Cross-dict entry-coverage divergence

Which headwords does each dictionary treat very differently from its peers? This page ranks lemmas by the spread in how many main entries each of the 7 core dictionaries devotes to the same headword — a structural proxy for editorial disagreement. High divergence = worth inspecting before citing across dictionaries.

Deep comparison uses validated feature adapters only. Broad coverage/overlap covers eligible local Sanskrit/BHS headwords; missing deep markup is not counted as zero evidence.

Trust Block

• Evidence: data/lexico/sense_divergence.json, generated from data/lexico/microstructure_profile.csv by python scripts/lexico/build_divergence_map.py. Core dicts: MW, AP, PWG, PWK, WIL, VCP, SKD.
• Metric: entry-count divergence = max(entries in any dict) − min(entries in any dict) for a given headword. Structural proxy only — true semantic sense divergence requires per-lemma sense-unit computation (deferred; see docs/RESEARCH_LAYER_ROADMAP.md).
• Limitations: only covers lemmas present in ≥2 of the 7 core dicts. Divergence = 0 does not mean the dicts agree semantically; it means they both give the headword exactly 1 entry. Headword splitting (homonym indices) varies by dict and inflates apparent divergence.
• Validation: python scripts/lexico/build_divergence_map.py regenerates; checked by npm run build.
• Owner repo: csl-atlas.
• Next use: inspect highlighted rows, then open exact dictionary source records before citing the pattern.

Generated by python scripts/lexico/build_divergence_map.py from data/lexico/microstructure_profile.csv. CC-BY-SA-4.0.