From 422dbe7e54d060d9be3effdecde26e72a3d2b670 Mon Sep 17 00:00:00 2001 From: Weksa Date: Fri, 6 Nov 2026 04:32:37 +0000 Subject: [PATCH] Zyphos briefs: revise against the Sa'ueia sound section Oluneihe withdrawn (forbidden h; Olune is Deru's root). Ilawerum is Deru's ila+werum, not a candidate. Ellumaiwa fails the no-doubled-letter rule. Dozkai/Druzkai is two edits. Drop the no-phonology caveat and the spelling state section (respell merged in Weksa #4, Zyphos #1). Co-Authored-By: Claude Sonnet 5.5 --- .../eusocial-interbeing/zyphos-name-briefs.md | 109 ++++++++---------- 1 file changed, 50 insertions(+), 59 deletions(-) diff --git a/settings/eusocial-interbeing/zyphos-name-briefs.md b/settings/eusocial-interbeing/zyphos-name-briefs.md index c81e6ca..aae4c2d 100644 --- a/settings/eusocial-interbeing/zyphos-name-briefs.md +++ b/settings/eusocial-interbeing/zyphos-name-briefs.md @@ -22,13 +22,12 @@ Both brief sets wait on Meta's answers to `vq-first-language` and - **Airawa** has a provisional list (R-A below). It is a tendency list, not a phonology: no vowel inventory, no photactic limits, no stress rule beyond "heavy on socially important forms". -- **Sa'ueia has no phonology section at all.** The only rules are negative and - comparative: softer than Airawa, "flow longer and wetter", route names read - smooth, and "avoid making native Airawa names read like smooth Sa'ueia route - names unless the form is explicitly a loan, exonym, or cross-continental - register". Briefs below that cite R-S1..R-S3 are citing that one paragraph. - Accepting or rejecting a Sa'ueia name on sound alone is therefore weak - evidence until someone writes a Sa'ueia sound section. +- **Sa'ueia has a working sound section** + ([Sa'ueia Sound](./language-design.md#saueia-sound-proposal-for-deru-to-cut), + enforced by `sa-ueia/check.py` against `sa-ueia/table.yaml`). It is a + proposal Deru approved in outline; only the species name is canon. Rule + labels R-S4..R-S8 below cite it. The consonant inventory, syllable shapes and + stress are still open, so sound arguments beyond the letter rules stay weak. - The **imperial Airawa register** has no sound rules either. It is only described as "an artificial quarter-continent clausiform exception imposed through biological-memetic control", not "named like a normal state with wet @@ -47,6 +46,11 @@ Rule labels used below (quoted or paraphrased from `language-design.md`): | R-S1 | Sa'ueia is softer than Airawa. | | R-S2 | Sa'ueia forms flow longer and wetter than Airawa. | | R-S3 | Sa'ueia route names read smooth. | +| R-S4 | Sa'ueia letters: vowels `a e i o u`; forbidden `h q y` (`k` where a `q` might have been). | +| R-S5 | Sa'ueia: short words, soft and open; at most 10 letters joined (warning, Weksa placeholder). | +| R-S6 | Sa'ueia: suffixing morphology; identical letters merge at a join (`weru`+`um` -> `werum`). | +| R-S7 | Sa'ueia: no doubled letter inside a morpheme (Weksa placeholder); a double arises only at a join, and the merge removes it. | +| R-S8 | Sa'ueia roots from Deru: `ila` walk, `weru` know, `olune` breeding ground (unanalyzed), `-um` verb suffix. Composed: `werum`, `ilawerum` (walk of knowledge). | | R-N | Naming rule: meaning decomposition first; define the speaker's scale, name type (endonym, exonym, ritual avoidance, technical, imperial administrative, route-name, child-name, trade gloss), what the speaker's ontology makes obligatory, and whether the name is stable, seasonal, situational, inherited or evasive. | Existing scratch evidence to keep in view: `Druzkai` (Face for a non-imperial @@ -104,8 +108,8 @@ namer: minus if it was only a sound resemblance. - `Dozkai`: closed first syllable, open second ending in a diphthong `ai`. Fits R-A1/R-A2 but the open `-ai` tail is the softer shape R-A4 contrasts - against. Also one phoneme away from the scratch name `Druzkai`; two names - with that near neighbourhood will blur unless one is explicitly derived from + against. Also two edits from the