How AI is learning to unlock the secrets of historical texts
Translated from German and summarized by DistantNews. Read the original for the full story.
At a glance
- Artificial intelligence and machine learning could help researchers examine large collections of previously inaccessible medieval documents.
- The St. Gallen Abbey Library preserves manuscripts, incunabula and early printed works spanning roughly from the eighth century to today.
- Historian Helmut Reimitz says AI will allow researchers to pursue long-standing questions at a scale that was unthinkable a decade ago.
Artificial intelligence could give historians access to the hidden contents of medieval archives, which hold vast collections of papyri, records and documents that researchers have not yet fully unlocked.
The St. Gallen Abbey Library is a central example. It is the only major early medieval monastic library whose holdings have remained relatively intact from the eighth century to the present. Its collections include manuscripts, incunabula and early printed works.
Helmut Reimitz, a medieval historian moving from Princeton University to the University of Vienna, is set to direct the Institute for Medieval Research at the Austrian Academy of Sciences. He remains cautious about claims that AI will bring a revolution to historical research. But he says the technology is already changing the scale of what scholars can investigate.
"We can pursue questions that have occupied us for a long time on a scale that was unthinkable ten years ago," Reimitz said. He also expects the technology to change the questions historians choose to ask. Machine learning, in this view, is not only a tool for reading old texts. It could influence the direction of historical research itself.
We can pursue questions that have occupied us for a long time on a scale that was unthinkable ten years ago.
Originally published by Der Standard in German. Translated, summarized, and contextualized automatically by DistantNews, with a note on how the source frames the story. Not individually reviewed before publishing. How this works.