The Disappearing Accent

Jessica Love’s last Psycho Babble column was published last week. This month we are re-posting several of our favorite columns.

Chat with someone who speaks another dialect of English, and it’s the differences you notice. Differences in your vowels, sure, but also differences in everything you’ve come to associate with those vowels: where you’ve lived, how you’ve lived, who you are. Over time, should you get close to this person, her dialect may become unremarkable. It may become just another thing about her. Rarely, though, will it go entirely unheard.

That’s why it’s so intriguing to learn that babies, as they become accustomed to accented speech, go through a developmental period where they seem to stop perceiving it as accented.

A study by researchers Christine Kitamura, Robin Panneton, and Catherine Best found that six-month-old infants paid more attention to speech in their native Australian English than to speech in an unfamiliar dialect—South African English. This makes sense. Infants like hearing language that sounds familiar to them. They’re drawn to it, a fortunate thing because a native language takes a lot of listening to unpack.

By nine months, however, infants were equally attentive to both their native dialect and South African English. A follow-up study suggests that this is because the nine-month-olds could no longer discriminate between the two dialects.  (American English, somewhat familiar to Australian infants thanks to the pervasiveness of American media, is lumped indiscriminately with Australian English at just six months old.)

Why the temporary deaf ear for accents? It helps to remember that we make sense of sounds by categorizing them, and categories are as much about similarities as they are about differences. Being able to distinguish t’s from d’s or pets from pits is critical for an English speaker. But so is ignoring the differences between one t and another, no two of which are ever the same. Likewise, we must de-emphasize the differences between individual voices to recognize that my pits should be categorized with your pits, even though I am female and you’re male, or I tend to speak quickly and you more slowly.

All this categorization actually changes our perception of these sounds. Meaningful differences between t’s and d’s are perceived in an exaggerated way, while less meaningful differences between two t’s or two d’s (even if the size of the differences are acoustically equivalent) are collapsed. What these infants appear to be doing is lumping dialects together—dialects they’ve previously perceived as distinct—once they detect the underlying patterns that South African English or American English share with their native Australian English.

“For adults,” the researchers reiterate, “regional accent functions as an important social marker enabling the listener to identify the speaker’s geographic and ethnic origin, and social status.” Even five-year-olds recognize that an unfamiliar accent marks a speaker as somehow different, and are less willing to learn things—like how to play with a toy—from him. But for infants, to whom these social categories are somewhat inscrutable, accented English is, for a while, just another idiosyncratic difference to ignore, like a lisp or the hoarseness of my voice after a late night.