Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/3/84; site cfa.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!bellcore!decvax!genrad!panda!talcott!wjh12!cfa!wyatt
From: wyatt@cfa.UUCP (Bill Wyatt)
Newsgroups: net.nlang
Subject: Re: Re: Semantic Reversals
Message-ID: <127@cfa.UUCP>
Date: Fri, 1-Mar-85 12:25:04 EST
Article-I.D.: cfa.127
Posted: Fri Mar  1 12:25:04 1985
Date-Received: Mon, 4-Mar-85 06:18:30 EST
References: <258@unm-cvax.UUCP> <486@ptsfa.UUCP> <22060@arizona.UUCP> <804@sdcsla.UUCP>
Organization: Harvard-Smithsonian Center for Astrophysics
Lines: 33

> In article <22060@arizona.UUCP> gary@arizona.UUCP (Gary Marc Levin) writes:
> >> One curious semantic reversal occurs in the expression:
> >> 	The exception proves the rule.
> >> Rob Bernardo, Pacific Bell, San Francisco, California
> >
> >...[ other responses ] ...
> Another simple reading of this is that the exception to the rule is
> what brings to light (or underscores) the fact that there is a rule.
> That is, you don't realize there is a rule at all until you come across
> something which violates it.
> --  Larry West, UC San Diego, Institute for Cognitive Science

(I think these comments have already been
made, but I had to put in my two cents...)

"Proves" in the above statement harks back to the earlier meaning
of "tests". The same meaning is behind "proof" in alcohol content of
liquor, wine, beer, etc. - the distiller would test the alcohol content
by seeing if it would burn, and what the color of the flame was. I'm not sure
how sensitive this test was, or how high the proof has to be for liquor
to burn, but I think it's 80+ proof, the traditional whiskey proof.

Anyway, "the exception proves the rule" really means that APPARENT
exceptions to a rule *test* that rule. If and only if you can explain
why that "exception" is not really an exception does the rule stand.
TRUE RULES DO NOT HAVE EXCEPTIONS! (except in politics, economics... :-))

The modern interpretation that this saying means "exceptions to some
hypothesized rule prove it really is a rule" is a misinterpretation.
Unfortunately too many people think it's somehow valid, and that's 
fuzzy thinking.
-- 
Bill  {harvard,genrad,allegra,ihnp4}!wjh12!cfa!wyatt