> there are obviously no testable hypotheses about subjective descriptions

Clearly false if you take just 30 seconds to scroll through this thread. Embarrassing.