 Your new post is loading...
|
Scooped by
Charles Tiayon
November 25, 2011 3:35 AM
|
The first pages... Total abyss :) You generally never know what you will come up. Even if you go over the whole text once before translating, it always turns out to be harder- at least for me. In t...
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
"That time “AI” translation almost caused a fight between a doctor and my parents Thom Holwerda 2025-06-02
What if you want to find out more about the PS/2 Model 280? You head out to Google, type it in as a query, and realise the little “AI” summary that’s above the fold is clearly wrong. Then you run the same query again, multiple times, and notice that each time, the “AI” overview gives a different wrong answer, with made-up details it’s pulling out of its metaphorical ass. Eventually, after endless tries, Google does stumble upon the right answer: there never was a PS/2 Model 280, and every time the “AI” pretended that there was, it made up the whole thing.
Google’s “AI” is making up a different type of computer out of thin air every time you ask it about the PS/2 Model 280, including entirely bonkers claims that it had a 286 with memory expandable up to 128MB of RAM (the 286 can’t have more than 16). Only about 1 in 10 times does the query yield the correct answer that there is no Model 280 at all.
An expert will immediately notice discrepancies in the hallucinated answers, and will follow for example the List of IBM PS/2 Models article on Wikipedia. Which will very quickly establish that there is no Model 280.
The (non-expert) users who would most benefit from an AI search summary will be the ones most likely misled by it.
How much would you value a research assistant who gives you a different answer every time you ask, and although sometimes the answer may be correct, the incorrect answers look, if anything, more “real” than the correct ones?
↫ Michal Necasek at the OS/2 Museum
This is only about a non-existent model of PS/2, which doesn’t matter much in the grand scheme of things. However, what if someone is trying to find information about how to use a dangerous power tool? What if someone asks the Google “AI” about how to perform a certain home improvement procedure involving electricity? What if you try to repair your car following the instructions provided by “AI”? What if your mother follows the instructions listed in the leaflet that came with her new medication, which was “translated” using “AI”, and contains dangerous errors?
My father is currently undertaking a long diagnostic process to figure out what kind of age-related condition he has, which happens to involve a ton of tests and interviews by specialists. Since my parents are Dutch and moved to Sweden a few years ago, language is an issue, and as such, they rely on interpreters and my Swedish wife’s presence to overcome that barrier. A few months ago, though, they received the Swedish readout of an interview with a specialist, and pasted it into Google Translate to translate it to Dutch, since my wife and I were not available to translate it properly.
Reading through the translation, it all seemed perfectly fine; exactly the kind of fact-based, point-by-point readout doctors and medical specialists make to be shared with the patient, other involved specialists, and for future reference. However, somewhere halfway through, the translation suddenly said, completely out of nowhere: “The patient was combative and non-cooperative” (translated into English).
My parents, who can’t read Swedish and couldn’t double-check this, were obviously taken aback and very upset, since this weird interjection had absolutely no basis in reality. This readout covered a basic question-and-answer interview about symptoms, and at no point during the conversation with the friendly and kind doctor was there any strife or modicum of disagreement. Still, being into their ’70s and going through a complex and stressful diagnostic process in a foreign healthcare system, it’s not unsurprising my parents got upset.
When they shared this with the rest of our family, I immediately thought there must’ve been some sort of translation error introduced by Google Translate, because not only does the sentence in question not match my parents and the doctor in question at all, it would also be incredibly unprofessional. Even if the sentence were an accurate description of the patient-doctor interaction, it would never be shared with the patient in such a manner.
So, trying to calm everyone down by suggesting it was most likely a Google Translate error, I asked my parents to send me the source text so my wife and I could pour over it to discover where Google Translate went wrong, and if, perhaps, there was a spelling error in the source, or maybe some Swedish turn of phrase that could easily be misinterpreted even by a human translator. After pouring over the documents for a while, we came to a startling conclusion that was so, so much worse.
Google Translate made up the sentence out of thin air.
This wasn’t Google Translate taking a sentence and mangling it into something that didn’t make any sense. This wasn’t a spelling error that tripped up the numbskull “AI”. This wasn’t a case of a weird Swedish expression that requires a human translator to properly interpret and localise into Dutch. None of the usual Google Translate limitations were at play here. It just made up a very confrontational sentence out of thin air, and dumped it in between two other sentence that were properly present in the source text.
