Canon, revision and retcon: how the Elder Scrolls record accumulated across decades of releases, and the discipline for handling material that later works contradict.
LESSON 01
A Layered Rather Than Flat Corpus
Establish the shape of the record.
Accumulation over decades
The material constituting this setting was produced across more than three decades, by different teams, in different formats, with later works freely revising earlier ones. It is not a single document written at one moment but a stratified deposit. Treating it as a flat corpus in which every statement carries equal weight guarantees contradictions that are artefacts of the method rather than of the setting.
Conflicts within the record have different origins. Some are in-world partisan disagreement between fictional authors. Some are revisions in which a later work supersedes an earlier one. Some are ordinary inconsistency. Some are the in-world Dragon Break claim. Diagnosing which is present determines the remedy, and misdiagnosis is the commonest failure in advanced Elder Scrolls scholarship.
Where a later work contradicts an earlier one, the Academy records the earlier as superseded and retains it. Retention matters because the earlier material remains informative about how the setting developed, and because deletion would make the history of the record itself unrecoverable. A superseded claim is evidence about a different question, not worthless evidence.
Provenance determines weight. Direct depiction in a released work carries more weight than in-world assertion; later material generally supersedes earlier where they conflict directly; and material from a work of narrower scope does not automatically override broader treatment. Every claim in this programme therefore carries its provenance, which is why citation is required rather than decorative.
A revision is a decision, and decisions are evidence. When a later work changes how the setting describes a region, a people or an event, that change indicates what the setting's makers found unsatisfactory or wished to develop differently. Studying the pattern of revisions is therefore a legitimate scholarly activity distinct from studying the setting's internal history.
The two must not be mixed. An in-world explanation for a discrepancy and an account of why the discrepancy exists in the record are answers to different questions. The jungle descriptions of Cyrodiil illustrate the danger: adopting an in-world transformation account as fact would conflate a documentary problem with a historical claim, which is why the Academy records both and asserts neither.
For any conflicting pair: identify the medium and date of each claim; determine whether the conflict is partisan, revisionary, erroneous or an asserted break; apply supersession where a later work directly overrides an earlier; record partisan disagreement as contested; and state the diagnosis explicitly. A reader should always be able to see which kind of conflict was found and what was done about it.
Resolving a conflict without disclosing it produces a smooth account that misrepresents the state of the evidence, and it removes the reader's ability to check the judgement. Silent harmonization is therefore treated in this programme as a more serious fault than reaching a debatable conclusion openly, because the latter can be corrected and the former conceals that correction is needed.