[86914] in tlhIngan-Hol
Re: The topic marker -'e'
daemon@ATHENA.MIT.EDU (Christopher Doty)
Sun Nov 22 13:51:30 2009
In-Reply-To: <249d5b950911220911u5cd335b7g38f46d46b5a30508@mail.gmail.com>
From: Christopher Doty <suomichris@gmail.com>
Date: Sun, 22 Nov 2009 10:48:37 -0800
To: tlhingan-hol@kli.org
Errors-to: tlhingan-hol-bounce@kli.org
Reply-to: tlhingan-hol@kli.org
> "mapum" doesn't mean 'fall'. It means "we fall" (or "we accuse"; "pum" is
> two different verbs). There is no point in losing information that is given
> in the original just because the translation is odd.
> In fact, "mapum Sor" could be interpreted as "We trees fall", although this
> use of a noun as subject with a non-third-person prefix is controversial at
> best.
I think this exactly what Tracy meant in saying that, for
ungrammatical (or "controversial") sentences, the machine translator
isn't going to work very well due to ambiguity. You posit three
possible interpretations of "mapum (Sor)" because of the ambiguity
found in an ungrammatical sentence. There seems little point in
having an automatic translator that could posit every single possible
esoteric meaning for anything ungrammatical...
Chris
