Let's talk about question polarity for a moment. Finite questions are questions that can be answered with yes or no. Other questions are non-finite and can take a range of answers. A question always has a corresponding normal statement: You have a pencil ⇔ Do you have a pencil?
You can ask a question in the positive, Do you have any pencils?, and you can ask a question in the negative, Don't you have an appointment this afternoon? Note that the statement corresponding to a negative question is itself negative, and further note that the negative statement has a corresponding positive statement.
Answering a positive finite question is easy; you answer yes if you agree with the positive statement and no if you do not. Everyone in every language does it the same way.
In English, answering a negative finite question has the interpretation that yes agrees with the corresponding positive statement and no disagrees. That is, answering yes to Don't you have any pencils? means I have some pencils. In other words, the polarity of your answer does not depend on the polarity of your question.
You can obviously use the opposite convention, where your answer to a negative finite question confirms or denies the underlying negative statement; that is, answering yes to Don't you have any pencils? means I do not have any pencils. It turns out that which convention you use is nearly a free choice for a language, which means that how you answer a negative finite question is one of the things you just have to learn when you learn a new language. (It is often quite difficult to learn to use the convention opposite to your native language, and getting it wrong marks you as a non-native speaker.)
Because the interpretation of the answer to a negative finite question depends on both speakers agreeing on convention, it is rarely the case that such a question is answered with just yes or no. In English, we'll usually add on a simple tag statement whose polarity confirms the polarity of our answer. E.g. Don't you have any pencils? is answered with Yes, I do or No, I don't. In that way, the choice of convention is reinforced.
English regularly phrases finite questions in a slightly different way, by making a statement and then adding a tag question at the end: You don't have any pencils, do you? or You have an appointment this afternoon, don't you? Note that the tag question inevitably has opposite polarity to the statement: It is inviting the speaker to agree or disagree with the statement.
The existence of such tag questions raises a chicken-and-egg problem with respect to English's choice of convention for answering negative finite questions. Opposite polarity tag questions mean that both polarity values of the statement are present in the question; only an answer that ignores the polarity of the question can be interpreted. But it could also easily be the case that the once English decided to use the positive polarity convention, opposite polarity tag questions because possible. I feel comfortable in assuming that linguists have figured out which is which, but I don't know the answer (yet).
So this one small corner of the world of questions all by itself presents some complicated issues for analysis. I hope this suggests to you how complicated, and interesting, a full grammatical analysis of a language can get.
Posted by Greg at March 8, 2007 9:36 AM
I'm amazed that you didn't mention the French "si", which is the answer to a negatively polarized question:
Ne m'aimez-vous plus? (Don't you love me anymore?)
Si! Je t'aime! (Yes! I love you!)
I don't know anything about language, and generally I stay out of Elmo's discussion thereof, but this reminded me of a joke.
A professor is teaching a class, and he begins with a discussion of how different cultures do or do not use double negatives. "Every culture has single positives, and single negatives. Yes and no, and they pretty much always mean the same thing. But some cultures use a double negative to imply a positive, while others use it to reinforce the negative. To my knowledge, however, there are no cultures that use a double positive to imply a negative." From the back of the room a student is then heard to sarcastically mutter, "Yeah, right."
Please, no comments on the correctness of the above joke.
Kevin:
Because I forgot about it. I didn't cite any examples of other languages' negative question conventions because I didn't want to do any research 8)
Folks, you're missing the big picture here:
GIANT FREAKING SQUIDS! And they'll be coming for us all soon...