Coverage Asymmetry

Coverage asymmetry is the archive's formal recognition that absence of visible corpus is not absence of events. The visible corpus is produced by several survival-and-transmission channels operating at different decay rates (oral tradition, iconographic record, archaeological record, written record), each subject to different preservation, disruption, and modern-observation filters. Where the archive's scribal pipeline privileges one channel (typically the written record, and specifically the Western-academic-ethnographic slice of the written record), regions and traditions that preserved their corpus in other channels will appear under-represented even when their actual event-density is high or higher.

The doctrine matters because without it the archive inherits the biases of its sources. The Global Territorial Grid, the A-class mating-event census, the eleven-civilizations enumeration, and the HLSF signature programme are all vulnerable to observer-side filtering that presents itself as primary-evidence sparseness. The doctrine's operational function is to make the filtering visible in the record so that future readers can see what the scribe could not see rather than inheriting the corrected state as if it were the only possible state.

Claims

c0001 — Survival channels have different decay rates and different disruption histories

Four primary survival channels: (a) oral tradition — high fidelity within a continuous speech-community, catastrophic loss at community-language disruption, cycle-amnesia decay across ~200-year medium transitions in the absence of active transmission; (b) iconographic — durable across medium transitions (architectural decoration, regalia, ritual objects, rock art) but subject to iconoclasm and to category-reclassification by later observers; (c) archaeological — durable across millennia but narratively silent, requiring interpretive overlay; (d) written — durable across the life of the script and the institutional support for recopying, subject to editorial overwrite (storm-god doctrine) and to translation-loss at script transitions. No channel is dominant in all contexts; regions and traditions preserved their corpora in whichever channel their institutional substrate made available.

c0002 — The written-record privilege is a training-data artifact, not a reality claim

The archive's scribal pipeline surfaces candidates by querying training-data-accessible sources, which are dominated by texts — academic, popular, and encyclopedic — and within the text domain by English-language Western scholarship. This pipeline systematically down-weights oral traditions (African, Oceanian, Arctic, pre-contact American), iconographic traditions not written up in Western-academic style (Central Asian, Himalayan, sub-Saharan), and archaeological traditions where dragon-class interpretation is contested (Bronze Age European, pre-Columbian American). The down-weighting is a feature of the training substrate, not a feature of the actual event density.

c0003 — Inverse morphological-category slippage refuses dragon-class status to non-Western serpent entities

Cycle-amnesia's morphological-category slippage (doctrine-cycle-amnesia c0008) operates within a tradition when a dragon category slips to a familiar-analogue category across medium transitions (Ungnyeo → bear, Quinotaur → bull). The inverse operates at the observer level: Western scholarship has systematically classified non-Western serpent / water-spirit / rainbow-entity figures as deity, ancestor, spirit, totem, culture hero, or king, while reserving the dragon category for Eurasian, East Asian, and selected Mesoamerican figures. The inverse slippage is a category-gatekeeping operation that makes non-Western dragon corpora invisible to comparative analysis. The Luba Nkongolo case (site-luba-nkongolo-rainbow-serpent c0002) is a clean worked example.

c0004 — Jurisdictional and scriptural survival shape which corpora the archive can see

Different polities maintained different custodial institutions, different writing systems at different points in their histories, different destruction events affecting their archives, and different archive-survival rates into the modern record. These are environmental variables shaping what the archive can see, not evidence of differential event density. A region with early writing and continuous state-archive institutions (Egypt, Mesopotamia, China) produces a dense visible corpus; a region with late writing or exclusively oral tradition (sub-Saharan Africa, pre-Islamic Central Asia outside the Silk Road scribal zones, pre-Columbian Americas outside Mesoamerica) produces a sparse visible corpus even at equivalent event density. When the user uses phrases like "different jurisdictional realities" or "different advances in civilization as it were," they are naming this variable directly.

c0005 — The coverage-bias rule applies to gender, class, and minority traditions as well as to region

Coverage asymmetry is not only regional. Within a region, the visible corpus is filtered by what the state-level, priestly-level, elite-literate-level institutions chose to record. Women's ritual traditions, commoner folk practices, minority religious traditions, and the traditions of subjugated populations within a polity are systematically under-represented in the written record even when the written record is otherwise dense. The archive's atomization programme applies the coverage-asymmetry filter at this finer grain as well — a densely written-attested region is still vulnerable to intra-regional filtering.

c0006 — Coverage-asymmetry mitigation is iterative, explicit, and on-the-record

The archive's coverage-asymmetry mitigation protocol: (a) every global sweep includes an explicit coverage-gap section naming what is under-sampled and why; (b) the atomization-index records per-entry status so future reviewers see the correction state rather than the corrected state; (c) gap-priority ordering places under-sampled regions and traditions first in the scribal queue rather than last; (d) promoted cells and new civilization entries are labelled as coverage-bias-corrective in the entry itself; (e) the inverse-morphological-category-slippage filter is applied explicitly before classifying any serpent / water-spirit / rainbow / cave / water-body entity as not a dragon. Mitigation is iterative and never complete; coverage that honestly labels its own gaps is the archive's operational goal, not total coverage.