You know what else scientists aren't? A group that can be generalized about that easily. Are you seriously willing to put forth scientists 'can tell no lies'? Do you have any evidence to support this generalization?
As an earlier poster has said, being a scientist doesn't make you any less human and subject to everything that goes along with that - emotions, irrationality, ego, and so on.
"not good liars" does not mean "can tell no lies"; in this case it means "has not had years of professional training and experience in debating/argument tactics and how to manipulate irrational audiences".
Heh. I did, actually. :) That's why I added the "irrational audiences" bit; an academic environment does train you to argue and debate effectively, but with other academically-minded, vaguely rational people. Playing politics with a broader audience requires two related skills that communicating in an academic require doesn't require as often or as strongly: communicating with people who have a much larger inferential distance from you (http://wiki.lesswrong.com/wiki/Inferential_distance), and dealing with people who don't subscribe to logic at all.
As an earlier poster has said, being a scientist doesn't make you any less human and subject to everything that goes along with that - emotions, irrationality, ego, and so on.