 Your new post is loading...
|
Scooped by
Charles Tiayon
December 22, 2025 6:51 PM
|
"Ces mots français venus du Moyen-Orient et de Méditerranée Des milliers de mots de la langue française proviennent de contrées lointaines de l’Hexagone. Parmi eux, des centaines ayant des liens directs ou indirects avec l’arabe mais également, à une moindre échelle, le persan. Publié le : 21/12/2025 - 12:04 Modifié le : 22/12/2025 - 10:10 Le mot français bougie provient de la ville algérienne éponyme aujourd’hui appelée Béjaïa. REUTERS/KIM HONG-JI Par :bAnne Bernas Monsieur Jourdain ignorait qu’il faisait de la prose, mais Molière savait-il que de nombreux mots dont il usait provenaient de contrées moyen-orientales et méditerranéennes ? L’arabe est la sixième langue la plus parlée au monde et est utilisée quotidiennement par plus de 400 millions de personnes, dont 200 millions de locuteurs natifs, selon les données des Nations unies. Une « Journée mondiale » lui est également consacrée chaque 18 décembre, date coïncidant avec le jour où, en 1973, l'Assemblée générale de l’ONU a adopté l'arabe comme sixième langue officielle de l'Organisation par la résolution 3190. D’une infinie richesse, l’arabe, classique mais aussi dialectal, a transmis à de nombreuses langues indo-européennes une somme de mots considérable. « Il y a beaucoup plus de mots d’origine arabe dans la langue française que de mots d’origine gauloise ! » rappelait en 2022 l’académicien Erik Orsenna. Plus de 800 références Dans le dictionnaire français, pas moins de 400 mots sont d’origine arabe, 200 autres sont des mots dérivés et quelque 800 portent la marque de la langue arabe. Depuis le Moyen Âge, des mots arabes voyagent jusqu’en France, par la terre et la mer, en faisant parfois escale dans d’autres pays. À ce titre, le mot « amiral », rang le plus élevé dans la marine française, provient de « Amir al-bahr», le « chef de la mer ». La dissémination de mots d’origine arabe se poursuit lors des conquêtes coloniales du XIXᵉ siècle en Afrique du Nord et jusqu'à plus récemment avec le langage des jeunes de banlieues. Pourtant, comme le rappelle Salah Guemriche, auteur du Dictionnaire des mots français d'origine arabe (Seuil, 2007), de nombreux mots français d’origine arabe n’ont pas toujours été « assumés » par Paris, à la différence des mots d’origine persane. « Aux XVIᵉ et XVIIᵉ siècles, l’origine ne dérangeait pas, elle était sans arrière-pensée idéologique, ce qui change au XIXᵉ, à l’heure de la colonisation. Comment voulez-vous dire d’un côté “on va apporter la civilisation” et de l’autre dire que la langue française contient déjà des mots arabes, des Arabes qu’on est allé coloniser. » Les termes issus de l’arabe sont si nombreux qu’il est impossible d’en dresser une liste, mais parmi les plus connus, dans le domaine des sciences et des mathématiques, à côté des centaines de noms d’étoiles, notons les célèbres « zéro », « chiffre », « zénith », etc. Mais aussi « algèbre », provenant de l'arabe « al-jabr »,« restauration » ou « réunion ». Ce mot figure initialement dans le titre d'un ouvrage du IXᵉ siècle consacré à la résolution d'équations, écrit par le savant persan Muhammad ibn Musa al-Khwarizmi, installé à Bagdad, et dont dérive le mot « algorithme ». Plus terre à terre. Le sud de l’Hexagone est réputé pour sa culture des abricots (« al barbuk »), mais les Bretons, dont l’artichaut est l’un de leurs symboles de fierté, savent-ils que ces « épines de la terre » se disent en arabe « ardi chawk » ? Savent-ils que l’arsenal de Brest puise son origine de l’arabe « dar al-sina » ou « maison de fabrication ». Et que dire du caban breton, mais aussi des écharpes, chemises, jupes, gilets, parfois en coton, lui aussi issu de l’arabe « kutun ». « Le français n'est pas français » À l’heure des vacances de fin d’année, certains pourraient être tentés de partir faire un safari, dont le terme vient de l’arabe « safar » signifiant « voyage », pour découvrir les girafes (de l’arabe « zarafa ») et les gazelles (« ghazal ») au milieu des baobabs ( « abu hibab »). Un voyage sans grosses tempêtes pour les voyageurs de l’océan Indien puisque la saison des moussons est terminée. La mousson provient de l’arabe « mausim », « saison ». Il arrive dans le vocabulaire français à la fin du XVIᵉ siècle après être passé par le néerlandais, le portugais et l’espagnol sous différentes formes. Pour fêter la fin de l’année, certains feront la nouba (de l’arabe algérien) et trinqueront avec un verre d’alcool, mot venant du mot « al-khol ». Il signifie à l’origine une fine poudre noire de sulfure de plomb et d’antimoine, un produit de beauté mais aussi un produit d’alchimistes. À la Renaissance, le chimiste allemand utilise le terme « alcohol vini » pour désigner « l’esprit de vin », le « produit d’une distillation de vin ». Quoi qu’il en soit, pour se remettre de toutes ces émotions, rien ne sera plus efficace qu’un petit « caoua ». D’autres mots sont directement entrés dans le vocabulaire, par exemple bougie (de la ville algérienne éponyme aujourd’hui appelée Béjaïa), casbah (directement venu d’Alger), bled, barda, chouia, et plus près de nous kiffer, dawa ou encore seum. Et puis il y a des « mots arabes clandestins » tels que les nomme le lexicographe Roland Laffitte qui raconte l’exemple de l’épithète « romaine » accolé à une balance, la fameuse balance romaine. « Romaine » vient de l’arabe rumāna signifiant « grenade ». Car cette balance possède sur un bras un dispositif d’équilibre qui lui ressemblait. La langue arabe, la plus parlée des langues sémitiques et la quatrième source d’emprunts de notre langue, après l’anglais, l’italien et l’allemand, a permis un dialogue des cultures le long des routes de la soie, de la côte de l'Inde à la Corne de l'Afrique. Les empires arabes successifs, après la fondation de l’islam, régnèrent sur un territoire qui s’étendait du Maroc à l’ouest jusqu’aux frontières de ce qui est aujourd’hui la Chine. La langue nous a rapporté d’Inde le mot sucre, issu du sanskrit et qui se dit en arabe « sukkar »... « À travers les siècles, l’arabe a été au cœur des échanges entre les continents et entre les cultures », déclarait Audrey Azoulay, ex-directrice générale de l’Unesco. La langue, ajoutait-elle, a été « maniée par tant de grands poètes, penseurs, scientifiques et savants ». Il n’y a pas de hasard (du mot arabe « az-ahar », « dé à jouer »), le génie de la langue française tient beaucoup à ces innombrables apports multiculturels. « Le français n'est pas français », s’amusait à dire le linguiste de renom et historien de la langue française Alain Rey. De quoi y perdre son latin. « Les mots ont beaucoup voyagé » À une moindre échelle, de nombreux mots issus du persan - tulipe, divan, nénuphar, babouches, bazar, etc.- ont enrichi le vocabulaire français depuis le Moyen Âge. Explications avec Agnès Lenepveu-Hotz, maître de conférences au département d'études persanes de l’université de Strasbourg. RFI : Comment expliquer que le nombre de mots français issus du persan (0,7%), langue plus ancienne que l’arabe, soit très inférieur à celui des mots issus de l’arabe (5%) ? Agnès Lenepveu-Hotz : On trouve moins de mots français issus du persan que de l’arabe pour des raisons historiques. Les royaumes arabes dans le sud de l'Espagne ont par exemple engendré des contacts et donc l'espagnol, comme le français du Sud- Ouest, a emprunté des termes à l'arabe. Aussi, beaucoup de mots persans ne sont pas arrivés directement jusqu’à nous. Assez souvent, ces mots ont transité par l'arabe, le turc, parfois le hindi. Quand le mot a été emprunté de l'anglais au hindi, il est parfois repassé en français, et souvent le hindi l'avait emprunté au persan avant. Ainsi, les mots ont beaucoup voyagé, comme le mot « châle ». Ce mot a été emprunté par les Anglais au hindi, mais l’hindi l'avait emprunté au persan. Le mot « orange » est passé par l'arabe et même par l'espagnol, mais au départ, c'est un mot persan. Pour certains mots, on peine vraiment à savoir si l’origine est d’abord persane ou arabe. Pour d'autres, c'est beaucoup plus clair, comme pour le mot « kiosque ». Ce mot est attesté dès le moyen perse, donc bien avant les emprunts que le persan a fait à l'arabe. Ce mot signifie « palais », « pavillon », « kiosque ». Le mot « bakchich », lui, a transité par le turc, mais c'est un verbe en persan qui veut dire « donner ». Le français l'a ensuite emprunté du turc, mais l'arabe et le turc l'avaient emprunté au persan." https://www.rfi.fr/fr/moyen-orient/20251221-ces-mots-fran%C3%A7ais-venus-du-moyen-orient-et-de-m%C3%A9diterran%C3%A9e #Metaglossia #metaglossia_mundus
Researchers across Africa, Asia and the Middle East are building their own language models designed for local tongues, cultural nuance and digital independence
"In a high-stakes artificial intelligence race between the United States and China, an equally transformative movement is taking shape elsewhere. From Cape Town to Bangalore, from Cairo to Riyadh, researchers, engineers and public institutions are building homegrown AI systems, models that speak not just in local languages, but with regional insight and cultural depth.
The dominant narrative in AI, particularly since the early 2020s, has focused on a handful of US-based companies like OpenAI with GPT, Google with Gemini, Meta’s LLaMa, Anthropic’s Claude. They vie to build ever larger and more capable models. Earlier in 2025, China’s DeepSeek, a Hangzhou-based startup, added a new twist by releasing large language models (LLMs) that rival their American counterparts, with a smaller computational demand. But increasingly, researchers across the Global South are challenging the notion that technological leadership in AI is the exclusive domain of these two superpowers.
Instead, scientists and institutions in countries like India, South Africa, Egypt and Saudi Arabia are rethinking the very premise of generative AI. Their focus is not on scaling up, but on scaling right, building models that work for local users, in their languages, and within their social and economic realities.
“How do we make sure that the entire planet benefits from AI?” asks Benjamin Rosman, a professor at the University of the Witwatersrand and a lead developer of InkubaLM, a generative model trained on five African languages. “I want more and more voices to be in the conversation”.
Beyond English, beyond Silicon Valley
Large language models work by training on massive troves of online text. While the latest versions of GPT, Gemini or LLaMa boast multilingual capabilities, the overwhelming presence of English-language material and Western cultural contexts in these datasets skews their outputs. For speakers of Hindi, Arabic, Swahili, Xhosa and countless other languages, that means AI systems may not only stumble over grammar and syntax, they can also miss the point entirely.
“In Indian languages, large models trained on English data just don’t perform well,” says Janki Nawale, a linguist at AI4Bharat, a lab at the Indian Institute of Technology Madras. “There are cultural nuances, dialectal variations, and even non-standard scripts that make translation and understanding difficult.” Nawale’s team builds supervised datasets and evaluation benchmarks for what specialists call “low resource” languages, those that lack robust digital corpora for machine learning.
It’s not just a question of grammar or vocabulary. “The meaning often lies in the implication,” says Vukosi Marivate, a professor of computer science at the University of Pretoria, in South Africa. “In isiXhosa, the words are one thing but what’s being implied is what really matters.” Marivate co-leads Masakhane NLP, a pan-African collective of AI researchers that recently developed AFROBENCH, a rigorous benchmark for evaluating how well large language models perform on 64 African languages across 15 tasks. The results, published in a preprint in March, revealed major gaps in performance between English and nearly all African languages, especially with open-source models.
Similar concerns arise in the Arabic-speaking world. “If English dominates the training process, the answers will be filtered through a Western lens rather than an Arab one,” says Mekki Habib, a robotics professor at the American University in Cairo. A 2024 preprint from the Tunisian AI firm Clusterlab finds that many multilingual models fail to capture Arabic’s syntactic complexity or cultural frames of reference, particularly in dialect-rich contexts.
Governments step in
For many countries in the Global South, the stakes are geopolitical as well as linguistic. Dependence on Western or Chinese AI infrastructure could mean diminished sovereignty over information, technology, and even national narratives. In response, governments are pouring resources into creating their own models.
Saudi Arabia’s national AI authority, SDAIA, has built ‘ALLaM,’ an Arabic-first model based on Meta’s LLaMa-2, enriched with more than 540 billion Arabic tokens. The United Arab Emirates has backed several initiatives, including ‘Jais,’ an open-source Arabic-English model built by MBZUAI in collaboration with US chipmaker Cerebras Systems and the Abu Dhabi firm Inception. Another UAE-backed project, Noor, focuses on educational and Islamic applications.
In Qatar, researchers at Hamad Bin Khalifa University, and the Qatar Computing Research Institute, have developed the Fanar platform and its LLMs Fanar Star and Fanar Prime. Trained on a trillion tokens of Arabic, English, and code, Fanar’s tokenization approach is specifically engineered to reflect Arabic’s rich morphology and syntax.
