Study of 1,700 languages finds shared grammatical rules across human speech

Monday 06 April 2026 - 16:35

By: Dakir Madiha

Study of 1,700 languages finds shared grammatical rules across human speech

Despite the striking diversity of the world's languages, a large share follow the same underlying grammatical patterns. A large-scale study analyzing more than 1,700 languages has found that roughly one third of long-proposed "linguistic universals" hold up under rigorous statistical testing, pointing to deep cognitive and communicative forces that shape how humans construct language.

The research, conducted by Annemarie Verkerk of Saarland University and Russell D. Gray of the Max Planck Institute for Evolutionary Anthropology, tested 191 proposed rules of universal grammar using Grambank, the largest database of grammatical features ever assembled. Rather than relying on selective sampling of distantly related languages, a common but limited approach in earlier work, the team applied spatio-phylogenetic Bayesian analyses that account for both shared ancestry and geographic proximity between languages.

The findings, published in Nature Human Behaviour, show strong support for recurring patterns in word order, such as whether verbs precede or follow objects, and in hierarchical structures governing how grammatical relationships are marked within sentences. These patterns appeared repeatedly across unrelated languages in different parts of the world.

"Given the immense linguistic diversity, it is intriguing that languages do not evolve randomly," Verkerk said. "Language change must be a central component in explaining universals."

The results suggest that shared human cognition and the demands of efficient communication steer languages toward a limited range of grammatical solutions, even when those languages share no historical connection. Gray noted that the team debated how to frame the findings, weighing whether to present them as a pessimistic account of how many proposed universals fail scrutiny, or as an optimistic one. Ultimately, he said, shared cognitive and communicative pressures push languages toward a limited set of preferred grammatical solutions.

By identifying which universals survive modern statistical methods, the study narrows the field of inquiry for linguists seeking to understand the origins of language structure. The work arrives in the middle of a broader disciplinary debate: a wave of research published in February 2026 by scholars including Steven Piantadosi of the University of California, Berkeley, argued that statistical learning and cultural transmission play a larger role than innate grammatical rules. The new findings do not settle that debate, but offer what the authors describe as clear evidence that language evolution is not random, and that its recurring patterns warrant continued investigation.