The take-home message from any perusal of the more accessible papers in robust statistics literature is that methods and data issues are inter-related ??? a point that you struggle against in constantly harping about whether something is a point about data or methods.