Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
Job manager crashed while running this job (missing heartbeats).
Error code: JobManagerCrashedError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
shift_id string | from_ipa string | to_ipa string | language string | language_family string | is_allophone bool | from_form string | to_form string | note string | certainty int64 | commonality int64 | url string |
|---|---|---|---|---|---|---|---|---|---|---|---|
a_nas_to_a | ã | a | Galician | Ibero-Romance | false | Old Galician-Portuguese lãa 'wool' (< Latin lāna) | Galician la [laː] | The nasal vowel lost its nasality and the two vowels merged into a long [aː]; Portuguese keeps the nasal in lã. | 4 | 3 | https://echodrift.lipu.dev/transform/a_nas/a |
a_nas_to_a | ã | a | Galician | Ibero-Romance | false | Old Galician-Portuguese irmãa 'sister' (< Latin germāna) | Galician irmá /iɾˈmaː/ | Standard Galician has an oral long vowel; Portuguese keeps irmã with a nasal vowel. | 4 | 3 | https://echodrift.lipu.dev/transform/a_nas/a |
a_nas_to_a | ã | a | Lithuanian | Baltic | false | Old Lithuanian žąsis with nasal ą [ãː] (< Proto-Balto-Slavic *źansís) | žąsis [ʒɑːˈsʲɪs] 'goose' | The ogonek letter ą once marked a nasal vowel; it now marks a long oral vowel. | 4 | 3 | https://echodrift.lipu.dev/transform/a_nas/a |
a_nas_to_a | ã | a | Lithuanian | Baltic | false | Old Lithuanian kąsti with nasal ą [ãː] (< Proto-Balto-Slavic *kanˀstei) | kąsti 'to bite', with oral long [aː] | The ogonek is kept in spelling as an etymological length mark. | 4 | 3 | https://echodrift.lipu.dev/transform/a_nas/a |
a_nas_to_au | ã | au | Portuguese | Ibero-Romance | false | Old Galician-Portuguese can /ˈkã/ (also cã) 'dog', from Latin canem | cão [ˈkɐ̃w̃] | [ɐ̃w̃] is a nasalized [au]-type diphthong, filed under atlas au (no nasal-diphthong cell). | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/au |
a_nas_to_au | ã | au | Portuguese | Ibero-Romance | false | Old Galician-Portuguese pan (also pã) 'bread', from Latin pānem | pão [ˈpɐ̃w̃] | Galician keeps can and pan without the diphthong. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/au |
a_nas_to_e_nas | ã | ẽ | Polish | West Slavic | false | Old Polish język 'tongue, language' with 14th-century short nasal /ã/ (< Proto-Slavic *ęzykъ) | Polish język, with ę [ɛ̃] | null | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/e_nas |
a_nas_to_e_nas | ã | ẽ | Polish | West Slavic | false | Old Polish węgiel 'coal' with 14th-century short nasal /ã/ (< Proto-Slavic *ǫgľь) | Polish węgiel, with ę | The long nasal /ãː/ went the other way, to ą [ɔ̃]. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/e_nas |
a_nas_to_o | ã | o | Old English | Anglo-Frisian | false | Anglo-Frisian *gą̄s 'goose' (< Proto-Germanic *gans) | gōs | Nasal spirant law gave long nasal *ą̄, which was raised and rounded to ō; Modern English goose. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o |
a_nas_to_o | ã | o | Old English | Anglo-Frisian | false | Anglo-Frisian *tą̄þ 'tooth' (< Proto-Germanic *tanþs) | tōþ | Modern English tooth; German Zahn keeps the nasal. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o |
a_nas_to_o | ã | o | Old English | Anglo-Frisian | false | Anglo-Frisian *brą̄htæ 'brought' (< Proto-Germanic *branht-) | brōhte | Here the nasal vowel goes back to the older Proto-Germanic loss of n before h. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o |
a_nas_to_o | ã | o | Old Frisian | Anglo-Frisian | false | Anglo-Frisian *tą̄þ 'tooth' | tōth | North Frisian tôs, toss continue the ō. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o |
a_nas_to_o | ã | o | Old Frisian | Anglo-Frisian | false | Anglo-Frisian *brą̄htæ 'brought' | brōchte | West Frisian brocht. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o |
