Simon Winchester

So research is a terribly imperfect science, and you learn an awful lot more after you've published a book, because people keep writing to you and saying, 'Oh, gosh, I was related to such and such a character and I have a letter in my possession.'

The nature of catastrophe is, after all, reasonably unvarying in the way it ruins, destroys, wounds and devastates. But if something can be learned from the event - not least something as profound as the theory of plate tectonics - then it somehow puts the ruination into a much more positive light.