There's a lively discussion going on at Usable Security about the most recent post, which draws connections between Stanley Milgram's experiment and the problems of usable security.
