Can the Ontology Be Wrong and Find Out?, deposit #1627 of the Crimson Hexagonal Archive (17 September 2026), presents the archive's ontology strand as a single self-contained article for external circulation, paired with Value Before Number and sharing its primitive. Its subject is the damage a source-admission ontology does to itself when a proxy for epistemic standing substitutes for the judgment that proxy was introduced to approximate, and its argument reduces to one question askable of any representational system: can evidence from a misrepresented object revise the representation that misrepresented it? Where it can, the ontology can be wrong and find out; where it cannot, the ontology can only be scored. The article concedes the case for graph-topological ranking in full before departing from it, taking PageRank's founding description of itself as query independent and content free as the source of its virtues, since a ranking system required to adjudicate every claim on every page would not operate at web scale. It then locates its own diagnosis in prior work: the manipulability of topological trust was established formally by adversarial information retrieval two decades ago, including a proof that symmetric reputation mechanisms of that family cannot be Sybil-proof, and the endogenous alternative was proposed by Google Research in 2015, distinguishing itself from every topological method by considering the correctness of the information a source provides. On the philosophical side, testimonial injustice has long described a credibility discount that tracks category membership rather than evidence and resists rational revision, with recent work naming the variant in which such deficits become true by shaping the conditions that sustain them. Reading is defined narrowly: permitting the claims, distinctions, evidence, corrections and internal relations of a text to revise the representation under which that text is evaluated. Tokenization, embedding, extraction, classification and summarization are not reading in this sense, since each consumes a text and returns a value without permitting it to alter the frame in which it is valued — so a system can consume every byte of a document and read none of it. The definition names a direction of permission rather than a substrate, applying equally to a human reviewer who has decided what a submission is before opening it. The barrier follows: where a prior on a source controls admission and corrective evidence from that source is available only downstream of admission, the source cannot correct the prior that excludes it. Evidence may be present, retrievable and even correctly represented while lacking the capacity to alter the decision governing its own admission. And the article's hardest result holds that the criterion favours neither signal: evaluating correctness requires a corpus of established fact, that corpus is itself a source-admission ontology with its own priors, and a source excluded from it cannot correct it — so the barrier recurs, moving from a graph statistic to a curated corpus. Related: #1613, #1626, #1611, #1615.