In pragmatics, which is the studies of the meaning of language that falls outside of semantic theory (word and sentence definition, word order etc), there is a system of maxims defined by American philosopher Paul Grice, which govern the rules by which we can understand each other. These are under the umbrella of the cooperative principle for the reason that people generally want to understand what another is saying, so naturally we adhere to some basic principles that allow us to do so.
1. Maxim of quantity - give the correct amount of information; not too little, not too much.
i.e. What are you reading?
no - A book. (although many people do this for comedic effect, or to be difficult)
no - A book, A Tale of Two Cities, which I checked out from the library on August 21st, 2008 at 12:30 pm, from the literature section of the Pilsen library in Chicago... etc (this is of course if they do not ask for more information)
yes - A Tale of Two Cities, by Charles Dickens.
2. Maxim of quality - give correct information; don't blatantly lie.
i.e. What is in that bowl on the table? (Spaghetti)
no - Corn (of course, this is also subject to jest/sarcasm)
yes - Spaghetti.
3. Maxim of relevance - giving a reply which fits or fulfills the question.
i.e. Where did you get that sweater?
no - There's a new issue of Superman coming out tomorrow.
yes - From the clearance rack at the Gap.
4. Maxim of manner - speak with clarity and order, which may otherwise confuse the listener.
i.e. no - I fell asleep and we talked about language after we ate dinner.
yes - After we ate dinner, we talked about language and then I fell asleep.
So these rules from first glance are extremely simple, common sense that we all possess and use without thought. However, the interesting part is breaking these rules.
We are so set in these principles and so trusting that people follow them, that when the maxims are broken we as language users assume information from or importance in these words.
The book that I'm reading gives an apt example:
"For example, if someone said: 'What's for supper?' and the reply was the superficially irrelevant one: 'Billy fell downstairs', the hearer is likely to assume that the information about Billy was somehow important, and will fill in the gaps with assumptions such as 'Since Billy was supposed to cook the supper, and he's fallen downstairs, I assume that there isn't any supper ready'."
So this is fascinating to me, for a number of reasons. Firstly, the explanation of how we can understand what we do from what is said, when it is not completely explicit (conversational spoken language is inherently inexplicit) is quite amazing. Secondly, I'd like to further explore the other intentional forms of rules breaking, such as sarcasm and humor, in which people break language rules to reach some end.
Poetry and metaphor even have their ways of confusing the systems of linguistics, especially in concern to the 'impossible sentence' in semantic theory.
Wow, who ever thought I was this much of a nerd.