Now, I can only guess at what happened here, but my guess is that the preceding sentence in the source readout was very similar to a ton of other sentences in medical texts ingested by Google’s “AI”, and in some of the training material, that sentence was followed by some variation of “patient was combative and non-cooperative”. Since “AI” here is really just glorified autocomplete, it did exactly what autocomplete does: it made shit up that wasn’t there, thereby almost causing a major disagreement between a licensed medical professional and a patient.
Luckily for the medical professional and the patient in question, we caught it in time, and my family had a good laugh about it, but the next person this happens to might not be so lucky. Someone visiting a foreign country and getting medicine prescribed there after an incident might run instructions through Google Translate, only for Google to add a bunch of nonsense to the translation that causes the patient to misuse the medication – with potentially lethal consequences.
And you don’t even need to add “AI” translation into the mix, as the IBM PS/2 Model 280 queries show – Google’s “AI” is entirely capable of making shit up even without having to overcome a language barrier. People are going to trust what Google’s “AI” tells them above the fold, and it’s unquestionably going to lead to injury and most likely death.
And who will be held responsible?"
https://www.osnews.com/story/142469/that-time-ai-translation-almost-caused-a-fight-between-a-doctor-and-my-parents/
#metaglossia_mundus
Pym, A. (2025). Deconstructing translational trust. Translation Studies, 1–16. https://doi.org/10.1080/14781700.2025.2476487
"ABSTRACT: Although trust seems germane to the post-Renaissance translation form, it is also important in numerous other kinds of service provision. The act of trusting is pertinent because it responds to uncertainty found in translation situations. One type of uncertainty ensues from the risks of non-aligned loyalties, giving rise to distrust of the traitor. A second type ensues from the nature of language use, where decisions are undetermined, different translators give different translations, and the client or user cannot verify the optimality of a translation. This means translators’ credibility claims cannot be empirically tested, familiarity cannot provide a sufficient foundation, and there are no grounds for accepting trustworthiness as an inherent virtue of translators. The blind trust idealistically invested in professionals may be instructively contrasted with the vigilant low trust with which automated translations can be received. Such deployment of low trust may become a viable ethical alternative to essentialist presuppositions."
https://www.tandfonline.com/doi/full/10.1080/14781700.2025.2476487
#metaglossia_mundus
"Prix Albertine Jeunesse is awarded to three works of children’s literature in translation
(c) Jennifer Knotts
By Villa Albertine
The Prix Albertine Jeunesse is awarded in 2025 to Never, not Ever! (Même pas en rêve) by Beatrice Allemagna, Aiko and the planet of dogs (Aiko et la planète des chiens) by Ainhoa Cayuso and Christoffer Ellegaard, and Shepherdess Warriors, volume 1(Bergères Guerrières, volume 1) by Jonathan Garnier and Amélie Fléchais. It’s been an enthusiastic 7th edition of the Prize: another year rich in amazing readings and debates, with nearly 15,596 students gathering to determine the annual reader’s choice award for the best Francophone children’s book in English Translation.
New York, June 2, 2025 — Villa Albertine, the French Institute for Culture and Education, the Embassies of France in the United States and Canada, Institut français du Canada, Agence pour l’enseignement français à l’étranger (AEFE) and Albertine Books are thrilled to announce the three winning titles of the 2025 Prix Albertine Jeunesse. Congratulations to these exceptional books and to the thousands of young voters who selected them!
Encouraging children ages 3-11 to choose their favorite book from several works of Francophone youth literature available in English translation, the Prix Albertine Jeunesse fosters a love of reading in both French and English. This year’s winners are: to Never, not Ever! (Même pas en rêve) by Beatrice Allemagna, translated by Jill Davis; to Aiko and the planet of dogs (Aiko et la planète des chiens) by Ainhoa Cayuso and Christoffer Ellegaard, translated by Irene Vázquez; and to Shepherdess Warriors, volume 1(Bergères Guerrières, volume 1) by Jonathan Garnier and Amélie Fléchais, translated by Ivanka Hahnenberger. These books have been honored for their exceptional literary and artistic value, the uniqueness and authenticity of the perspectives they offer to young readers on the world, and their captivating appeal.
Created in 2018, the annual Prix Albertine Jeunesse promotes recently published books translated from French in the US and Canada. This year, 15,596 students (672 classes) in North America participated, reading several books from a bilingual shortlist of various youth titles. Young voters assessed the quality of each book in both French and English languages.
