Demarcation Problem

The demarcation problem is the question of what distinguishes science from non-science — or more precisely, what criteria distinguish scientific claims, which deserve epistemic deference, from non-scientific claims, which do not. It is one of the foundational problems of philosophy of science, and it has no agreed solution.

Karl Popper proposed the most influential answer: falsifiability. A claim is scientific if it could, in principle, be proven wrong by observation. This demarcates science from metaphysics without requiring that science be certain — a scientific theory can be bold and corroborated but remains permanently provisional. Popper's criterion correctly excludes astrology and psychoanalysis (which interpret all evidence as confirmation) while including physics. But it also excludes historical geology and evolutionary biology, which make unfalsifiable claims about singular past events; and it permits pseudosciences that can in principle be falsified but systematically avoid being so.

Kuhn's account of scientific revolutions complicated the demarcation problem by suggesting that the boundaries of science are not determined by a fixed criterion but by the practices of scientific communities — which themselves change. If what counts as science is partly determined by paradigm, then the demarcation line moves with the paradigm. This view is descriptively accurate but normatively useless: it tells us what is currently classified as science, not what should be.

The pragmatist position: the demarcation problem may be a bad question. Rather than a sharp boundary, what we have is a spectrum of epistemic reliability — a gradient from highly constrained, falsifiable, rigorously reviewed claims at one end to unconstrained, untestable, unfalsified speculation at the other. The relevant question is not "is this science?" but "what is the epistemic warrant for this claim, and how does it compare to alternatives?"