Become a member

Language Magazine is a monthly print and online publication that provides cutting-edge information for language learners, educators, and professionals around the world.

― Advertisement ―

― Advertisement ―

Aligning AI-Powered Writing Practice with Common Core Standards

LearningWrite helps ELL and world language educators provide students with fast, fair, and frequent writing support, through an online platform powered by AI. It...

Anonymous Edvisors

Emboldening Indian Youth

HomenewsResearchUT Austin Statisticians Win Prestigious Award for Language Learning Research

UT Austin Statisticians Win Prestigious Award for Language Learning Research

UT Austin statisticians Giorgio Paulon and Abhra Sarkar have been awarded the prestigious Mitchell Prize for their paper on tone learning in adults.

The Mitchell Prize, jointly sponsored by the American Statistical Association, the International Society for Bayesian Analysis, and the Mitchell Prize Founders’ Committee, is awarded annually to the author(s) of an outstanding paper that utilizes Bayesian analysis to solve an important applied problem.

Paulon and Sarkar’s study centered around the biological changes that take place in the brain when non-native English-speaking adults learn another language’s tonal differences.

Because of its extensive use of tonal variations, Paulon and Sarlar selected Mandarin Chinese as the target language for their study.

For example, there are four ways to pronounce “ma” in Mandarin Chinese and each way has a completely different meaning; pronouncing it one way means “mother” and another way means “horse.”

At the outset of the experiment, 20 non-native English-speaking subjects were taught how to differentiate four tonal variations in Mandarin Chinese.

In the days that followed, subjects were tasked with listening to native Mandarin Chinese speakers and reporting on the tones they heard.  

The study yielded two significant findings: 1) tones 1 (high level) and 3 (low dipping) were the easiest for subjects to learn to tell apart, and 2) the greater a subject’s aptitude for language learning, the quicker they process audio information when learning tonal differences.

Paulon and Sarkar believe the implications of their findings could be far-reaching. “…this could help eventually develop precision learning strategies for different people depending on how their individual brains work,” said Sarkar.

Additionally, Sarkar believes the statistical model his team used could assist clinicians in understanding why an individual has a speech or hearing disorder, and help neuroscientists studying other kinds of decision making.

Paulon and Sarkar’s paper appeared in the Journal of the American Statistical Association in September 2020.

Previous article
Next article
Language Magazine
Send this to a friend