This year, a network of 70 accredited French schools and public bilingual/dual language schools across North America have integrated the Prix Albertine Jeunesse books into their curricula. This Prize is a tool to strengthen bilingual and multilingual education, fostering connections between French and English. Participating schools receive complimentary educational resources and lesson plans for teachers made available for free, here. Additionally, Villa Albertine organized a series of 32 online author events in March 2025, providing students with direct interactions and discussions with the authors of the shortlisted books.
The 2025 shortlist was selected by a committee of experts representing Villa Albertine, the Embassies of France in the United States and Canada, Albertine Books, and representatives of the AEFE network of North America.
The forthcoming 2026 Prix Albertine Jeunesse shortlist will be available by the beginning of summer on the Albertine website. School classes may register for the 8th Prize edition in fall 2025.
The winners of this year’s prize are available for purchase on the Albertine website, alongside all the nominees.
AWARDED BOOKS:
3-5 years old: Never, not Ever! (Même pas en rêve)
by Beatrice Allemagna
translated by Jill Davis
Harper Collins (L’école des loisirs)
A laugh-out-loud tribute to little kids everywhere who would prefer not to leave home on the first day of school. The other animals are marching dutifully to school, but Pascaline could care less. “Never, not ever!” she declares. And nothing—not even her parents pulling her by her feet—will change her mind. She shrieks so loudly that her parents shrink down to the size of peanuts—becoming just the right size to fit snugly under Pascaline’s wing. Now they can all go to school together!
6-8 years old: Aiko and the planet of dogs (Aiko et la planète des chiens)
by Ainhoa Cayuso and Christoffer Ellegaard
translated by Irene Vázquez
Levine Querido (Les fourmis rouges)
Aiko is a courageous astronaut, specially trained to brave the extremes of space. The whole of humanity is counting on her success. But on a planet that shows signs of life, something goes awry, and when she wakes up, she finds . . . a pack of dogs? And . . . they can talk? Aiko is delighted. This discovery will make her the most famous astronaut on Earth! The dogs are… less delighted. They’re going to keep her prisoner on their planet rather than let humanity find them again.
9-11 years old: Shepherdess Warriors, volume 1 (Bergères Guerrières, volume 1)
by Jonathan Garnier and Amélie Fléchais
translated by Ivanka Hahnenberger
Ablaze (Glénat)
Shepherdess Warriors is the odyssey of a young heroine living great adventures in a medieval-fantasy universe inspired by Celtic legends. It’s been 10 years since the men of the village left to fight in The Great War. But the women quickly took charge of village matters. This is how the Order of Shepherdess Warriors was formed, a group of female fighters chosen among the most courageous and acrimonious, to protect not only their flocks, but also the village! Molly is happy because as soon as she turns 10, she can finally start training, and if she’s good enough, she can join the order. This quest will take her far beyond the boundaries of the land she knows, into the sorcerers’ forest and on the trail of her missing parents."
https://villa-albertine.org/va/press-release/prix-albertine-jeunesse-is-awarded-to-three-works-of-childrens-literature-in-translation/
#metaglossia_mundus
"Mexican novelist Guadalupe Nettel (The Accidentals), Argentine novelist Gabriela Cabezón Cámara (We Are Green and Trembling) and Uruguayan author Fernanda Trías (Pink Slime) offer sharp commentary on Latin American politics, gender roles, and environmental and political corruption in their novels.
As part of the PEN World Voices Festival, these three authors were joined by writer and curator Lily Philpott for a discussion entitled “Bold Voices: Latin American Writers in Conversation” to reflect on Latin American literature, literature in translation, and language.
On Language
Latin America is a melting pot of languages, dialects, and accents. During the panel, the authors discussed the languages they grew up hearing, and speaking, and how these languages impacted their writing.
Gabriela Cabezón Cámara: “When I think of languages, I don’t think of something fixed, but of something living. As a girl, my grandfather spoke the Spanish of Spain, which is very different from Argentinian Spanish. I learned English from music and school and tried to learn Chinese. My neighbors spoke Italian and French. The only languages that I didn’t hear growing up were Indigenous languages, which parents did not teach to their children back then. But now we are reclaiming these languages. So in a way, language is always evolving and being revised for me.”
Guadalupe Nettel: “I grew up in a neighborhood full of mixed languages, accents, and dialects from South America.…There are some Mexican writers who believe that we should only write in Mexican Spanish [because we are Mexican]. But Spanish is such a rich language…that I believe that writers should be able to make full use of different phrases, insults, and languages; not just those that we as authors, in theory, correspond with.”
