# Language Instruction ## expand\_more We work with language instructors both to create materials and tools for their students and to help students use AI in ways that support language learning. Instructors can use AI to generate sample dialogues and other course materials, or build interactive websites for language practice. We also work with faculty on teaching students to use these tools: students might code their own learning apps or learn to prompt and practice with models that can communicate in multiple languages. Part of that learning is investigating what the models do well and where they fall short, so students can judge where to trust them and where to be more critical. ## Hands-on AI training for language instructors The Learning Lab runs hands-on AI workshops for language instructors: sessions for a single department, and sessions with the Language Center that are open to instructors from language programs across Harvard. One sequence, led by Madeleine Woods and Jungyoon Koh, was shaped with instructors in Romance Languages and Literatures through preparatory conversations and built on sample lessons contributed by RLL faculty across multiple languages and levels of instruction. Instructors learned to generate dialogues and reading passages keyed to a given week's vocabulary, to develop new exercises and comprehension questions, to surface patterns in student errors for individual students and for whole sections, and to test how well AI voice models handle their particular language, since quality varies sharply across languages. In small groups, instructors then applied the tools to lessons from their own languages and levels, and the session closed with discussion of the affordances, limits, and ethics of AI in language teaching. Throughout, the aim was culturally relevant, level-appropriate materials matched to what a specific class needs, grounded in the instructor's own pedagogical judgment, at a fraction of the preparation time. To arrange an AI training for your language program or department, contact the Bok Center. ## Students build their own practice tools (Modern Hebrew BB) Students in Modern Hebrew BB built their own AI-powered practice tools, such as interactive flashcards and vocabulary drills populated with their own course materials, in a workshop Hebrew instructor Romy Neumark designed with the Learning Lab. Students worked from a short reading list on how people learn languages, covering spaced repetition, comprehensible input, and deliberate practice, and designed tools grounded in that research and aimed at what each of them needed to practice. Students already use AI to study; a workshop like this teaches them to do it in ways that follow the evidence on how language learning works. To run a workshop where your students build study tools grounded in learning research, contact the Bok Center. ## A faculty-built research site for French students (French 64) After attending the Bok Center's summer workshop series on Claude for teaching, course development, and research, Claire-Marie Brisson, Preceptor in French, built Le Sillage in Claude Code: a bilingual website where her French 64 students read, transcribe, and map the written record of French North America. That record is scattered across dozens of libraries, and locating it can take most of a semester, so students in a language course rarely get to work with it. Le Sillage puts the documents in one place: images of manuscripts, early printed books, and maps called up live from the libraries that hold them, with citations attached automatically when a transcribed passage goes into an essay. The scholarly work stays with the students. They read the pages in the French in which they were written, transcribe them, decide which places and people to tag on the course's shared map, and write their essays on the site alongside their sources and notes. Le Sillage grew out of research Brisson has pursued since her dissertation; the workshop series gave her the means to turn years of her own scholarship into a place where students can work with it. To learn to build tools for your own course, contact the Bok Center. ## AI literacy in a translation course (FRENCH 16) As students increasingly turn to AI platforms for translation, Professor Nicole Mills, in collaboration with Matthew Rodriguez, redesigned FRENCH 16: Reading, Understanding and Translating Written French for Research. Building on its traditional focus on academic, literary, and philosophical translation, the course now foregrounds AI literacy as a core learning objective: students learn to translate with attention to meaning, argumentation, tone, and cultural context, while also evaluating and revising AI-generated translations, identifying bias, experimenting with prompt engineering, and making informed post-editing decisions. The course connects these skills to students’ diverse disciplinary interests and research needs, and now includes sequenced AI-enhanced learning modules, comparative translation and bias-identification activities, and new assessment rubrics for fidelity, clarity, tone, and disciplinary alignment. To support these objectives, the Learning Lab designed a hands-on session at the outset of the semester introducing students to how large language models work, how tokenization shapes translation, and where and why AI tools succeed or struggle. Students experimented with designing custom AI tools to translate French texts, comparing how different approaches handled meaning, context, and ambiguity, and considered the changing role of translation in an era of machine-generated text. The workshop emphasized to students that AI makes close reading and linguistic expertise more essential, not less, and that evaluating a machine translation requires returning to the original and recognizing what may have been lost, altered, or obscured. To plan a class session on how AI models work and where they fail in your field, contact the Bok Center.