[UncleMeat]

> First, it needs to be shown that the changes proposed will not be worse than the original condition.

I'm not sure this is always true. Or at least, it is important to specify what confidence you demand here. If you demand an infinite regress of dissertations before taking any action, you can perform a denial of service attack on all progress. Almost any policy change can have chaotic and unpredictable effects and demanding flawless prediction of the outcome is demanding the impossible.

And further, "some bad outcome will come from this change" should not be sufficient to stop progress. You cannot make perfect be the enemy of good, and net progress can be achieved even if there are some things that regress with a new policy.

At almost every step of major social progress, agitators were told that they were unruly and moving too fast. In hindsight, can we really say that women's liberation or civil rights or gay rights movements actually moved too quickly? We can even have a historical example of aggression being more effective than respectability politics. Women's rights movements largely failed in the late 19th century and it wasn't until the "disrespectful" suffragette movement ("suffragette" was a derogatory term to contrast with the more respectful activists of the prior generation) that real progress was made.