2 of 17

Opening Sources in Breton Language: Offering the “Minoritized’’ Language to the Majority

Presentation plan

DEVRI: illustration of an open sources tool in Breton

Martial Ménard [1951-2016]
DEVRI: The diachronic dictionary

WIKImammenn: challenging the open sources

Digitize the corpora
Some good fresh news

Without open sources: no A.I.!

Feeding the Beast
Political issues for digital survival

Minorized language :

result of a voluntary process, product of linguistic colonialism and not “minority’’ as state of affairs, by essence or by nature of the languages themselves.

3 of 17

I) DEVRI:

illustration of an open sources tool in breton

4 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

Martial MÉNARD [1951-2016]

breton activist
political prisoner
editor of An Here/The Harvest [1983-2003]
lexicologist

« We must take what we are entitled to, not hold out our hands. »

M. Ménard

5 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

Some elements of the lexicographer’s work:

Monolingual dictionary (1995), 1 232p.
Breton dictionary of erotism (1995), 528p.
French->Breton dictionary (2012), 1463p.
Monolingual dictionary (2001), 1 436p.
Breton chronicles (2021), 960p.

S.A.B.

Stourm ar brezhoneg (Struggle for the Breton language) is a a collective demanding the officialization of Breton (1980s)

6 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

Roparz HEMON [1900-1978]

founder of Gwalarn (1925-1944)
novelist, short story writer, poet
grammarian lexicologist
translator (Andersen, Marlowe, etc.)
A historical morphology and syntax of Breton

GWALARN/ (Northwest)

First monolingual literary review, then literary movement which emanated from it

7 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

69,333 entries

40 years of work

words from 1 100

to 2 010

8 of 17

II) WIKIMAMMENN:

Challenging the open sources

9 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

WIKImammenn (WIKIsource)

Already 10 000 pages in breton monolingual

191 book
552 index
333 songs
186 stories
811 poems
50 plays
1 118 readable listenable music scores
written by 180 authors

10 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

Some good fresh news : Transkribus !

OCR: Optical Character Recognition

HTR: Handwritten Text Recognition

11 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

Some good fresh news : Transkribus !

Highlighted content

12 of 17

III) Without open sources: no A.I.

13 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

Feeding the Beast

14 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

unified language and spelling
grammar / dictionary
monolingual dictionary
oral historical corpus
written historical corpus
corrector
international translator (google.traduction)

Written recognition (on the way !)
LLM (Large Language Model)
vocal recognition
G.P.S.
automatic subtitling
chatbots
else …

Pragmatical tools for breton language TODAY

> still quite improvable !

15 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

Political issues for digital survival

"META has already created speech recognition systems for a thousand languages’’
"It would be enough to finance’’
"It is not the job of the Public Office of the Breton Language to work for a private company that earns billions every year’’
“We need free corpora, in line with compliance with the FAIR principles […] to operate this online distribution »
NLP

rich corpora
make it accesses / visible by LDC
fight for « digital surveyval »

No official status

= no strong linguistic policy

= no financial means

16 of 17

Opening Sources in Breton Language: Offering the Minoritized Language to the Majority

SOURCES

INTERVIEWS:

interview with Hervé Le Bihan, 18/07/2024

interview with Cédric Choplin, 17/09/2024

interview with Kristian Hamon, 18/09/2024

interview with Fañch Jestin, 18/09/2024

interview with Hervé Baudry 18-20/09/2024

WEBSITES:

ARBRES :

https://arbres.iker.cnrs.fr

Ressources numériques du breton u.t.d.:

https://entrelangues.modyco.fr/index.php/Breton#Ressources_num%C3%A9riques

Traitement automatique des langues -TAL- u.t.d.:

https://arbres.iker.cnrs.fr/index.php?title=Traitement_automatique_des_langues_-_Breton

Premier corpora Universal Dependencies (2018) :

https://github.com/UniversalDependencies/UD_Breton-KEB

Plateforme de classement de structures orales : COllections de COrpus Oraux Numériques u.t.d. :

https://cocoon.huma-num.fr/exist/crdo/

Initiative de collecte META :

https://ai.meta.com/research/no-language-left-behind/

Correcteur - traducteur Korvigelloù an Drouizig :

https://drouizig.org/br/

Kristian Hamon - Herve Baudry :

https://www.transkribus.org/model/breton-prints-17th-19th-centuries

Alan Enterm/Brendan-Budok Durand-Le Ludec : traducteur local (OPAB)

https://niverel.brezhoneg.bzh/fr/troer/

17 of 17

Thanks for your attention!

Get in touch with me:

tristan.loarer@univ-rennes2.fr

CELTIC BLM Research lab

https://sites-recherche.univ-rennes2.fr/celtic-blm/

Credits

For sources and license rights concerning the illustrations above, cf. the complete bibliography/websitography which follows (3p)

Dr. Tristan LOARER

CELTIC BLM

University Rennes 2 (Brittany)

https://perso.univ-rennes2.fr/tristan.loarer