Quod erat demonstrandum
When working at Guardian Unlimited's Film desk, I once made an Excel spreadsheet to predict the Oscar winner for best picture. It worked fantastically well until last year when it predicted The Aviator, rather than the actual winner, Million Dollar Baby. It's probably unrelated, but I was moved off the film desk the next day.
This week a colleague of mine blogged about my abject failure when the Golden Globe winners were announced, and kindly excused me on the basis that Million Dollar Baby was "the exception that proved the rule". Of course I was touched, but more importantly I was reminded that set expressions can have their meaning reversed by changes to the meanings of one of their constituent words.
"Prove" comes into English sometime before 1200, via French, and ultimately from Latin "probus" meaning good or trustworthy, whence our "probity". "Probare" in Latin meant to examine or test something or someone for its worthiness or fitness. From there the meaning extended to "demonstrate its fitness", and from there to recommending it - whence our "approve". Chambers Etymological Dictionary gives two forms for the first uses in English, of which the first "pruven", meant test, although the second "proven" also meant to demonstrate (using a test) that something was adequate or true. "The proof of the pudding" which we understand to reside in the eating thereof could thus be one of two things: a test to ascertain whether the pudding was any good, or a demonstration that it was (confusingly,"to prove the pudding" would also mean to allow its constituent dough to rise before baking). Similarly, I have been assured - and I encourage contradiction from better informed readers, as I haven't been able to source this - that "the exception that proves the rule" originally meant "the exception which exposes the rule to a test" - and which by implication finds it wanting. Certainly, anyone putting my Oscar predictor through "a test for fitness" would have to fail it.
The other example of this, and one on which I'm very keen, is the old saw that "you feed a cold and starve a fever", which is to say the remedy for a slight cold is to tuck in, while a serious fever is treated through abstinence. Starve comes from Old English steorfan, which meant simply to die or kill. If that sounds odd, "starve" can still mean either "die of hunger" or "kill by depriving of food". Only in later Middle English did the word specifically apply to death through malnutrition. "Sterben" in German still has the more general meaning.
Again, I would love to know of documentary proof (=demonstration) for or against this one, but I have been told by a reliable source that the original saying was "you feed a cold and starve [ie. die] of fever", which, you'll note sounds exactly the same as the version we know, though one of the words is different. That is to say, the original advice was - very strongly - whatever you do, don't pig out when you've got a sniffle, unless you want to develop a serious illness and shuffle off. We know that some phrases can stay in English long after the original meanings of the words in them has changed, or even beyond the point when those words have any meaning. Who now knows what a bushel is? And yet we all strive not to hide our lights thereunder. (Readers are politely thanked for not writing in with explanations about the meaning of the word bushel.)
For once, there are important life lessons here.They are: 1. Don't trust old wive's tales like "feed a cold and starve a fever". Like "wine after beer will make you feel queer" but "beer before wine will make you feel fine", they're often so opaque or self-contradictory as to be meaningless. 2. Don't trust rules with exceptions. 3. Don't trust my Oscar predictor.
