Skirmishing toward a general theory of evidence and proof

Abstract

Traditional probability fundamentally assumes bivalence and additivity: there is only truth and falsity, whose odds add to one. The consequence is many problems and paradoxes for factfinding, all attributable to the assumptions’ exclusive focus on random uncertainty. By contrast, multivalent belief theory abjures those two assumptions, thereby allowing consideration of epistemic uncertainty. This theory divides the factfinder's state of mind into three, not two, gradated concepts: belief, uncommitted belief representing epistemic unknowns, and disbelief. This theory utilises a more general but perfectly valid logic that accounts for all relevant kinds of uncertainty and so explains the law's wise practices. In practice, legal factfinding rightly rejects probabilism. It favours instead the multivalent theory of evidential proof based on inner convictions or beliefs. Other theories compete with multivalent belief theory in trying to overhaul probabilism to fit practice, but they prove clumsy in achieving that goal.