Digested from Wikipedia:
Willard Van Orman Quines Indeterminacy of
translation considers the methods available to a
field linguist attempting to translate a hitherto
unknown language. He notes that there are always
different ways one might break a sentence into words,
and different ways to distribute functions among
words.
I've wondered about the arbitrariness of this. Are
some of the distinctions among language types really
do to spelling conventions?
See Spanish daré vs. Old Spanish dar he, both
pronounced the same yet one is analytic while the
other is inflected; one is 2 words while the other is
infinite verb stem + bound morpheme for 1st person
singular future tense.
Fine, but in French and Italian the infinitive +
avoir, avere starts becoming truly inflected
nous parlerons vs. parler [av]ons --maybe as
apocopated verb
parlaremmo vs. parlare abbiamo --here abbiamo is
completely unrecognizable.
My question is how much latitude do we have at the IE
level between words and morphemes?
Was Quine on the right path as essentially seeing
morphemes as "demoted" words? At least that's what I
get out from his remarks.
Can we posit all morpheme endings as originally
separate words?
This, of course, is also where he refers to his
"gavagai" thought experiment. We see a rabbit go by
and a local yells "gavagai". What are we to think?
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com