Have linguists measured the rate of evolution of a language by analyzing the rate of change of the language's words' usages over time? Is there a term for this sort of measurement?

For example, suppose a language has N words (or usages of a word) W0, W1, W1, …, Wn. For each Wi one could associate a rate of increase or decrease Ri (e.g., determined with Google's N-grams or, better, using some other smarter means which can distinguish usages). Then the rate of change of the language could be estimated by taken some kind of average of all the Wi.

  • It is called glottochronology.
    – amegnunsen
    Commented Dec 20, 2018 at 20:56
  • What do you mean by "usage"? For example "meat" was formerly used broadly to refer to food, but now is used primarily to refer to animal flesh as food. "Sick" is now used (qua slang term) to mean something good.
    – user6726
    Commented Dec 20, 2018 at 20:59
  • @user6726 "meat" as animal flesh, "meat" as food, "sick" as good, and "sick" as ill would all be different Wᵢ's.
    – Geremia
    Commented Dec 20, 2018 at 21:21
  • 1
    It is called Rate of replacement and is symbolised by L. The formula is L = 2ln(r), where r is called Glottochronological constant.
    – amegnunsen
    Commented Dec 20, 2018 at 22:20
  • 4
    It is not known whether r exists or not. I.e, whether the rate of lexical replacement is a constant or not is unknown, and what the value or r ought to be in any particular context is usually not known either.
    – jlawler
    Commented Dec 21, 2018 at 0:50

1 Answer 1


Robert B. Lees devised a "glottochronological constant" in:


It is shown that a linguistic dating system can be set up on the basis of several explicit assumptions about morpheme decay. Thirteen sets of data, presented in partial justification of these assumptions, serve as a basis for calculating a universal constant to express the average rate of retention k of the basic-root-morphemes: k = 0.8048 ± 0.0176 per millennium, with a confidence limit of 90%. Finally an expression is derived for the sampling-error to be expected in the calculated time-depths of related dialects.

  • 2
    The assumption of a glottochronological constant is challenged by many linguists, Daniel Nettle argues that language evolution is faster paced for small speech communities, see also this answer linguistics.stackexchange.com/questions/20971/… Commented Dec 21, 2018 at 10:44
  • 1
    I doubt that small speech communities evolve more rapidly. On contrary, they are more conservative and even more if they are isolated.
    – amegnunsen
    Commented Dec 21, 2018 at 15:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.