, , , , ,


As we all know, Korean sounds nothing like French. Except for when it does. Since the first decades of the twentieth century linguists have noted that a certain syllable structure is favored across Native American and Indo-European languages. It is so common that linguists have long wondered whether all human languages are cut from the same cloth. As it turns out, baby talk—or, more to the point, the words babies prefer—may provide clues to this most fundamental of linguistic puzzles.

147353-149904The favored structure identified by linguists is this: A syllable begins and ends with sounds that are very different from the sounds that occupy its middle. In other words, variety is the spice of language.

Take “bat” for example. “B,” and “t,” like “p,” don’t really make strong sounds; mostly they stop air flow in its tracks (which is why they are called “obstruents”). In the structure found in many languages, obstruents don’t normally share the beginning or end of a syllable with other obstruents. They’re far too alike to provide the variety that humans seem to prefer. And it’s at least true of English that there are no “tb” pairings, for example. However, two obstruents often are positioned at opposite ends of a syllable, in which case they usually have a very different sort of sound in between them. Vowels fit that bill; entirely unlike obstruents they depend on smooth air flow. “Bat” is an ideal syllable by this linguistic theory. So are “tot,” “pup,” “top,” “poop,” and “boot.” Even “psst” passes the test because, while “s” is not a vowel, it is a long, audible, and decidedly non-obstruent sound.

So, what is the universal cloth from which languages are cut? Is a partiality to variety within syllables actually hard-wired into the human brain? Fascinated by such questions, an international team of psychologist and linguists has now tested whether a preference is evident in newborn babies. They’ve done so by monitoring language-induced chemical changes in the blood of babies 2-5 days old.

The researchers focused on blood in language processing centers of the brain, and used a non-invasive neuroimaging technology to measure newborns’ reactions to hearing extremely well-formed (“blif-like”) syllables, extremely ill-formed (“bdif-like”) ones, and middlingly-formed (“lbif-like”) ones. (“L” is more sonorous than b, but not particularly so. According to the theory of ideal syllabic structure, its pairing with b at the start of a syllable is technically allowable but leaves much to be desired.)

As the researchers report in this April’s Proceedings of the National Academy of Sciences, in two separate language-controlling areas of the babies’ brains, blif-, bdif-, and lbif-like syllables each elicited distinct chemical changes in blood. What’s more, those differences mirrored documented linguistic preferences in adults.

Do these findings in newborns show that language preferences are actually coded into human DNA? The researchers suggest as much . But even they admit that the ultimate answer to this question is still up for grabs.

This is because newborns may not be the blank human slates we sometimes think they are. Fetuses can hear through the uterine wall. At birth a baby recognizes and prefers the rhythms and melody of Mommy’s voice. Yes, the chances seem slim that, in the weeks before birth, fetuses hear specific consonants and vowels through the thick uterine wall, especially given the heart pounding, digestive gurgling, and liquid whooshing that is the soundtrack of their oh-so-cozy lives. But it could be that they do. And if they do, at birth, babies may already have firm ideas about how the well-formed syllable should sound.


Iris Berent, Tracy Lennertz, Jongho Jun, Miguel A. Moreno, and Paul Smolensky, “Language Universals in Human Brains,” Proceedings of the National Academy of Science of the United States, http://www.pnas.org/content/105/14/5321.