Darren LaScotte

Darren Lascotte
Darren LaScotte
Teaching Specialist, Minnesota English Language Program
College of Continuing & Professional Studies
UMN Twin Cities

Generative AI and L2 Writing: How can AI Tools Promote Language Learning?

Project Overview 

This generative AI project is aimed at assisting multilingual students in academic writing. It addresses a common challenge these students face: being instructed to write originally and critically while adhering to unfamiliar genre-specific conventions. Traditionally, writing instructors tackle this by providing example essays, analyzing their structure, and giving feedback on students' papers. However, this project explores additional strategies using generative AI tools such as ChatGPT.

A key focus of the project is 'textual borrowing', defined as adopting features from model texts at various linguistic levels to create meaning. This approach helps bridge educational and cultural gaps in learning environments by using textual borrowing to help multilingual students develop their academic writing styles, particularly in adapting to patterns typical in U.S. academic genres.

The project builds upon existing research on textual borrowing, investigating how second-language (L2) English learners can use the 'voices' of others and AI tools to enhance their language proficiency and writing skills. The project explores using generative AI for idea generation, source identification, and discourse analysis. Specifically, it involves questioning how the AI responds to prompts, evaluates the relevance of its evidence or examples, identifies missing information or additional sources, and analyzes the AI's rhetorical moves and organizational patterns.

Ultimately, the project aims to integrate these AI tools into various stages of the teaching and learning cycle, particularly in the writing process, to enhance the educational experience of multilingual students.