a_nas_to_o | ã | o | Old Saxon | West Germanic | false | Pre-Old Saxon *ą̄þar 'other' (< Proto-West Germanic *anþar) | ōthar | Variant forms āthar, andar show that the rounding was not general in Old Saxon. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o |
a_nas_to_o | ã | o | Old Saxon | West Germanic | false | Pre-Old Saxon *tą̄þ 'tooth' (< Proto-West Germanic *tanþ) | tōth | Attested beside tand, with the nasal consonant restored. | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o |
a_nas_to_o_nas | ã | õ | Polish | West Slavic | false | Old Polish long nasal /ãː/ (< Proto-Slavic *ǫ in *zǫ̑bъ 'tooth') | ząb /zɔmp/ 'tooth' | In the 15th century the short nasal was fronted to [ɛ̃] (ę) and the long one was backed to [ɒ̃~ɔ̃] (ą) and lost its length. That is why ą and ę alternate within one paradigm, as in ząb 'tooth' vs. gen. sg. zęba and nom. pl. zęby. Modern ą is decomposed: [ɔm] before labial stops (hence /zɔmp/) and [ɔw̃] before fricative... | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o_nas |
a_nas_to_o_nas | ã | õ | Polish | West Slavic | false | Old Polish long nasal /ãː/ in mąż 'husband' (< Proto-Slavic *mǫžь) | mąż 'husband', with ą = nasal /ɔ̃/ | The oblique stem has the short-nasal reflex ę (gen. sg. męża, dim. mężulek). | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o_nas |
a_nas_to_o_nas | ã | õ | Old English | West Germanic | false | Proto-Germanic *þanhtaz [θɑ̃ːxtɑz] 'thought' | Old English þōht 'thought' | The long nasalized *ą̄, from *anh, was not fronted by Anglo-Frisian brightening. It was raised to ǭ and later denasalized to ō, whereas Old High German kept ā (gidāht). | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o_nas |
a_nas_to_o_nas | ã | õ | Old English | West Germanic | false | Proto-West Germanic *fą̄han 'to seize' | Old English fōn (earlier *fōhan) | null | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o_nas |
a_nas_to_o_nas | ã | õ | Old English | West Germanic | false | Proto-West Germanic *hą̄han 'to hang' | Old English hōn | null | 4 | 2 | https://echodrift.lipu.dev/transform/a_nas/o_nas |
a_to_a_nas | a | ã | Portuguese | Ibero-Romance | false | Latin lāna 'wool' | lã [ˈlɐ̃] | Intervocalic /n/ nasalized the preceding /a/ and was lost (Old Galician-Portuguese lãa > lã); the modern vowel is a raised nasal [ɐ̃]. | 5 | 4 | https://echodrift.lipu.dev/transform/a/a_nas |
a_to_a_nas | a | ã | Portuguese | Ibero-Romance | false | Latin germāna 'sister' | Old Galician-Portuguese irmãa, Portuguese irmã | Intervocalic /n/ was absorbed into the preceding /a/ as nasality and then lost. | 5 | 4 | https://echodrift.lipu.dev/transform/a/a_nas |
a_to_a_nas | a | ã | Hindi | Indo-Aryan | false | Sanskrit danta 'tooth' (Prakrit daṃta) | dā̃t /d̪ɑ̃ːt̪/ | The nasal of the cluster -nt- was absorbed into the preceding vowel, which became long and nasalized. [ɑ̃ː] = atlas ã; Hindi long ā is phonetically back [ɑː]. | 5 | 4 | https://echodrift.lipu.dev/transform/a/a_nas |
a_to_a_nas | a | ã | Hindi | Indo-Aryan | false | Sanskrit candra 'moon' (Prakrit caṃda) | चाँद cā̃d 'moon' | The nasal of the cluster -nd- was absorbed into the preceding vowel, which became long and nasalized. | 5 | 4 | https://echodrift.lipu.dev/transform/a/a_nas |
a_to_a_nas | a | ã | Bengali | Indo-Aryan | false | Sanskrit candra 'moon' | চাঁদ cãd 'moon' | The chandrabindu marks the nasal vowel left by the lost cluster nasal. | 5 | 4 | https://echodrift.lipu.dev/transform/a/a_nas |
a_to_a_nas | a | ã | Bengali | Indo-Aryan | false | Sanskrit danta 'tooth' | দাঁত dãt 'tooth' | This is the same nasal absorption as in Hindi dā̃t. | 5 | 4 | https://echodrift.lipu.dev/transform/a/a_nas |
a_to_a_nas | a | ã | Lithuanian | Baltic | false | Proto-Balto-Slavic *źansís 'goose' | Old Lithuanian žąsis with nasal ą | Tautosyllabic *an became a nasal vowel, written with the ogonek; it later denasalized to long [aː]. | 5 | 4 | https://echodrift.lipu.dev/transform/a/a_nas |
