I feel a little silly, but also proud to say that I was able to, at least minimally, keep my language learning streak alive while I took a vacation to Europe last month. Since the throes of COVID, I’ve been using the gamified aspects of the DuoLingo app to become acquainted with German. It’s the native tongue (Muttersprache) my immigrant great-grandmother offered me for free as a child and that I foolishly declined.

Common wisdom has it that childhood is the best time to learn a new language (Snow and Hoefnagel-Höhle 1978; Johnson and Newport 1989; Snow 2014), although some research has called this belief into question (Schouten 2009).

Sadly, I am an adult, so I asked a friend who is a language professor. She suggested that no, the owl was not going to make me fluent, but it was a good way to get familiar at least with the sounds of the language so I could more effectively pick it up from other places as well. 

It necessarily begins from a small set of vocabulary and builds, so my DuoLingo-informed view of Germans includes having a lot of barbecues and cleaning the apartment, as well as working as a civil servant, lawyer, doctor, or actor. There are also some truly bizarre sentences that are fodder for social media accounts.

For a long time now, automated translation like Google Translate has made it difficult for language instructors to persuade students to learn the language the hard way (Jolley and Maimone 2022). As AI support for translation has improved, it’s easier than ever to travel without making an effort to learn the local language at all. You can say what you want to express into your phone, and the phone can translate out loud for the person with whom you are speaking. There are even these earphones, evocative of Star Trek’s universal translator. 

So what are the implications of AI-assisted language translation on a high-impact practice like study abroad?

The Center for Engaged Learning’s seminar on Global Learning in Higher Education’s statement on integrating global learning notes “Our research groups define global learning as a lifelong developmental process in which the learner engages with difference and similarity and develops capabilities to interact equitably in a complex world.”

If learners can navigate while abroad without learning the language, I worry that they will miss opportunities to learn about the culture in a way needed for intercultural understanding. There are a few reasons why.  

First, is the experience I’ve had taking my own students on short-term study abroad where some know the language from previous experience and others don’t. I have generally found that students who know some of the language learn things verbally, as they concentrate on what people are saying. Students who don’t know the language learn things in other ways. Unable to understand the words, they tend to develop rich insights from what people do and from the environment. They have seen the way people drive while those with prior language knowledge have been focused on decoding the advertising on the side of the road, for example. In this faculty-led program, these different lessons make for rich group discussions that put the culture and language in a larger context. For students who are not faculty-led, if a reflection piece is missing, that opportunity to see the big picture may fade. 

Second, generative AI is only as good as its training data. If that data is biased towards groups that have been social influential, which is common, those who depend on AI translation may only get pieces of a complex culture as their understanding is always presented through translation with a majority lens. 

Contemporary higher education enrollment pressures are reducing faculties in world language departments. At the same time, intercultural competence is treated as a necessary strategic advantage for national security and economic success. In the face of these realities, it’s easy to see a future where students are sent abroad to absorb those cultural lessons without the struggle to learn to communicate first, relying on AI support to do the translation. More study would be useful to see how this affects student learning. 

POSTSCRIPT – AI as a Language Tutor

One of the things some educators are excited about is the ability to have generative AI act as a custom tutor for learners. I realized I don’t know the German prepositions very well at all, so I created the following prompt for Chat GPT 4.0 and found it to work pretty well:

I want you to act as my German tutor and help me learn to use the prepositions in German. When I say begin, start offering me simple, 1-sentence prompts using A1 level vocabulary and present tense grammar where there are 5 asterisks in the places where prepositions go. Limit the number of prepositions to 1 per sentence. At the conclusion of the prompt, ask me to answer with what the correct pronoun is. Offer me feedback on my answer, indicating if it is correct or wrong. If it is wrong, offer a brief and simple understanding of what the correct preposition is and why. Please offer me 10 prompts. After the 10th, please ask me if I want to continue. Begin.

References

Johnson, Jacqueline S., and Elissa L. Newport. 1989. “Critical Period Effects in Second Language Learning: The Influence of Maturational State on the Acquisition of English as a Second Language.” Cognitive Psychology 21 (1): 60-99.

Jolley, Jason R., and Luciane Maimone. 2022. “Thirty Years of Machine Translation in Language Teaching and Learning: A Review of the Literature.” L2 Journal: An Electronic Refereed Journal for Foreign and Second Language Educators 14 (1).

Schouten, Andy. 2009. “The Critical Period Hypothesis: Support, Challenge, and Reconceptualization.” Studies in Applied Linguistics and TESOL 9 (1).

Snow, Catherine E., and Marian Hoefnagel-Höhle. 1978. “The Critical Period for Language Acquisition: Evidence from Second Language Learning.” Child Development, 49: 1114-1128. 

Snow, Catherine E. 2014. “Relevance of the Notion of a Critical Period to Language Acquisition.” In Sensitive Periods in Development (pp. 183-209). Psychology Press.

About the Author

Amanda Sturgill, Associate Professor of Journalism at Elon University, is the 2024-2026 CEL Scholar, focusing on the intersection of artificial intelligence (AI) and engaged learning in higher education. Connect with her at asturgil@elon.edu.

How to Cite this Post

Sturgill, Amanda. 2024. “Using Generative AI to Support Global Learning Could Leave Cultural Lessons Lost in Translation.” Center for Engaged Learning (blog), Elon University. August 20, 2024. https://www.centerforengagedlearning.org/using-generative-ai-to-support-global-learning-could-leave-cultural-lessons-lost-in-translation/.