India has emerged as a major hub for AI localization. In 2024, the government launched BharatGen, a public-private initiative funded with 235 crore (€26 million) initiative aimed at building foundation models attuned to India’s vast linguistic and cultural diversity. The project is led by the Indian Institute of Technology in Bombay and also involves its sister organizations in Hyderabad, Mandi, Kanpur, Indore, and Madras. The programme’s first product, e-vikrAI, can generate product descriptions and pricing suggestions from images in various Indic languages. Startups like Ola-backed Krutrim and CoRover’s BharatGPT have jumped in, while Google’s Indian lab unveiled MuRIL, a language model trained exclusively on Indian languages. The Indian governments’ AI Mission has received more than180 proposals from local researchers and startups to build national-scale AI infrastructure and large language models, and the Bengaluru-based company, AI Sarvam, has been selected to build India’s first ‘sovereign’ LLM, expected to be fluent in various Indian languages.
In Africa, much of the energy comes from the ground up. Masakhane NLP and Deep Learning Indaba, a pan-African academic movement, have created a decentralized research culture across the continent. One notable offshoot, Johannesburg-based Lelapa AI, launched InkubaLM in September 2024. It’s a ‘small language model’ (SLM) focused on five African languages with broad reach: Swahili, Hausa, Yoruba, isiZulu and isiXhosa.
“With only 0.4 billion parameters, it performs comparably to much larger models,” says Rosman. The model’s compact size and efficiency are designed to meet Africa’s infrastructure constraints while serving real-world applications. Another African model is UlizaLlama, a 7-billion parameter model developed by the Kenyan foundation Jacaranda Health, to support new and expectant mothers with AI-driven support in Swahili, Hausa, Yoruba, Xhosa, and Zulu.
India’s research scene is similarly vibrant. The AI4Bharat laboratory at IIT Madras has just released IndicTrans2, that supports translation across all 22 scheduled Indian languages. Sarvam AI, another startup, released its first LLM last year to support 10 major Indian languages. And KissanAI, co-founded by Pratik Desai, develops generative AI tools to deliver agricultural advice to farmers in their native languages.
The data dilemma
Yet building LLMs for underrepresented languages poses enormous challenges. Chief among them is data scarcity. “Even Hindi datasets are tiny compared to English,” says Tapas Kumar Mishra, a professor at the National Institute of Technology, Rourkela in eastern India. “So, training models from scratch is unlikely to match English-based models in performance.”
Rosman agrees. “The big-data paradigm doesn’t work for African languages. We simply don’t have the volume.” His team is pioneering alternative approaches like the Esethu Framework, a protocol for ethically collecting speech datasets from native speakers and redistributing revenue back to further development of AI tools for under-resourced languages. The project’s pilot used read speech from isiXhosa speakers, complete with metadata, to build voice-based applications.
In Arab nations, similar work is underway. Clusterlab’s 101 Billion Arabic Words Dataset is the largest of its kind, meticulously extracted and cleaned from the web to support Arabic-first model training.
The cost of staying local
But for all the innovation, practical obstacles remain. “The return on investment is low,” says KissanAI’s Desai. “The market for regional language models is big, but those with purchasing power still work in English.” And while Western tech companies attract the best minds globally, including many Indian and African scientists, researchers at home often face limited funding, patchy computing infrastructure, and unclear legal frameworks around data and privacy.
“There’s still a lack of sustainable funding, a shortage of specialists, and insufficient integration with educational or public systems,” warns Habib, the Cairo-based professor. “All of this has to change.”
A different vision for AI
Despite the hurdles, what’s emerging is a distinct vision for AI in the Global South – one that favours practical impact over prestige, and community ownership over corporate secrecy.
“There’s more emphasis here on solving real problems for real people,” says Nawale of AI4Bharat. Rather than chasing benchmark scores, researchers are aiming for relevance: tools for farmers, students, and small business owners.
And openness matters. “Some companies claim to be open-source, but they only release the model weights, not the data,” Marivate says. “With InkubaLM, we release both. We want others to build on what we’ve done, to do it better.”
In a global contest often measured in teraflops and tokens, these efforts may seem modest. But for the billions who speak the world’s less-resourced languages, they represent a future in which AI doesn’t just speak to them, but with them."
Sibusiso Biyela, Amr Rageh and Shakoor Rather
20 May 2025
https://www.natureasia.com/en/nmiddleeast/article/10.1038/nmiddleeast.2025.65
#metaglossia_mundus
"Le Japon investit dans la traduction de mangas assistée par l'IA.
Chaque année, seulement 10 % environ des mangas publiés au Japon sont traduits en anglais. Ce pourcentage est encore plus faible pour les autres langues.
Le gouvernement japonais a désigné le manga comme un nouveau secteur clé. - Photo : Romancing Japan
Selon Nikkei Asia , le Japon soutiendra l'expansion de la distribution de mangas sur les marchés internationaux en formant des ressources humaines capables de traduire rapidement des bandes dessinées à l'aide de l'intelligence artificielle (IA).
Il s'agit d'une nouvelle initiative du gouvernement japonais visant à empêcher les internautes d'accéder à des sites web piratés. Auparavant, l'Agence japonaise des affaires culturelles avait également pour objectif de développer un système d'intelligence artificielle capable de détecter automatiquement les sites web enfreignant le droit d'auteur.
Prévenir le piratage des mangas dû au manque de traductions.
Les mangas sont très populaires sur les marchés étrangers, mais de nombreux lecteurs choisissent encore de lire des versions copiées et mises en ligne illégalement.
Selon Yukari Shiina, chargée de cours à l'Université des Arts de Tokyo, la principale raison est que le rythme de traduction des mangas n'a pas suivi la demande des lecteurs, créant ainsi les conditions propices à la propagation des violations de droits d'auteur.
Une enquête menée par l'organisation anti-piratage Authorised Books of Japan (ABJ) a révélé qu'environ 900 sites Web pirates spécialisés dans la publication de mangas ont enregistré 2,8 milliards de visites en provenance de 123 pays et territoires rien qu'en juin 2025.
Face à cette situation, l'Agence japonaise des affaires culturelles soutiendra des programmes de formation à la traduction de mangas par intelligence artificielle, à hauteur de 100 millions de yens par projet. Ce soutien devrait inclure des formations intensives en traduction et des méthodes d'application efficaces de l'IA.
Shogakukan, l'éditeur de mangas à succès comme Doraemon et Conan, soutient également le développement de logiciels de traduction de mangas basés sur l'intelligence artificielle. - Photo : Toho
Parallèlement, des entreprises privées accélèrent également le développement d'outils de traduction de mangas basés sur l'IA. Mantra, une start-up affiliée à l'Université de Tokyo, a développé un outil capable de traduire des œuvres complètes, y compris le style des dialogues et le contexte narratif.
Cet outil prend en charge 18 langues et permet de réduire de moitié le temps de traduction par rapport aux méthodes traditionnelles. Actuellement, Mantra traite environ 200 000 pages par mois, soit l’équivalent de 1 000 volumes de manga.
« L’IA réduit considérablement les tâches simples comme le remplacement de mots, mais les humains jouent toujours un rôle clé pour garantir l’exactitude, le naturel de l’expression et le contexte », a déclaré Shonosuke Ishiwatari, PDG de Mantra.
L'interface de traduction de mangas de Mantra, basée sur l'IA - Photo : Nikkei Asia
Depuis deux ans, la maison d'édition Shogakukan utilise des outils de traduction développés par différents spécialistes de l'IA, dont Mantra. Ces outils sont capables de reconnaître le texte des images de manga et d'effectuer les traductions. Lorsque le texte traduit est trop long pour la bulle de dialogue, l'IA ajuste automatiquement la taille de cette dernière.
La relecture, essentielle à l'amélioration de la qualité des traductions, reste encore une tâche humaine. Nobumasa Sawabe, PDG de Shogakukan, estime que l'IA accélère considérablement le processus de traduction. Shogakukan ambitionne de porter la part de son chiffre d'affaires réalisé à l'international à 10 % d'ici quatre ans, contre 3 à 4 % actuellement.
Par ailleurs, Mme Yukari Shiina a souligné que la qualité des traductions sur les sites de mangas piratés est souvent inégale. « Développer les activités de traduction est essentiel, non seulement pour garantir les droits légitimes des auteurs et des éditeurs, mais aussi pour offrir aux lecteurs des œuvres de meilleure qualité », a-t-elle insisté."
Báo Tuổi Trẻ
04/01/2026
https://www.vietnam.vn/fr/nhat-ban-dau-tu-cho-nhan-luc-dich-manga-bang-ai
#Metaglossia
#metaglossia_mundus
une application de traduction bidirectionnelle khmère-anglais gratuite, désormais accessible sur Play Store, App Store et via le site www.translate.kh.
Le ministère des Postes et des Télécommunications (MPTC) du Cambodge marque un tournant décisif dans la transformation numérique nationale. Le 2 janvier, il a dévoilé TranslateKH 2.0, une application de traduction bidirectionnelle khmère-anglais gratuite, désormais accessible sur Play Store, App Store et via le site www.translate.kh.
Cette initiative, qui succède à la première version lancée le 25 novembre 2024, s’appuie sur l’intelligence artificielle avancée, le deep learning et un modèle linguistique massif pour des traductions contextuelles, surpassant les approches purement syntaxiques.
Au-delà de la commodité quotidienne, TranslateKH 2.0 s’inscrit dans une stratégie économique ambitieuse : doper la compétitivité du Cambodge sur la scène régionale et mondiale, où le bilinguisme anglais-khmer devient un atout stratégique.
Révolution technologique au cœur du projet
Dotée d’une compréhension contextuelle fine, l’application excelle dans les nuances culturelles et idiomatiques propres au khmer, une langue à tonalité complexe. Contrairement aux outils génériques comme Google Translate, TranslateKH privilégie la précision sémantique, idéale pour les documents officiels, contrats commerciaux ou échanges touristiques.
Le MPTC met en avant une interface intuitive, un mode hors ligne partiel et des mises à jour régulières basées sur les retours utilisateurs, positionnant le Cambodge comme pionnier en IA localisée en Asie du Sud-Est.
Enjeux économiques : un levier pour les PME et le commerce régional
Dans un Cambodge où le numérique explose – avec 75% de pénétration internet en 2025 –, TranslateKH agit comme un catalyseur économique. Les PME, pilier de l’économie (représentant 60% du PIB), gagnent en fluidité pour négocier avec des partenaires RCEP (Chine, Vietnam, Thaïlande).
Prenons l’exemple des exportateurs de textiles et d’agroalimentaire : des contrats en anglais clair accélèrent les deals, réduisant les coûts de traduction externe de 30 à 50%. Le tourisme, contributeur majeur (12% du PIB pré-pandémie), bénéficie aussi : guides et hôteliers communiquent sans barrière.
Focus sectoriel : Les entreprises du garment, concentrées à Phnom Penh et autour de Sihanoukville, intègrent déjà l’outil pour former leurs équipes bilingues, alignant compétences locales sur les standards ASEAN.
Impact éducatif : démocratiser l’anglais, former la jeunesse
L’éducation supérieure cambodgienne, avec ses 200 000 étudiants, souffre d’un déficit linguistique : seul 20% maîtrise l’anglais courant. TranslateKH 2.0 pallie cela en servant d’auxiliaire pédagogique – pratique interactive, quiz intégrés –, favorisant l’accès aux ressources MOOC internationales (Coursera, edX).
Pour les universités techniques (automobile, IT, construction), cela accélère la curriculum internationalisation, alignée sur les besoins industriels. Résultat projeté : hausse de 15% des embauches qualifiées d’ici 2027, selon les projections du ministère de l’Éducation.
Témoignage fictif illustratif : « En tant qu’enseignant en comptabilité, TranslateKH transforme mes cours : mes étudiants saisissent désormais les normes IFRS en un clin d’œil », confie un professeur de l’Institut Royal d’Économie.
Inclusion numérique et attractivité touristique
L’inclusion reste clé : l’application cible les zones rurales (40% de la population), via des partenariats avec les opérateurs télécoms pour des forfaits data bonifiés. Côté tourisme, Phnom Penh, Siem Reap et Sihanoukville attirent 7 millions de visiteurs prévus en 2026. Des menus bilingues instantanés ou des visites guidées en temps réel valorisent le patrimoine khmer (Angkor, gastronomie), multipliant les upsells dans l’hôtellerie-restauration.
Vers une souveraineté linguistique numérique
https://www.cambodgemag.com/post/traduction-connect%C3%A9e-le-cambodge-propulse-le-bilinguisme-num%C3%A9rique-avec-translatekh-2-0
#Metaglossia
#metaglossia_mundus
Les limites de la traduction des symboles
"Dans le cadre des "Grandes conférences" était diffusée sur France Culture au cœur de l'été 1971, une allocution de Roger Caillois intitulée "Les limites de la traduction des symboles". Auteur littéraire et aussi essayiste, c'est en tant que traducteur qu'il intervenait à cette occasion.
“Seul de tous les arts, la littérature souffre de la malédiction de Babel.” C’est par cette sentence introductive que Roger Caillois résume l’enjeu du problème de la traduction, dans la conférence prononcée quelques mois seulement après son élection à l’Académie française au début de l'année 1971.
La tâche ingrate du traducteur
En cette année 1971, Caillois est un intellectuel reconnu depuis déjà de nombreuses années. Ses essais, ses critiques littéraires, son travail poétique lui valent une reconnaissance sinon du grand public du moins des cercles intellectuels. Mais Roger Caillois est aussi un traducteur lui-même et non des moindres puisqu’il a traduit en français l’œuvre du maître argentin Jorge Luis Borges.
C'est donc en connaissance de cause qu'il évoque dans cette intervention la difficulté de transposer d’une langue à l’autre ce qui est parfois au-delà des mots. Il y explique ce qu'est selon lui une traduction fidèle, et quelles en sont les difficultés alors que "dans chaque langue, aucun mot même le plus ordinaire ne constitue l'équivalent exact du mot étranger correspondant."