a_to_a_nas | a | ã | Lithuanian | Baltic | false | Proto-Balto-Slavic *kanˀstei 'to bite' | Old Lithuanian kąsti with nasal ą | Proto-Slavic *kǫsati shows the parallel Slavic nasal vowel. | 5 | 4 | https://echodrift.lipu.dev/transform/a/a_nas |
a_to_ai | a | ai | Icelandic | North Germanic | true | hagi 'pasture' (/a/ before ⟨gi⟩) | [ˈhaijɪ] | Before ⟨gi⟩ [jɪ] within the same morpheme, /a/ is short and becomes [ai]; East Skaftafell speakers keep [ˈhaːjɪ]. | 3 | 2 | https://echodrift.lipu.dev/transform/a/ai |
a_to_ai | a | ai | Icelandic | North Germanic | true | magi 'stomach' (/a/ before ⟨gi⟩) | [ˈmaijɪ] | null | 3 | 2 | https://echodrift.lipu.dev/transform/a/ai |
a_to_ai | a | ai | Icelandic | North Germanic | true | agi 'discipline' (/a/ before ⟨gi⟩) | [ˈaijɪ] | null | 3 | 2 | https://echodrift.lipu.dev/transform/a/ai |
a_to_ash | a | æ | Old English | Anglo-Frisian | false | Proto-West Germanic *fader 'father' | fæder /ˈfæ.der/ | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Old English | Anglo-Frisian | false | Proto-West Germanic *glas 'glass' | glæs /ɡlæs/ | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Old English | Anglo-Frisian | false | Proto-West Germanic *watar 'water' | wæter /ˈwæ.ter/ | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Old English | Anglo-Frisian | false | Proto-West Germanic *hwat 'what' | hwæt /xwæt/ | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Old English | Anglo-Frisian | false | Proto-West Germanic *fast 'firm' | fæst /fæst/ | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Old English | Anglo-Frisian | false | Proto-West Germanic *þak 'roof' | þæc /θæk/ | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Old English | Anglo-Frisian | false | Proto-West Germanic *dag 'day' | dæġ /dæj/ | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Old English | Anglo-Frisian | false | Proto-West Germanic *lātan 'to let' | lǣtan | Anglo-Frisian brightening fronted short *a to æ and long *ā to ǣ except before nasals and (in Old English) before back vowels; Old Saxon keeps a (fadar, glas, watar, lātan). | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Inari Sami | Samic | false | Proto-Samic *ājmē 'needle' | äimi [æ] | Umlaut: Proto-Samic first-syllable *ā was fronted to ä /æ/ before a second-syllable *ē, but stays á before *ā, *ō and gives a before *ë (Northern Sami áibmi keeps á). | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Inari Sami | Samic | false | Proto-Samic *sāppē 'gall' | säppi | Same umlaut before *ē. | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Inari Sami | Samic | false | Proto-Samic *čālē- 'to slit' | čäälliđ 'to write' | Same umlaut before *ē. | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_ash | a | æ | Inari Sami | Samic | false | Proto-Samic *cākē- 'to push' | cäähiđ | Same umlaut before *ē. | 5 | 3 | https://echodrift.lipu.dev/transform/a/ash |
a_to_au | a | au | Icelandic | North Germanic | false | Old Norse ár 'year' | ár [ˈauːr] | null | 5 | 1 | https://echodrift.lipu.dev/transform/a/au |
a_to_au | a | au | Icelandic | North Germanic | false | Old Norse máni 'moon' | máni [ˈmauːnɪ] | null | 5 | 1 | https://echodrift.lipu.dev/transform/a/au |
a_to_au | a | au | Icelandic | North Germanic | false | Old Norse láta 'to let' | láta [ˈlauːta] | null | 5 | 1 | https://echodrift.lipu.dev/transform/a/au |
a_to_au | a | au | Icelandic | North Germanic | false | Old Norse bátr 'boat' | bátur [ˈpauːtʏr] | null | 5 | 1 | https://echodrift.lipu.dev/transform/a/au |
a_to_au | a | au | Icelandic | North Germanic | false | Old Norse dáð 'deed' | dáð [ˈtauːð] | Old Norse long á /aː/ became the diphthong /au/ in Icelandic, still written á; the mainland languages have å /oː/ in the same words. | 5 | 1 | https://echodrift.lipu.dev/transform/a/au |
