Baby Talk, Here and Abroad
By: Hannah Austin
Posted: November 6, 2012
Sometimes life can change with a question.
Dr. Rory DePaolis was entering his tenth year of experimental research on adult acoustic speech at Southeastern Louisiana University when one of world’s leading phonologists, Dr. Marilyn Vihmaen, joined the Communication Sciences and Disorders team. His new co-worker had a question that fascinated DePaolis; a question he is still trying to answer seventeen years later.
“She wanted to know, how does an infant’s early production of babble affect how they perceive the world?” explains DePaolis. “She had done a lot of observational work, trying to understand how infants come to language, but she hadn’t done much experimental, where you actually bring the infant in. Experimental was my strong point, but it had always been with adults. Since I had done zero observational research at this time, it was a wonderful collaboration.”
Vihmaen was especially interested in studying infants cross-culturally and had already gathered a massive database of infant babble. The English, French, Swedish, and Japanese recordings had never been analyzed experimentally, a task DePaolis was eager to complete. The duo traveled to the University of Wales to begin their project, study Welsh, and add Finnish to the infant database. The project was moving ahead smoothly; however, when results started coming in, it looked like something had gone seriously wrong.
“We were trying to replicate in Wales work that had been done in the U.S., but we weren’t getting the same results,” said DePaolis. “It seemed like the British infants weren’t segmenting words at the same time as the Americans. We found out it was way more complicated than that once we started analyzing moms and babies both in Britain and the U.S.”
Segmenting words is the ability to pick out a word as a separate entity within a phrase, a skill babies learn as they are learning language. Vihmaen and DePaolis wanted to know, when does an infant hear, for example, the word “truck” and know it is not connected to other words in the sentence? They suspected the differences in their results from the results of American based research might be related to a cultural variation in the manner in which mothers spoke to their babies. However, they would not get to test their theory until 2008 and 2009, with funding from the Economic and Social Research Council and the British government.
In the meantime, DePaolis was able to continue his work at JMU, where he began teaching in the Communication Sciences and Disorders department in 1998. In 2004, the college gave him a semester long sabbatical, during which he returned to the University of Wales to continue his work with infant babbling. While there, he was awarded the Marie Curie Fellowship from the European Union, a grant that encourages cross-border mobility for researchers, and decided to stay another year. It was during this time that he met Tamar Keren-Portnoy of Israel, who was in Britain on the same fellowship. She became a third member on the project, eventually helping add Lebanese infant babble to the ever-growing database.
When the team finally received the funding necessary to cross- test the British and American babies, DePaolis set up a system that would allow him to observe them in their normal environment and the lab. For the home portion of the experiment, vests with battery-operated recorders sewn into the lining were purchased from a non-profit organization called Lena. On a Friday, participating mothers took the vest and a picture book home, with the instruction to read the book to their baby several times and record for a total of 16 hours over two days. When returned, DePaolis downloaded the audio through automatic processing, but explained that the system does not provide for the nuances of language.
“We have to transcribe a lot of it by hand, especially the bits like the book reading which we look into a great deal. It sounds easy, but you have to keep in mind that the way we speak to infants is often broken,” said DePaolis. “For example, ‘how are you doing’ becomes ‘h’ya’ and we as researchers have to decide if that was really a repetition of the same phrase or not.”
Although the babies who participated in the experiment are now preschool age, there are over three hundred hours of audio tapes that still need to be analyzed and transcribed, providing many JMU graduate and undergraduate students hands on research experience.
“Working with students is great,” said DePaolis. “We go through this process together and I see them learn so much, but I also learn from them. One of my students, Julie Vest, recently made a list of rules for classifying repetitions in speech on the tapes. She did it all on her own, and it’s actually proving to be a great way to approach the difficulties of transcribing.”
After years of looking at infant babbling cross-linguistically, some preliminary differences have emerged. For example, the researchers have found that Welsh babies pronounce hard consonant sounds such as f and v far earlier than their American counterparts, but that American babies are skilled at final consonants before the French. When it comes to differences in caretaker’s speech, American moms often say key words in isolation or at the end of utterances, while British moms place them within longer phrases.
“This doesn’t sound that important, but it’s huge from the baby’s perspective,” said DePaolis. “For example, an American mom was more likely to say, ‘See the ball?’ while a British mom might change that to, ‘It’s a ball, isn’t it?’ These findings are still preliminary as we have a lot more data to go through, but if it is true, it could point to an important difference in how babies come to language. I’ve been really fortunate that I’ve been able to look through all these languages and try to understand the differences and similarities.”
With the help of undergraduate and graduate students at JMU, DePaolis is still searching for the answer to Vihmen’s question. The team spends long hours diligently listening to the recordings in cramped quarters, but it may be in this tiny lab on JMU's campus that the early babble of infants is finally understood.
- January 12
Classes meet as scheduled
- January 19
Martin Luther King, Jr. Holiday
Classes do not meet.
- March 9-13