Retrofit manual Part VI: Mastery

[dcellison]

Summary

This completes the SEM-09D contextual retrofit of all twenty files in manual/part6_mastery/. Eighteen lessons are revised against current canon and the completed lexicon; the two source-material chapters that already matched the present language are reviewed and preserved.

The pass also advances the active teaching-corpus ledger and project handoff material to manual/part7_reference/, and refreshes the phonetic-neighbour baseline for changed corpus-attestation counts.

Social context

• Teach kona as direct address without promising care or ceremony, and use allative kau for physical motion towards a person.
• Keep pi as an explicit courtesy marker while separating it from consent, authority, and safe refusal.
• Present sa, ni, and le as relationship claims without pretending that rank-neutral grammar removes unequal social power.
• Rebuild the register lesson around recoverable subjects, referents, source marking, address, and pace without treating any of them as moral guarantees.
• Restore direct kia as the ordinary greeting, distinguish polite questions from directives, and remove the false use of ho as an uncertainty softener.

Transmutation

• Bring the method under the completed lexicon: direct words, registered compounds, transparent composition, and coinage are compared on their merits.
• Make the Five Pillars guides to editorial judgement rather than effects supposedly guaranteed by grammar.
• Correct the anger study so it names its retained accusation, changed aim, and added request.
• Keep all news claims reportative, use kau for evacuation to safety, preserve the unnamed repairer with the passive, and give exact source notation its own record.
• Add a close Basho rendering with base tuleno (frog), tupi (jump), and shonuwa (sound), while retaining the creature-arrival-voice poem as an explicitly different transmutation.
• Teach registered phelu pula (holding-wish) in the jealousy exercise and test whether its possession relation fits the particular source.

Living practice and source material

• Replace clock quotas with natural points in the day and use direct norawhi for dusk rather than the established expression for west.
• Remove invented practitioner and community evidence, privacy-by-obscurity claims, and claims that grammatical fluency supplies mindfulness.
• Keep Buddhist mindfulness and Right Speech explicit in the meditation chapter, with the current texts/metta_sutta.md location and no prescribed duration.
• Label optional Commons kowela and give base sila lona as the alternative for a community meeting.
• Keep exact technical notation in source records while using base themio for reflection on a measurement.

Lexical result

• No root is added and no vocabulary decision is opened.
• Part VI needs one optional module word, Commons kowela, which is labelled and paired with a base alternative.
• Direct tuleno, tupi, shonuwa, norawhi, rinu, whemura, and registered phelu pula replace stale gaps or workarounds.
• The phonetic-neighbour audit still contains 250 pairs. Only corpus-attestation counts change.

Humanizer and voice audit

The separate Humanizer pass found contract-style boundary prose, bold inline labels used as pseudo-headings, and the stamped phrase "a close translation would" across three lessons. The final pass states social boundaries once in positive terms, converts the outline labels into real subsections, and varies the repeated translation language. Mechanical scans found no new dash punctuation, curly quotes, banned inflation vocabulary, -bearing constructions, copula-avoidance phrases, or inline-header lists.

Checks

• python3 scripts/validate_examples.py (1,275 entries; 0 errors; 0 warnings)
• python3 scripts/test_vocabulary_schema.py (22 tests)
• python3 scripts/content_vocabulary_decisions.py --check (63 batches; 235 decisions)
• python3 scripts/test_content_vocabulary_decisions.py (6 tests)
• python3 scripts/test_name_forms.py (20 tests)
• python3 scripts/audit_phonetic_neighbors.py --output /tmp/phonetic_neighbors_baseline_final.txt plus clean diff against the tracked baseline
• python3 scripts/generate_reference.py plus clean diff for generated Part VII lexicon and compound references
• python3 scripts/build_site.py (145 manual sections and all other site collections built)
• git diff --cached --check