Prof. Nathan Schneider receives NSF Career award to develop algorithms for analyzing metalinguistic text
December 19, 2022
The National Science Foundation has awarded Nathan Schneider a grant of more than $500,000 to develop algorithms to analyze metalinguistic text. Schneider, whose research lies at the intersection of computation and human language, is investigating how natural language processing (NLP) can incorporate explicit descriptions of how languages work by linguists, judges and lay speakers.
“The grant is focused on the concept of metalanguage,” Schneider explains. “We talk using language and we talk about language. Language-about-language shows up in writing too: in dictionaries, textbooks, scientific articles, web forums – even legal cases. I want to develop technologies that help us learn about language from those texts.”
The grant is an NSF CAREER award, which supports early-career faculty “who have the potential to serve as academic role models in research and education and to lead advances in the mission of their department or organization.”
Schneider’s study of metalanguage encompasses multiple genres. One is the extensive collection of writings, opinions and rulings that constitute the American system of justice. Laws are written down, and when their meaning is contested in court, it is up to judges to articulate an interpretation.
“Some of the clearest statements of language don’t come from the classroom, but from the courtroom,” says Schneider. “So we’re starting with court cases that leverage this expert interpretation of language.”
Parts of the internet, too, are abuzz with language commentary. “I’m also interested in discussions by language learners and writing that is aimed at those learners,” Schneider says. “Online forums where learners post questions about English, for example, are a gold mine for NLP algorithms.”
The research – conducted in partnership with doctoral students in Schneider’s lab and in consultation with experts including Kevin Tobia, a Georgetown Law professor – could result in publicly available code and tools that empower academics and members of the public to augment their reading and writing. Other areas stand to benefit as well.
“Metalanguage understanding is an important approach for learning how language constructs relate across different taxonomies, a foundational issue for social science and public policy research,” says Lisa Singh, Professor of Computer Science and Director of the Massive Data Institute.
-by Hayden Frye (C’17)