Fernanda Trías: “I have lived in many countries over the past twenty years, including Colombia, Chile, and France. I ask myself what country I really belong to, because I sound like a foreigner wherever I go, including Uruguay. Even my mother tells me I have an accent. From moving, I have learned that language is a way to understand the place that you are in, and this has deeply affected me as a writer.”
From moving, I have learned that language is a way to understand the place that you are in, and this has deeply affected me as a writer.– Fernanda Trías
On Literature in Translation
All three authors’ works were written in Spanish and translated into other languages, including English, French, and Italian. Here, Nettel and Trías discuss what it’s like to have their work translated and what it is like to enter into a relationship with a translator.
Guadalupe Nettel: “I always feel tension when translating my work. It’s that you are not just translating into another language, but another culture. A good translator is a writer and poet too, one that does not cut out irony, nuance, and poetry from the book when the other language doesn’t have those exact sounds or plays on words, but recaptures them for another culture in the literary production.”
Fernanda Trías: “I was very involved when my novel was translated into English and French, which I speak. But there are translations of my book that happened into languages I do not speak, and some translators had a lot of questions about the text while others had none. For those where the translator did not have questions, I must accept that those translated texts are different bodies of work, different novels altogether.”
A good translator is a writer and poet too, one that does not cut out irony, nuance, and poetry from the book when the other language doesn’t have those exact sounds or plays on words, but recaptures them for another culture in the literary production.-Guadalupe Nettel
Check out the translations panelists’ books:
The We Are Green and Trembling by Gabriela Cabezón CámaraThe Girl From the Orange Grove by Gabriela Cabezón Cámara, translated by Robin Myers
Pink Slime by Fernanda Trías, translated by Heather Cleary
The Accidentals by Guadalupe Nettel, translated by Rosalind Harvey"
Allison Lee
June 2, 2025
https://pen.org/latin-american-writers-pen-world-voices-festival/
#metaglossia_mundus
"Complete New World Translation Released in Two Languages During May 2025
Kisi
On May 25, 2025, Brother Winston Bestman, a member of the Liberia Branch Committee, released the complete New World Translation of the Holy Scriptures in Kisi. The announcement was made to an audience of 1,029 gathered for a special program held in Guéckédou, Guinea. An additional 2,467 tied in to the program via videoconference from satellite locations in Foya, Liberia, and Koindu, Sierra Leone. Printed copies were distributed to those in attendance. The release was also immediately made available for download from jw.org and in the JW Library app. Additionally, it was announced that audio recordings of the New World Translation in Kisi will gradually be made available for download.
Nearly one million people speak Kisi in what is known as Kissiland, a region where Guinea, Liberia, and Sierra Leone share a common border. More than 1,200 brothers and sisters currently serve in 29 Kisi-language congregations and one group throughout the branch territory. These Kisi-speaking Witnesses rejoice at having an accurate translation of the complete Bible that uses God’s personal name, Jehovah.
Solomon Islands Pidgin
On May 25, 2025, Brother Jeffrey Winder, a member of the Governing Body, released the complete New World Translation of the Holy Scriptures in Solomon Islands Pidgin. The announcement was made during a special program in Honiara, the capital of the Solomon Islands. A total of 2,317 attended the program in person. Another 1,695 tied in via videoconference from two remote locations on the island of Malaita. All in attendance received a printed copy of the New World Translation in Solomon Islands Pidgin. The release was also made available for download from jw.org and in the JW Library app.
The good news first reached the Solomon Islands in 1953. Today, this island nation is home to approximately 757,000 people who speak more than 70 indigenous languages. Although English is the country’s official language, Solomon Islands Pidgin is the most widely spoken language in the Solomon Islands. Currently, some 2,100 brothers and sisters who speak Solomon Islands Pidgin serve in 38 congregations throughout the country."
JUNE 2, 2025
GLOBAL NEWS
https://www.jw.org/en/news/region/global/Complete-New-World-Translation-Released-in-Two-Languages-During-May-2025/
#metaglossia_mundus
"12e édition du Prix de la Francophonie pour jeunes chercheurs 2025
L’Agence Universitaire de la Francophonie (AUF) lance la douzième édition du Prix de la francophonie pour jeunes chercheurs. Ce prix est ouvert tous les deux ans et couvre les deux champs disciplinaires suivants : Sciences et Technologies et Sciences humaines et sociales.
Le Prix de la Francophonie pour jeunes chercheurs vise à valoriser le mérite et la qualité des travaux de recherche de quatre chercheur·ses francophones tout en prenant en compte la diversité de l’espace universitaire francophone, notamment celle des pays en développement. Ce prix distingue des chercheur·ses qui se sont illustrés dans leur domaine et ont obtenu une reconnaissance scientifique pour les avancées significatives réalisées.
