Zipf's law

Zipf's law is an empirical regularity stating that the frequency of any word in a natural language corpus is inversely proportional to its rank in the frequency table. The most frequent word occurs approximately twice as often as the second most frequent word, three times as often as the third, and so on. Formally, if words are ranked by frequency f(r), then f(r) ∝ 1/rᵃ, where the exponent α is typically close to 1. This is a specific instance of a power-law distribution applied to ranked categorical data.

The law is named after linguist George Kingsley Zipf, who observed it in multiple languages, but similar patterns appear in city populations (where it is called the rank-size rule), website traffic, company sizes, and other domains. Zipf's law has been explained through multiple mechanisms: least-effort principles in language production, random typing processes, and optimization under information-theoretic constraints. None of these explanations is universally accepted, and the law's ubiquity remains partly mysterious.

What is clear is that Zipf's law emerges in systems where entities compete for a shared resource — attention, space, or energy — and where early advantage compounds. It is the linguistic signature of the same positive-feedback dynamics that produce scale-free networks in topology and Pareto distributions in wealth.

Zipf's law is often presented as a curiosity of language statistics. The deeper pattern is that any system in which entities compete for limited attention will produce a power law in the ranking. Language is just one instance. The law tells us not about words but about competition: when winners take more than their share, the rank-frequency relationship becomes inevitable.