Response to the Discussion of Belief Functions1

The present discussion is primarily about how to interpret belief functions, and the discussants have a wide variety of views on this topic. I am somewhat disappointed at this, both because I would like to see work on belief functions move on from interpretation to application, and also, of course, because I have been explaining my own views on the interpretation of belief functions for many years, and I had hoped to be more persuasive. Yet the diversity of views is understandable, for belief functions use probability, and there is no consensus on how to interpret probability.