AI pragmatists: How language teachers are navigating AI with nuance
Summary
A study at the University of Ottawa's Official Languages and Bilingualism Institute (OLBI) surveyed 24 English and French as a second language instructors to understand their attitudes and use of AI-assisted tools, achieving a 40% response rate. The findings challenge the "resistant teacher" narrative, revealing that most instructors are "pragmatists" who acknowledge genAI's potential but await pedagogical evidence for full adoption. The survey also uncovered a "hidden AI" problem, where instructors use tools like Grammarly or DeepL without recognizing their underlying generative AI components, distinguishing between AI that refines existing work (assistive) and AI that produces new content (generative). Crucially, instructors primarily use genAI for administrative efficiency, such as generating lesson plans, but express reluctance to integrate these tools into student learning due to concerns about cognitive offloading and its impact on "desirable difficulties" essential for language acquisition.
Key takeaway
For university administrators developing AI education policies, you should prioritize consulting educators to ensure policies are pedagogically sound and not merely administratively convenient. Trust your instructors to define the terms of AI engagement, focusing on critical AI literacy that distinguishes between assistive and generative AI functions, and protects the learning process by valuing drafting and revision over final product generation.
Key insights
Educators are pragmatic about AI, distinguishing between assistive and generative functions, and prioritize learning process over efficiency for students.
Principles
- Teachers resist tools that don't solve problems.
- Learning requires "desirable difficulties."
- Policy must follow pedagogy.
Method
A bottom-up institutional survey of ESL/FSL instructors examined attitudes and current use of AI-assisted tools, yielding qualitative insights into nuanced decision-making.
In practice
- Distinguish AI by function: assistive vs. generative.
- Design assessments valuing process, not just product.
- Reposition instructors as learning architects.
Topics
- AI in Language Education
- Generative AI Adoption
- Cognitive Offloading
- Pedagogical Evidence
- Assistive AI
Best for: Research Scientist, Domain Expert, Policy Maker
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial intelligence (AI) – The Conversation.