Baby studies

By studying social interaction and learning as it occurs in moment-to-moment interactions, we connect specific mechanisms of perceptual and cognitive development with social influences on the acquisition of speech, words, and language.

Taken together, our studies show that the infant’s social world is a crucial component of the learning process. When studied as a form of social interaction, vocal precursors such as babbling constitute a crucial, formative phase in the development of communication. The social environment is structured by consistent behavior that captures infants’ attention and increases arousal. Infants are active participants in these interactions, creating new patterns of vocalizing that catalyze developmental changes in speech and language.

Our research on socially guided learning of speech, words, and language uses multiple paradigms, including playback of digital stimuli to parents, controlled social feedback to prelinguistic vocalizations, forced-choice measures of word comprehension, looking time measures of lexical and statistical learning, and eye tracking using a new system that is wearable and wireless, allowing infants to freely explore and interact as we measure moment-to-moment changes in visual attention.

The major projects in the lab are:

1. Social influences on prelinguistic vocal learning and phonological development. We are finding that infants can learn to produce novel speech patterns if social feedback is both contingent and contains sufficient statistical variability (i.e. multiple exemplars). There are limits, however, to what speech patterns babies can learn. How does social learning and motoric constraints interact? We are currently running studies to find out.

In a series of studies that link prelinguistic vocal learning with early word learning, we are studying the function of object-directed vocalizations (ODVs), which are babbles that are directed at objects. We found that labeling an object contingently on an ODV facilitates learning word-object associations. It seems that ODVs signal a state of focused attention and arousal. ODVs are the auditory equivalent of a furrowed brow. We are currently studying the effects of ODVs on the behavior of caregivers, and the ability of infants to learn from caregivers’ responses to their ODVs. Thus we are elucidating the earliest social interactions that support the association of sounds and referents.

2. Longitudinal studies of vocal learning over the first year. We recently found that 5-month-olds have learned that their early non-cry vocalizations are effective on adults, and the strength of this early learning predicts language comprehension at 13 months. Five-month-olds don’t change the form of their babbling like the 9-month-olds do, but they will increase the amount of babbling if a social partner stops responding. We are currently running a longitudinal study to assess relations between vocal learning in 5-month-olds, the more advanced forms of vocal learning in 9-month-olds, and language development in the second year. We are also using non-social learning tasks to see if the ability to learn from social feedback is grounded in more general learning mechanisms.

3. Experiments on the role of syntactic and lexical variability in caregivers’ speech on early noun and verb learning. Speech to toddlers typically includes groups of sentences that have partially overlapping syntactic structure, e.g. “Roll the ball. Roll it. Can you roll the ball?” The partial redundancy across related sentences, known as “variation sets” may facilitate noun and verb learning. These studies are ongoing, with more studies planned to determine the effects of infant-directed speech, coordinated object motion, and variation sets on word learning. We are also investigating the role of action participation in learning new verbs. We plan to use the eyetracker to understand how infant attention is being organized by social cues as nouns and verbs are being learned.

4. Investigations of caregiver responsiveness to infant babbling. In a series of studies, we play back prerecorded (and carefully controlled) infant behavior to adults, to determine the role of caregiving experience, infant vocalizations, and context (e.g. the presence of a toy) on adult responsiveness to infant behavior. We are planning many additional playback studies to look at caregiver responsiveness as it develops across the birth transition and with additional caregiving experience. We are also planning studies that would validate the computer-based playback paradigm with measures of caregiver responsiveness when interacting with their own infants.