semantics - Trouble with second conditionals (with “could”)

I would do B if you could do A.

This is a statement which has been bothering me for quite a while. I come across such statements often and, to me, they make no sense. Could is the subjunctive of can — it is often used to express possibility (or anything else along those lines). Sometimes, could is also used as the "past form of can". Now, the second conditional is used to express unreal(istic) situations in the present, we construct it using the simple past form of the verb at hand and instigate a would in the "response sentence".

Now, does this could then indeed relate to the capability and not to possibility of someone doing A? And, furthermore, isn't the condition fulfilled only by the fact there exists this capability (or possibility)? That's the part that makes no sense for me.

You don't actually have to "do A", you just must have the possibility/capability to "do A", for the if clause to return true and make the speaker "do B".

I would do B if you did A.

does the proper job of the original idea that the speaker intended.

And since the second conditional indeed is used for unreal situations in the present, why do we so often use it to express conditions to other people?

Wouldn't it be more prudent to actually state a requirement in the present, which deals with "real situations":

I will do B if you do A.

What's the difference?

Answer

Very often would and could are used not to express possibility/capability but to express willingness to do a thing. It is a polite way of asking.

If you could get me a beer, I'd be much obliged.

Obviously, anyone who can walk to the fridge could get you a beer, but you are asking whether the person is inclined to do you that small favor.

So the statement

I would do B if you could do A.

more likely means "I'd be willing to do B if you'd be willing to do A." And it's nothing to be worried about.

Addendum (Responding to the OP's comment question)

Specifically, you can't apply mathematical strictness to human language. If you look at my profile, you'll see that one of the quotes I cite is Frank McWhorter's keen observation that "no language makes perfect sense."

Much of what happens in language, especially speech, is rife with overtones and undercurrents and connotations; very seldom is it the strict chop-logic imparting of fact. You know this is true, for without that overloading of meaning how could something like irony exist? Example:

You say a person can't use two affirmative words to express a negative? Yeah, right.

"Yeah" and "right" both represent affirmations. And taken together they can also be an affirmation. But said skeptically, they can mean the person speaking the words doesn't believe whatever he's being asked to believe. It can be a very emphatic way to say "That's not true at all."

Math is clean, pristine, beautiful; human language is messy, murky, and also beautiful. You just have to understand that they're beautiful in different ways.

Yeah, right. (And I mean that in the positive sense this time.)