I like to do my principal research in bars, where people are more likely to tell the truth or, at least, lie less convincingly than they do in briefings and books.