I hypothesise that there's a meme going around that says: "It's bad to be wrong. Only say things that are right." I think this is bad.
A friend was mentioning a recent result which is a retraction of a slightly less recent result suggesting that neutrinos could travel faster than light. Apparently there was some call for the head of this experiment to resign, and he did.
What folly, I say. Not of this guy, but of his peers, and us, to shun him this way.
If we restrict ourselves to only publishing and talking about things that are 100% certain to be correct, first, we'll publish and talk about nothing. It kills me that the end of that article quotes this guy saying "Now we are 100% sure that the speed of light is the speed of neutrinos." Really? After all that, you're now 100% sure of something? I feel fine being pretty damned sure of something, but never 100% certain.
Second, if we restrict ourselves to only publishing and talking about things that we're so sure about that we're willing to bet our careers on them, then we'll publish and talk about practically nothing. And the things we do talk about will take us inch by inch, slower and slower, toward some local maximum that we're pretty close to already.
I feel like the path toward breakthroughs in science involves wrongness. Wrongness, conversation, refinement, more wrongness, and eventually astounding rightness of a superior kind to our current rightness -- and even that is worthy of being questioned, because it too is probably wrong.