The Claim-Function: Correction and the Structure of Defeat

Journals and other record-keeping institutions sometimes revise a published claim by issuing a public correction notice, and sometimes by replacing the text without marking the earlier claim as erroneous. The corrected content may be the same in both cases. This paper argues that the difference is not a matter of content, truth-tracking, or sincerity. It is a difference in normative structure. In practices that distinguish correction from replacement, a correction is a public, attributable, fault-marked supersession of a prior claim that withdraws entitlement to rely on that claim and licenses future challenges to reliance on it or on claims whose entitlement depended on it. I call the resulting burden a corrective liability. The liability is enforceable by participants in the practice, temporally extended, and locally scoped to a dependency-neighborhood. Replacement, by contrast, is supersession without this fault-marked challenge license. The paper also defends a locus condition: corrective liability can operate only where there remains some attributable locus of answerability, either an ongoing speaker or an archived corpus. This condition is not career survival, institutional resilience, or social permission to apologize. It is the minimal structural requirement that there be something to which the liability can attach. The account explains why transparent versioning, silent updates, errata, and retractions differ not merely in editorial presentation but in what they authorize others to do with the prior claim.