Scientific progress results from conflicts about the existence of new phenomena. Two theorems are formulated: Whatever is predicted by a sufficiently famous theorist will be discovered independently of whether it exists or not. Thus the criterion for existence of a phenomenon is not its discovery but rather its reproducibility! Whatever is "discovered" by a sufficiently famous experimentalist will be explained independently of its existence. Thus the criterion for understanding a phenomenon is not an explanation but rather the unanimous acceptance of one single explanation. It is argued that the optimal intention of scientists is to be critical but not rigid and at the same time open but not credulous. Since the intentions "open" and "critical" are in opposition, the historical path of science often oscillates between the two extremes "credulous" and "rigid". This is demonstrated by historical examples and applied to the special case of water.