confusion in that if you treat lambda calculus as computation instead of "mathematical truths", then following the proof in the opposite direction requires conj

confusion in that if you treat lambda calculus as computation instead of "mathematical truths", then following the proof in the opposite direction requires conjuring out of thin air a complicated construction that isn't obvious from the state you'd have