in their language.


The story was picked up by Agence France-Presse and also appears in the Australian News in Science and the South African IOL. However, this version contains the somewhat implausible suggestion that:

The result was that the ducks' "accents" mimicked those of the humans in their home region.
That isn't what Dr. De Rijke said. If you read through the direct quotes in the story in The Guardian, what she said is that the London ducks "speak" as they do because, in contrast to the Cornish ducks, they have to make themselves heard over the noise of city life. She then compared this to the difference between the Cockney and Cornish accents in English. Her thesis is that both ducks and humans are responding to differences in their environment. I'd need more evidence to persuade me that this is true of the differences in English, but it is very different from the hypothesis that the ducks have mimicked human regional accents. She didn't say that; Agence France-Press did.

The problem of attributional abduction has been discusssed a number of times here on Language Log. When a news item attributes to someone a view that we find strange, we need to be careful not to rush to attribute it to the putative source; in practice the strangeness is often introduced by a reporter or editor. In this case we can be sure that the error was introduced by Agence France-Presse because we know which version is the original and which the derivative. The journalist may have misunderstood, but it is possible that the problem here is a poor choice of words, namely mimicked, which implies copying, rather than something like paralleled, which does not.


Posted by Bill Poser at June 5, 2004 12:02 PM