Same But Different: Systems Of Smoothing Noun Entropy In Communication

Language is a defining human characteristic, and it shapes our cultures. But how does language actually work? Traditionally, people have thought that the meaning of sentences comes from the words it contains and how grammar organizes them. This view suggests that humans are born with much of their language ability, and that some language features—like grammatical gender in German—don’t really serve a purpose.

Our research challenges that idea. We argue that the main goal of communication isn’t just to transfer meaning from one person to another, but to reduce uncertainty about what someone means. From this perspective, children don’t need a perfect linguistic environment to learn a language. Instead, their learning abilities naturally tune into the patterns and rules of language.

We apply this idea to grammatical gender, which is often considered to be a redundant feature of language. We suggest that gender systems actually help people process words more efficiently. In particular, they make it easier to predict and understand nouns—the least predictable part of speech. By comparing German (which uses gender) and English (which doesn’t), we aim to show that gender systems reduce uncertainty in communication. This may help explain how all languages are structured to solve similar problems, revealing important insights into how human language works.