3/20: Computers and language learning: where we started, where we're going

Laura Romig, Class of 2025, Language Ambassador

If you're learning a language in 2023, you probably use some kind of technology to assist with your studies. Whether that's having assignments posted on Canvas, listening to audio clips online, or participating in your language class through Zoom, technology has become a cornerstone of language education. And as the field of artificial intelligence has grown substantially, AI has also been recruited as a potential tool or alternative for language learning. Researchers at top universities, for example, have createdan artificial intelligence system capable of learning linguistic rules and nitpicking grammar textbooks, and even the US government has signaled that AI has a role to play in our education-technology. 

Given this context, it seems likely that future language learning experiences will incorporate AI, though it's not yet clear how. One aspect to pay attention to—in all fields, not just language learning—is whether AI-centric technology seeks to replace the face-to-face, human aspect of interaction. This enthusiastic piece, for example, argues that the instant feedback and lack of judgement from AI tools will facilitate language learning. If language learners are reliant on technology not only as an auxiliary reference, but as their source of knowledge, feedback, and encouragement, it might eliminate the impact of learning from another human. Multilingual environments and second language learning have been shown to increase empathy, by some measures, but the strongest effects seem to come from when a child is exposed to a second language since a young age, often from family members or an immersive environment. Can these kind of holistic benefits be replicated with technology-first language learning? It's unclear.

AI discussion aside, the history of technology in language learning, or Computer-Assisted Language Learning (CALL), as the term was coined in the 1950s, has humbler origins, beginning with drill-based methods on slow, soundless computers. This type of language learning mostly involved using the computer as a tool for repeated and formulaic input, essentially aiming for memorization. It corresponded with pedagogical theories at the time, which postulated that learning meant a person changing their behavior in response to external stimuli. But this theory of learning, as well as the available technology, were soon to change; a second stage of CALL emerged, in which the learner's own discovery and critical thinking mattered more. Updated technology also meant that students could use computers as assistants in producing their own work, such as by typing in their second language and getting grammar edits. 

Finally, the era of CALL we are most familiar with is the era of multimedia and the Internet. This era has allowed for the integration of multimodal stimuli into the same activity on a computer—graphics, sound, video, text—which we take for granted now. The openness of the Internet also means we have more rapid access to many native speakers of a foreign language than ever before. Additionally, current theories of learning often emphasize the importance of how individual learners construe and construct the world, which has become part of the structure of today's CALL. Assignments might be more open-ended, project-based, and interpretative, and one-to-one translation is less emphasized. 

This broad history leads us back to present day, a time in which future education researchers might be likely to distinguish a new phase of CALL, based on the rising prevalence of AI alone. Already, some researchers have discussed and tested the possibilities of robot-assisted language learning, for one example. The original paper that divided CALL into those three phases was published back in 1998, before apps, social media, and much of the modern Internet, so an updated structure seems past due.

So far, the AI age of CALL seems to be focused on more interactive, more instantaneous, and more individualized language learning. Education theorists have also suggested the possibilities to further integrate reading, writing, speaking, and listening to simulate a more immersive environment. Virtual reality will probably be recruited as a language tool, and language corpora (large collections of real spoken or written language, not unlike the datasets used to train AI models) might be used to make input for learners more authentic. The end of this article details more predictions for this age of CALL.

As we come to a possible new stage in its development, it's helpful to know the broader context of where language learning technology has been in the past. For more, check out this article about what the Metaverse has to do with language education or this recent review of technologies for collaborative writing in language learning.