I like to do some plain ole' dissolving of my unupdated concept of the world and asking "What did I value about X (in the unupdated version)?" and compare the result to see if those features are withstanding or not in the updated version. And oftentimes I only care about that which is left unchanged, since my starting-point is often how normality come about rather than what normality is. Come to think of it, this sounds somewhat like a re-phrasing of EY:s stand on reductionism(?).