scratch name `Druzkai` (insert `r`, `o` to + `u`); two names with that near neighbourhood will blur unless one is explicitly derived from the other. - `Tegzhir`: `t-g-zh` cluster, closed final syllable, clean fit. The only candidate with a front vowel in the stressed syllable, which may or may not @@ -253,7 +257,8 @@ Common context: the Sa'ueia civilization is mobile, family-unit based sacred communal breeding grounds and reciprocal ecological partners. A name earns its place by revealing the ecological deal, reproductive pressure, memory technology or route consequence (`language-design.md`, design target). -The sound rules available are only R-S1..R-S3 and R-A5; see the caveat above. +The sound rules available are R-S1..R-S8 and R-A5; the inventory, syllable +shapes and stress are still open. ### S1. Sa'ueia self-name (candidates: `Ellumaiwa`, `Sauwele`) — DRAFT @@ -279,11 +284,14 @@ target: vault as the species/continent ethnonym; the candidates may be (a) family-unit *kind* names, (b) nicknames for the people derived from it, or (c) a competing ethnonym. This needs `vq-name-scope` before judging. -- **Sound rules that apply:** R-S1, R-S2; R-A5 as a negative check. -- **Per-candidate read (judgment):** - - `Ellumaiwa`: long, open, liquid-heavy (`l`, `m`, `w`); matches R-S1/R-S2 - strongly. Four syllables. - - `Sauwele`: short, open vowels and `w`, `l`; matches R-S1/R-S2, but it +- **Sound rules that apply:** R-S1, R-S2, R-S4, R-S5, R-S7; R-A5 as a negative check. +- **Per-candidate read (judgment, then the letter rules):** + - `Ellumaiwa`: long, open, liquid-heavy. Fails R-S7: `ll` is a doubled + letter inside a morpheme, with no join to merge it, unless the form + is parsed as `el`+`lumaiwa` (then the merge rule would give `elumaiwa`). + Also 9 letters, near the R-S5 bound. Needs a respelling or a parse, not a + patch. + - `Sauwele`: short, open vowels and `w`, `l`; passes R-S4/R-S5/R-S7, but it begins `Sa-u-` exactly like `Sa'ueia`. Reads as a respelling or derivation of the existing name rather than an independent self-name. - **Reject if:** the owner has not decided the target scale; `Sauwele` is @@ -293,7 +301,7 @@ target: "nomad" is a camera-angle word); it cannot take a family-specific modifier when a particular unit needs naming. -### S2. Breeding ground (`Oluneihe`) — DRAFT +### S2. Breeding ground (`Olune`, Deru's root; `Oluneihe` withdrawn) — DRAFT ```yaml target: @@ -317,16 +325,21 @@ target: these visibly different). If the name is a *kind* name, it has to be generic enough to name every ground; if it names one ground, it needs a proper-name layer. -- **Sound rules that apply:** R-S1, R-S2, R-S3. -- **Read:** `Oluneihe` is long, liquid, open, ends on the `-ihe` vowel - sequence. Strong fit to R-S1/R-S2. The `-eihe` tail does not resemble `-ia` - of `Sa'ueia`; no collision. -- **Reject if:** it decomposes to "home", "village", "city" or "capital" (the - breeding ground failing as a capital is a stated system failure); it equals - or contains the route word (S3) or the family self-name (S1), which would - merge distinct targets; it is used for a camp. +- **Sound rules that apply:** R-S1, R-S2, R-S3, R-S4, R-S8. +- **Read:** the target already has a name from Deru: `olune`, breeding ground + (R-S8), unanalyzed. The earlier candidate `Oluneihe` is withdrawn: it + contains `h`, which R-S4 forbids, and its stem `Olune` is Deru's own root, so + the candidate only added a suffix to a word that already answers the brief. + What remains open for this brief is whether `olune` is the kind-word, a + single ground's name, or needs a proper-name layer; and its morpheme + boundaries, which Deru has not given. +- **Reject if:** any proposed form contains `h`, `q` or `y`; it decomposes to + "home", "village", "city" or "capital" (the breeding ground failing as a + capital is a stated system failure); it equals or contains the route word + (S3) or the family self-name (S1), which would merge distinct targets; it is + used for a