Nothing generates email like a discussion of what constitutes proper English usage. My interpretation of the common phrase "the exception that proves the rule" is that a fact that is known to be exceptional counts as evidence that the opposite statistical tendency is true. For example, that Spud Webb was famous for being a 5'-7" NBA player suggests that most NBA players are tall..
In contrast, I've never heard of a soccer player being famous for being 5'-7", so I presume that soccer players are not extraordinarily tall on the whole.
But, quite a few readers have offered objections falling into two classes, linguistic and legal:
"I had always heard that this was a mistranslation of the Latin "probat" which can mean either proves or tests. The (seeming) example tests the rule means it puts the rule to the test -- if you can't come up with an explanation then the rule has to be thrown out."
So, I guess, the exception probes the rule.
Okay, but people have been using the phrase in English to mean the opposite for centuries.
Another offers the legal interpretation:
The phrase has been shortened (and its meaning corrupted) from a legal maxim - that fact that certain exceptions are explicitly made proves that a rule is true in all other cases.
In other words, if the sign says, "No Parking 4-6pm" that likely means the rule is that you can park at other times. (Here's a "Straight Dope" column reviewing the controversy.)
Much of the disagreement stems from what is meant by "rule." I'm going to review this in some depth because it's directly related to the problems I've run into in discussions over the existence of race and my definition of it as a partly inbred extended family. I've often been told that race can't possibly be of any importance in the modern world because, "Look at Tiger Woods." I reply that the fact that Tiger Woods is famous for being highly multiracial suggests that an awful lot of people aren't terribly multiracial. Tiger Woods is an exception that proves the rule.
The first alternative definition (the corruption of the Latin phrase about exceptions testing the rule) comes from the math/physics tradition, which (deservedly) has immense prestige. If the orbit of Mercury doesn't quite fit Newton's Law of Gravity, you aren't supposed to say, "Oh, well, that's just the exception that proves the rule." You're supposed to say, "Hmmhmm, maybe Newton's Law of Gravity is wrong, maybe instead we need a ... General Theory of Relativity!"
The second alternative definition above comes from the also prestigious (but perhaps less deservedly so) legal tradition, where rules are supposed to be as precise as possible to avoid confusion. If you build a casino west of Las Vegas to catch weary drivers from Los Angeles, it very much matters whether your slot machines end up located on the Nevada or California side of the the very distinctly defined stateline. They can be either in Nevada or in California, but the rules don't allow for some vague Calivada transition zone.
The problem is that the glamour of these two traditions frequently blinds us to how messy most aspects of reality are. In truth, we really don't have all that much trouble thinking with rough accuracy about concrete things in terms of averages, probabilities, and tendencies.
What we have more difficulty with is thinking about thinking about statistical realities. The relative nature of most of reality seems to strike many Westerners rather lowbrow and unintellectual. No matter how fashionable relativism is in terms of the higher moral thought, Platonic essentialism retains its vast prestige in the intellectual world when it comes to describing humanity. My empirical relativism-squared (I'm a relativist about the importance of blood relatives in social life) simply fails utterly to register with 99% of intellectuals.
In the third tradition, one of empirical observations useful in fields such as biology, sociology, and psychology, the word "rule" means, more or less, a statistical tendency. That's the usage found in the very common phrase "as a general rule," of which Google finds more than 13 million examples. A "general rule" sounds like it ought to be super-universal omnipresent: e.g., "as a general rule, gravity is in operation throughout the galaxy," or "as a general rule, the Pythagorean theorem has been true since the Universe began."
But that's not how it's used at all. Instead, "as a general rule" means that a tendency exits although it's probably not universal. Here are some examples of how "as a general rule" is used in a probabilistic sense:
"And since, as a general rule, the effects of causes are far more accessible to our study than the causes of effects, it is natural to think that this method has a much better chance of proving successful than the former." John Stuart Mill, A System of Logic
"In a future part of this work we shall show that, as a general rule, groups of allied species gradually appear and disappear, one after the other, on the face of the earth, like the individuals of the same species: and we shall then endeavour to show the probable cause of this remarkable fact." - Charles Darwin, 1844
"As a general rule, people marry most happily with their own kind. The trouble lies in the fact that people usually marry at an age where they do not really know what their own kind is." - Robertson Davies, novelist
"As a general rule, the most successful man in life is the man who has the best information." - Benjamin Disraeli, prime minister and novelist
In the last example, Jimmy Carter's tenure as President was a notorious exception to Disraeli's rule. Carter became bogged down as a decisionmaker by the quality and quantity of information available to him once he became President. But that doesn't mean Disraeli's rule should be discarded. Indeed, it offers evidence for its broad applicability.
"It may be that the race is not always to the swift, nor the battle to the strong - but that is the way to bet." - Damon Runyon.