Concept · linguistics

Zipf's law

In natural text, a word's frequency falls off in proportion to its rank.

In a long text, the most common word appears about twice as often as the second, three times as often as the third, and so on. Natural languages follow it closely; so, it turns out, do many artificial texts, so a Zipfian curve is weak evidence that a text is language and a non-Zipfian one is stronger evidence that it isn't. The site's unread scripts are tested against it as a first filter, not a proof.

Cite it

Mysteries Department. (2026). Zipf's law [Concept CPT0DECVZZZVP1M]. https://mysteriesdept.com/r/CPT0DECVZZZVP1M

@misc{mysteriesdept_cpt0decvzzzvp1m,
  title     = {Zipf's law},
  author    = {{Mysteries Department}},
  year      = {2026},
  note      = {Concept, catalog key CPT0DECVZZZVP1M},
  url       = {https://mysteriesdept.com/r/CPT0DECVZZZVP1M}
}