Somebody said -- and it is true -- that the key to long-run academic success is to create a research program. Being brilliant and insightful won't do it, unless you give your graduate students something to do.
That seems to be true of all academic disciplines, whether scientific, social scientific, or whatever.
One puzzle for those of us who tend to think that science reveals the truth about the world is why it should do so. It is just another social practice, after all. You constantly run into Popperians on the Internet, but philosophers of science will tell you the positivist project of finding some extra-scientific criterion for why science works hit a dead end.
Could it be that any research project will eventually, through some collective Bayesian process, either die out or become a science? Is that crazy?