AIM Framework glossary
Reciprocity
Canonical definition
Reciprocity is the condition in which two Individuals arrive at convergent reciprocity-percepts through coupled Bayesian ground-resolution operating over a shared Hypothesis Space, where each Individual's posterior about the fairness of the exchange is updated by the counterpart's expressed posterior functioning as social-proof evidence, and where the matching of explanations stabilises both posteriors at a coupled-equilibrium value. Reciprocity is therefore a property of the convergence between two I-resolutions, not a property of the underlying value transfer, which proceeds independently of, and unmeasured by, the percept.
Explanation
When two people walk away from an interaction both believing the exchange was fair, what they share is not a measurement of what was actually transferred. It is a matched conclusion, reached by each person's brain doing the same Bayesian inference over the same set of available explanations, with each person's expression of their conclusion functioning as evidence that updates the other's. Reciprocity is the agreement between the two conclusions, not a property of the underlying transfer, which the architecture has already integrated and stripped of its source tags before either conclusion is formed.
Epistemic status
Framework-Specific Construct.