| // Native canonical-IPA data manifest for Basque (eu) — euskara, a LANGUAGE ISOLATE (no living relatives), |
| // ~750k speakers, Latin script. The orthography is highly phonemic, so the engine (basque.ts) is a |
| // left-to-right greedy scan over a digraph + letter table with two context rules (the ⟨r⟩ tap/trill split |
| // and the sibilant places). |
| // ⚠⚠ THE HALLMARK — the THREE-WAY SIBILANT / affricate system, a laminal vs apical vs postalveolar contrast: |
| // fricatives ⟨z⟩→[s̻] (laminal) ⟨s⟩→[s̺] (apical) ⟨x⟩→[ʃ] (postalveolar) |
| // affricates ⟨tz⟩→[t͡s̻] ⟨ts⟩→[t͡s̺] ⟨tx⟩→[t͡ʃ] |
| // (zu 'you'→[s̻u] vs su 'fire'→[s̺u]; atzo 'yesterday'→[at͡s̻o] vs hots 'sound'→[ot͡s̺]). The laminal/apical |
| // distinction is neutralised in the western dialects and the wikipron referee records it INCONSISTENTLY |
| // (both s̺/s̻ for ⟨z⟩), and the eval backbone strips the apical/laminal diacritics anyway — but we EMIT |
| // the standard three-way. |
| // ⚠ ⟨r⟩ is a TAP [ɾ] between vowels, a TRILL [r] word-finally / before a consonant / doubled ⟨rr⟩ |
| // (hartu→[artu], udare→[udaɾe], agirre→[aɡire]) — code, in basque.ts. ⚠ ⟨j⟩→[x] (the southern/Gipuzkoan |
| // standard; [j]~[ɟ]~[d͡ʒ] dialectally). ⟨h⟩→[h] (northern/careful standard; the southern majority drops |
| // it — the referee lists both, and the h-form scores higher). Palatal digraphs ⟨tt⟩→[c], ⟨dd⟩→[ɟ], |
| // ⟨ll⟩→[ʎ], ⟨ñ⟩→[ɲ]. ⟨g⟩→[ɡ] always (no soft g). |
| // Validated vs wikipron eus_latn broad (20114) + narrow (20079). |
| { |
| "language": "eu", |
| "name": "Basque (euskara)", |
| "script": ["Latin"], |
|
|
| // Digraphs, longest-first (ordered pairs). The three affricates carry the sibilant place; ⟨tt dd ll rr⟩ |
| // are the palatal/geminate pairs. |
| "digraphs": [ |
| ["tx", "t͡ʃ"], ["tz", "t͡s̻"], ["ts", "t͡s̺"], ["dd", "ɟ"], ["tt", "c"], ["ll", "ʎ"], ["rr", "r"], |
| ], |
|
|
| // The five VOWEL LETTERS — Basque has no accented or extra vowels. This is the environment for the one |
| // context rule in basque.ts: a single ⟨r⟩ is the TAP [ɾ] only BETWEEN two of these, and the trill [r] |
| // word-initially, word-finally or next to a consonant. |
| "vowelLetters": ["a", "e", "i", "o", "u"], |
|
|
| // Single letters. ⟨z⟩→[s̻] laminal, ⟨s⟩→[s̺] apical, ⟨x⟩→[ʃ]; ⟨j⟩→[x]; ⟨g⟩→[ɡ] always. Loan letters |
| // ⟨c q v w y ç⟩ map to their nearest Basque value. ⟨h⟩ and ⟨r⟩ are handled in the scan (basque.ts). |
| "letters": { |
| "a": "a", "e": "e", "i": "i", "o": "o", "u": "u", |
| "b": "b", "d": "d", "f": "f", "g": "ɡ", "j": "x", "k": "k", "l": "l", "m": "m", "n": "n", |
| "ñ": "ɲ", "p": "p", "s": "s̺", "t": "t", "x": "ʃ", "z": "s̻", |
| "c": "k", "q": "k", "v": "b", "w": "w", "y": "j", "ç": "s̻", |
| }, |
|
|
| // The Basque VIGESIMAL (base-20) number words. 0-19 are listed; the tens are built on scores of 20 — |
| // 20 hogei, 40 berrogei (2×20), 60 hirurogei (3×20), 80 laurogei (4×20) — with the connective ⟨-ta⟩ |
| // suffixed for a remainder (hogeita hamar = 20+10 = 30). Hundreds prefix the score system (ehun, |
| // berrehun…) and take the free connective ⟨eta⟩; likewise ⟨mila⟩ (thousand), ⟨milioi⟩ (million). ⟨eta⟩ |
| // is placed once, before the final sub-100 group (the common Euskaltzaindia convention) — composition |
| // in basque.ts. Not referee-validated (the wikipron dump has no composed numbers) — the component |
| // words ARE (bat→bat, hiru→hiɾu, hogei→hoɡei̯…). |
| "numbers": { |
| "ones": [ |
| "zero", "bat", "bi", "hiru", "lau", "bost", "sei", "zazpi", "zortzi", "bederatzi", |
| "hamar", "hamaika", "hamabi", "hamahiru", "hamalau", "hamabost", "hamasei", "hamazazpi", |
| "hemezortzi", "hemeretzi", |
| ], |
| "scores": ["", "hogei", "berrogei", "hirurogei", "laurogei"], // multiples of 20 |
| "hundreds": ["", "ehun", "berrehun", "hirurehun", "laurehun", "bostehun", "seiehun", "zazpiehun", "zortziehun", "bederatziehun"], |
| "thousand": "mila", |
| "million": "milioi", |
| "and": "eta", |
| }, |
| } |
|
|
