Data Week 2026: Museum of Languages

Date: June 3rd. 10:30am - 8:15pm

Part of Data Week June 1st - June 5th

Location: TBC

The most tumultuous period for a language is often the moment when all its speakers are bilingual. Before that point, there was always a reason to speak it: to hear stories from elders, to trade with neighbours on distant farms, to share in-jokes within close-knit communities. Once everyone is bilingual, the language no longer needs to be spoken; it survives only if people choose to speak it. Many languages disappear soon after their speakers reach universal bilingualism. Sometimes, though, the opposite happens: bilingualism makes the less-spoken language newly visible, turning it into a source of pride, a cultural treasure that enriches people's sense of the world and a heritage they feel called to preserve. Sometimes the moment the population became bilingual does not herald a language's demise but, instead, marks the beginning of its revival.

We are approaching a similar tipping point, not for one language, but for all of them! Machine learning tools are making translation effortless. Machine-assisted subtitling lets us watch films and television without language barriers; we email or text in our own language, confident that recipients can read it in theirs through machine translation. Soon, instantaneous translation will make real-time conversation possible between people with no shared tongue. At the same time, learning a new language is easier than in the past. Modern learning apps are a vast improvement on the tape-and-book format of twenty years ago, and they are beginning to include realistic and carefully staged conversational practice with AI avatars. Online platforms also mean that in-person lessons have become cheaper and more accessible.

We do not yet know what all this will mean for linguistic diversity. Will machine translation mark the end for many languages, or the beginning of a revival? Will we all speak one of a handful of world languages, or will our current lingua franca splinter into dozens of new tongues? Will we all speak one language, or will each household have a language of its own? Does this matter? Is a language a way of understanding and interpreting the world, built on the lived experience of its speakers and the history of a culture, or does the success of machine translation demonstrate the primacy of meaning over expression? Is language diversity a cultural treasure or an impediment to shared culture? Does a world disappear when a language dies, or is nothing lost?

This workshop will consider some of these questions, but the last is easy to answer: when a language dies, something is lost. The craft of a writer lies in the precise choice of words, in the precise use of language; the art is in the words themselves, not just in the facts they refer to. Our own speech and writing are distinct not only in content but in style; the manner of our speech and the idiosyncrasies of our writing help define our identity. These choices, the choices writers make in writing and the choices we make in speech, occur within the context of the language we use, and this distinct use of words is lost in translation. A translation may preserve the facts; it may be as artful and as beautiful as the original, but it is always a new text. In this way, when a language dies something is lost. By the same measure, however, a world without a lingua franca, a world full of languages may lack any common understanding since all communication would exist in translation.

How, then, do we preserve the written and spoken culture of the past as the languages in which they were composed change and disappear? This question is particularly acute now, when machine learning brings us to this tipping point for language survival, and machine translation may transpose facts without translating expression. If machine translation means the death of lesser-spoken languages, how do we preserve the corpus they produced? Conversely, if it destroys the very notion of a lingua franca and frees languages to diversify, how will we sustain a shared culture or preserve the literature of the past as language continuously shifts and changes?

The workshop

A Museum of Languages, if one could be imagined, would not be a monument to what has been lost but a living space for what endures, the human instinct toward the creation of languages. Such a museum would invite us to imagine a future of communication, a future in which technology and tradition shape one another rather than compete. This workshop aims to consider what a Museum of Language might be and to be itself a transient Museum of Language: to host a discussion of language diversity and preservation in the age of machine learning, and to exist, briefly, as an act of remembrance and celebration of languages.

The workshop will have three elements:

• A set of academic talks and panel discussions bringing together people with interests in machine learning, translation studies, linguistics, languages and museums.
• Ten five minute ``Ignite-style" talks introducing ten different languages, each speaker will be asked to teach something distinctive or beautiful about the language they speak.
• A key-note evening talk aimed at a broader public addressing a topic related to language diversity, language preservation or language learning. 

Indicative timetable:

10h30-11h00: Coffee!

11h00-11h10: Introduction to the Museum of Languages.

11h10-11h20: Two language introductions

11h20-12h00: Two talks about language translation.

12h00-13h00: Lunch!

13h00-13h10: Two language introductions.

13h10-13h50] Two talks about language diversity and preservation.

13h50-14h00: Two language introductions.

14h00-14h30: Coffee!

14h30-14h40: Two language introductions.

 14h40-15h20: Four short talks on ``What a Museum of Languages might be".

15h20-16h00: Panel discussion and wrap up.

16h00-19h00: Break!

19h00-19h10: Introduction.

19h10-19h20: Two language introductions.

19h20-20h00: Keynote talk.

20h00-20h10: Wrap up.