Repository navigation
Conversation
Thirteen dictionary examples disagreed with their own Phi or with canon word order. Where the Phi was sound, the English now follows it. Where the Phi broke a ruling or used the wrong word, the Phi moved. Two module guides quoted the old sentences and change with them. The manual and the Short Road said thirty-five particles where the inventory holds thirty-eight. Three manual sections sent readers to chapter 23 for the journal practice, which lives in chapter 22. The quick reference, the preposition reference, and the opening of chapter 19 left whau out of the clause relators. Exact counts of legal word forms give way to round figures in the book, the protocol, and the word-shape rationale.
A vowel pair is two different vowels. The phonology reference said so and the lexicon kept to it, but neither the word check nor the name check enforced the rule, so kaali passed as a productive name. Both checks now reject a vowel that follows itself. Canon records the ruling, with decision D120 in the development log. The phonology reference, the protocol checklists, the manual's chapter on hiatus, and the glossary state the rule.
…rd_possibility_frame
`po ma` denies possibility, so Phi had no settled way to say that a negative is possible. The maintainer chose a frame over reordering: the negated clause stands inside `tha ... tho`, and `po` stands outside it on `nai`, as in `tha shia so ma shua tho po nai`. The construction already parses, so no grammar or parser rule changes. Regression tests pin it and keep `ma po` rejected. Canon, the grammar references, the manual, the Three Slots pamphlet, and the `po`, `ma`, and `peshiro` entries now teach it, and D121 records the decision.
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
po madenies possibility ("cannot"), and negation is the last Slot 1 rank, so Phi had no settled way to say that a negative is possible. The maintainer ruled the gap real on 2026-11-19 and chose a frame over reordering. The negated clause stands insidetha ... tho, andpostands outside it onnai:No grammar changes. The parser already accepts the sentence, because
naiasserts that a condition holds and a declarative frame supplies one.ma postays unlicensed. English "may not" as refused permission is stillpo ma. D121 records the decision.Stacked on #718 and #719
This branch contains both open branches. It restores a dictionary example that #718 corrected, and its log row follows D120 from #719. Until those two merge, the diff here also shows their changes. Once they are in, only the files below remain.
Changes by file
canon.mdma po, and the note that no particle, rank, or parser rule changes.project/development_log.mddocuments/grammar/particle_reference.mddocuments/grammar/complementizer_reference.mdnai, with the possibility frame as its example.documents/reference/particle_finder.mdmanual/part4_grammar/ch15_voice_possibility/05_negation_ma.mdmanual/part4_grammar/ch15_voice_possibility/03_possibility_po.mdpo mato the negation section.manual/part4_grammar/ch09_particle_system/04_slot_1_verb_phrase.mdmanual/part7_reference/quick_reference_grammar.mdmanual/part7_reference/modules/medical_and_bodily_care.mdpeshiroexample block restored to "may not be contagious" with the frame.pamphlets/three_slots/05_slot_1_evidence_modality.mdpamphlets/three_slots/08_the_interaction_tables.mdpamphlets/three_slots/10_common_errors.mdpo maas "may not"; the former Error 10 becomes Error 11.pamphlets/three_slots/12_appendix.mdvocabulary/function/particle/pot.jsonvocabulary/function/particle/neg.jsonpoentry for "may not".vocabulary/content/infection.jsonscripts/test_sentence_validator.pytest_possibility_frame: six accepted sentences and twoma porejections (PHS100).site/manual_editorial.json,site/pamphlet_editorial.jsondocuments/validation/phonetic_neighbors_baseline.txtJudgment calls to check
peshiroexample returns to its original meaning. Correct dictionary examples and stale counts #718 had reduced it to "is not contagious" becausepo macould not say "may not". With the frame settled, the entry and the medical guide say "may not be contagious" again.ka maentry as Error 10, and the Slot 2 entry moves to Error 11 so the page keeps its closing line. No other page cites these two errors by number.po maandna mabefore the reader has mettha ... tho, so the frame has no place there yet. A later chapter could carry it.nai. With an established topic, the same words could be heard as "[the topic] may be that ...", the shape of the Manifesto'sha moenu remole tha ... tho nai. None of the examples here sets such a topic.Voice audit
The Humanizer pass and the batch scan changed two things after the first write. The manual's possibility page and the pamphlet's error entry both ended a sentence on "
po manever ..."; the pamphlet sentence was rebuilt around a learner thinking in English. Chapter 9 repeated chapter 15's sentence about the frame and the particle outside it almost word for word; it now says only that the claim takes two clauses, one for each particle. The additions contain no em dashes outside the manual's protected example form.Validation
All run as separate commands, each exiting 0:
validate_examples.py(0 errors, 0 warnings),validate_sentences.py(0 errors),test_sentence_validator.py,test_vocabulary_schema.py,content_vocabulary_decisions.py --check,test_content_vocabulary_decisions.py,test_name_forms.py,test_translation_layers.py,translation_process_status.py --check,test_translation_process_status.py,source_reconstruction.py,test_source_reconstruction.py.generate_reference.py,build_site.py,vocabulary_prose_coverage.py, and the neighbor baseline were regenerated; only the baseline changed.