Separation of Causal Effectiveness From Prior Truth: Why an Account That Changes What Follows Need Not Have Been True

This conceptual paper introduces Separation of Causal Effectiveness From Prior Truth, the structural constraint that an account’s effects, taken alone, neither establish nor refute its accuracy at the time it described. Responses can create or remove the described condition later without altering whether it held at the earlier reference time. The paper assesses accuracy and causal influence as distinct questions that may draw on overlapping evidence. Later outcomes can inform judgments about earlier conditions when that connection is justified after accounting for the account’s effects. Forecasts concern a later reference time and can accurately predict outcomes they help produce. The discussion relates the constraint to Merton’s self-fulfilling prophecy and examines the Treasury Inspector General’s review of IndyMac’s 2008 failure. The review treated the deposit run as contributing to the timing of closure while identifying unsafe operations as the underlying cause. Evidence predating the public letter supported a separate assessment of prior distress. Part of the Constitution, History, and Standing series within On Human Modeling, the paper offers questions about reference times, induced changes, and the evidence supporting claims of accuracy. Its framework can inform human–AI research on whether outcomes influenced by a system’s outputs support claims about the conditions those outputs originally described.