Non-Justificationist Learning

Non-justificationist learning is the Critical Rationalist position that we never positively support, prove, or raise the standing of our ideas; we only advance by spotting and discarding errors. It is the constructive flip side of rejecting justificationism. We brainstorm unjustified guesses and improve them through negative criticism, keeping ideas that survive — the conjecture-and-refutation process Popper identified as literally evolutionary.

The motivating point is an asymmetry: one counter-example can refute a universal claim, but no quantity of confirming instances can prove it, or even hint that it is probably true. A thousand black ravens give no support to “all ravens are black”; one orange raven kills it. What people mistake for positive support is merely failure to contradict — and infinitely many rival ideas are equally compatible with any evidence. Demanding justification also triggers a regress of authorities, so CR drops justification entirely and relies on error correction instead.

CF (Critical Fallibilism) accepts this fully and tightens it. Where CR still permits ranking surviving ideas by degree via critical preferences, CF rejects degree arguments too, judging each idea-purpose pair as refuted or non-refuted in a binary way. CF adds that many seeming pros are convertible into cons of rivals, which explains why positive arguments feel like they work. It pairs with learner-driven learning: knowledge is built by the active guesser, not poured in. The opposed view is justificationism’s accumulation of “goodness” toward higher confidence.