This is a little bit out there from a rationalist point of view, but I think there are domains in which not all of the requisite concepts for the scientific met

This is a little bit out there from a rationalist point of view, but I think there are domains in which not all of the requisite concepts for the scientific method actually exist. An example that comes to mind is the nature of conscious experience - having spoken to a lot of meditators, for instance, I'm convinced that there really are commonalities to it that you can experience yourself by sitting down and paying attention. But I don't think there's anything to really falsify here, because there's nothing to measure! You can't tell someone "here's this totally objective way you can pick apart your conscious experience into a bunch of statistics, and when you do change X I predict you'll see the statistics vary like Y". You can kinda just point their brain in a direction and hope that they experience it for themselves. Another example, I think, is domains that are under optimisation pressure or control of some kind (like an ecosystem, or much more simply a thermometer). There was a post on HN a few days ago about causality, and how the standard statistical methodologies for determining causality kinda break down when you start measuring systems like this. Correlation is not causation, but in these systems causation no longer necessitates correlation either! Perhaps a different kind of epistemology would be useful here too. Anyway, not sure this is making sense. Would be curious to hear your thoughts.