names translations

[Logical Language Group <cbmvax!uunet!cuvmb.cc.columbia.edu!lojbab>]

Mark writes:
>And we here say "New Delhi", "The Dead Sea", etc, etc.  Belgium isn't
>the only place with two native names; what about lakes/rivers/oceans
>that form borders between two countries with different languages?  Those
>certainly are local to both (or all).  And as you say, what of the moon?
>I fear the only way I see is the informal, patchwork, inconsistent
>manner of naming found in current NLs.  Open to new ideas, tho.

My hypothesis is that names, and words you want to borrow for that
matter, may be loan-translated iff the etymological components are
semantically meaningful in some obvious way; i.e. the name is
'descriptive'.  The name "The Dead Sea" describes a feature of the body
of water.  It is known, and not just locally, that this description
applys in this way.  Likewise the Great Salt Lake of Utah.  New Delhi
can be mix translated iff there exists a Delhi, which New Delhi is a
rebuilding of.  Mark's West Orange works, because (I believe) there is
an East Orange, and maybe a couple of other Oranges nearby.  Thus, if
the group of them is called (as it might be) "The Oranges", then if the
town to the west is West Orange, Lojban-translating the West is legit.

New York does not have an Old York nearby to be contrasted, and the
listener may not know of the one in England, and can't find it on a
typical map showing New York (the map must be in Lojban of course, or
this wouldn't be an issue).  Meanwhile the locals will not recognize it
in translation either - so nothing is gained.

When in doubt, Lojbanize from the local name, not loan-translate, but
there are rare cases where loan-translation is valid.

Only my opinion
lojbab