I am going to tell you three very different stories from my career, with my own role and shortcomings in these conflicts unconcealed, and four times during the presentation, I am going to seek your views. What to do when you see fraud, assessing if it is intentional or accidental, and trying to decide whether to contact the author, the journal, or the author's institution, are all tough decisions. The personal risks and costs of trying to expose or correct fraud are extremely high, particularly if you are a coauthor, young scientist, or a student. I will argue that, despite what the University or AGU might say, fraud is slippery and gradational, not black and white. What’s essential is to be alert in our own work to the temptation to make the ambiguous unequivocal, and to remember that our job as scientists is to build and destroy our theses before the eyes of our readers.