I always assumed that it was more a case of behavior selection:
Yes, people did drink water frequently, but those societies and groups that regularly consumed fermented drinks like wine or beer, had on average a lower chance of consuming polluted water.
As a result behavior that favored drinking slightly alcoholic drinks became more widespread by selection.
This doesn’t require that people actually knew their water sources were polluted.
I’ve become a lot more circumspect about pushing through Chesterton fences as I’ve gained more experience in life.
I used to believe that we could determine a basis for most decisions from first principles. However, that requires a level of complete a priori knowledge that’s simply unattainable except for extremely niche situations.
Yes, people did drink water frequently, but those societies and groups that regularly consumed fermented drinks like wine or beer, had on average a lower chance of consuming polluted water.
As a result behavior that favored drinking slightly alcoholic drinks became more widespread by selection.
This doesn’t require that people actually knew their water sources were polluted.