Lexicon Valley

What’s the Point of Baby Talk?

A version of this post originally appeared on The Week.

Don’t talk down to me.



Do you know what I’d like to know? Do you? Dooo you? Yes! Yeeeesss! Yes, you dooo! Don’t you?


Well, if you don’t, I’ll tell you. What I’d like to know is: What’s the point of baby talk?

This is a bit of a disingenuous question, because I’m going to give you some possible answers. But I’m not entirely sure which of them, if any, is right. Linguists don’t agree on this. Seriously: You can take courses from different professors who will tell you directly contradictory things with equal certainty.

You may not know the term motherese, but I’m sure you’ll recognize what it is when I tell you it’s that way many people speak to infants (also to dogs, sometimes to cats, and sometimes even to senior citizens). Linguists have several terms for this. The classical term is motherese, which has fallen out of fashion a bit in recent years. Caregiver speech is sometimes seen, though more technical-sounding options are infant-directed speech and child-directed speech.


It turns out that motherese is fairly similar in most languages: Shorter phrases, simpler grammar, different versions of words, repetitions, slower speech, higher pitch, very exaggerated pitch contours (just think of someone saying to a baby, “Who’s that? Say hiiii!”), and more exaggerated vowel contrasts (wider-open “ah”s, tighter “ee”s, and so on).


The idea, according to some, is that all of this is important to help babies learn language.

But is motherese important in child language learning, really?

Well, here’s the thing: In some societies, they not only refrain from using baby talk, but also refuse to speak to babies at all. Linguists who hold no truck with motherese will quickly point to work by Elinor Ochs of UCLA in which she contrasted the helpful, encouraging, accommodating, motherese-using approach common in America with that of societies such as Western Samoa, where children who can’t talk aren’t spoken to directly.


Furthermore, when Samoan children are learning language, if they say something unclear, the parent doesn’t guess what it is and help them along—he or she just tells the child it can’t be understood. And how does this affect Samoan language learning? Well, obviously the children learn language. After all, the parents were once infants, too.

So does that mean motherese has no effect on children’s language learning? In fact, there is some evidence that it may help.

First of all, babies—especially ones half a year old or less—seem to prefer motherese. This isn’t really surprising; babies also like bright colors. Bigger contrasts seem to appeal to them more: Big rises and falls in pitch, mouths open wide or closed tight when speaking, etc.


Second, some aspects of motherese do seem to correlate with better language learning in infants. Studies have found that infants appear to detect such things as syllable and phrase boundaries better when hearing motherese, and that infants spoken to with motherese appear to be better at identifying differences between consonants.

A common hypothesis is that better ability to discriminate sounds leads to better ability to learn words and grammar. Some studies have indicated that how mothers talk to infants before they’re old enough to speak can have an effect on the children’s language abilities a year or so later.

However, the studies that show positive effects only focus on specific areas. Also, they don’t show overwhelming effects and other studies have not found any meaningful effects at all.


There are also factors that are very hard to filter out. For starters, we have to remember that most of the speech an infant hears is normal speech, not motherese, because most of the speech it hears is not directed towards it. And not all the speech directed to the infant will be motherese, either—it’s unlikely that every person who speaks to the baby will use motherese. Finally, women who use motherese will hopefully speak differently when speaking to adults, and the babies will overhear that, too. Not to mention that motherese isn’t likely to help children learn much about grammar directly, considering it’s so often grammatically incomplete or plain wrong.


So if motherese has positive effects, they are not enough to make the difference between a child genius and a semiliterate Neanderthal, and may not even be enough to make the difference between a B+ and an A– in English class years later.


And that’s if it has positive effects. The evidence is inconclusive enough so far that it will continue to make developmental linguists’ careers for decades to come. But I know two things that shape my opinion on whether to use motherese:

First, two of the most highly literate, grammatically acute people I know, with the largest vocabularies, grew up with linguist parents who never used motherese with them.

Second, motherese drives me absolutely nuts. Yesss! Yes it does! Oh yeeeeeesssss! Doesn’t it drive you nuts too?! Doesn’t it? Doesn’t it?!

More from The Week:

15 words etymologically inspired by animals
Capital letters: FTW or WTF?
26 of Noah Webster’s spelling changes that didn’t catch on