Common Voice Scripted Speech 25.0 - Aragonese

License:

CC0-1.0

Steward:

Common Voice

Task: ASR

Release Date: 3/22/2026

Format: MP3

Size: 358.09 MB

Description

A collection of read speech recordings in Aragonese (Aragonés).

Specifics

Licensing

Creative Commons Zero v1.0 Universal (CC0-1.0)

https://spdx.org/licenses/CC0-1.0.html

Considerations

Restrictions/Special Constraints

None provided.

Forbidden Usage

It is forbidden to attempt to determine the identity of speakers in the Common Voice datasets. It is forbidden to re-host or re-share this dataset.

Processes

Intended Use

This dataset is intended to be used for training and evaluating automatic speech recognition (ASR) models. It may also be used for applications relating to computer-aided language learning (CALL) and language or heritage revitalisation.

Metadata

Aragonés — Aragonese (`an`)

This datasheet is for cv-corpus-25.0-2026-03-09 of the Mozilla Common Voice Scripted Speech dataset for Aragonese [Aragonés - an]. The dataset contains 14040 clips representing 17.83 hours of recorded speech (16.86 hours validated) from 53 speakers, recorded from a text corpus of 10,811 sentences.

Language

Aragonese is a Romance language spoken in several dialects by about 8,000—12,000 people in the northern areas of Aragon, Spain, primarily in the counties of Chacetania, Alto Galligo, Sobrarbe and Ribagorza.

Accents

Code	Accent	Clips	Speakers
-		7,456 (53.1%)	21 (39.6%)

Demographic information

The dataset includes the following self-declared age and gender distributions. A coverage summary is shown below each table.

Gender

Self-declared gender information. The table shows clip and speaker counts with percentages. Speakers who did not declare a gender are listed as Unspecified. A dash (-) indicates zero.

Code	Gender	Clips	Speakers
male_masculine	Male, masculine	2,040 (14.5%)	1 (1.9%)
female_feminine	Female, feminine	2,036 (14.5%)	12 (22.6%)
transgender	Transgender	-	-
non-binary	Non-binary	-	-
do_not_wish_to_say	Prefer not to say	5 (0.0%)	1 (1.9%)
-	Unspecified	9,959 (70.9%)	41 (77.4%)

Gender declared: 4,081 of 14,040 clips (29.1%), 12 of 53 speakers (22.6%)

Age

Self-declared age information. The table shows clip and speaker counts with percentages. Speakers who did not declare an age are listed as Unspecified. A dash (-) indicates zero.

Code	Age	Clips	Speakers
teens	Teens	5 (0.0%)	1 (1.9%)
twenties	Twenties	293 (2.1%)	4 (7.5%)
thirties	Thirties	2,854 (20.3%)	7 (13.2%)
fourties	Fourties	4,446 (31.7%)	6 (11.3%)
fifties	Fifties	1,580 (11.3%)	8 (15.1%)
sixties	Sixties	1,721 (12.3%)	2 (3.8%)
seventies	Seventies	539 (3.8%)	2 (3.8%)
eighties	Eighties	-	-
nineties	Nineties	-	-
-	Unspecified	2,602 (18.5%)	31 (58.5%)

Age declared: 11,438 of 14,040 clips (81.5%), 22 of 53 speakers (41.5%)

Data splits for modelling

Clip buckets

Bucket	Clips
Validated	13,277 (94.6%)
Invalidated	758 (5.4%)
Other	5 (0.0%)

Training splits

Split	Clips
Train	4,190 (31.6%)
Dev	2,872 (21.6%)
Test	2,920 (22.0%)

Training split coverage: 9,982 of 13,277 validated clips (75.2%)

The dataset contains 13277 validated, 758 invalidated, and 5 unresolved clips. The average clip duration is 4.574 seconds.

Text corpus

Validated sentences: 10,454

Category	Count
Unvalidated sentences	357
Pending sentences	14
Rejected sentences	343
Reported sentences	8

The corpus contains 10,811 sentences: 10,454 validated and 357 unvalidated (14 pending review, 343 rejected), with 8 reported for review.

Writing system

In 2023, the Academia Aragonesa de la Lengua, a public institution created as language regulatory body, established the Orthography of Aragonese as the official spelling for the language. The spelling system is flexible enough to allow writing all Aragonese varieties.

All the submitted sentences have been adapted to the official orthography, although errors can be found occasionally.

Sample

There follows a randomly selected sample of five sentences from the corpus.

Y qué ye una candeleta?
He rascluciau un poquet per astí.
Imos a fer un mueso.
Lo feban pa fer-lo mover y asinas colarse'n de gratis.
En que le va pasar per debant, va brincar d'el cado.

