A subjective probabilistic attitude that an agent assigns to a proposition representing their degree of belief; credence is numerically represented on a [0,1] scale and functions as a decision-guiding probability distinct from long-run frequency or objective chance.

Principle

Credences are rationally constrained by the probability axioms (coherence) and by rules for updating (e.g., conditionalization) when new evidence is acquired; failure to maintain coherence leads to Dutch‑book vulnerability in decision contexts.

Demonstration

Illustrative scenario → An agent assigns credence 0.7 to ‘it will rain tomorrow’ based on current evidence (Situation). Presented with a reliable new forecast, the agent updates that credence to 0.9 (Recognition → Action). The updated credence then guides choices such as carrying an umbrella (Consequence).

Misapplication

Confusing credence with objective frequency or treating it as an empirical property rather than an agent-relative degree of belief is a category error; likewise interpreting incoherent numeric assignments as merely imprecise beliefs misses normative constraints on rational credence.

Consequence

Credences determine rational decisions under uncertainty, influence betting behaviour and expected‑utility calculations, and serve as inputs for belief‑revision procedures; their calibration affects the quality of decisions across repeated situations.

Reversal

In some contexts (e.g., intersubjective settings, legal standards, or objective propensity interpretations) agents’ credences may be constrained or pooled by institutional rules, calibration requirements, or objective evidence, reducing pure subjectivity.

Boundary

Clearly within: an individual's probability assignment about an upcoming event. Boundary case: reported scientific ‘probability’ derived from both objective data and prior modelling choices (mix of credence and frequency). Clearly outside: empirical frequency measured over many trials.

Semantic Tension

Credence ↔ Objective Chance — practical decision guidance based on subjective degrees of belief can conflict with interpretations that treat probabilities as physical propensities or long‑run frequencies.

Synthesis

Credence is best understood as an agent‑relative, normative device for guiding choice under uncertainty: it must respect coherence and updating rules but remains distinct from empirical frequencies and theoretical chances.