Cette "opération ingrate, complexe, traitresse" n’est pas pour autant impossible pour le traducteur mais sa tâche est colossale. Il lui faut selon les mots de Caillois "inventer le texte qu’aurait écrit l’auteur traduit si sa langue maternelle avait été celle du traducteur et non la sienne".
Retrouvez l'ensemble de la Nuit "Roger Caillois, arpenteur de l’imaginaire"
Par Denise Parent
Présentation Gérard Mourgue
Avec Roger Caillois
Les grandes conférences - Les limites de la traduction des symboles, par Roger Caillois : Le 21/05/1971 à Kaboul (1ère diffusion : 30/07/1971)
Indexation web : Documentation sonore de Radio France
Archive Ina-Radio France"
https://www.radiofrance.fr/franceculture/podcasts/les-nuits-de-france-culture/roger-caillois-arpenteur-de-l-imaginaire-16-18-la-traduction-ideale-selon-roger-caillois-2631438
#Metaglossia
#metaglossia_mundus
"For centuries, the Arapaho have called Colorado and Wyoming home. The tribe gave names to places like the Kawuneeche Valley, the Never Summer Mountains, and Mount Blue Sky.
But the language the Arapaho have spoken for centuries is at risk of disappearing, as fewer members of the tribe have learned the language.
A team of language experts at the University of Colorado Boulder is working to change that. They’re compiling an online database that includes recordings of the Arapaho language and can be used as a learning and teaching tool.
Andrew Cowell is a linguistics professor at CU, and faculty director of the Center for Native American and Indigenous studies. He helped launch this project more than two decades ago.
He joined Erin O’Toole to talk about how he hopes the digital database helps future generations learn and continue to speak the Arapaho language.
You can access the Arapaho Language Project here.👇🏿👇🏿
https://www.kunc.org/podcast/inthenoco/2026-01-02/the-arapaho-language-is-endangered-a-cu-professor-hopes-this-project-will-help-preserve-it
#Metaglossia
#metaglossia_mundus
"Con el objetivo de impulsar la inclusión y mejorar la comunicación con personas sordas en distintos entornos sociales y educativos, los estudiantes José Eduardo Méndez Merino y Luis Óscar Aguilar Cota trabajan en el desarrollo del Proyecto Signa, una innovadora propuesta tecnológica enfocada en la creación de un traductor de lenguaje de señas.
El proyecto, actualmente en su versión beta, consiste en una plataforma tipo curso interactivo que traduce al Lenguaje de Señas Mexicano (LSM). A diferencia de las plataformas tradicionales que traducen entre idiomas orales, Signa está diseñada para facilitar la comprensión y el aprendizaje del lenguaje de señas, buscando impactar positivamente en la vida de personas con discapacidad auditiva.
“El proyecto se llama Signa, es un traductor de lenguaje de señas que actualmente se encuentra en su versión beta. Desarrollamos una plataforma tipo curso interactivo que traduce al lenguaje de señas en lugar de a otro idioma”, explican sus creadores.
Actualmente consiste en una plataforma web, pero en el futuro los jóvenes quieren implementar una versión también en aplicación móvil. “Funciona mediante clases para los usuarios que gusten aprender la lengua de señas, poniéndoles actividades dependiendo el nivel que elijan.
De momento, el sistema inicia con tres cursos básicos, que son “Abecedario”, “Alimentos” y “Animales”. Estos módulos buscan introducir al usuario en los fundamentos del LSM a través de animaciones e interfaces accesibles. Además, presenta un sistema de evaluación para determinar si han aprendido los cursos.
Otra mejora en la que se encuentran trabajando es en implementar el reconocimiento de voz, para que mediante un creador de imágenes con Inteligencia Artificial pueda generar en automático la seña de lo que la persona diga. Para ello, detallan, tuvieron un acercamiento con un ingeniero de Google, quien les dio consejos y opciones para implementar esta tecnología a su programa.
“Cuando observamos los noticieros en televisión, es común ver a una persona en un recuadro pequeño interpretando el contenido en lenguaje de señas. Nuestro propósito es innovar este proceso incorporando tecnología; buscamos que, cada vez que una persona esté hablando, aparezca una animación generada con inteligencia artificial que traduzca simultáneamente a lenguaje de señas”.
Méndez Merino y Aguilar Cota señalaron que uno de los principales propósitos es generar empatía y reducir la discriminación que enfrentan las personas sordas en los entornos laborales y educativos.
“Lo que buscamos es generar empatía hacia las personas que utilizan el lenguaje de señas, a fin de disminuir la discriminación que enfrentan en los entornos laborales y educativos. Queremos invitar a la comunidad a mostrarse más abierta e incluyente con las personas usuarias del lenguaje de señas”.
Para garantizar precisión y confiabilidad, los estudiantes obtuvieron información sobre LSM a partir de fuentes oficiales del gobierno, recordando que cada país cuenta con su propio lenguaje de señas.
De esta manera, los jóvenes universitarios buscan romper la barrera de comunicación que enfrentan las personas con discapacidad auditiva, promoviendo mayor accesibilidad, comprensión y participación en los distintos ámbitos de la vida cotidiana."
1 de enero de 2026
NBCS Noticias
https://nbcs.mx/?p=159640
#Metaglossia
#metaglossia_mundus
"Google has started rolling out a new version of its Translate app with a feature that allows you to create more accurate Gemini AI-assisted translations, 9to5Google reported. The feature appears as an AI model picker at the top of the app, allowing you to choose between "Fast" and "Advanced" translations. It's appeared for some users on iOS but not Android to date, and the Advanced mode only translates between English and French, and English and Spanish.
Google Translate now offers Gemini-assisted translations https://www.engadget.com/ai/google-translate-now-offers-gemini-assisted-translations-151008774.html
#Metaglossia
#metaglossia_mundus
"On December 30, 2025, Tencent Hunyuan announced the open-sourcing of Hunyuan Translation Model 1.5 (HY-MT1.5), available in 1.8B and 7B variants. Supporting mutual translation across 33 languages and 5 ethnic/regional dialect conversions, the models are now live on the Hunyuan website, GitHub, and Hugging Face for direct developer access.
Key highlights:
HY-MT1.5-1.8B targets edge deployment: quantized version runs smoothly on 1GB memory, enables offline real-time translation with average 50-token processing in 0.18 seconds (faster than major commercial APIs), matching most commercial tools—ideal for instant messaging and smart customer service.
HY-MT1.5-7B upgrades the WMT25 30-language champion model, improving accuracy and reducing annotation carryover and language mixing for professional scenarios.
Both models support edge-cloud hybrid deployment and three practical features: custom terminology libraries for fields like medicine and law, context understanding for coherent long-text translation, and format-preserving translation for webpages and documents. Built with large-model distillation, they deliver high performance in small sizes, already deployed in Tencent Meeting and Enterprise WeChat, with compatibility for Arm, Qualcomm, Muxi, and other platforms.
Source: IT Home"
https://pandaily.com/tencent-hunyuan-open-sources-translation-model-1-5-runs-on-1-gb-phone-memory-outperforms-commercial-ap-is
#Metaglossia
#metaglossia_mundus
"Building a national Large Language Model: Opportunities and challenges
Imagine interacting with a virtual assistant that understands you perfectly – whether you’re speaking fluent Arabic, simplified English, or a natural mix of both. Within seconds, it delivers a response that isn’t just accurate but also culturally attuned, recognizing regional nuances and context. This is the promise of AI agents, intelligent digital assistants powered by cutting-edge Large Language Models (LLMs).
LLMs are transformative AI systems capable of understanding and generating human-like language. They serve as the brains behind AI agents, enabling them to process complex queries, adapt to different dialects, and engage in meaningful conversations. Whether assisting with customer service, translating in real-time, or offering legal and financial guidance, these agents are redefining human-computer interactions.
While models of private-sector leaders like OpenAI’s GPT 4.5, Google’s Gemini 2.0 and DeepSeek R1 lead the global AI race and more advanced AI agents are evolving (such as Manus AI), a new movement is taking shape: national LLMs. Countries are developing their own AI models tailored to their linguistic, cultural and economic realities.
These homegrown models not only ensure greater data sovereignty but also provide AI systems that truly understand local dialects, traditions and societal values – bridging gaps left by global models and shaping the future of AI on their own terms.
For the UAE and the broader Middle East, the potential of Large Language Models (LLMs) extends far beyond technological innovation. They represent a strategic opportunity to drive economic growth, enhance digital sovereignty, democratize AI access and foster inclusivity. While developing an advanced LLM comes with technical complexities, it also involves key economic decisions, workforce development and long-term strategic planning.
What gives the UAE an edge is its strong political determination, investment capacity and rapidly growing AI ecosystem. With a clear vision to lead in AI, the country is not just adopting global models but actively shaping the future of LLMs through the development of models like Jais (a 13-billion parameter model trained on a dataset containing 395 billion tokens of Arabic and English data).
By building models that understand the region’s linguistic diversity, cultural nuances and economic needs, the UAE can create AI solutions that are not only locally relevant but also globally competitive—setting a new benchmark for AI in the Arabic-speaking world and beyond.
The case for a national LLM
The demand for LLMs is undeniable, with applications spanning customer service, education, healthcare, coding, and even governance. Globally, over 30 nations are developing their own LLMs, embedding their unique linguistic and cultural preferences into AI systems. That’s not just about national pride but also about developing independent capabilities.
National LLMs allow countries to reduce reliance on foreign models by ensuring data sovereignty and security as well as full control over model parameters and biases.
Significant steps in this direction have already been taken with Jais in the UAE, one of the most advanced Arabic LLMs, demonstrating the country’s commitment to AI leadership and laying the groundwork for future innovations.
Arabic is a rich and complex language, infused with layers of dialects depending on the region and social context. Often, global models struggle with these intricacies. LLMs developed locally could further bridge this gap, fostering AI systems that understand and adapt to Emirati nuances while extending this capability to the wider Arabic-speaking world.
The region’s bilingualism – most prominently the interplay between Arabic and English – creates yet another incentive. Think of a business traveler at Dubai International Airport or a student accessing academic materials; an LLM that intuitively understands the cultural and linguistic expectations in such scenarios could transform user experiences. By enabling these services locally, the UAE can reduce reliance on foreign-produced AI systems, keeping data safe within national borders.
Beyond immediate use cases, a successful national LLM enhances the UAE’s competitiveness in the global AI race. It positions Emirati institutions as essential players in shaping the future of AI, fostering not just innovation but economic growth through AI-enabled industries.
The economic upside
Building a national LLM also becomes an economic catalyst beyond a technological milestone. Having an LLM available locally means government agencies, universities, and private enterprises can integrate advanced language understanding into their systems. This reduces time, effort, and costs previously spent outsourcing or adapting foreign tools.
But the economic benefits extend far beyond ease of use. A national LLM becomes the foundation for an ecosystem of innovation. It empowers local tech startups to build applications on top of the language model – from chatbots and translation aids to advanced analytics tools and agents. This boosts local entrepreneurship and economic diversification, a critical pillar of the UAE’s long-term development strategy.
National LLMs also create jobs, not only in the fields of AI model development but also in linguistics, ethics, and policymaking. Expanding regional AI expertise fosters human capital, ensuring that regional talents lead the way, rather than relying on external specialists.
The challenges
At the same time, developing an LLM is a high-stakes endeavor. Building state-of-the-art AI systems requires vast amounts of high-quality data, computing power, and financial investment.
Data availability and quality are arguably the most pressing challenges. Arabic dialects vary drastically not just between countries like Morocco and Iraq but even intra-nationally. Sourcing digitized, representative datasets for these diverse linguistic forms is daunting. Large portions of Arabic text available online – social media posts and blogs – might not meet the quality standards required for AI training.
Resources are another challenge. Training LLMs requires massive computing power and high-performance GPUs. The UAE’s growing investments in cloud infrastructure provide a strong foundation, while efficient approaches like DeepSeek’s offer a path to more affordable models. But technology alone isn’t enough – a national LLM holds value only if widely adopted. Without strong buy-in from government agencies, businesses, and educators, it risks becoming an underutilized investment.
Lastly, the talent question is critical. The UAE has made significant strides in cultivating AI-centric education, but the reality of a global AI talent shortage remains. Attracting world-class experts while simultaneously developing homegrown capabilities will be essential to leading the way in the future.
For the UAE and regional players, the choice to invest in a national LLM must be guided by purpose. Does everyone build from scratch, like GPT-NL in the Netherlands? Or do we layer Arabic-language capabilities onto an existing foundational model (RAG-layering), a cost-effective approach?
Each decision has trade-offs. Building a new foundational model offers exceptional contextualization but requires ongoing financial commitment for updates and maintenance. RAG-layering limits the scope of flexibility but ensures faster rollouts and lower initial costs.
Stakeholders must also decide how to prioritize applications. Should early focus be on public sector adoption, such as using LLMs in smart government services? Or should innovation centers work closely with private enterprises looking to solve commercial challenges? These choices will define the trajectory of the UAE’s AI ambitions.
Conclusion
The drive to build a national LLM isn’t merely a technical pursuit; it’s a strategic response to a rapidly shifting world. For the UAE and its neighbors, the opportunity lies not just in offering localized capabilities but in galvanizing economic innovation through AI tools that resonate globally and regionally.
The challenges, no doubt, are significant. But with visionary leadership, strategic public-private partnerships, and a commitment to developing both talent and infrastructure, the UAE can emerge not just as a participant but as a pioneer in the global AI landscape.
