Binary Epistemology

Binary epistemology is Critical Fallibilism’s claim that evaluating ideas is a digital problem with two outcomes, not an analog one measured by degrees. Temple distinguishes digital issues (which map to counting numbers and are often binary—true/false, refuted/non-refuted) from analog issues (which map to the real numbers and have infinitely many gradations). His central original claim is that error correction requires digital issues: you cannot definitively eliminate an error by nudging an analog “goodness” score; you can only flip a discrete judgment from fail to pass.

This makes “stronger” versus “weaker” arguments—the smuggling of analog scales into epistemology—a mistake. CF replaces it with binary evaluation: a criticism either explains why an idea fails at its purpose or it does not, so what matters is whether a decisive criticism exists, not how impressive the surrounding arguments feel. An idea you should act on is simply one with exactly one non-refuted option among contradicting alternatives.

The main opposition is probabilistic and justificationist epistemology, which assigns credences or degrees of certainty to ideas. CF answers that such numbers, outside genuinely statistical domains like dice or measurement, are made up, and that lowering confidence from 100% to 90% merely avoids confronting the actual criticism. Assigning certainties to ideas also triggers an infinite regress, since each certainty judgment is itself an idea needing its own certainty—a regress that binary judgment, being open to refutation rather than quantified, escapes. Probability belongs to physics and math, not to judging which ideas are true.