a_to_caret | a | ʌ | Godié | Eastern Kru | false | Proto-Eastern Kru *mla² 'swallow' | Godié mlʌ² [mʌňʌ] | Superscript digits are Marchese Zogbo's tone levels; the reconstruction rests on Dida, Neyo, Guibéroua and Daloa Bété forms. | 4 | 2 | https://echodrift.lipu.dev/transform/a/caret |
a_to_caret | a | ʌ | Godié | Eastern Kru | false | Proto-Eastern Kru *ka³ 'have' | Godié, Gbawale kʌ³ | Kouya ka³. | 4 | 2 | https://echodrift.lipu.dev/transform/a/caret |
a_to_caret | a | ʌ | Godié | Eastern Kru | false | Proto-Eastern Kru *ga³ 'be awake' | Godié gʌ³ | null | 4 | 2 | https://echodrift.lipu.dev/transform/a/caret |
a_to_caret | a | ʌ | Godié | Eastern Kru | false | Proto-Eastern Kru *sa 'pick up' | Godié sʌ | Marchese Zogbo finds few conditioning factors; Godié keeps peripheral /a/ alongside the new central vowel /ʌ/. | 4 | 2 | https://echodrift.lipu.dev/transform/a/caret |
a_to_e | a | e | Old French | Romance | false | Latin mare 'sea' | Old French mer | Stressed Latin a in an open syllable became e. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old French | Romance | false | Latin nāsus 'nose' | Old French nés | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old French | Romance | false | Latin prātum 'meadow' | Old French pred, pré | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Uyghur | Turkic | false | Proto-Turkic *bālïk 'fish' | بېلىق bëliq | Uyghur umlaut: first-syllable *a is raised to e before a high vowel /i/ (from *ï) in the next syllable; Turkish keeps balık. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Uyghur | Turkic | false | Proto-Turkic *yaŋï 'new' | يېڭى yëngi | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Uyghur | Turkic | false | Proto-Turkic *siarïg 'yellow' | سېرىق sëriq | cf. Old Turkic sarïɣ, Turkish sarı. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Wu Chinese (Shanghainese) | Sinitic | false | 三 MC sam | se (Wugniu) | Nasal coda lost and the vowel raised; Mandarin sān, Cantonese saam. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Wu Chinese (Shanghainese) | Sinitic | false | 反 MC pjonX | fe (Wugniu) | Mandarin fǎn. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Proto-Nakh to Plains Chechen | Nakh-Dagestani | false | *maqe 'harrow' | meqi, meqa | Palatal umlaut: a first-syllable *a fronts before a second-syllable *e or *i, which was later reduced | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Proto-Nakh to Plains Chechen | Nakh-Dagestani | false | *nace 'moth' | neci, neca | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Proto-Samic (from Proto-Uralic), unstressed syllables | Samic | false | Proto-Uralic *kala 'fish' | Proto-Samic *kuolē | Reconstructed stage. After the first-syllable metaphony, unstressed *ā became *ē unless followed by *ë in a later syllable (*kalanï > *kuolānë 'my fish'); before *j it gave *i. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Proto-Samic (from Proto-Uralic), unstressed syllables | Samic | false | Proto-Uralic *taka 'back, rear' | Proto-Samic *tuokē | Reconstructed stage. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Proto-Samic (from Proto-Uralic), unstressed syllables | Samic | false | Proto-Uralic *sarka 'branch' | Proto-Samic *suorkē > Northern Sami suorgi | Reconstructed stage; the *ē later became i in Northern Sami. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Proto-Samic (from Proto-Uralic), unstressed syllables | Samic | false | Proto-Uralic *säppä 'gall' | Proto-Samic *sāppē | Reconstructed stage; front *ä of the second syllable merged with *a before the change. