🚀 go-pugleaf

In article <1992Oct15.224329.20578@agt.uucp>, Brian F G Bidulock <bbidulo1%agt@cs.ualberta.ca> writes:
|> In article <rkaivola.718902434@mits> Risto Kaivola,
|> rkaivola@mits.mdata.fi writes:

|> How about:
|> ~A --> ~B
|> or
|> "If we don't invite Tom, he will not visit us."  == "Tom will visit us
|> only if we invite him."

In logic courses (where I tended to spend a *lot* of time on issues
of translation to and from English) I emphasized as a fundamental
principle of translation that one should attempt to have the structure
of the formal representation mirror (as closely as possible) the
structure of the English.  In particular, one should not end up with
a formal representation containing negations if there were no explicit
negations in the original English.  Cleaving to this principle avoids
a number of nasty confusions and problems.  Among other things, it
tends to remove *inference* from the translation process and keep
it where it belongs (post-translation, in the formalism).  And this
is particularly important as the complexity of the sentence to
be translated increases.

Accordingly, a good translation of

	If we don't invite Tom, he will not visit us.

is

	~A -> ~B

since the two negations are explicit in the English.  But this is
*not* a good translation of

	Tom will visit us only if we invite him.

since there are no explicit negations in the English.  These are,
of course (truth-functionally) equivalent, but it is *not* the
case that any logical equivalent to a good translation is itself
a good translation.
-- 
Gary H. Merrill  [Principal Systems Developer, C Compiler Development]
SAS Institute Inc. / SAS Campus Dr. / Cary, NC  27513 / (919) 677-8000
sasghm@theseus.unx.sas.com ... !mcnc!sas!sasghm