An article from Javier Alvarez and Fred Liebler. Both are senior members of FTI Consulting."
31 December 2025 Consultancy-me.com
https://www.consultancy-me.com/news/12262/building-a-national-large-language-model-opportunities-and-challenges
#Metaglossia
#metaglossia_mundus
"Forget Keanu: Ulster Scots translation of Beckett classic takes on spate of celebrity Godots
Tragicomedy will be performed outdoors in Northern Irish countryside as part of new festival celebrating Irish playwright
Beneath a stark steel tree in a bleak upland bog, a literary masterpiece is set to assume a different linguistic mantle.
Samuel Beckett’s enigmatic tragicomedy Waiting for Godot will make its world premiere in Ulster Scots, a moment described as a “coming of age” for the minority language, and the antithesis of the trend for celebrity Godots.
On Good Friday, after an uphill trek of about 3km, the audience will arrive at a spot in the vast volcanic Antrim Plateau in Northern Ireland, if not footsore then certainly empathic to the physical discomfort of Estragon struggling to remove his ill-fitting boots.
The “existential landscape of heath, moss and bog” in County Antrim lends itself to a script “peppered with exterior references”, said Seán Doran, of festival organiser Arts over Borders, which is staging the production as part of a major new arts festival, the Samuel Beckett Biennale.
But while there have been previous outdoor productions, it will be the “forceful pronunciation and sound” of delivering it in Ulster Scots, or Ullans, for the first time and in a region where the language is spoken, that will “bring a whole new total register” and change the whole performative aspect of the play, said Doran.
In October, a commissioner for Ulster-Scots was appointed under the Identity and Language Act in Northern Ireland to act as cultural safeguard for the language, which has its roots in the early 17th -century plantations of Scots speakers to the north of Ireland.
Against this backdrop, Frank Ferguson, who is translating the play, hailed its performance as “a major coming of age moment”. “It shows a confidence in what Ulster-Scots can do as a language, because you take on one of the major global dramatic phenomenon and you place it within its Ulster-Scots translation.”
The working title is Ettlin Fur Godot, and its famous stage directions of “A country road. A tree. Evening” will translate as “A loanen. A tree. Dailygan”. Ferguson, research director of the Centre for Irish and Scottish Studies at Ulster University, suggested that audiences may develop an ear for Ulster Scots, which has many words with similar roots to English – but that translations might be provided.
Ferguson considers it a language, not dialect, one that in the wake of the Good Friday agreement is “discovering itself and trying to find its way in the world”. And it works “beautifully” in the Godot setting, he said.
Not least because of “ the sense of waiting and hoping and longing for something; all minority languages are longing for that sort of moment of salvation, that moment of revelation. So, that looking, and hoping, and wishing for a Godot-like figure or moment works very well I think with Ulster Scots because in a sense it’s waiting for its moment to live and find itself again”.
It will be performed on Good Friday, 3 April 2026 (Beckett was born on a Good Friday), and is part of a new Beckett Biennale, which over the next 10 years will experiment with unexpected approaches including translations in Aboriginal Noongar, Sami and Inuit , and productions starring homeless actors.
Keanu Reeves (right) appearing as Estragon in Waiting for Godot, is the latest celebrity to star in the play. Photograph: Andy Henderson/AP
Arts Over Borders, producers of cross-border north-south arts festivals, aims to bring Godot back to its original roots when it premiered in French in Paris in 1953 and in London and Dublin two years later. The biennale aims to be the antithesis of the current trend for productions starring big names.
Keanu Reeves is the latest Hollywood star, currently performing on Broadway; double-acts such as Patrick Stewart and Ian McKellen, Bill Paterson and Brian Cox, Ben Whishaw and Lucian Msamati, and Robin Williams and Steve Martin are others who have risen to its challenges.
Doran appreciates “celebrity Godot” as a vehicle that spreads the word more effectively than anything else. But he believes it can detract from other perspectives, possibilities and insights. “And that’s clearly what we’re trying to do through the different languages, the outdoor setting, the homeless actors.”
The Samuel Beckett Biennale will be set in rural and urban settings in Northern Ireland, the Republic of Ireland and England in 2026, returning in 2028.
This article was amended on 2 January 2026."
https://www.theguardian.com/culture/2026/jan/01/samuel-beckett-waiting-for-godot-new-ulster-scots-translation
#Metaglossia
#metaglossia_mundus
We are building a world that treats translation as a problem to be solved. But translation was never just a technical challenge — it is an act of witnessing.
"When an interpreter breaks, they are not breaking down. They are breaking open — making room for unbearable truth to enter, and for all of us to see it.
Flynn Coleman is an international human rights attorney. She is a visiting scholar in the Women, Peace, and Leadership Program at Columbia University’s Climate School and the author of “A Human Algorithm.”
Roman Oleksiv was 11 years old when he stood before the European Parliament and, in a calm voice, described the last time he saw his mother. She was under the rubble of a hospital in Vinnytsia, Ukraine, hit by Russian missiles in July 2022. He could see her hair beneath the stone. He touched it. He said goodbye.
That’s when Ievgeniia Razumkova, the interpreter translating his words, stopped mid-sentence. Her eyes filled with tears, she shook her head. “Sorry,” she said. “I’m a bit emotional as well.”
A colleague then stepped in to finish, as Ievgeniia, still crying, placed her hand on the boy’s shoulder. He nodded and continued on.
That moment is what makes us human.
A translation algorithm would not have stopped. It would have rendered Roman’s testimony with perfect fluency and zero hesitation. It would have delivered the words “the last time I saw my mother” just as it would the sentence “hello, my name is Roman.” Same tone. Same rhythm. No recognition.
Today, we are building a world that treats translation — and increasingly everything else — as a problem to be solved. Translation apps now handle billions of words a day. Real-time tools let tourists order coffee in any language. Babel, we are told, is finally being fixed.
All of this has its place. But translation was never just a technical challenge. It is an act of witnessing.
An interpreter does not merely convert words from one language to another. They carry meaning across the chasm between us. They hear what silences say. They make split-second ethical and semantic decisions over which synonym preserves dignity, when a pause holds more truth than a sentence, whether to soften a phrase that would shatter a survivor.
When Ievgeniia broke down in Brussels, she was not failing. She was doing her job. Her face told a room full of diplomats what no algorithm could: “This matters. This child’s suffering is real. Pay attention.”
I have spent years working in international human rights law, war crimes tribunals, genocide prevention — all the imperfect architecture we try to rebuild after atrocity. In these spaces, everything hinges on language. One word can determine whether a survivor is believed. The difference between “I saw” and “I was made to see,” or between “they did this” and “this happened.”
Roman Oleksiv has undergone 36 surgeries. Burns cover nearly half his body. He was 7 years old when that missile hit. And when he described touching his dead mother’s hair, he needed someone in that room who could hold the weight of what he was saying — not just linguistically but humanly. Ievgeniia did that. And when she could not continue, another person stepped forward.
There is a reason interpreters in trauma proceedings receive psychological support. The best ones describe their work as a sacred burden. They absorb something. They metabolize horror, so it can cross from one language to another without losing its force.
Interpreters are not alone in this either. There are moments when trauma surgeons pause before delivering devastating news, journalists choose to lower their cameras, and judges listen longer than procedure requires. These are professions where humanity is not a flaw — it is the point.
This is not inefficiency. It is care made visible.
Algorithms process language as pattern, not communion. They have no understanding that another mind exists. They do not know that when Roman said goodbye, he was not describing a social gesture — he was performing the final ritual of love he would ever share with his mother, in the rubble of a hospital.
Translation apps do serve real purposes, and generative AI is becoming more proficient every day. But we should be honest about the trade we are making. When we treat human interpreters — and any human act of care — as inefficiencies to be optimized away, we lose that pause before “the last time I saw my mother.” We lose the hand on the shoulder. We lose the tears that say: “This child is not a data point. What happened to him is an atrocity.”
My work studying crimes against humanity has taught me that some frictions should not be smoothed. Some pauses are how we recognize one another as human. They are echoes in the dark, asking: “I am still here. Are you?”
When an interpreter breaks, they are not breaking down. They are breaking open — making room for unbearable truth to enter, and for all of us to see it.
Roman deserved someone who could help us stand in his deepest pain, so that we might all lift it together.
A machine could not do that. A machine, by design, does not stop."
January 1, 2026 4:00 am CET
By Flynn Coleman
https://www.politico.eu/article/when-the-interpreter-wept-what-automation-erases-inside-europes-institutions/
#Metaglossia
#metaglossia_mundus
"More than spelling: why language & identity are the heart of Bill 218
Ensuring language consistency, Senator Shelly Calvo introducing Bill 218, legislation aimed at standardizing the official use of the spelling “Chamoru” in Guam law and government documents.
Calvo noting that the government currently uses three different spellings to refer to the same indigenous people and language—an inconsistency she says is unusual in government practice and creates unnecessary confusion in public records, education, and policy.
The senator emphasized that the effort behind the bill is not new.
Discussions on orthography, pronunciation, and cultural accuracy have been underway for years among educators, language experts, and cultural institutions. She said the bill represents a continuation of that work—not its beginning."
Friday, January 2nd 2026, 2:17 PM ChST
By Destiny Cruz-Langas
https://www.kuam.com/story/53350045/more-than-spelling-why-language-and-identity-are-the-heart-of-bill-218
#Metaglossia
#metaglossia_mundus
"New York Gov. Kathy Hochul this month signed a bill requiring hospitals to provide language assistance services to patients. Translators who work with refugee and immigrant communities say translation services currently available can sometimes be confusing and unhelpful to patients.
WAMC’s Sajina Shrestha spoke to Yousaf Sherzad, a lead translator at Advanced Translations Services in Albany, about the new law and how existing translation services can be improved."
WAMC Northeast Public Radio | By Sajina Shrestha
Published December 31, 2025 at 1:15 PM EST
https://www.wamc.org/news/2025-12-31/translators-hope-new-nys-bill-will-improve-on-existing-translation-services
#Metaglossia
#metaglossia_mundus
System 1 thinking is a near-instantaneous thinking process while System 2 thinking is slower and requires more effort.
Where do translators/interpreters fit? System 1 or/and System 2 Thinking?
"System 1 and System 2 thinking describes two distinct modes of cognitive processing introduced by Daniel Kahneman in his book Thinking, Fast and Slow. System 1 is fast, automatic, and intuitive, operating with little to no effort. This mode of thinking allows us to make quick decisions and judgments based on patterns and experiences. In contrast, System 2 is slow, deliberate, and conscious, requiring intentional effort. This type of thinking is used for complex problem-solving and analytical tasks where more thought and consideration are necessary...
The Basic Idea
When commuting to work, you always know which route to take without having to consciously think about it. You automatically walk to the subway station, habitually get off at the same stop, and walk to your office while your mind wanders. It’s effortless. However, the subway line is down today.
While your route to the subway station was intuitive, you now find yourself spending some time analyzing alternative routes to work in order to take the quickest one. Are the buses running? Is it too cold outside to walk? How much does a rideshare cost?
Our responses to these two scenarios demonstrate the differences between our instantaneous System 1 thinking and our slower, more deliberate System 2 thinking.
However, even when we think that we are being rational in our decisions, our System 1 beliefs and biases still drive many of our choices. Understanding the interplay of these two systems in our daily lives can help us become more aware of the bias in our decisions—and how we can avoid it.
“The automatic operations of System 1 generate surprisingly complex patterns of ideas, but only the slower System 2 can construct thoughts in an orderly series of steps.
Key Terms
System 1 Thinking: Our brains’ fast, automatic, unconscious, and emotional response to situations and stimuli. This can be in the form of absentmindedly reading text on a billboard, knowing how to tie your shoelaces without a second thought, or instinctively hopping over a puddle on the sidewalk.
System 2 Thinking: The slow, effortful, and logical mode in which our brains operate when solving more complicated problems. For example, System 2 thinking is used when looking for a friend in a crowd, parking your vehicle in a tight space, or determining the quality-to-value ratio of your take-out lunch.
Automatic Thinking: An unconscious and instinctive process of human thinking. This term can be used interchangeably with System 1 thinking.
Reasoning: Consciously using existing information to logically make a decision or reach a conclusion, a key feature of System 2 thinking.
Dual Process Model: A theory in psychology that distinguishes two thought processes in humans by describing them as unconscious and conscious, respectively."
System 1 and System 2 Thinking - The Decision Lab https://share.google/tK9xREA3ne7hVjx4X
#Metaglossia
#metaglossia_mundus
La langue et la terre sont indissociables
Une langue ancienne éclaire l’adaptation climatique dans la région la plus chaude des États-Unis
Voir plus
Il y a deux ans, une source naturelle à Pueblo Zunile Nouveau-Mexique, l'une des régions les plus sèches des États-Unis, a commencé à disparaître. C'était une source que de nombreuses personnes, dont Jim Enote, un membre de la tribu Zuni, utilisaient pour arroser leurs champs et leurs jardins. Enote avait cultivé la région pendant 68 années consécutives, alors lorsque ses voisins ont tenté de rétablir l'approvisionnement en eau crucial, il les a rejoints.
Malgré tous leurs efforts, la source s'est tarie, victime du pire mégasécheresse vécue dans le Sud-Ouest depuis plus d'un millénaire. Pourtant, Enote pensait qu'il devait y avoir un moyen d'étancher la soif de sa communauté. Enfant, il avait appris les noms traditionnels des lieux locaux auprès de son grand-père. Et il se souvint qu'une zone située à seulement 30 mètres de l'un de ses champs portait un nom Zuni qui faisait référence à l'eau.