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old High German | High German | false | Proto-West Germanic *sattjan 'to set' | sezzen | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old High German | High German | false | Proto-West Germanic *talljan 'to count' | zellen | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old High German | High German | false | Proto-West Germanic *hari 'army' | heri | Primary i-umlaut: short *a became e before i, ī or j in the next syllable (German setzen). | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Norse | North Germanic | false | Proto-Germanic *satjaną 'to set' | setja | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Norse | North Germanic | false | Proto-Germanic *taljaną 'to count' | telja | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Norse | North Germanic | false | Proto-Germanic *harjaz 'army' | herr | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Norse | North Germanic | false | Proto-Germanic *gastiz 'guest' | gestr | i-umlaut fronted *a to e before a following *i or *j, which was then usually lost. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Frisian | Anglo-Frisian | false | Proto-West Germanic *fast 'firm' | fest | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Frisian | Anglo-Frisian | false | Proto-West Germanic *glas 'glass' | gles | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Frisian | Anglo-Frisian | false | Proto-West Germanic *watar 'water' | weter | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Frisian | Anglo-Frisian | false | Proto-West Germanic *þak 'roof' | thek | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Frisian | Anglo-Frisian | false | Proto-West Germanic *fader 'father' | feder | null | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_e | a | e | Old Frisian | Anglo-Frisian | false | Proto-West Germanic *lātan 'to let' | lēta | Anglo-Frisian brightening went further in Frisian than in English: short *a became e (Old English æ) and long *ā became ē (Old English ǣ); compare Old English fæst, glæs, wæter, lǣtan. | 3 | 3 | https://echodrift.lipu.dev/transform/a/e |
a_to_empty | a | ∅ | Latvian | East Baltic | false | Proto-Balto-Slavic *wilkás 'wolf' (Lithuanian vilkas) | vilks | Latvian syncope of the short vowel in the final syllable | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Latvian | East Baltic | false | *śímta 'hundred' (Lithuanian šimtas) | simts | null | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Kuuk Thaayorre | Southwestern Paman | false | Proto-Paman *kuna 'excrement' | kun 'excrement' | Final vowels of the proto-forms are lost; Yir-Yoront also has kun. | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Kuuk Thaayorre | Southwestern Paman | false | Proto-Paman *pula 'they (dual)' | pul 'they (dual)' | null | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Kuuk Thaayorre | Southwestern Paman | false | Proto-Paman *paca- 'bite' | path 'bite' | null | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Trinitario (Mojeño) | Bolivia-Paraná | false | Proto-Mojeño *kaˈʧiru 'ant' (Ignaciano kaˈʧiru) | Trinitario ˈkʧiru | null | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Trinitario (Mojeño) | Bolivia-Paraná | false | Proto-Mojeño *haˈnore 'toucan' (Ignaciano haˈnare) | Trinitario ˈhnore | null | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Trinitario (Mojeño) | Bolivia-Paraná | false | Proto-Mojeño *kaˈhoʔo 'north' (Ignaciano kaˈhaʔa) | Trinitario ˈkhoʔo | The unstressed vowel in the weak position of the first iambic foot was syncopated; word-final syllables were never affected. | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Proto-Tungusic to Even | Northern Tungusic | false | *ŋāla 'hand' | Even ӈа̄л | Evenki ӈа̄лэ, Nanai ӈа̄ла keep a final vowel. | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Proto-Tungusic to Even | Northern Tungusic | false | *gara 'branch' | Even гар | Evenki гара. | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Proto-Tungusic to Even | Northern Tungusic | false | *buga 'place' | Even буг | Evenki буга. | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Proto-Tungusic to Even | Northern Tungusic | false | *darama 'waist' | Even дарам | Evenki дарама. | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_empty | a | ∅ | Proto-Tungusic to Even | Northern Tungusic | false | *ńama 'warm' | Even ням | Evenki няма. | 3 | 4 | https://echodrift.lipu.dev/transform/a/empty |
