--- In
cybalist@yahoogroups.com, "ghozzis" <ghozzis@...> wrote:
> Hello!
> Do you know a way to measure how far a language is from another?
> And I have another question similar to the previous one: do you
know
> a way to quantify the "rate of evolution" from a language from a
> language it is derived from.
> I remember reading once that French was 44% different from Latin,
> whereas Italian was just 17% or something in this area.
> Thanx for your answers!
These numbers probably come from the Swadesh 100-word list and the
200-word lists (several variants - at the Rosetta Project,
www.rosttaproject.org , you'll consequently see something like 208
words in the 'Swadesh list'). The words on the 100-word list are
generally less likely to be replaced than the others on the 200-word
list, and the 20 best conserved words (word-conserving meanings, to
be more precise) vary widely in their replacement rates. How well
speed of change in the lexicon matches other changes is another
matter; the lexicon is just easier to quantify.
I'm not sure how much stability was taken into account in drawing up
the 200-word list. The first one was drawn up as a list of meanings
for which every language should have a word.
Richard.
P.S. Do list and web searches before you even ask about
lexicostatistics.