Rudyard Kipling once said “What do they know of England, who only England know?”, and what he meant was, not only do you not know much of other countries if you only know England, you don't know so much of England either without knowing other countries with which to make comparisons.
Lovers sometimes say that of beloveds too - they love them not just by knowing the beloved and all the qualities she has, but by knowing how the qualities and faults of others give further exhibition to what the beloved has to offer.
I feel this is also very true in debates too; it is important to understand the position of your opponent in order to understand your own position properly too. In the days of debating on forums a few years ago, if I could sense an opponent hadn't got a cognitive purchase on his (or her) argument I would try to persuade him to write a post as though he was arguing passionately and intelligently for the other side.
From what I recall, no one ever took me up on my advice, but I think they missed out. Because I think in terms of probability, two things hold most of the time. If someone can accurately and comprehensively explain a position with intelligent reasoning but continue to think that position is wrong, there is quite a high probability that it is wrong. And if someone can accurately and comprehensively explain a position with intelligent reasoning and agree with it, there is quite a high probability that it is right.
Now for that good idea I was talking about: I just discovered today that economist Bryan Caplan has combined notions similar to what I just said above with the Turing test, which for those that don't know, is a test whereby a machine is required to convince a neutral judge that it could pass as being indistinguishable from a human. He calls his version of this The Ideological Turing Test.
This test tries to determine whether someone with a particular view or belief adequately understands the arguments of his or her intellectual opponents. The test is that the individual is challenged to write an essay posing as his opposite number, and if neutral judges cannot tell the difference between the partisan's essay and the answers of the opposite number, the candidate is judged to correctly understand the opposing side.
Here's what I think would be an interesting social experiment; put five randomly chosen socialists/atheists/young earth creationists and one libertarian/Christian/evolutionist in a forum and let other socialists/atheists/young earth creationists ask them questions for an hour. At the end of the hour, the socialist/atheist/young earth creationist questioners have to vote on which one they think is a libertarian/Christian/evolutionist.
Then put five randomly chosen libertarians/Christians/evolutionists and one socialist/atheist/young earth creationist in a forum and let other libertarians/Christians/evolutionists ask them questions for an hour. At the end of the hour, the libertarian/Christian/evolutionist questioners have to vote on who they think is a socialist/atheist/young earth creationist.
After repeated experiments, that could give a good indication of who understands the psychology of the other group best. The same could be tried for any polarised group you like. It won't surprise you to know that in the above scenarios I think the libertarians, Christians and evolutionists would do far better than their opponents, as well as being able to write far more comprehensive essays on their opponents' positions than would be the case the other way round.