Loading datasets...
Task: N/A
Release Date: 9/15/2026
Format: WAV, TextGrid, MD, CSV
Size: 54.60 MB
This collection contains 150 isolated lexical items recorded and segmented for phonetic analysis. Each word has its own short WAV file and corresponding Praat TextGrid with phonetic segmentation. The same questionnaire was used for the Awajún (collection 18) and Kakataibo (collection 19) materials, making the three collections a controlled comparative dataset: the same set of concepts and recording protocol applied to three languages from different language families (Aymaran, Jivaroan, and Panoan). This makes the collection a valuable resource for comparative phonetic research on Peruvian languages. The items cover basic vocabulary, including body parts, animals, plants, everyday actions, kinship terms, and spatial expressions, with labels in Spanish. Documentation was carried out by Heriberto Avelino, Roberto Zariquiey, and Jorge Iván Pérez as part of the project Phonetic Documentation of Three Peruvian Languages (PUCP–MPI-EVA).
Licensing
Licencia Chana 2.0 — Licence for Peruvian Indigenous Language Documentation Collections
https://github.com/rzariquiey/licencia-chanaRestrictions/Special Constraints
This dataset is released under the Licencia Chana 2.0, not under a Creative Commons licence. Download is direct and unrestricted, but use is subject to the following terms.
PERMITTED WITHOUT FURTHER AUTHORISATION
REQUIRES PRIOR WRITTEN AUTHORISATION FROM THE DONOR
ATTRIBUTION Cite as: Zariquiey, Roberto (donor). Aymara: Phonetic Word List. Mozilla Data Collective. RIGHT OF WITHDRAWAL Speakers retain the right to withdraw their recordings from this collection at any time. This is the principal reason a Creative Commons licence was not used: CC licences are irrevocable and cannot accommodate this commitment.
Contact for authorisations: [email protected]
Forbidden Usage
Note on machine learning: this is not a blanket prohibition on ML research. Evaluating and benchmarking existing models on this data is permitted, and so is non-production academic experimentation, provided the model is not deployed, the weights are not released, and the results cite the collection. What requires authorisation is production training and weight release. The reason is that the consent forms signed before approximately 2020 do not mention AI training and cannot reasonably be read as covering it; the communities have not yet been consulted on this point.
Contact for authorisations: [email protected]
Ethical Review
All recordings in this collection were made with the informed consent of the speakers, obtained in the field at the time of recording. Signed consent forms are held by the donor at the Pontificia Universidad Católica del Perú and are available to the MDC team on request. Speakers are identified by code (S01, S02...) rather than by name, in file names, in the annotation tiers of the ELAN files, and in all accompanying documentation. Contributors are acknowledged collectively in each collection's documentation, without linking any name to any specific recording or code. A small number of individuals appear under their own names because they asked to be credited as authors of the material; this is stated explicitly in the documentation of the collections concerned. Material judged sensitive was withheld from deposit entirely rather than published under restriction. Collections involving funerary practice, medicinal knowledge, ritual knowledge, or extended life histories of identifiable individuals are published under restricted access with individual review of each request. One point remains open and is declared here rather than glossed over: the consent forms signed before approximately 2020 do not mention the use of recordings for training artificial intelligence models. They cannot reasonably be read as covering it. The communities have not yet been consulted on this specific question. Until that consultation takes place, production model training is not authorised, and the licence reflects this.
Intended Use
Isolated lexical items with hand-checked Praat segmentation, part of a controlled three-language comparative set with Awajún and Kakataibo. Intended for contrastive phonetic research, for evaluating forced aligners and phone recognisers on three-way stop contrasts and uvular consonants, and as pronunciation reference material for Aymara teaching.
1. Donante — Roberto Zariquiey Biondi, PUCP, [email protected]
2. Lengua — Aimara · ISO 639-3 aym · familia aimara
3. Otras denominaciones — Aymara, Aymar aru
4. Descripción general
Ciento cincuenta ítems léxicos aislados, grabados y segmentados para análisis fonético. Cada palabra tiene su propio archivo WAV corto y su TextGrid de Praat con la segmentación correspondiente.
