Jump to content

ZIPF LAW

From glossaLAB
Charles François (2004). ZIPF LAW, International Encyclopedia of Systems and Cybernetics, 2(2): 3804.
Collection International Encyclopedia of Systems and Cybernetics
Year 2004
Vol. (num.) 2(2)
ID 3804
Object type General information, Methodology or model

A numeric correlation between the frequency of occurence of different words in a language.

Zipf showed that the most common word (“the” in English) occurs 10 times more than the 10th more common word, 100 times more often that the 100th most common word, and so forth.

This intriguing feature seems to be a quite general property, probably correlated to fractals and renormalization properties.

Recently, some genetists found that non-coding DNA (so called “”junk“ DNA) follows the ZIPF relation, i.e. has the structure of a language (Ph. YAM, Scientific American. vol 272, 3, 1995)

Zipf Law is obviously a power law , as it applies to the same general type of phenomena , for instance in the use of words in a language , or in the number of cities as ranked by their population . It reflects a collective property emerging from a statistical behavior or relation among numerous elements of similar individual characteristics within a class or aregion .

This website only uses its own cookies for technical purposes; it does not collect or transfer users' personal data without their knowledge. However, it contains links to third-party websites with third-party privacy policies, which you can accept or reject when you access them.