On-Faith Observation

The epistemic status of an Observation whose claim rests on an adopted source the author accepts provisionally without independent verification. An On-Faith Observation declares has_epistemic_status::[[On-Faith Observation]] and carries a Grounds section that names the source adopted, the reason for accepting it provisionally, and — per the form's requirement — what it would take to promote the claim to Empirical or Retrospective status.

"On faith" is the honest label, not a pejorative. Many load-bearing claims in any graph come from sources the author has not personally verified: a researcher's published finding, a colleague's reported experience, a standard that names a behavior as true. The graph does not refuse to carry such claims; it asks that they be labeled for what they are, so that a later reader knows which claims rest on adopted authority and which rest on directly accessible evidence.

The promotion-path requirement is distinctive to this status. An On-Faith Observation MUST name what would turn it into Empirical or Retrospective grounding — typically "a measurement of X would confirm this," or "inspecting the audit log would verify this event happened." The promotion path is the author's accountability: future work CAN upgrade the grounding, and the form names the next step.

The four epistemic statuses — [[Empirical Observation]], [[Retrospective Observation]], On-Faith, [[Contested Observation]] — come from Classify Observations by Epistemic Status With Matched Grounds. On-Faith is distinguished from Contested by what the author knows about the claim's status: On-Faith means the author has not personally verified a source they accept; Contested means the author knows the claim is disputed and the dispute itself matters.