HPMOR is making me rethink human nature -- because of how people react to it. This is a story full of cunning disguises, and readers seem reluctant to see past those disguises. RL rkcerffrq chmmyrzrag ng ubj many readers took forever to decide Quirrell = Voldemort; I think I now know why.

I suggest that humans are instinctive "observation consequentialists." That is, we think someone is competent and good if the observed results of their actions are benign. We weigh what we observe much more strongly than what we merely deduce. If we personally see their actions work out well, we'll put aside a great deal of indirect evidence for their incompetence or vileness.

In HPMOR, Quirrell's directly observed actions are mostly associated with Harry getting to be more of what he thinks he wants. Even rescuing Bellatrix amounts to Harry getting to save a broken lovelorn creature in terms of what we directly observe. To believe Quirrell evil we have to bring in all kinds of expected consequences to weigh against those immediate positive observations.

Does the resistance to saying Quirrell=Voldemort maybe reflect a broader unwillingness to overlook what we directly witness in favor of abstract deduction? If it does, this implies some interesting predictions about human behavior:

• if you can be kind and moderate in your personal behavior, you can get away with incredible amounts of institutionally-mediated violence and extremism, especially to anyone who feels like they "know" you. Hypothesis: the most dangerous people are those who can give us the illusion of "knowing" them while they command an institution whose internal operations we don't see.

• More generally, an institution "wired" to do us harm can get away with it much longer than an individual doing it personally and directly. Faceless corporate evil, faceless societal evil, and faceless government evil are much more deadly than our emotional impulses realize. Hypothesis: we are biased to confuse the institutions with our attitude toward their leaders, or to refuse to act against the institutions because of the outward manners of their leaders.

• if this 'observation consequentialism' bias is heuristic, then maybe it evolved as an anti-gossip function. In that case we should expect that people are too quick to believe outrageous things about people they can't observe. Hypothesis: the further away someone is from your understanding, the less likely you are to think of them as mostly a typical human being, and the quicker you are to believe them a saint, a monster, or something similarly exciting.

• And, alas for EY, hypothesis: telling a story about cunning disguises, in which the protagonist of the story does not see through those disguises, is almost always going to lead to lots of readers also not seeing through those disguises.