L'endroit était couvert d'herbe sèche, mais Enote décida de commencer à creuser. Moins d’un pied plus bas, il a heurté de l’eau stagnante. « Je viens d'un endroit où la terre parle une langue plus ancienne que n'importe quelle nation, où les noms des montagnes, des sources et des canyons sont des instructions, des histoires et des souvenirs », a déclaré Enote. « En tant que Zuni et agriculteur, je sais que la langue et la terre sont indissociables. »
Enote pratique une méthode autochtone de culture de fruits et de légumes appelée jardinage de gaufres. Il s’agit de créer des dépressions en forme de blocs bordées de murs en terre, formant une grille qui maximise la conservation de l’eau. Il a transporté cette eau de source nouvellement découverte, « avec précaution avec chaque goutte », dans son jardin.
Bientôt, Enote remarqua un wapiti visitant son point d'eau. Des asclépiades poussaient autour. Des libellules et des papillons monarques sont apparus. Aujourd'hui, son jardin n'est pas le seul à prospérer ; le paysage environnant aussi. Enote pense que ce type « d’adaptation climatique éprouvée » a aidé ses ancêtres à survivre à des cycles de sécheresse extrêmes, comme en témoignent les anciens tessons de poterie qu’il a trouvés autour de la source. « La langue est le point de rencontre de la terre et de la mémoire, et ce respect n'est pas de la nostalgie, c'est une responsabilité. Chaque nom de lieu dans nos langues est une leçon de survie et de respect », a-t-il déclaré.
Le consensus parmi les climatologues est clair : le Sud-Ouest connaît depuis un quart de siècle la période la plus sèche de la région depuis 1 200 ans. En conséquence, les niveaux d’eau des lacs Mead et Powell, les plus grands réservoirs du pays, sont tombés à des niveaux historiquement bas. Des méga-incendies, tels que l'incendie de Calf Canyon-Hermit's Peak au Nouveau-Mexique en 2022, ont fait rage dans la région, intensifiés par les conditions sèches. Et les responsables du Nouveau-Mexique rapportent que l'État a perdu jusqu'à 52 pour cent de ses zones humides.
En plus de ces pressions, les aquifères locaux continuent d'être épuisés par l'expansion de villes telles que Phénix et Tucson, tandis que le changement climatique réduit la fonte des neiges dans les montagnes, une source clé de reconstitution des eaux souterraines. Les fermes, les plus grands utilisateurs d’eau en Arizona et au Nouveau-Mexique, ont été contraintes de laisser leurs champs en jachère.
Mais de nombreux peuples autochtones de la région savent que cela n’a rien de nouveau. Les traditions tribales, les pétroglyphes et les études scientifiques, y compris les données sur les cernes des arbres, racontent que les Puebloans ancestraux ont subi plusieurs sécheresses extrêmes, dont une qui a conduit à l'abandon de colonies majeures. Ces ancêtres ont transmis des connaissances et des outils qui ont permis à leurs descendants de résister aux épreuves climatiques d’aujourd’hui.
« Les connaissances traditionnelles autochtones ont une vision du climat. La science du climat ne devrait donc pas seulement inclure les connaissances autochtones, mais commencer par les considérer comme une base pour comprendre la façon dont le monde change », a déclaré Enote. « Il ne s'agit pas seulement de résilience climatique. J'utilise le terme respect du climat passer du langage technique au langage relationnel. De cette façon, nous honorons le climat et la terre comme des éléments apparentés et non comme des marchandises. »
Enote estime que les connaissances autochtones pourraient contribuer à atténuer la crise de l'eau et à promouvoir la durabilité dans le Sud-Ouest, où de plus en plus de sources et de rivières s'assèchent, notamment la rivière Jemez, à l'est de Zuni Pueblo. Imaginez, dit Enote, si les tribus étaient consultées sur l'emplacement des sources historiques pour éviter qu'elles ne soient détruites par le développement.
Enote, qui est également fondateur et directeur exécutif de la Colorado Plateau Foundation, qui aide à financer la conservation et la préservation culturelle menées par les autochtones dans cette région désertique, a déclaré que les conditions de méga-sécheresse catalysent les solutions intertribales. Le Flower Hill Institute, basé à Jemez Pueblo, est l'un des groupes soutenus par la fondation. En mai, il a lancé un programme visant à honorer les agriculteurs Pueblo et à promouvoir une combinaison de pratiques traditionnelles et modernes, notamment pour la conservation de l'eau. Michael Kotutwa Johnson, agriculteur Hopi des zones arides et expert en ressources naturelles, a qualifié le rassemblement d'« événement historique », étant donné que les tribus Pueblo « n'ont pas été unies autour d'une question clé comme celle-ci depuis la révolte des Pueblo de 1680 ».
Bryn Fragua, coordinateur agricole, technique et tribal de l'institut, a déclaré qu'il était puissant d'entendre différentes langues Puebloan parlées « en harmonie » alors que les dirigeants et d'autres membres de la communauté faisaient des offrandes et des prières pour le travail sur le point d'être accompli. Les jours précédant et pendant l’événement, le temps était venteux, poussiéreux, sec et chaud. « Après cette réunion, il semblait que ces paroles prononcées et ces prières faites nous apportaient des nuages de pluie », a déclaré Fragua. « Depuis, il pleut. »
« Dans nos langues, le climat n'est pas quelque chose d'extérieur », a déclaré Enote. « C'est le souffle du vent, le moment de la fonte des neiges et le silence avant les moussons d'été. Le plateau du Colorado nous enseigne que le respect n'est pas passif. C'est un soin actif, c'est une cérémonie et c'est choisir d'écouter. »"
Nicolas Guillot
27.12.2025 à 21h23
https://www.especes-menacees.fr/actualites/exploiter-les-connaissances-autochtones-peut-nous-aider-a-naviguer-dans-un-climat-changeant/
#Metaglossia
#metaglossia_mundus
"Redefining AI for underserved cultures and languages
THE rapid evolution of large language models has achieved grammatical correctness across major languages, yet this represents marketing inclusivity rather than true linguistic equality. While today’s AI can construct sentences in Tagalog, Swahili or Mongolian, the output remains inconsistent and inferior to English performance. Worse, the proliferation of wrapper applications without proper localization amplifies these deficiencies, potentially causing more harm than having no AI support at all.
Our experience building Egune AI in Mongolia revealed that perfect grammar is merely the starting point. The deeper challenge is not making AI speak every language; it is making AI think in every culture. When children grow up interacting with AI that fundamentally operates in English thought patterns, they gradually adopt linear, analytical reasoning, losing the circular, contextual cognition that characterizes many Asian and Indigenous languages. This represents cognitive homogenization disguised as technological progress.
The implications extend far beyond individual users. When governments process documents through foreign AI systems, health care providers rely on AI trained on different medical traditions, and educational institutions deploy AI that misunderstands local pedagogical approaches, entire nations become digitally dependent. They participate in the AI revolution as data providers rather than value creators, feeding sensitive information into systems that extract economic benefit without proportional return.
Digital sovereignty has become essential for nations serious about their technological future. This transcends data localization requirements, though those are important. It is about controlling the AI systems that increasingly mediate citizen interactions with essential services. When critical infrastructure depends on foreign entities subject to external laws and priorities, true independence becomes impossible.
Building AI for languages with limited digital presence requires fundamental rethinking. Mongolian, spoken by three million people, represents less than 0.01 percent of internet content. Web scraping yields mostly translated content reflecting foreign thought patterns rather than authentic expression. We discovered that a million translated sentences teach less about genuine Mongolian thinking than a thousand authentic native conversations.
Our solution involved building dedicated data teams to create quality datasets. We digitized books, transcribed audio and video content, collected academic writing, and manually cleaned and corrected everything. While major AI systems train on trillions of tokens of varying quality, we focused on millions of carefully curated, culturally authentic examples. Every piece underwent native-speaker verification for natural expression. This approach took longer and cost more, but the results justified the investment: our AI understands not just Mongolian words but Mongolian thinking patterns.
Sovereign AI development enables capabilities impossible with foreign systems. During Mongolia’s harsh winters, we immediately update models with emergency protocols. When legislation changes, government agencies integrate updates within days rather than months. This agility becomes impossible when dependent on global release cycles prioritizing larger markets.
Local development ensures consistent alignment with national priorities. Security vulnerabilities specific to local infrastructure receive immediate attention. Critical bugs affecting local users receive priority treatment rather than languishing in global backlogs. Most importantly, economic value from AI development remains domestic, creating jobs and building expertise rather than extracting value abroad.
Our recent launch of Egune Chat on iOS and Android platforms demonstrates practical sovereign AI implementation. Users select their geographic region, accessing AI trained on specific cultural and administrative contexts. Rural Mongolians receive responses that consider traditional practices and limited infrastructure. Urban professionals get advice relevant to modern city resources. This is not translation; it is fundamentally different training for different realities...
Building national AI capabilities no longer requires competing with trillion-parameter models. Focused, culturally optimized systems serve specific populations better than generic global solutions while requiring a fraction of the resources. A $50 million investment in sovereign AI creates more local value than $500 million spent on foreign AI subscriptions.
International collaboration need not conflict with digital sovereignty. Nations can share open-source frameworks and methodologies while maintaining control over implementations. Mongolia’s experience can inform efforts in the Philippines, while Filipino innovations in multi-dialect processing might benefit Indonesia. This creates a network of diverse, interoperable AI systems rather than monolithic global platforms.
The future of inclusive AI requires recognizing that each culture offers unique problem-solving patterns worth preserving in the digital age. True inclusivity demands AI systems that understand local context, respect sovereignty, and serve actual population needs rather than assuming universal solutions exist..."
By Badral Sanlag
December 28, 2025
https://www.manilatimes.net/2025/12/28/business/sunday-business-it/inclusive-ai-from-the-ground-up-redefining-ai-for-underserved-cultures-and-languages/2249949
#Metaglossia
#metaglossia_mundus
"Vivre l’Afrique au Maroc: quand le football devient langage africain
Lahcen Haddad.
ChroniqueLa CAN version Maroc 2025 a commencé bien avant le sifflet du premier match opposant le Maroc aux îles Comores. On la sent dans l’air, dans les rues qui se parent de couleurs, dans les drapeaux africains éclatants qui embellissent les carrefours, dans les accents et les langues africaines qui se croisent partout. On la lit aussi sur les maillots: des jerseys africains riches de traditions, de motifs, d’histoires et de savoir-faire millénaires.
Par Lahcen Haddad
Le 29/12/2025 à 16h00
L’Afrique est rassemblée sans sommet, sans tribune officielle, sans discours. Elle s’assemble par l’émotion. Le football est sans doute la plus grande sensation commune partagée par tous les Africains.
Mais les Africains ont, depuis longtemps, fait parler le football à leur manière. Le football est une langue africaine à part entière. Les rythmes, l’improvisation, l’intensité physique, l’intelligence collective en constituent la grammaire commune, partagée entre les Africains. Chaque match se lit comme un texte, comme un palimpseste: avec son intertextualité, sa propre rhétorique, ses signifiants. Le style de chaque équipe révèle son identité. Le football africain est reconnaissable au-delà des résultats. Les stades deviennent ainsi des espaces de production de sens culturel et interculturel. Et l’interculturel, en Afrique, n’a jamais été une exception: il a toujours été une évidence. L’Afrique parle et se parle via le jeu, à travers le football.
Mais l’organisation de la CAN au Royaume chérifien, terre de baraka et d’empires mythiques, n’est pas un simple exercice de soft power. C’est une expérience vécue, célébrée, chantée et dansée par des foules venues de tous les coins d’une Afrique fière de son africanité. L’explication par le «soft power» ne rend pas justice à la réalité de ce qui se joue. Le soft power est un regard externe, une lecture exogène — une forme d’objectification, au sens lacanien du terme.
La CAN relève au contraire d’un regard interne: une reconnaissance de soi, de ses capacités, de ses rêves et de ses ambitions. Son audience n’est pas d’abord le monde, mais les Africains eux-mêmes. Le football y devient une connexion horizontale, traversant espaces, communautés et sociétés; il n’est pas une projection, mais une présence. Les Africains ne jouent pas pour être regardés: ils jouent pour se reconnaître.
Reconnu comme l’un des pays les plus hospitaliers au monde, le Maroc se distingue en Afrique non pas comme un simple carrefour, mais comme un véritable centre de l’africanité. Son ouverture, son soutien constant et son sens profond de l’identité sont forgés par une longue histoire, depuis les caravanes des IXème et Xème siècles qui transportaient non seulement l’or et le sel, mais aussi des cultures, des religions, des vivres et un art de vivre partagé.
«le football opère comme un espace de décolonisation symbolique. Il traverse les langues imposées, court-circuite les bureaucraties postcoloniales et ignore les cartographies héritées»
— Lahcen Haddad
Dans ce contexte, le stade devient une demeure continentale: la culture marocaine y joue le rôle de pont, de trait d’union au sens le plus authentique du terme. Il n’y a pas ici de hiérarchie des identités africaines, ni d’ethnocentrisme hérité du regard occidental, mais une appartenance horizontale, multiple, multiculturelle et plurielle. Être hôte devient alors un acte de reconnaissance envers la terre africaine et sa mosaïque de géographies symboliques, de mémoires croisées et de récits partagés.