Sources

Sentences have been obtained from the works of two writters whose works are already in the public domain: Dámaso Carrera and Domingo Miral, as well as by living authors who have submitted sentences from their own works.

1400 sentences from Diccionario chistabino. Author. F. Blas / F. Romanos. Chistabin Aragonese.
251 sentences from stories. Author: D. Carrera (public domain). Ribagorzan Aragonese.
814 sentences from stories. Author: M. Castán. Benás (patués) Aragonese.
360 sentences from stories. Author: F. González. Ribagorzan Aragonese.
1389 sentences from stories and plays. Author: E. Gusano. Ansotano Aragonese.
997 sentences from linguistic surveys. Author: Ch. Lozano. Belsetan Aragonese.
625 sentences from linguistic surveys. Author: Ch. Lozano. Tella Aragonese.
870 sentences from a story book. Author: J.C. Marco. Ribagorzan Aragonese.
732 sentences from stories and articles. Author: M. Marín. Cheso Aragonese.
1309 sentences from scripts of the "Pizarra y Clarión" section of the Aragonese TV Show "Charrín Charrán". Author: J.P. Martínez. Variants: Standard Aragonese and all the main variants.
441 sentences from theater play Qui bien fa nunca lo pierde. Author: D. Miral (public domain). Cheso Aragonese.
209 sentences created for Aragonese course. Author: J. Orduna. Standard Aragonese.
444 sentences from surveys and dictionary examples. Author. F. Romanos. La Fueva Aragonese.
529 sentences from stories and articles. Author: J.M. Satué. Sobrepuerto Aragonese.
100 sentences from stories. Author: Chorche. Panticuto Aragonese.

Source	Sentences
Self	2,420 (23.1%)
self	1,743 (16.7%)
self (ansotano)	1,389 (13.3%)
self (chistabín)	1,387 (13.3%)
self-citation (JPM)	1,007 (9.6%)
Self cheso	716 (6.8%)
self (JMS)	536 (5.1%)
Taken from: Domingo Miral (1872-1946). Qui bien fa nunca lo pierde. Published in 1903.	444 (4.2%)
Autocita	342 (3.3%)
Other	470 (4.5%)

Text domains

Code	Domain	Clips	Speakers
general	General	13,638 (97.1%)	53 (100.0%)
agriculture_food	Agriculture and Food	-	-
automotive_transport	Automotive and Transport	-	-
finance	Finance	2 (0.0%)	2 (3.8%)
service_retail	Service and Retail	2 (0.0%)	2 (3.8%)
healthcare	Healthcare	-	-
history_law_government	History, Law and Government	-	-
media_entertainment	Media and Entertainment	-	-
nature_environment	Nature and Environment	-	-
news_current_affairs	News and Current Affairs	-	-
technology_robotics	Technology and Robotics	-	-
language_fundamentals	Language Fundamentals	-	-

Fields

Clips

Each row of a tsv file represents a single audio clip, and contains the following information:

client_id - hashed UUID of a given user
path - relative path of the audio file
text - supposed transcription of the audio
up_votes - number of people who said audio matches the text
down_votes - number of people who said audio does not match text
age - age of the speaker1
gender - gender of the speaker1
accents - accents of the speaker1
variant - variant of the language1
segment - if sentence belongs to a custom dataset segment, it will be listed here
prompt_upvotes - number of upvotes the sentence prompt received
prompt_reports - number of reports the sentence prompt received
is_edited - whether the clip's transcription has been edited

`validated_sentences.tsv`

The validated_sentences.tsv file contains one row per validated sentence in the text corpus:

sentence_id - unique identifier for the sentence
sentence - the sentence text
variant - the variant of the language
sentence_domain - the domain(s) the sentence belongs to
source - the source the sentence was collected from
is_used - whether the sentence is still in circulation for recording
clips_count - number of clips recorded for this sentence

`unvalidated_sentences.tsv`

The unvalidated_sentences.tsv file contains one row per unvalidated sentence in the text corpus:

sentence_id - unique identifier for the sentence
sentence - the sentence text
variant - the variant of the language
sentence_domain - the domain(s) the sentence belongs to
source - the source the sentence was collected from
up_votes - number of upvotes the sentence received
down_votes - number of downvotes the sentence received
status - current status of the sentence (pending or rejected)

Get involved

Community links

Discussions

Contribute

Acknowledgements

Datasheet authors

Juan Pablo Martínez

Licence

This dataset is released under the Creative Commons Zero (CC-0) licence. By downloading this data you agree to not determine the identity of speakers in the dataset.

Footnotes

For a full list of age, gender, and accent options, see the demographics spec. These will only be reported if the speaker opted in to provide that information. ↩ ↩2 ↩3 ↩4