Modalités de candidature
Les candidatures doivent être portées et présentées par un·e dirigeant·e (Recteur·ice, Président·e, Directeur·ice Général·e…) des institutions d’enseignement supérieur et de recherche membres de l’AUF.
Le Prix s’adresse à des candidat·es âgé·es de 40 ans au plus à la date de clôture de l’appel à candidatures, titulaires d’un doctorat, pouvant justifier d’une activité de recherche importante et innovante et rattachés à des établissements membres de l’AUF.
Le dépôt des dossiers de candidatures doit être fait impérativement avant le 27 juin 2025 (18h GMT) par l’intermédiaire d’un formulaire à renseigner en ligne sur la plateforme de l’AUF.
Remise du prix et récompenses
L’annonce et la remise du Prix Jeunes chercheurs 2025 aura lieu lors de la 9e Assemblée générale et de la 5e Semaine Mondiale de la Francophonie Scientifique (SMFS) qui se tiendront à Dakar (Sénégal) du 3 au 7 novembre 2025.
Les lauréat·es recevront une certification officielle et une dotation monétaire de 5 000€ (cinq mille euros) chacuns·es, afin de les encourager à poursuivre leurs activités
Contact et règlement
Pour toutes les informations concernant les critères de sélection, les conditions d’éligibilité et le contenu du dossier de candidature, veuillez consulter l’appel à candidatures.
Pour toute précisions ou informations supplémentaires, écrivez à activites-instances@auf.org
Date de publication : 21/05/2025
Date limite : 27/06/2025"
https://www.auf.org/nouvelles/appels-a-candidatures/12e-edition-du-prix-de-la-francophonie-pour-jeunes-chercheurs/
#metaglossia_mundus
"Ouverture des Candidatures pour la 18ème Edition du ″Prix de la Traduction Ibn Khaldoun-Senghor″ - 2025
L’Organisation Arabe pour l'Education, la Culture et les Sciences (ALECSO) et l'Organisation Internationale de la Francophonie (OIF) ont le plaisir d'annoncer l'ouverture des candidatures pour la 18e édition du « Prix Ibn Khaldoun-Senghor de la traduction″ - 2025.
Le prix a pour objectif de récompenser les traducteurs qui excellent dans la traduction d’un ouvrage de l'arabe vers le français ou du français vers l'arabe dans le domaine de la littérature ou des sciences humaines et sociales.
La date limite de dépôt des candidatures est fixée au 30 juin 2025.
https://www.alecso.org/nsite/fr/newscat/4661
#metaglossia_mundus
"Kenya : Ngugi wa Thiong’o, l’homme qui voulait décoloniser l’esprit par la langue
La Rédaction 2 juin 2025
Le célèbre écrivain kényan Ngugi wa Thiong’o est décédé mercredi 28 mai au matin à l’âge de 87 ans. Auteur d’une œuvre ancrée dans sa terre natale, il a refusé de suivre la tradition occidentale en écrivant dans sa langue, le Kikuyu. C’était pour lui, un moyen plus puissant de décoloniser l’esprit de son peuple et de faire une littérature par lui et pour lui.
Le célèbre écrivain kényan Ngugi wa Thiong’o nous a quittés le mercredi 28 mai à l’âge de 87 ans. C’est sa fille qui a annoncé son décès sur Facebook. « C’est avec le cœur lourd que nous annonçons le décès de notre père, Ngugi wa Thiong’o, ce mercredi matin », a écrit Wanjiku Wa Ngugi, ajoutant qu’« il a vécu une vie bien remplie et s’est bien battu. ».
Un militant des droits de l’homme de la démocratie au Kenya
Plusieurs personnalités ont salué sa mémoire, mais également des organisations non gouvernementales. C’est le cas d’Amnesty International, dont la branche kényane écrit sur X : « Ayant déjà gagné sa place dans l’histoire du Kenya, il passe de la mortalité à l’immortalité ». Ngugi wa Thiong’o s’est plusieurs fois distingué par son combat pour la liberté et les droits de l’homme dans son pays. Il a été emprisonné par les autorités kényanes en 1977 et 1978, notamment pour avoir écrit des pièces de théâtre qui s’attaquent aux élites du pays.