El cuestionario es el mismo que se aplicó en awajún (colección 18) y en kakataibo (colección 19), lo que convierte a las tres en un conjunto comparativo controlado: mismo inventario de conceptos, mismo protocolo de grabación, tres lenguas de familias distintas (aimara, jíbaro y pano). Es un recurso poco frecuente para estudios fonéticos contrastivos sobre lenguas peruanas.
Los ítems cubren vocabulario básico —partes del cuerpo, animales, plantas, acciones cotidianas, términos de parentesco, ubicación espacial— con etiquetas en castellano.
5. Participantes — hablantes identificados con código (S01, S02…); ver la sección
«Participantes y seudonimización» más abajo.
6. Documentación — Heriberto Avelino, Roberto Zariquiey y Jorge Iván Pérez, en el marco del proyecto Documentación fonética de tres lenguas peruanas (PUCP–MPI-EVA)
9. Citación
Zariquiey, Roberto (donante). Listas léxicas para análisis fonético: aimara. Archivo Digital de Lenguas Peruanas. Lima, PUCP.
10. Ortografía
Inventario fonológico del aimara y su correspondencia ortográfica, según el cuestionario del proyecto Documentación fonética de tres lenguas peruanas (PUCP–MPI-EVA).
Consonantes
El aimara distingue tres series en las oclusivas y africadas: simple, aspirada y glotalizada.
| Simple | Aspirada | Glotalizada | |
|---|---|---|---|
| Bilabial | /p/ ‹p› | /pʰ/ ‹ph› | /p'/ ‹p'› |
| Alveolar | /t/ ‹t› | /tʰ/ ‹th› | /t'/ ‹t'› |
| Palatal (africada) | /t͡ʃ/ ‹ch› | /t͡ʃʰ/ ‹chh› | /t͡ʃ'/ ‹ch'› |
| Velar | /k/ ‹k› | /kʰ/ ‹kh› | /k'/ ‹k'› |
| Uvular | /q/ ‹q› | /qʰ/ ‹qh› | /q'/ ‹q'› |
| Fonema | Grafema | Fonema | Grafema |
|---|---|---|---|
| /m/ | ‹m› | /s/ | ‹s› |
| /n/ | ‹n› | /χ/ | ‹x› |
| /ɲ/ | ‹ñ› | /h/ | ‹j› |
| /ɾ/ | ‹r› | /j/ | ‹y› |
| /l/ | ‹l› | /w/ | ‹w› |
| /ʎ/ | ‹ll› |
Vocales
| Frontal | Central | Posterior | |
|---|---|---|---|
| Alta | /i/ ‹i› | /u/ ‹u› | |
| Baja | /a/ ‹a› |
Sistema de tres vocales. La fricativa uvular /χ/ se escribe ‹x› y la glotal /h/ se escribe ‹j›, al contrario de lo que sugeriría la ortografía castellana.
150 ítems · 0.25 h de audio · 150 TextGrid · 0.16 GB
18_agr_listas-lexicas-foneticas — mismo cuestionario en awajún19_cbr_listas-lexicas-foneticas — mismo cuestionario en kakataiboLos hablantes aparecen con código, no con nombre, tanto en los nombres de archivo como en los tiers de los ELAN. Códigos usados en esta colección:
S01
Nivel 1 — Acceso abierto
| Licencia | Licencia Chana — ver LICENCIA_CHANA.md |
| Acceso en MDC | Descarga directa, sin solicitud previa |
| Uso académico, docente y de revitalización | Permitido con atribución |
| Uso comercial | Requiere autorización previa y por escrito |
| Entrenamiento de modelos de IA | Requiere autorización previa y por escrito (cláusula 5) |
No se usa Creative Commons en ninguna colección: CC BY permite el entrenamiento de modelos, no admite excepciones y es irrevocable, lo que anularía el derecho de retirada. Este nivel de acceso se asignó porque es elicitación léxica controlada: palabras aisladas, sin contenido personal ni narrativo.
Ver LICENCIA_CHANA.md para los términos y 00_LICENCIAS_Y_ACCESO.md para el esquema de acceso.