Listen carefully to your two-year-old. What may sound like a simple request for another cookie is, in reality, a complex utterance packed with audible geographic cues, according to recent research conducted by a UVM undergraduate.

The ability to detect regional accents in small children is the subject of senior communication sciences major Katherine Sadis’ research project: "Learning to Talk Native: Listeners’ Perception of Speech from Three Dialect Areas." The project, which documents the ability of adults to accurately identify New York, Tennessee and Vermont accents in native speakers aged two, four and adult, was winner of the Charles A. Ferguson Prize for Best Student Poster at NWAV 34 (New Ways of Analyzing Variation), a linguistics conference held at New York University last fall.

The results are groundbreaking in the field of linguistics, a rare accomplishment for an undergraduate. "That (children) can physically reproduce the sounds when they are just two-years-old is remarkable," says Julie Roberts, professor of communication sciences and Sadis’ advisor for the project. As toddlers, most children are still acquiring single words and are not yet able to string together complete sentences, so the possibility of detecting such nuances as dialect features in their speech has captured the interest of scholars in the field.

Sadis began studying accents while enrolled in Roberts' "American English Dialects" class as a sophomore. Her thick Long Island accent, which became a recurring topic of conversation in the class, may have also propelled her to take on the project, which she began the following year after contacting the communication sciences professor to enquire about continuing her studies.

"It's very unusual for an undergrad to seek a professor out," Roberts says. "I think she's incredibly ambitious and motivated."

Poster prize
Ambition and motivation were certainly necessities for an undertaking that would occupy her entire junior year. "It was a lot of reading, a lot of trying to understand the field. Because it is so new, the research available is difficult to find," Sadis explains. Not only did she complete extensive readings to expand her comprehension of speech and linguistics, she spent months devising, implementing and then analyzing the research project itself.

Beginning with a set of recordings Roberts had made of speakers from Vermont and Tennessee, Sadis created her own recordings of native Long Islanders and created a project that would test the perception of accents using the three sets of acoustical data. With money she acquired through an Undergraduate Research Endeavors Competitive Award, Sadis was able to offer $10 per person to a group of 150 undergraduate students, aged 18 to 25, to listen to recordings of nine speakers, determine each speaker's native region and indicate how sure they were of their answer by choosing one of four certainty levels, ranging from unsure to certain.

After entering the data into spreadsheets, it became clear to Sadis that adults were able to correctly identify the native region of each of the three age groups based on speech, two-year-olds most notably. Not only were their answers correct when it came to the two-year-olds, but the majority of participants indicated that they were very sure to certain of their answers.

The novelty of the research project kept Sadis busy at NWAV. "Everyone was really surprised by the results. For three hours I stood around while people were coming up to me, genuinely interested in the project. It was hard to get a chance to walk around," she jokingly complains.

Among those endless visitors to her table were scholars listed in Sadis’ literature cited, an exciting experience for a student of any level, let alone an undergrad who wasn't expecting to be accepted to the conference in the first place. As an undergraduate with a research project focused more on speech than linguistics, Sadis submitted her project to the conference with an outside chance of being accepted. Winning the award for best poster — while competing primarily against doctoral students — "was unreal."

Bolstered by her success at NWAV, Sadis plans to take her project to the New York State Speech-Language-Hearing conference in April and has been contacted by the University of Pennsylvania Working Papers in Linguistics to publish her findings. With a publication in a scholarly journal lined up for next fall, Sadis hopes to receive several acceptance letters from graduate programs. Among the schools to which she’s currently sending applications, a list that includes New York University, Boston University, Northeastern, Emerson and others, UVM is her top choice. "The amount of support and guidance I’ve received from my professors here — I just couldn’t say enough wonderful things about them."

Sadis remains humble after the stir caused by her research and poster.

"Academically I was a little lost for the first couple years here. I was in communication sciences, but I wasn’t grounded in the field," she recalls. "I’m very average academically. I'm not an honors student… but give me a project and I’ll show you what I can do. The entire experience has given me such a sense of confidence."