La condition postcoloniale de l’Afrique n’a pas seulement fragmenté les territoires; elle a désaccordé les sensibilités politiques, désarticulé les temporalités et installé une dissonance durable entre les peuples et les formes héritées de l’État. Les frontières tracées à la règle par l’histoire coloniale ont laissé des cicatrices mémorielles, des blocages politiques et des blessures existentielles encore actives.
Face à ces lignes imposées, le football opère comme un espace de décolonisation symbolique. Il traverse les langues imposées, court-circuite les bureaucraties postcoloniales et ignore les cartographies héritées. Il ne reconnaît ni centre métropolitain ni périphérie subalterne. L’unité africaine ne se proclame pas dans les sommets ni les communiqués technocratiques: elle se fabrique par le bas, dans les tribunes, dans les corps en mouvement, dans les chants partagés. C’est là que s’exprime une fierté nationale décolonisée — ni folklorisée, ni mimétique du nationalisme occidental, mais vécue, située, charnelle.
L’Afrique se rassemble davantage dans le chant que dans le discours, dans le rythme que dans l’institution, dans le jeu que dans la norme. Cependant, il ne faut pas idéaliser cette unité: elle demeure temporaire, imparfaite, parfois même contradictoire. Mais c’est précisément parce qu’elle est vécue qu’elle est vraie. La puissance de l’instant peut, parfois, dépasser celle de la permanence. Nous sommes ici face à un moment postmoderne, presque uncanny: un moment qui ouvre un horizon de rêve, même s’il est éphémère, fragile et voué à disparaître.
Bien sûr, la mémoire demeure. Elle s’attache aux smartphones autant qu’à l’imaginaire collectif. Mémoire des moments vécus, parfois au point de se substituer à la chose elle-même — non comme falsification, mais comme refiguration du réel. Ce jeu de la mémoire — fait de buts devenus anthologiques, comme celui d’Ayoub El Kaabi, d’efforts titanesques portés par des équipes entières, de gestes justes et de prouesses incarnées par de grands joueurs tels que Riyad Mahrez, Victor Osimhen, Omar Marmoush, Ademola Lookman ou Achraf Hakimi — s’inscrit durablement dans la conscience des supporters, même lorsque certains n’ont pas encore pleinement pris part au récit sur le terrain.
Si l’union vécue se dissipe avec le temps, la mémoire, elle, persiste par la narration. Elle se nourrit de répétitions, de partages, de récits réactivés. Une communauté mémorielle se constitue ainsi, non pour figer l’instant, mais pour lui donner une continuité symbolique. Au sens ricœurien, le football devient récit: il articule l’événement, la mémoire et l’identité. L’unité ne survit pas comme fait politique durable, mais comme horizon narratif — maintenu vivant par l’imaginaire collectif qui insiste, obstinément, sur la possibilité de l’union.
Le football ne résout pas les problèmes de l’Afrique, mais il montre quelque chose d’essentiel: une imagination partagée, une capacité d’être en commun, de vivre ensemble, d’être de bons voisins. La CAN est un miroir, pas un blueprint. Le Maroc 2025-2026 est un moment charmeur où l’Afrique se regarde dans le miroir du terrain et croit en elle-même.
Dans un monde divisé, fragmenté et privé de boussole, la CAN rappelle que l’Afrique existe déjà — chaque fois qu’un ballon circule librement entre ses peuples. Malgré les frontières ouvertes ou fermées, les espaces aériens libérés ou verrouillés, malgré les paroles médiatiques tantôt intégratrices et tolérantes, tantôt saturées de haine, le ballon défie les rancœurs et les calculs mesquins. Il circule à travers les périples, en dépit des obsessions fragmentaires et fragmentantes de certains leaders.
Le ballon crée le rêve, ouvre un horizon de coexistence: celui du bon voisinage, d’une unité africaine profondément désirée — non par les institutions, mais par les peuples africains eux-mêmes.
Par Lahcen Haddad
Le 29/12/2025"
https://fr.le360.ma/economie/vivre-lafrique-au-maroc-quand-le-football-devient-langage-africain_TPMCMZRR6ZFMZNYFYCIPFO2EOM/
#Metaglossia
#metaglossia_mundus
"A row erupted over the rising cost of translators in the UK benefits system after far-right, anti-Islam activist Tommy Robinson accused the government of wasting millions of pounds of public money on people who "can't speak English.
In a post on X, Robinson criticised the use of taxpayer-funded interpreters for migrants and pushed for their deportation: “If they can't speak English, then they shouldn't be here anyway. Absolute p***take. Deport.”
— TRobinsonNewEra (@TRobinsonNewEra)
His comments came as a new report called for migrants to be banned from using free translation services when claiming benefits. The study was published by the Policy Exchange think tank, according to the Daily Mail.
It argues that the ability to speak English should be a basic requirement or bare minimum for accessing the welfare system.
The report follows a surge in benefit claims and says the government should stop offering free translators in most civil cases. It describes the benefits system as part of a “social contract” with society and its people, adding: “Part of this is the ability to converse in the official, national language.”
According to official figures cited in the study, spending on translation services in civil cases rose by 80 per cent in the three years after the Covid-19 pandemic, reaching £12.8 million last year.
Claimants are currently entitled to free translation to help them appeal decisions denying benefits such as Personal Independence Payment and Employment and Support Allowance.
Should claimants pay for their own translators?
The report also says that in future, claimants who cannot speak English should be expected to pay for their own interpreters.
It recommends that free translation should remain only for deaf people and for criminal cases, where “freedom and liberty are on the line”.
The proposal is to curb the growing welfare bill and reduce the influence of courts over benefit rules. The report argues that courts have gradually expanded the rules on who qualifies for benefits, driving up overall costs. It says Parliament has failed to assert control, with MPs unwilling to reverse decisions made by judges...
Tommy Robinson is known for his strong opposition to immigration, especially from Muslim-majority countries. He argues that mass migration compromises British identity, security, and social cohesion. In September, Robinson led major anti-immigration rallies in London that included tens of thousands of supporters and sometimes turned violent.
About the Author
TOI World Desk"
https://timesofindia.indiatimes.com/world/uk/if-they-cant-speak-english-tommy-robinson-slams-12-million-going-to-translators-for-migrants-in-uk/articleshow/126230426.cms
#Metaglossia
#metaglossia_mundus
"Uber has introduced a new in-app translation tool designed to help drivers and passengers communicate more easily, even when they don’t share a common language.
According to a report by Mashable, the ride-hailing platform has launched an automatic translation feature that supports more than 100 languages, allowing messages sent through the Uber app to be translated in real time. The feature is aimed at improving communication around pick-ups, directions and delays, particularly in busy or unfamiliar locations.
Uber says the tool works directly within the app’s messaging system, meaning neither drivers nor passengers need to switch to external translation apps. Messages are translated automatically, helping both sides understand key information such as where to meet, which entrance to use, or whether there are any issues with traffic or access.
The company has positioned the update as part of a wider push to improve the overall ride experience, especially in cities and regions where language barriers are more common.
What the new translation tool does
The new feature allows:
Automatic translation of in-app messages
Support for over 100 languages
Clearer communication during pick-ups and drop-offs
Uber believes this will reduce confusion, missed connections and cancelled trips caused by misunderstandings between drivers and passengers.
DM News Commentary
From a private hire driver’s point of view, better communication tools are always welcome. Language barriers can be a genuine challenge, particularly in city centres, airports and tourist hotspots where pick-up instructions can make or break a job.
That said, while translation tools may help smooth out day-to-day issues, they don’t address some of the wider frustrations drivers often raise — such as pay rates and service fees, etc.
For passengers, especially visitors to the UK, clearer communication could make app-based travel less stressful — provided the translations are accurate and reliable in real-world conditions."
matthewDec 29, 2025
https://dmnews.co.uk/uber-rolls-out-real-time-translation-feature-to-bridge-language-gaps-for-drivers-and-passengers/
#Metaglossia
#metaglossia_mundus
"De Christine de Pizan à Simin Daneshvar, quarante traductrices pionnières
Elles décident de signer leurs traductions au lieu de rester anonymes, comme il est d’usage à l’époque dans un monde régi par les hommes. Loin de se confiner au travail intellectuel que représente la traduction et l’écriture (la plupart sont aussi écrivaines), elles voyagent, luttent pour les droits des femmes et mènent une vie défiant les conventions de leur temps.
Quelles œuvres traduisent-elles? Des œuvres littéraires, bien sûr, mais aussi des publications scientifiques et politiques. Pourquoi traduisent-elles ces œuvres?
À quels défis sont-elles confrontées pendant cinq siècles? Quels sont leurs accomplissements envers et contre tout? Quelle est leur influence sur la société en tant que figures littéraires, scientifiques et politiques? Quelle est leur contribution à la cause des droits des femmes?
Voici les courtes biographies de quarante traductrices/autrices rédigées avec l’aide de Wikipédia - ces femmes disposent toutes d’une page Wikipédia (en français ou dans d’autres langues) grâce aux multiples collaborateurs/collaboratrices de l’encyclopédie en ligne. Elles méritent maintenant des biographies complètes, des documentaires, des expositions et des œuvres d’art en tous genres. Place à un voyage virtuel aussi passionnant qu’incomplet.
Crédits illustration : Les chimistes Claudine Picardet (qui tient un livre à la main) et Marie-Anne Paulze Lavoisier (à sa droite), dont les traductions sont essentielles lors de la Révolution chimique menée par André Lavoisier (assis) entouré de ses collègues. Portrait de groupe anonyme disponible dans Wikimedia...
Le 29/12/2025
par Marie Lebert
Contact : marie.lebert@gmail.com
https://actualitte.com/article/128408/auteurs/de-christine-de-pizan-a-simin-daneshvar-quarante-traductrices-pionnieres
#Metaglossia
#metaglossia_mundus
"Abstract
Background: Psychiatric interpreters in Japan-many of whom work on a voluntary basis-play a vital role in bridging critical language barriers for the country’s immigrants while facing unique psychological challenges within a non-certified system.
Aim: This qualitative study aimed to explore the psychological experiences and support needs of volunteer psychiatric interpreters in Japan.
Methods: Semistructured interviews were conducted with 15 medical interpreters recruited via a national dispatch organization (10/15, 66.7%), local networks (3/15, 20%), and hospital referrals (2/15, 13.3%). Data were analyzed using the constant comparative method.
Results: Four major themes emerged: inhibiting factors (e.g., emotional exhaustion, lack of support), facilitating factors (e.g., agency assurance, patient approval), interpreters’ personal qualities (e.g., resilience, altruism), and the pursuit of a professional foundation (e.g., desire for formal training). Participants perceived an emotional toll greater than in general medical interpreting, intensified by Japan’s volunteer-based, under-supported system. None of the participants were full-time psychiatric interpreters, limiting the applicability to highly specialized settings.
Conclusions: Structured psychological training and support networks are recommended to enhance interpreter well-being and improve access to mental health services for immigrants, thereby contributing to psychiatric care, public health outcomes, and cross-cultural communication."
https://www.cureus.com/articles/445123-psychological-experiences-and-support-needs-of-volunteer-psychiatric-interpreters-in-japan-a-qualitative-study#!/
#Metaglossia
#metaglossia_mundus
"Decolonising the Mind: The Politics Of Language In African Literature
Mehnaz Ali Jan
“Every language has a right to be spoken, even if only five people speak it, because it is democratic and a human right.” (Ngugi wa Thiong’o)
The book “Decolonising the mind: the politics of language in African literature” is written by a Kenyan writer, Ngugi wa Thiong’o, who was born in January 5, 1938. He is one of the most influential writers, novelists, playwrights, essayists and a post-colonial theorists. He firmly upholds writing in African native language such as his own Gikuyu rather than colonial languages. His various publications include Weep Not Child, The River Between, A Grain Of Wheat, Petals Of Blood, Devil On The Cross, Homecoming and many more.
The book debated book contains 114-120 pages, written in English and has been translated into several languages including French, Spanish, Portuguese, Italian, Arabic and some African and Asian languages and got published in 1986.
Furthermore, to fully grasp the book’s title, one must first understand the concept of colonisation.
Colonisation refers to the subjugation (process of bringing something under control) of one country or nation by the other militarily advanced and a powerful country or nation. The colonizer usually rule over the weak nations, form colonies and exploit their resources for its own benefits. And the colonized countries or nations undergo various changes such as political, economic, cultural and social changes in the subjugated country.
Decolonisation is its opposite which is the process by which colonies become independent of the colonizing country. It is about cultural, economic and psychological freedom of the native people with the achievement of national sovereignty.
Moreover, the book is the combination of essays regarding language and its constructive role in national development, culture, history and identity. The book is a collection of four essays namely:
1- The Language Of African Literature
2- The Language Of African Theatre
3- The Language Of African Fiction
4- The Quest For Relevance
1. The Language of African Literature:
In this essay, the writer argues that language is deeply tied to culture and identity. He explains how colonial education system pushed African writers to use European languages, which gradually distanced them from their own people and traditions. Ngugi believes that writing in colonial languages reinforces mental domination, while using African languages helps reclaim cultural pride and restore a true African perspective in literature.
2. The Language of African Theatre:
The author focuses on theatre as a communal art form that directly engages the people. He describes his experience with community theatre in Kenya, where performances were done in local languages and reflected real social issues. This essay shows how African theatre could become a powerful tool of resistance and education when it speaks the language of the people, rather than the language of the colonizers.
3. The Language of African Fiction:
Here, Ngugi examines African novels written in European languages and questions whom these works truly serve. He argues that when African stories are told in foreign languages, they often cater more to international audiences than local readers. By advocating for African languages in fiction, Ngugi emphasizes the importance of authenticity, accessibility, and preserving indigenous storytelling traditions.
