I like this bit:

"Even more invidious than simple error was the way that evidence was, at times, misused: cited out of context in a way that distorted the reading; used to confirm pre-existing biases; or treated with increasing certainty without additional corroboration. At a recent history festival, a scientist asked me if historians triangulate: if we look specifically for data or evidence that might undermine our hypotheses, as well as confirm them; I replied that I sincerely hoped so."

Susan