By Laura Fenton
Every sound a baby makes is fascinating to D. Kimbrough Oller and Eugene Buder. The
two U of M professors of Communication Sciences and Disorders observe the ways infants
squeal, growl and make vowel sounds to develop tools that can determine early risk
markers for communicative disorders such as autism, hearing impairment or language
IVOC research team members (from left to right): Dr. Kim Oller, Chia-Cheng Lee, Neeraja
Rangisetty, Yuna Jhang, Dr. Eugene Buder and Anne Warlaumont.
Oller and Buder are Directors of the U of M’s Vocal and Speech Development Project.The
research team will be working with a new $2.7 million grant from the National Institutes
of Health over the next five years, a follow-up to prior funding from the same agency
to the Memphis team.
The fundamental intention of the research is to determine how human infants come to
develop the sounds of speech, said Oller, who holds the Plough Chair of Excellence
in the School of Communication Sciences and Disorders.
“Across the first months of life, infants develop sounds that are foundations for
speech, and they use these baby sounds to communicate with their caregivers,” Oller
said. “They do so in a way that reveals a key feature of speech.
“Anything you can say, you can say with different emotional content. For example,
you can speak the words happily or angrily, using different facial expressions and
intonation. You can express a variety of emotional attitudes regardless of the words
or sentences you use.
“Babies show they can also express themselves flexibly with squeals, growls and vowels
in the first months. It seems no other primate has that kind of vocal flexibility
at any age. That’s one reason we think the infant sounds are important foundations
The intense longitudinal study will involve hundreds of hours of recording on each
of 12 normally developing babies across a 30-month period for each baby. Three baby
and mother pairs are already being recorded. When mothers and babies come to the laboratory,
they find a room set up as a standard playroom or nursery.
“Babies wear a vest with a small wireless microphone in a pocket on the chest area,”
said Oller. “This allows for crisp recordings of the child’s sounds without interfering
during the child’s movement. Because the microphone is close to the infant’s mouth,
the recordings are much better than if the microphone were somewhere else in the room.
Mothers and babies are recorded several times a month in the laboratory.”
During the recordings, mothers also wear microphones so there are two channels of
audio being recorded at all times.
“This allows for better judgments on exactly who is speaking in the audio recordings,
an important feature in the procedure because the mothers often imitate the babies
and can sound remarkably like them,” said Buder.
Eight cameras in the room offer various angles for video recording. Two cameras are
selected at each moment by remote control from an adjacent room to record both a close-up
of the baby’s face and a view of the interaction between baby and mother. In the lab,
mothers are instructed to think of the recording room as a home away from home. All
interactions are naturalistic with only minimal protocol restrictions. For example,
in some sessions the mother is instructed to interact normally with the child, while
other times she is asked to read alone and allow the baby to play in the room.
“(Mothers) have a lot of intuitive awareness of what the baby’s sounds are about and
what they express emotionally,” Oller said. “They may not remember this as time passes
and the child goes on to real speech. But they react to baby sounds very effectively
while they are going on. They seem to know how to communicate with the baby and how
to interpret the sounds of the infants emotionally. Mothers often mimic the sounds
of the baby and gauge the baby’s emotional state from the sounds and the associated
Sometimes another adult will enter the room and speak with the mother. While mother
and the other adult talk, the baby may make sounds, occasionally vocalizing what seems
to be a bid to enter into the conversation. Mothers react naturally to baby’s bids
for attention and may engage the baby in dialogue.
“Mother and the infant negotiate about the functions of the infant sounds, even before
the baby is really using speech and language,” said Buder.
Buder added that the grant also focuses on how mothers initiate and respond vocally
and on how the interactions change according to the types of vocalizations the baby
In addition to the laboratory recordings, all-day home recordings (using a battery-powered
recorder that fits into the infant’s clothing) are also made regularly to complement
the laboratory work. Online diaries are completed by the mothers to give even more
information about how the babies are learning to vocalize and communicate with sound.
Information gathered from these intensive observations and automated analysis procedures
developed in the Memphis laboratory will be applied to information in a database from
the LENA Research Foundation in Boulder, Colo., which has 80,000 hours of recordings made in the homes of hundreds
of infants and children. The collaborative work aims to create automated tools to
determine risk markers for important disorders of communication.
“The Vocal and Speech Development Project is primarily basic research on infant development,
but the findings may soon be applicable to the work of pediatricians, speech-language
pathologists, and other health care workers,” Oller said.
Parent and infant participants are recruited from birthing classes at local hospitals,
from among prospective mothers who are already pregnant. More participants are needed.
Prospective parents interested in participating with the research study should contact
the Project Coordinator, Dr. Edina Bene at firstname.lastname@example.org or they can call the laboratory at (901) 545-8850.