a_to_eps | a | ɛ | Attic-Ionic Greek | Hellenic | false | Proto-Greek *mā́tēr, Doric μάτηρ mátēr | Attic μήτηρ mḗtēr [mɛ̌ːtɛːr] 'mother' | Proto-Greek long ā became η /ɛː/ in Attic-Ionic. Ionic made the change everywhere; Attic kept ā after e, i and r (Attic χώρα khṓrā vs Ionic χώρη khṓrē 'place, country'). | 5 | 3 | https://echodrift.lipu.dev/transform/a/eps |
a_to_eps | a | ɛ | Attic-Ionic Greek | Hellenic | false | Proto-Greek *dā́mos (Mycenaean da-mo, Doric δᾶμος) | Attic δῆμος /dɛ̂ː.mos/ 'people, district' | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/eps |
a_to_eps | a | ɛ | Latin | Italic | false | factus 'done' (Old Latin *perfaktos) | perfectus [pɛrˈfɛk.tʊs] 'completed' | In early Latin, short a in a closed non-initial syllable was raised to e, while in open syllables vowels were reduced toward i (cecidī). | 5 | 3 | https://echodrift.lipu.dev/transform/a/eps |
a_to_eps | a | ɛ | Latin | Italic | false | factus 'done' | īnfectus [ĩːˈfɛk.tʊs] 'not done' | null | 5 | 3 | https://echodrift.lipu.dev/transform/a/eps |
a_to_eps | a | ɛ | Mbabaram | Southern Paman | false | */djana/ 'sit' | /nɛ-/ 'sit' | Second-syllable *a fronted to [ɛ] when the word began with *ɟ; the first consonant and vowel were then dropped, so /ɛ/ became phonemic. | 5 | 3 | https://echodrift.lipu.dev/transform/a/eps |
a_to_eps | a | ɛ | Mbabaram | Southern Paman | false | */djiba/ 'liver' | /bɛ/ 'liver' | Alpher lists Mbabaram be 'liver' under Proto-Paman *cipa. | 5 | 3 | https://echodrift.lipu.dev/transform/a/eps |
a_to_eps | a | ɛ | Mbabaram | Southern Paman | false | Proto-Paman *cana 'they (pl.)' | ne 'they (pl.)' | Mbabaram has no /e/, so the e in Alpher's spelling is /ɛ/, matching Dixon's /nɛ-/ 'sit'. | 5 | 3 | https://echodrift.lipu.dev/transform/a/eps |
a_to_ie | a | ie | Old French | Gallo-Romance | false | Latin cārum 'dear' | Old French chier | Stressed open-syllable a after the palatalized velar (Latin k > tʃ before a) became ie by Bartsch's law; modern French cher [ʃɛʁ] later reduced the diphthong. | 4 | 1 | https://echodrift.lipu.dev/transform/a/ie |
End of preview.
EchoDrift atlas of phonetic shifts
An open dataset of 762 phonetic shifts (sound changes and allophones) between pairs of IPA sounds, with 5,759 word examples from 930 languages in 268 language families and branches.
Browse it at https://echodrift.lipu.dev; source and corrections at https://github.com/komapc/phonomorph. This copy is updated automatically whenever the atlas changes.
What a record is
Each shift is one from→to change between two IPA sounds (for example [p] → [f]). For every language it lists word pairs whose from form is the attested ancestor (a proto-form or an older stage, never a sister language) and whose to form is the descendant. Shifts are one step: chains with an attested intermediate stage are split into separate shifts.
Files
echodrift-examples.csv— one row per word example. Columns:shift_id,from_ipa,to_ipa,language,language_family,is_allophone,from_form,to_form,note,certainty,commonality,url.echodrift-atlas.json— every shift with its full record: summary, phonetic effects, per-language examples, certainty and commonality (1–5), sources, tags and related shifts.
Scores
- certainty 5 = textbook-standard, 1 = contested.
- commonality 5 = found in many unrelated families, 1 = one language or family.
License and citation
CC BY 4.0. Please cite as: EchoDrift atlas of phonetic shifts (2026), https://echodrift.lipu.dev, doi:10.5281/zenodo.23760955.
- Downloads last month
- 67