simardcasanova’s avatarsimardcasanova’s Twitter Archive—№ 19,400

          1. There’s nothing to disagree here: the *fact* is a lot of pills in medicine are already causally evaluated without knowledge of their chemical and biological effects (=mechanisms) That being said, medicine can do that thanks to double blinded trials @ethanbdm/1214941654437826561
        1. …in reply to @simardcasanova
          W/o double blinded trials, I am not sure it would be that easy What I’m saying is: - in theory, you don’t necessarily need to know the mechanisms to establish causality - but you probably need the right methodology to do that with a high enough confidence level
      1. …in reply to @simardcasanova
        Without something as methodologically clean as double blinded trials, I guess it’s not impossible but just harder, and maybe the increased difficulty is high enough to switch to a different regime where you need some knowledge of the mechanisms to direct your investigation
    1. …in reply to @simardcasanova
      Symmetrically though, knowing the mechanisms can sometimes allow to rule out a causality, or at least assess this causality is very unlikely For instance with homeopathy, that is based on outdated principles we know are not true (like the memory of water)
  1. …in reply to @simardcasanova
    But even in this case, you still need to do some empirical confirmation because maybe homeopathy is working but for completely different reasons (spoiler: homeopathy still doesn’t work)
    1. …in reply to @simardcasanova
      That was my 5 cents on this fascinating discussion There’s also a lot to discuss on what a "mechanism" really is, but I’m hungry and I’m going to eat breakfast