4. The Quest for Relevance:
This essay discusses the responsibility of African writers and intellectuals in post-colonial societies. Ngugi stresses that literature should not exist in isolation but should address the real struggles of ordinary people, such as oppression, inequality, and cultural loss. He calls on writers to remain relevant by engaging with their communities and contributing to social change rather than seeking approval from Western institutions.
However, the main arguments of the book cover:
•Language as a tool of colonial domination
• Colonial Alienation
• Reclaiming identity through indigenous (native) languages
Language as a Tool of Colonial Domination:
Ngugi argues that colonial powers didn’t just occupy land but they also used language to shape thought and identity. He mentions that when African intellectuals write in European languages (English, French, Portuguese), they are participating in a continuation of colonial influence because these languages carry the worldviews and values of the colonizers.
Colonial Alienation:
He introduces the idea of “colonial alienation” that the psychological estrangement that occurs when a people’s native languages and cultural frames are displaced by colonial languages. This alienation, he claims, leads to a loss of cultural grounding and self understanding.
Reclaiming Identity Through Indigenous Languages:
Ngugi passionately defends the use of African languages in literature, education, and culture, arguing that they are essential for genuine cultural expression and liberation. Language, for him, is not neutral instead it’s deeply political and tied to a people’s history, memory, and resistance.
Apart from this, this book talks about the implementation of English as the official language by the British in the colonized countries like Africa, and its impacts as this broadens the gap between the use of mother tongue Gikuyu and English. He started writing his works in Gikuyu since 1977. Since then he was confronted with a question, particularly, in Europe about the reason for his writing in Gikuyu language. His writing in mother tongue was perceived as an abnormal act. He shares incidents of humiliating experiences undergone by the Gikuyu speaking children in school and different types of punishments were given.
This reflects how colonialism shaped African thinking through language and that colonial rule didn’t end with political independence as it survived in people’s minds – especially through education system that dominated European languages and cultures over African ones. For him, language is not merely a means of communication but also it carries culture, identity, memory and history. To decolonise the minds, therefore, means reclaiming African language using them to tell stories.
In short, Ngugi believes his writing in his native language is a part of the anti imperialistic struggles of Africans, especially Kenyan people. He does not want to see the future generation, specifically, school children growing up in the imperialistic tradition. He desires to see Kenyan national languages having a literature which reflects the rhythms of the languages as well as the native social nature, plus their struggle with nature and culture. Because it will benefit the future generations of Kenya to live in harmony; they can learn other languages and could enjoy revolutionary elements in the other languages without any complex about their mother tongue, environment and themselves.
Thus, “Decolonising the Mind” is a must-read for students, writers, scholars, political workers and others who are interested in understanding the lasting effects of colonialism on language and culture. We can touch it with the ongoing crises in Balochistan as well."
https://www.thebalochnews.com/2025/12/29/decolonising-the-mind-the-politics-of-language-in-african-literature/?amp=1
#Metaglossia
#metaglossia_mundus
"Rise in Demand for Quran Translations in Swedish after Quran Desecration
Culture & KnowledgeCulture9:41 - December 28, 2025
News ID: 3495872
IQNA – Following an incident of Quran desecration in front of the Stockholm Grand Mosque, demands to buy translations of the Holy Quran in Swedish have increased significantly.
According to a report by Alkompis, after the Quran burning in front of the Stockholm Grand Mosque, requests to purchase a translation of the Holy Quran in Swedish increased significantly, and perhaps this is one of the positive effects of this tragic incident in Sweden.
There have been several Quran translations in Sweden over the years, and many people have made great efforts to present the concepts of the Holy Quran to the Muslims of the European country.
Currently, there are 8 different translations of the Holy Quran in Swedish, the first translation being done in the 18th century and the last in the past few years.
It came out after the establishment of the Scandinavian Education Foundation on December 3, 2022, which took steps towards translating the Holy Quran.
Each translation of the Quran into Swedish has its own differences from the other, and it is surprising that more than half of the existing translations were made by non-Muslim Swedes and followers of other religions, including Judaism and Christianity, because they were very interested in learning about the Quran and also felt responsible for their community, so they translated the Holy Quran into Swedish and published it.
Qurans translated by Swedish orientalists
1. The first famous translation of the Quran was made by Bishop Johan Adam Tingstadius (1748-1827), but it was never published.
2. The first published translation into Swedish was made by Johan Fredrik Sebastian Kronstolp (1801-1882), which was published in 1843.
3. Between 1873 and 1874, Carl Johan Turnberry (1807-1877) presented a new translation.
4. In 1917, a complete translation of the Quran into Swedish was published.
5. In 1961, a Swedish translation of parts of the Quran by Ake Ollmark was published.
6. A new translation of the Quran with commentary, Koranen Budeskap or Messages of the Quran, was published in 1998 by former Swedish diplomat Mohammed Knut Brannström, followed by a second edition in 2000.
This Swedish translation is distinguished from other translations by receiving official approval from the Department of Islamic Research, Writing and Translation at Al-Azhar University.
In addition to the commentary, Brannström also included the original Arabic text of the Quran. This translation contains many commentaries, footnotes and explanations that contradict Islamic teachings.
Dr. Abdullah Al-Suwaidi, a graduate of the Islamic University of Medina, has described in detail the doctrinal errors in this translation in his master’s thesis.
7. Another translation into Swedish was done by Quneita Sadaqah, entitled “The Holy Quran”. In this translation, each page is divided into two columns, with the left column containing the translation of the Quran in Swedish and the right column containing the text of the Quran in Arabic.
8. The Scandinavian Education Foundation also published a translation entitled DEN ÄDLA KORANEN in 2022, which is the latest translation of the Holy Quran in Swedish.
How is the Quran translated?
Jan Eyarpe, a professor of the history of religions in Sweden, says, “The main duty of a believer in his religious life is to recite the Quran with humility from its Arabic text with Tajweed. Therefore, the text of the Quran, which is written in Arabic script, is a sacred text.”
The Holy Quran is recited in Tarteel and in eloquent Arabic on various occasions, but most Muslims in the world do not understand it although it has amazing beauty. The text of the Quran must be translated and interpreted in order to be understood.
In an article titled “Translating the Quran into Swedish”, published in 2005, Christopher Toll took a comprehensive look at the history and personalities who translated the Quran into Swedish. Academic articles and theses in Sweden also looked at parts of the Quran translated into Latin or Swedish.
From the end of the 18th century to the first decades of the 19th century, these theses formed the basis of a number of master’s theses and were used by a very limited group of teachers and students.
The Arabic language articles were related to the subject of Oriental languages (which also included Greek), and Arabic was essential for comparison with the Semitic languages, especially Hebrew and Syriac, which were used for the interpretation of the Bible and the study of church bishops.
Complete translations of the Quran into Latin, English, German and French were available to the public. In the field of Quranic studies, the Latin translations made by Ludovico Marracci in 1698 were also of great importance for a very long time.
Over time, the Swedish government showed greater interest in Oriental studies, culminating in the International Oriental Conference in Stockholm, which was held with the participation of King Oscar II in 1889.
At the same time, universities began to take a broader approach to the field. During this period, Carl Johannes Thornberg, professor of Oriental languages at Lund University, published his Swedish translation of the Holy Quran in pamphlets in 1873–74.
Thornberg had been closely associated with European Orientalists in the mid-19th century. He made some interpretative additions to the original text of his translation, which contained precise and close interpretations of the Quranic text, and also added explanations.
Zetterstéen’s translation
Karl Vilhelm Zetterstéen, professor of Oriental languages at Lund University from 1895 to 1904, and then professor of Semitic languages at Uppsala until 1931, produced a remarkable translation.
Translations of the latter half of the 20th century
Ake Ullmarks rewrote a translation of parts of the Quran in 1961 based on Ludwig Ullmann’s German translation of 1840. These translations are unreliable to a considerable extent.
A Swedish translation was also produced in 1988 based on the beliefs of the Ahmadiyya, reflecting their interpretation of the Quran’s content, and having a limited audience. Swedish Islamic authorities consider it heretical or un-Islamic.
Sweden now has a large population of Muslim background, which was not the case in the 20th century. A translation of the Quran into Swedish, which is now highly secular and at the same time multi-religious, has a different audience.
Bernström, a Swedish diplomat, has served twice as ambassador in Madrid and Rabat. Bernström, who converted to Islam in the mid-1980s, saw society in need of a new translation of the Holy Quran into Swedish. When Zetterstéen began translating the Quran in the 20th century for students and scholars, Bernström wondered how modern Muslims could understand the concepts of the Quran, which originated in the 6th century in the Arabian Peninsula. Bernström’s target audience was Swedish Muslims.
Bernström’s religious upbringing further explains this problem. He was born in 1919 into a Protestant-Catholic family and converted to Catholicism at a young age. But in the 1960s, when the Second Vatican Council chose the motto “to update,” Catholicism became problematic for him. This motto conflicted with the idea that religion was something fixed and permanent.
Bernström’s stay in Morocco was very important. He found there the permanence of Islam and the Quran, and its Arabic text, which is eternally sacred. But how could he understand this sacred text?
The book “The Message of the Quran” by Muhammad Assad (a convert from Judaism to Islam) in 1980 greatly helped Bernström in translating the Quran. In “The Message of the Quran”, Assad provided an edited and completed commentary.
According to Jan Eyarpe, a Swedish professor of the history of religions, this is not a translation, but a rewriting."
https://iqna.ir/en/news/3495872/rise-in-demand-for-quran-translations-in-swedish-after-quran-desecration
#Metaglossia
#metaglossia_mundus
Outcry over £12m cost of translators for migrants in claim disputes.
Migrants should be banned from using taxpayer–funded translators to claim benefits, a report says.
The study by the Policy Exchange think–tank, which follows a surge in claims, says tribunals considering benefit cases should no longer offer free translation services.
The report says the benefits system should be seen as 'part of a social contract that one has with society at large', adding: 'Part of this is the ability to converse in the official, national language.'
The study cites official figures showing the cost of providing translation services in civil cases jumped by 80 per cent in the three years following the pandemic, to reach £12.8million last year.
Claimants are entitled to free translation to help them challenge decisions refusing them benefits such as Personal Independence Payment (PIP) and Employment and Support Allowance (ESA).
The report says that in future they should be expected to pay for their own interpreters if they cannot speak English. It says that free translation services should only continue for the deaf and for criminal cases where 'freedom and liberty are on the line'.
The recommendation is part of a wider report urging ministers to take back control of the benefits system from the courts.
The study finds that the criteria for granting benefit claims has been systematically widened by the courts for years, helping to fuel the huge rise in the welfare bill.
There have been calls for tribunals to consider that benefit cases should no longer offer free translation services...
Tribunals currently deal with thousands of benefits appeals every year. The study found that the majority are granted, with up to 75 per cent of cases overturned in some courts.
Former Lord Chancellor Sir Robert Buckland said the discretionary power allowing judges to interpret the law 'should be returned to lawmakers' to help contain the 'breathtaking amount of public expenditure', adding: 'The subcontracting of decision–making of this nature and scale should concern all of us who care about the proper role of the courts.'
The report says £750million a year could be saved just by tightening the rules allowing people to claim PIP if they are unable to plan and undertake a journey because of mental health conditions.
Report author Jean–Andre Prager said criteria for benefit claims should be automatically reviewed every two years.
He also called for courts to issue 'suspended' judgments to give parliament six months to clarify the law before they came into effect, adding: 'The appeals system has become a substitute for good decision–making rather than a check on it. We need to restore parliamentary authority, improve accountability and use technology to deliver a fairer, more sustainable system..."
By JASON GROVES, POLITICAL EDITOR
23:51 GMT 28 Dec 2025 https://www.dailymail.co.uk/news/article-15418205/12million-cost-translators-migrants-benefit-claim-disputes.html?ito=native_share_article-nativemenubutton
#Metaglossia
#metaglossia_mundus
"The Judicial Council of California has approved recommendations from a study on how to expand access to the pool of court interpreters able to assist the state’s nearly 6.4 million limited-English-proficient (LEP) residents and potential court users.
Study considerations for the council include reviewing interpreter examination standards and content, adopting carryover scoring to aid candidates, and creating structured career pathways—including tiered systems and apprenticeship programs, particularly for Spanish interpreters. The study also advises the council to work with courts to develop a statewide directory for interpreters trained in video remote interpreting and to continue expanding access to American Sign Language and emerging languages.
The study also calls for action by the legislature, courts, and experienced court interpreters, which includes recommendations to continue adequately funding interpreter services, review and improve compensation packages, and develop mentorship programs.
“There was broad agreement among the stakeholders for the recommendations and considerations,” said Court Executive Officer Anabel Romero, vice-chair of the council’s Court Interpreters Advisory Panel. “The study offers a roadmap for the branch to improve interpreter services for the public and to ensure our goals of high-quality equitable access for all Californians, regardless of the language they speak.”
California Courts Newsroom (12/12/25) By Blaine Corren"
https://www.atanet.org/industry-news/judicial-council-to-continue-efforts-to-improve-the-availability-of-court-interpreters/
#Metaglossia
#metaglossia_mundus
"Why 2026 belongs to multimodal AI
The text-based era of interacting with AI is over.
For the past three years, AI’s breakout moment has happened almost entirely through text. We type a prompt, get a response, and move to the next task. While this intuitive interaction style turned chatbots into a household tool overnight, it barely scratches the surface of what the most advanced technology of our time can actually do.
