By Michael Hill
Rutgers mathematics and computer science Associate Professor Patrick Shafto says baby talk has benefits to infants learning.
“What we’re getting at is not whether this is necessary, but whether it might help facilitate learning a little bit more quickly than you might otherwise expect,” Shafto said.
That defies how pediatricians say we should speak to babies: speak to them as you would an adult. Some parents follow that advice, but Shafto says it seems natural to speak to babies in a sing-songy voice.
“We do it intuitively. There’s no sense in which we’re intentionally trying to teach the child the vowels of the language or any other property. It’s just a thing we do naturally. And that’s what I think is really quite interesting about it is that we might actually be built in some way or know implicitly how to help children,” said Shafto.
Shafto says the value that sing-songy talk to babies is dragging out the sounds of vowels. His team used sophisticated math and computer science models to break down a litany of language learning literature to describe how those vowels differ from one another.
“And what we were interested in is describing how the vowels that we speak — U and I, for example — differ from what you speak to a child, for example. So it’s been well established that they do differ and what we were trying to understand is why they might differ,” Shafto said. “Our argument is that, in fact, the deformations in the language that we’re introducing when we speak to children are consistent with what you would expect if you were trying to teach the child the vowel categories. So, for example, certain vowels that you might confuse tend to get further apart from one another so they’re more easy to discriminate in affected speech.”
Asked to demonstrate he said, “Well I’m not sure I can actually do that on camera, that’s a tricky one. I have small children so if you put my children in front of me I could do it.”
Shafto says we exaggerate vowel sounds to help foreign language speakers understand us. We do the same with our pets, but out of cuteness, not for them to understand vowel sounds.
The team’s research appears in Psychological Review. Shafto says the mathematical models can be applied to teach other subjects even though the educational benefits of baby talk are hard to probe because infants are too young to speak. Nevertheless…
“Infants receive atypical speech but it’s speech that appears, at least in our analysis, to be able to help them learn the actual vowels that they’ll experience as adults,” Shafto said.
A conclusion contradicting conventional wisdom.