AAI News Hub
ResearchWed, August 5, 2026·Aug 52 sources corroborating

ChronoLens tracks language change across five parliaments

The framework uses frozen multilingual models to compare linguistic change from 1803 to 2026.

Why it matters

The work offers a method for studying language change across languages and linguistic levels using multilingual language models. It may help researchers compare how linguistic systems evolve without relying on incompatible representations.

The key points

  • 1.ChronoLens analyzes 44.98 million parliamentary documents.
  • 2.Sparse representations outperformed dense embeddings on linguistic-statistic alignment.
  • 3.Languages differed in timing, magnitude and direction of change.

Researchers introduced ChronoLens, a framework for measuring historical language change across morphology, syntax, semantics and pragmatics in a shared analytical space. They applied it to 44.98 million documents and about 17.2 billion tokens from five parliamentary traditions spanning 1803 to 2026. Its sparse representations aligned more strongly with linguistic statistics than dense embeddings or a pooled sparse autoencoder, with ρ=0.72 versus 0.29 and 0.28.

Try this today

Read the paper before using multilingual embeddings to study historical language change.

Sources & original reporting

This brief summarizes and links to reporting from the publishers below.

Enjoyed this brief? Get the next one in your inbox.

More in Research