This disconnect has created a significant gap in how consumers utilize AI. While the underlying models are rapidly becoming multimodal—capable of processing voice, visuals, and video in real time—most consumers are still using them as a search engine. Looking toward 2026, I believe the next wave of adoption won’t be about utility alone, but about evolving beyond static text into dynamic, immersive interactions. This is AI 2.0: not just retrieving information faster, but experiencing intelligence through sound, visuals, motion, and real-time context.
AI adoption has reached a tipping point. In 2025, ChatGPT’s weekly user base doubled from roughly 400 million in February to 800 million by year’s end. Competitors like Gemini and Anthropic saw similar growth, yet most users still engage with LLMs primarily via text chatbots. In fact, Deloitte’s Connected Consumer Survey shows that despite over half (53%) of consumers experimenting with generative AI, most people still relegate AI to administrative tasks like writing, summarizing, and researching.
Yet when you look at the digital behavior of consumers outside of AI, it’s clear consumers crave immersive experiences. According to Activate Consulting’s Tech & Media Outlook 2026, 43% of Gen Z prefer user-generated platforms like TikTok and YouTube over traditional TV or paid streaming, and they spend 54% more time on social video platforms than the average consumer, abandoning traditional media for interactive social platforms.
This creates a fundamental mismatch: Consumers live in a multi-sensory world, but their AI tools are stuck delivering plain text. While the industry recognizes this gap and is investing to close it, I predict we’ll see a fundamental shift in how people use and create with AI. In AI 2.0, users will no longer simply consume AI-generated content but will instead leverage multimodal AI to bring voice, visuals, and text together, allowing them to shape and direct their experiences in real time.
MULTIMODAL AI UNLOCKS IMMERSIVE STORYTELLING
If AI 1.0 was about efficiency, AI 2.0 is about engagement. While text-based AI is limited in how deeply it can engage audiences, multimodal AI allows the user to become an active participant. Instead of reading a story, you can interact with a main character and take the plot in a new direction or build your own world where narratives and characters evolve with you.
We can look to the $250 billion gaming industry as the blueprint for the potential that multimodal AI has. Video games combine visuals, audio, narrative, and real-time agency, creating an immersive experience that traditional entertainment can’t replicate. Platforms like Roblox and Minecraft let players inhabit content. Roblox alone reaches over 100 million daily users, who collectively spend tens of billions of hours a year immersed in these worlds; engagement that text alone could never generate.
Shaping the mine of the future: The engineering challenge of a lifetime
With the rise of multimodal AI, users everywhere will be able to create these types of experiences they’ve loved to participate in through gaming. By removing technical barriers, multimodal allows everyone to build experiences that not only feel authentic to the real world but also actively participate in them. Legacy media is also responding to this trend. Disney recently announced a $1 billion investment in OpenAI and a licensing deal that will let users create short clips with characters from Marvel, Pixar, and Star Wars through the Sora platform.
WHY MULTIMODAL AI CAN BE SAFER FOR YOUNGER USERS
As AI becomes part of everyday life, safety—particularly for younger users—has become one of the most critical issues facing the industry.
Moving from open-ended chat to structured, multimodal worlds allows us to design guardrails within the gameplay. Instead of relying on continuous unstructured prompts, these environments are built around characters, visuals, voices, and defined story worlds. Interaction is guided by the experience itself. That structure changes how and where safety is designed into the system.
Educational AI demonstrates this approach. Platforms like Khan Academy Kids and Duolingo combine visuals, audio, and structured prompts to guide learning. The AI isn’t trying to be everything; it focuses on one task well. As multimodal AI evolves, one of its most meaningful opportunities may be this ability to balance creative freedom with thoughtful constraint. AI 2.0 presents a design shift that could give builders, educators, and families new ways to shape safer, more intentional digital spaces for the next generation.
WHY MULTIMODAL AI IS THE NEXT FRONTIER
In 2026, I predict that consumers won’t be prompting AI; it will be a more immersive interactive experience. This excites me because users won’t just passively receive outputs; they’ll actively shape experiences and influence how AI evolves in real time. We could see users remixing the series finale of their favorite TV show, or students learning history not by reading a textbook, but by actively debating a historically accurate AI simulation.
For founders and creators, the next step is to stop building tools only for efficiency and start building environments for immersion and exploration. The winners of the next cycle won’t be the ones with the smartest models, but the ones who make AI feel less like a utility and more like a destination for rich, interactive experiences."
Why 2026 belongs to multimodal AI - Fast Company
BY Karandeep Anand, CEO of Character.AI
https://www.fastcompany.com/91466308/why-2026-belongs-to-multimodal-ai
#Metaglossia
#metaglossia_mundus
|
"Ces mots français venus du Moyen-Orient et de Méditerranée
Des milliers de mots de la langue française proviennent de contrées lointaines de l’Hexagone. Parmi eux, des centaines ayant des liens directs ou indirects avec l’arabe mais également, à une moindre échelle, le persan.
Publié le : 21/12/2025 - 12:04
Modifié le : 22/12/2025 - 10:10
Le mot français bougie provient de la ville algérienne éponyme aujourd’hui appelée Béjaïa. REUTERS/KIM HONG-JI
Par :bAnne Bernas
Monsieur Jourdain ignorait qu’il faisait de la prose, mais Molière savait-il que de nombreux mots dont il usait provenaient de contrées moyen-orientales et méditerranéennes ?
L’arabe est la sixième langue la plus parlée au monde et est utilisée quotidiennement par plus de 400 millions de personnes, dont 200 millions de locuteurs natifs, selon les données des Nations unies. Une « Journée mondiale » lui est également consacrée chaque 18 décembre, date coïncidant avec le jour où, en 1973, l'Assemblée générale de l’ONU a adopté l'arabe comme sixième langue officielle de l'Organisation par la résolution 3190.
D’une infinie richesse, l’arabe, classique mais aussi dialectal, a transmis à de nombreuses langues indo-européennes une somme de mots considérable. « Il y a beaucoup plus de mots d’origine arabe dans la langue française que de mots d’origine gauloise ! » rappelait en 2022 l’académicien Erik Orsenna.
Plus de 800 références
Dans le dictionnaire français, pas moins de 400 mots sont d’origine arabe, 200 autres sont des mots dérivés et quelque 800 portent la marque de la langue arabe. Depuis le Moyen Âge, des mots arabes voyagent jusqu’en France, par la terre et la mer, en faisant parfois escale dans d’autres pays. À ce titre, le mot « amiral », rang le plus élevé dans la marine française, provient de « Amir al-bahr», le « chef de la mer ». La dissémination de mots d’origine arabe se poursuit lors des conquêtes coloniales du XIXᵉ siècle en Afrique du Nord et jusqu'à plus récemment avec le langage des jeunes de banlieues.
Pourtant, comme le rappelle Salah Guemriche, auteur du Dictionnaire des mots français d'origine arabe (Seuil, 2007), de nombreux mots français d’origine arabe n’ont pas toujours été « assumés » par Paris, à la différence des mots d’origine persane. « Aux XVIᵉ et XVIIᵉ siècles, l’origine ne dérangeait pas, elle était sans arrière-pensée idéologique, ce qui change au XIXᵉ, à l’heure de la colonisation. Comment voulez-vous dire d’un côté “on va apporter la civilisation” et de l’autre dire que la langue française contient déjà des mots arabes, des Arabes qu’on est allé coloniser. »
Les termes issus de l’arabe sont si nombreux qu’il est impossible d’en dresser une liste, mais parmi les plus connus, dans le domaine des sciences et des mathématiques, à côté des centaines de noms d’étoiles, notons les célèbres « zéro », « chiffre », « zénith », etc. Mais aussi « algèbre », provenant de l'arabe « al-jabr »,« restauration » ou « réunion ». Ce mot figure initialement dans le titre d'un ouvrage du IXᵉ siècle consacré à la résolution d'équations, écrit par le savant persan Muhammad ibn Musa al-Khwarizmi, installé à Bagdad, et dont dérive le mot « algorithme ».
Plus terre à terre. Le sud de l’Hexagone est réputé pour sa culture des abricots (« al barbuk »), mais les Bretons, dont l’artichaut est l’un de leurs symboles de fierté, savent-ils que ces « épines de la terre » se disent en arabe « ardi chawk » ? Savent-ils que l’arsenal de Brest puise son origine de l’arabe « dar al-sina » ou « maison de fabrication ». Et que dire du caban breton, mais aussi des écharpes, chemises, jupes, gilets, parfois en coton, lui aussi issu de l’arabe « kutun ».
« Le français n'est pas français »
À l’heure des vacances de fin d’année, certains pourraient être tentés de partir faire un safari, dont le terme vient de l’arabe « safar » signifiant « voyage », pour découvrir les girafes (de l’arabe « zarafa ») et les gazelles (« ghazal ») au milieu des baobabs ( « abu hibab »). Un voyage sans grosses tempêtes pour les voyageurs de l’océan Indien puisque la saison des moussons est terminée. La mousson provient de l’arabe « mausim », « saison ». Il arrive dans le vocabulaire français à la fin du XVIᵉ siècle après être passé par le néerlandais, le portugais et l’espagnol sous différentes formes.
Pour fêter la fin de l’année, certains feront la nouba (de l’arabe algérien) et trinqueront avec un verre d’alcool, mot venant du mot « al-khol ». Il signifie à l’origine une fine poudre noire de sulfure de plomb et d’antimoine, un produit de beauté mais aussi un produit d’alchimistes. À la Renaissance, le chimiste allemand utilise le terme « alcohol vini » pour désigner « l’esprit de vin », le « produit d’une distillation de vin ». Quoi qu’il en soit, pour se remettre de toutes ces émotions, rien ne sera plus efficace qu’un petit « caoua ».
D’autres mots sont directement entrés dans le vocabulaire, par exemple bougie (de la ville algérienne éponyme aujourd’hui appelée Béjaïa), casbah (directement venu d’Alger), bled, barda, chouia, et plus près de nous kiffer, dawa ou encore seum. Et puis il y a des « mots arabes clandestins » tels que les nomme le lexicographe Roland Laffitte qui raconte l’exemple de l’épithète « romaine » accolé à une balance, la fameuse balance romaine. « Romaine » vient de l’arabe rumāna signifiant « grenade ». Car cette balance possède sur un bras un dispositif d’équilibre qui lui ressemblait.
La langue arabe, la plus parlée des langues sémitiques et la quatrième source d’emprunts de notre langue, après l’anglais, l’italien et l’allemand, a permis un dialogue des cultures le long des routes de la soie, de la côte de l'Inde à la Corne de l'Afrique. Les empires arabes successifs, après la fondation de l’islam, régnèrent sur un territoire qui s’étendait du Maroc à l’ouest jusqu’aux frontières de ce qui est aujourd’hui la Chine. La langue nous a rapporté d’Inde le mot sucre, issu du sanskrit et qui se dit en arabe « sukkar »...
« À travers les siècles, l’arabe a été au cœur des échanges entre les continents et entre les cultures », déclarait Audrey Azoulay, ex-directrice générale de l’Unesco. La langue, ajoutait-elle, a été « maniée par tant de grands poètes, penseurs, scientifiques et savants ».
Il n’y a pas de hasard (du mot arabe « az-ahar », « dé à jouer »), le génie de la langue française tient beaucoup à ces innombrables apports multiculturels. « Le français n'est pas français », s’amusait à dire le linguiste de renom et historien de la langue française Alain Rey. De quoi y perdre son latin.
« Les mots ont beaucoup voyagé »
À une moindre échelle, de nombreux mots issus du persan - tulipe, divan, nénuphar, babouches, bazar, etc.- ont enrichi le vocabulaire français depuis le Moyen Âge. Explications avec Agnès Lenepveu-Hotz, maître de conférences au département d'études persanes de l’université de Strasbourg.
RFI : Comment expliquer que le nombre de mots français issus du persan (0,7%), langue plus ancienne que l’arabe, soit très inférieur à celui des mots issus de l’arabe (5%) ?
Agnès Lenepveu-Hotz : On trouve moins de mots français issus du persan que de l’arabe pour des raisons historiques. Les royaumes arabes dans le sud de l'Espagne ont par exemple engendré des contacts et donc l'espagnol, comme le français du Sud- Ouest, a emprunté des termes à l'arabe.
Aussi, beaucoup de mots persans ne sont pas arrivés directement jusqu’à nous. Assez souvent, ces mots ont transité par l'arabe, le turc, parfois le hindi. Quand le mot a été emprunté de l'anglais au hindi, il est parfois repassé en français, et souvent le hindi l'avait emprunté au persan avant. Ainsi, les mots ont beaucoup voyagé, comme le mot « châle ». Ce mot a été emprunté par les Anglais au hindi, mais l’hindi l'avait emprunté au persan. Le mot « orange » est passé par l'arabe et même par l'espagnol, mais au départ, c'est un mot persan.
Pour certains mots, on peine vraiment à savoir si l’origine est d’abord persane ou arabe. Pour d'autres, c'est beaucoup plus clair, comme pour le mot « kiosque ». Ce mot est attesté dès le moyen perse, donc bien avant les emprunts que le persan a fait à l'arabe. Ce mot signifie « palais », « pavillon », « kiosque ». Le mot « bakchich », lui, a transité par le turc, mais c'est un verbe en persan qui veut dire « donner ». Le français l'a ensuite emprunté du turc, mais l'arabe et le turc l'avaient emprunté au persan."
https://www.rfi.fr/fr/moyen-orient/20251221-ces-mots-fran%C3%A7ais-venus-du-moyen-orient-et-de-m%C3%A9diterran%C3%A9e
#Metaglossia
#metaglossia_mundus