Why do kids say "dog" months before they say "good"? Both are short and simple, yet one comes much later. In this episode, we look at a new study that tested one possible reason, using a computer model of the brain.

A team in Berlin built a small brain model and taught it two kinds of ideas: things like hammers, where every example looks alike, and things like beauty, where the examples have nothing that they all share. The model learned what a hammer is on its own, before it ever heard the word. But it could not learn what beautiful means until it heard the word beautiful again and again.

So for some ideas, the word may not just be a name. It may help build the idea itself.

To be clear: this is a computer model. No child was tested, and the authors say this is probably not the only way we learn such words.

The research was funded by the European Research Council, so it was paid for with public money.

THE PAPER

"What is so hard about abstract words, anyways? A neuromechanistic explanation using brain-constrained neural network models"

Fynn R. Dobler, Lorenzo Stroppa, Friedemann Pulvermüller

Cognitive Neurodynamics (2026)

https://doi.org/10.1007/s11571-026-10542-z

Owladio — Science. Translated.

One research paper per episode, explained in plain language.

Watch the video version: https://youtu.be/HOwRyQ-AqFc

Website: https://owladio.com

Instagram: https://www.instagram.com/owladio.science

Facebook: https://www.facebook.com/owladio

Can you remember learning a word like beauty or fair? Tell us in the comments or on Instagram!