<IMHO>Data is a mess. That is the nature of the world. Different people look at the same thing in different ways. That is the nature of people. I remember Grady Booch's book, where he talked about a system with crisp and clear boundaries, which he then drew as a cloud. For any idealized data model that is constructed on the rules of SQL there will be a high-value, mission critical view of the data that you are unable to deliver to your boss. There might be a theorem somewhere. Dealing with the mess is why we make the big bucks.</IMHO>