Learning to look for language

Picture a toddler playing with a toy train. Her mom wants to name it for her. If the toddler is and mom are both hearing and use a spoken language, the mom says "train" while the child stares at the train, and the word and the object arrive together, one through the ears, one through the eyes.

Now imagine the toddler is deaf and the language is ASL. The sign TRAIN and the train itself both have to come in through the eyes, and eyes can only point one direction at a time. To connect the sign to the object, the child has to look at the train, look up at mom's hands, and hold the two together across that gap. Learning a visual language means constantly deciding where to look, and when. Deaf kids get to be spectacularly good at this really quickly. Studying how they do this is important for two main reasons. First, by understanding how kids learn sign language, we are better prepared to support families who are new to sign language. But visual attention also a kind of deaf superpower that is helpful for research about how all kids learn language. It isn't always obvious what a hearing child is paying attention to. As any parent of a hearing child can attest, you can talk to them all you want and they might not actually pay attention. For deaf kids, following their eyes gives us a clue about what they're paying attention to.

Here is what we've learned about studying visual attention in language learning.

The puzzle

In foundational work on joint attention, Amy Lieberman and colleagues found that deaf children and their parents spend only about 22% of playtime in mutual gaze — kids spend most of their time looking at toys, not faces. Every sign an adult produces has to land inside those windows, or they have to engineer the interaction to create a window. Language learning in a visual language is, among other things, about attention-management.

How do children manage visual attention?

Babies as young as 9 months shift gaze efficiently between toys and their parent's hands. Infants raised with signing develop precise control of their eye gaze remarkably early, and deaf infants of deaf parents are nearly twice as likely as hearing infants to follow an adult's gaze. By 21 months, deaf toddlers in signing preschools are skilled at getting attention too: tapping shoulders, waving, and strategically switching tactics when a first attempt fails.

How do adults manage visual attention?

The other half of the picture is what signing adults do to make it possible for kids to see sign language. There is a whole repertoire at work when adults interact with deaf signing kids. They often get the child's attention before signing by tapping the child on the leg, waving into the field of view, holding an object up near the face. Parents also time their naming to land inside moments of joint attention, give extra care and repetition to signs that are new to the child, sign on or near objects of interest, and keep moving themselves into the child's line of sight. They follow the child's focus more than they redirect it, fitting language around whatever the child is interested in, the responsive style that word-learning research finds most effective. They point constantly and strategically: in our studies, half of parents' points functioned as pure traffic direction for the child's gaze. When adults do pair a point with a sign, they sometimes sequence the two so both can be caught in a "sandwich": sign the label, point, sign it again, giving the child a second chance at the sign as their eyes come back up, which might be an especially helpful strategy. Encouragingly, the hearing parents in our studies who were still learning ASL pointed, paced, and sequenced in largely the same ways deaf parents did.

How does visual attention matter for learning language?

Because a signing child's attention is visible, we can connect where a child looked to what they learned. In our fast-mapping experiments, deaf children learned brand-new signs whether an adult pointed at an object before the sign or after it, adapting their looking to whatever the moment served up. How a child moved their eyes during those exposures predicted whether they learned the new sign or not.

The same kind of pattern was true in whole interactions too. In our joint attention work, the quality of shared-attention episodes between signing caregivers and children was correlated with the size of children's vocabularies. Children who spend more time in well-built moments of shared focus know more signs.

Visual attention supports language learning, and the reverse is also true: as children's language grows, their attention gets smarter. When watching an ASL sentence, preschoolers glance at the right object before its sign even appears, but only when the sentence makes it predictable. They pull their eyes away from the signer exactly when it's safe to.