camp; it replaces `olune` without Deru's say. -### S3. Route as obligation (`Ilawerum`) — DRAFT +### S3. Route as obligation (no candidate) — DRAFT ```yaml target: @@ -348,15 +361,16 @@ target: distinguishes *the way* from *the debt*. If both senses are one word, the language is making a commitment about their identity that the vault does not make. -- **Sound rules that apply:** R-S2, R-S3 (route names read smooth). Also R-A5: +- **Sound rules that apply:** R-S2, R-S3 (route names read smooth), R-S4..R-S8. Also R-A5: Sa'ueia route names are what Airawa native names must not accidentally resemble. -- **Read:** `Ilawerum` is smooth through most of the word but ends in `-rum`, - a closed syllable with a nasal coda, which is the one place it trends toward - the clipped shapes the Airawa rule assigns to the other language. Not a - failure by R-S3 (no stated rule against closed syllables in Sa'ueia), but - worth a note: the project should state whether Sa'ueia permits final nasals. -- **Reject if:** it is a plain toponym (no debt sense); it is the same word as +- **Read:** `Ilawerum` is not a candidate. It is Deru's own composition, + `ila` (walk) + `werum` (knowledge, verb form `weru`+`um`), glossed "walk of + knowledge" (R-S8). It names a different thing, and the checker reproduces it + as a fixed composition. Any route-as-obligation word should be built from + Deru's roots or new roots she accepts, and must not be confused with, or + collide with, `ilawerum`. +- **Reject if:** it is `ilawerum` or only a respelling of it; it is a plain toponym (no debt sense); it is the same word as the breeding ground (S2) or shares its stem without a morphological reason; the owner decides route and obligation are two words and this is only one. @@ -368,32 +382,9 @@ target: - **First language:** Airawa vs imperial Airawa vs Sa'ueia vs a narrow naming-brief interlingua (`vq-first-language`; `state/map.yaml`). The Airawa briefs are better supported by existing sound evidence; the Sa'ueia briefs - have no phonology to defend them. -- **A Sa'ueia sound section** in `language-design.md` is the missing piece for - S1-S3 to be judged on more than the one paragraph of contrast. + have only the working sound section, which leaves the inventory open. +- **Sa'ueia open items** (inventory, syllable shapes, stress, apostrophe role, + `weru` versus `wer-`) are Deru's; S1-S3 are judged on the letter rules only + until she cuts them. - **Imperial register sound rules** for A4. - **`kh` in `Kharzdun`**: add to the Airawa inventory or respell. - -## Spelling state: Sa'auei'a vs Sa'ueia (as of 2026-11-03) - -The rename is incomplete in both repos. Observed, not fixed here: - -- **Zyphos vault** (`GameCult/Zyphos`, head `7661e1f`): the vault's own git log - has the rename ("Renamed mother trees and Sa'ueia", `3afb94f`, 2026-07-01). - Prose still mixes both spellings: roughly 45 `Sa'ueia` occurrences against 41 - `Sa'auei'a`, across 27 files that still contain the old form (including - `AGENTS.md`, `notes/*`, and mid-sentence mixtures such as `World/index.md`, - which writes `[[Sa'ueia Continent]]` next to "Sa'auei'a nomadic - civilization"). One filename keeps the old form: - `Technologies/Sa'auei'a Mobile Technologies.md` (and its index entry). - Five filenames already use `Sa'ueia`. -- **Weksa** (this repo): `settings/eusocial-interbeing/language-design.md` (6 - occurrences) and `README.md` (2) and `state/map.yaml` (2) still use - `Sa'auei'a`; `.voidbot/state/weksa-birth-seed-operations.json` quotes it in - summaries. This brief set uses `Sa'ueia` (the spelling the follow-up uses) - and leaves `language-design.md` alone, so the rule table above quotes rules - that the source file spells `Sa'auei'a`. -- The two spellings differ by more than an apostrophe: the old form has an - internal glottal (`Sa'auei'a`), the new one does not (`Sa'ueia`). Any - phonological argument that depended on the glottal (the old name's - syllable-break) has no support in the new spelling.