I consider Alan Jacobs' point about the manner in which survey responses are scored as 'correct' or 'incorrect' to reveal two facts. The first, which Jacobs notes, is that the usual belief that Americans are poorly informed on basic facts about political structures, etc, is wrong: people know the important things tolerably well. What they don't know is the precise terminology for these things (a point that applies to the people scoring the surveys just as much as the people answering them). The second point is then that mastery of particular knowledge that is difficult, but rather the ability to take a lot of information and discern its structure. In grad school, I had a professor argue very convincingly that this was the real advantage possessed by academics who teach in the humanities and the social sciences. If the comparative advantage which provides a reason to hire and retain you is that you know what it says in Book V of the Republic, you can be replaced by Wikipedia. Mastery of the web of allusions, references, and analogies in the politics of that day and the present day, the ability to distinguish the important and central from the accidental, the ability to see implications of word choice and positions: that's the hard thing to learn. And so in one sense you teach not so much the information, but a range of things that can be done with it.