The Angry Red Geek

Updated At one time it was believed that true language was limited to humans, with other animals limited to instinctive communication of the most basic messages, such as a warning cry or a mating call. Even though it ultimately developed that claims for language acquisition by chimpanzees such as Washoe were overblown, over time, it's become clear that the original idea was seriously in error. It's not that other animals don't have language at least by a more generous meaning of the term, it's rather that we're just so much better at it than any other animal.

This turns out to be an advantage because it means that by studying the language-related abilities of other animals we can gain some insight into how our own abilities developed. Here's a report on a recent study that did just that.

Reuters, January 9 - Rats can use the rhythm of human language to tell the difference between Dutch and Japanese, researchers in Spain reported Sunday.

Their study suggests that animals, especially mammals, evolved some of the skills underlying the use and development of language long before language itself ever evolved, the researchers said.

It is the first time an animal other than a human or monkey has been shown to have this skill.

"These findings have remarkable parallels with data from human adults, human newborns, and cotton-top tamarins," the researchers wrote in their report, published in the Journal of Experimental Psychology: Animal Behavior Processes, which is published by the American Psychological Association. ...

They used Dutch and Japanese because these languages were used in earlier, similar tests, and because they are very different from one another in use of words, rhythm and structure.

The rats were trained to respond to either Dutch or Japanese using food as a reward.

They were then tested four different ways to see if under different conditions they could recognize the difference in the patterns of the language.

"Results showed that rats could discriminate natural sentences when uttered by a single speaker and not when uttered by different ones, nor could they distinguish the languages when spoken by different people," the researchers wrote.

That confusing description is clarified some in New Scientist's coverage, which says the team

trained rats to recognise either Dutch or Japanese - by pressing a lever in response to a short sentence - and then exposed them to sentences they had not heard before, in both languages.

They found that the rats responded significantly more often to the language they had been trained in - as long as the sentences were computer-synthesised or both languages were spoken by the same person. However, the rats could not tell the difference if the sentences were played backwards or were spoken by different people.

That is, if the same person read a sentence in Japanese and one in Dutch, the Japanese-trained rats responded to the Japanese sentence but not the Dutch one, and vice versa. But if one person read the Japanese sentence and another read the Dutch sentence, the rats had trouble. That suggested the rats could tell the difference between the languages but adding the extra information of the variations between the two speakers was more than they could handle.

What makes this significant is that newborn humans have exactly the same ability and difficulty, strongly suggesting that this ability to recognize the pattern of a language, so important in humans for learning one (or more) of them, is an evolutionary precursor - that is, it's part of the puzzle of how human language came to be.

Footnote: The study can be found here in .pdf format. You might also notice at the New Scientist site some links to earlier stories about language, two of which I've already noted. I particularly liked the one about Rico. So did my own border collie when I read it to her.

Updated to add the link to an earlier post.