phonological learning

Not too many parents name their baby, “Bleeber,” but for English speakers, even though “Bleeber” is a strange name, most parents would agree that it is a better name than “Tlibit.” One reason that “Tlibit” is worse than “Bleeber” is because it violates the sound regularities of English. In English, words don’t begin with the “tl” sound sequence even though it shows up in other places, for example, at the end of the word “little.” Regularities in the location of sounds for a particular language have been shown to influence adult language processing.

We have been investigating when and how infants learn these types of regularities. For example, we had 10-month-old infants listen to words from a miniature language (“fes,” “kep,” “bes,”…) that displayed regularities in the position of consonants—for example, whenever “f” was heard, it was heard at the beginning of a word, but whenever “p” was heard, it was heard at the end of a word. After listening to this language for a couple of minutes, infants heard new words that they hadn’t heard previously. These new words either followed the same sound regularities as the words from the miniature language (e.g., “fip”) or violated those sound regularities (e.g., “pif”). Using the headturn preference procedure, we looked to see if infants could tell these two types of words apart, and they can, suggesting that after listening just a couple of minutes, infants are already picking up on the regularities present in the language. Infants seem to have very powerful abilities to learn patterns in their environment.

the developing lexicon

As adults, we know what the words “taxicab” and “airplane” mean, but at some point in our lives, we didn’t. Even as adults, we run into new words that we have never heard before—words we have to look up in the dictionary (or type into Google). Infants are constantly hearing sequences of sounds that are new and have no meaning for them. Do these sequences have any effect on words that infants do know?

To address this question, we had infants listen to nonwords that displayed regularities. For example, when “b”s occurred they were at the beginning of nonwords, and when “m”s occurred, they were at the end of nonwords (“bem”,“bip”,”kem”,..). Then infants saw pictures of two common objects that could be labeled with words they probably know, like “bear” and “mouse”. The label for one object was consistent with the regularities displayed by the nonwords (e.g., “bear” and “bem” both have “b” at the beginning) while the label for the other object was inconsistent with the same regularities (e.g., “mouse” and “kem” don’t have “m” in the same location). When children were asked to find one of the objects (e.g., "Where's the bear? Can you find it?"), they looked at the object more quickly when the label was consistent rather than inconsistent with the regularities displayed by the nonwords. This result demonstrates that nonwords and words mingle, and it starts to address important issues about when new linguistic information is integrated with established linguistic knowledge.