[susan53]

As I said before I think that most cases of "exceptions" are really cases wheren the rule has been formulated wrongly in the first place - and unfortunately that's true of a lot of so-called rules in coursebooks. How many coursebooks do you know for instance that tell you some is used in affirmative sentences, and any in negatives and questions? So sentences like And of course that's not the rule at all. So examples like Would you like some wine? are presented as "exceptions". Rubbish. Explain the rule correctly and all the examples fit in. (I'm leaving out of the discussion here whether it's not sometimes useful to present simplified rules to students, even though we know they're not fully correct. But the important thing is that the teacher knows the correct rule, and that the student understands it's a simplification and that s/he'll get a fuller story at a later level).

Then there are other examples of "historical hangovers" - ie things which were perfectly regular 500-700 years ago, but have now mostly died out in modern English, leaving just one or two examples - which again are presented as exceptions. Well, in a sense they are. But not in the sense that they're completely random and inexplicable.

And of course, there's the opposite. As language evolves new rules emerge, and there's a moment in the middle where the old and the new exist side by side - the new often being seen as a "mistake" or "exception". But 100 years on, that "mistake/exception" may have become the standard rule.

Get into the area of lexis, syntactic patterns etc and there are certainly things which are not rule governed - for example a recent discussion here was looking at "why" you can say (eg) I advised him to go, but you can't say I suggested him to go. The meaning of the verbs is comparable, but they take different structures and there seems to be no rule to say why. But it's not a matter of "exceptions".

Having said all that a) I know I'm going against a lot of current thinking in ELT by arguing this (proponents of the Lexical Approach wouldn't agree with me at all), b) I'm certainly not going to claim I've looked at every single structure in the language. But I know from experience that nine times out of ten when a teacher presents an "exception", it's actually because s/he's presenting an inaccurate or incomplete rule.

What would you put forward as "genuine" exceptions in the language?

PS - las night I came across this is Terry Brookes Armageddon's Children (I don't have the book here, but the quote's more or less right) : I don't know why you're telling me all this but I'm not understanding it. Which got me thinking that understand can, of course, be used dynamically when the context is right. Another example would be : He was having a lot of problems with his maths course at the moment, but he's understanding things a lot better now. It all comes back to what I've said before : grammar is meaning. The grammar of the language gives us the choice between presenting something as permanently true (simple) or limited to a time period around a certain point. We choose the form dependent on how we want to present the information - what we want to "mean". There's no such thing as a verb which "can't" take the continuous - again, another example of an inaccurate rule.