Ngugi wa Thiong’o était un prétendant régulier au Prix Nobel de littérature
Ngugi wa Thiong’o était surtout connu comme un grand écrivain d’Afrique de l’est. Il a même été pressenti à plusieurs reprises pour remporter le Prix Nobel de littérature. S’il l’avait gagné, le Kényan aurait été le sixième africain à recevoir cette récompense suprême après le Nigérian Wole Soyinka (1986), l’Égyptien Naguib Mahfouz (1988), les Sud-Africains Nadine Gordimer (1991) et J.M. Coetzee (2003) et le Franco-Mauricien J.M.G. Le Clézio (2008). Il a écrit une trentaine de romans, pièces de théâtre, recueils de nouvelles, essais et livres pour enfants.
Un auteur peu traduit en Afrique francophone
S’il est très célèbre dans le monde anglophone, Ngugi wa Thiong’o reste peu connu en Afrique francophone. Et pour cause, seules cinq de ses œuvres ont été traduites à ce jour en français. Ce sont : Et le blé jaillira (Julliard, 1969), Enfant, ne pleure pas (Hatier, 1983), Pétales de sang (Présence africaine, 1985), La Rivière de vie (Présence africaine, 1988) et Décoloniser l’esprit (La Fabrique, 2011). Encore faut-il les trouver en librairie pour découvrir ce talentueux auteur, l’un des rares en Afrique à s’être approprié sa langue maternelle.
Ngugi wa Thiong’o a fait le choix d’écrire en Kikuyu
En effet, alors qu’il s’était déjà fait un nom dans la sphère littéraire avec quatre romans et deux pièces de théâtre, Ngugi wa Thiong’o a fait le choix en 1980 d’écrire en Kikuyu. Aussi désigné par Gikuyu, le Kikuyu est une langue parlée par le peuple du même nom au Kenya. Il s’agit d’une langue bantoue, du vaste phylum Niger-Congo. Dans cette langue qui possède une grammaire bien structurée, il produira notamment les romans Devil on the cross (1980), Matigari (1986) et Wizard of the crow (2004).
Une littérature pour le peuple et par le peuple
En 1986, Ngugi wa Thiong’o consacre un livre entier à sa décision d’écrire en Kikuyu, intitulé Decolonising the Mind (Décoloniser l’esprit). Il y relève la nécessité de s’affranchir de la langue des colons pour exprimer une expérience, des représentations et une culture authentiquement africaines pour ainsi toucher un lectorat populaire. « Dans la mesure où elles sont celles du peuple, les langues africaines ne peuvent qu’être ennemies de l’État néocolonial », écrit-il. Pour lui, il fallait non seulement une littérature du peuple mais également par le peuple, afin que celui-ci s’identifie et contribue plus facilement à l’œuvre de création.
Ngugi wa Thiong’o n’est pas le seul écrivain africain à avoir écrit dans sa langue maternelle
En 2024, dans un entretien à l’Agence France-Presse depuis la Californie, où il vivait en exil, Ngugi wa Thiong’o affirmait encore qu’il croyait « tellement en l’égalité des langues », qu’il se sentait « complètement horrifié » par leur hiérarchisation. S’il a abondamment écrit dans sa langue maternelle, il faut toutefois noter que l’écrivain kényan n’est pas le premier ni le dernier à le faire en Afrique. D’autres auteurs ont relevé le défi. Ils étaient essentiellement anglophones. Cela s’explique par le fait que, contrairement à la politique coloniale en Afrique francophone, le système colonial anglophone était plus ouvert à l’apprentissage des langues locales et à leur usage dans les œuvres littéraires qui étaient acceptées par les éditeurs britanniques."
https://wayeno.net/2025/06/02/kenya-ngugi-wa-thiongo-lhomme-qui-voulait-decoloniser-lesprit-par-la-langue/
#metaglossia_mundus
"La fonction de traduction IA de Zalo permet aux utilisateurs de traduire rapidement des messages anglais-vietnamien et vietnamien-anglais, se connectant facilement avec des amis et partenaires étrangers sans se soucier des barrières linguistiques.
Zing News
02/06/2025
En tant que l'une des principales plateformes de messagerie au Vietnam avec 78 millions d'utilisateurs mensuels, Zalo met régulièrement à jour et améliore les fonctionnalités d'IA pour améliorer l'expérience utilisateur. En utilisant une technologie recherchée et développée par Zalo AI elle-même, la fonction AI Translate de Zalo a montré sa capacité à comprendre la langue vietnamienne. Cela permet de traduire naturellement, avec précision et une grande précision, même avec des termes vietnamiens ou de l'argot.
Fonctionnalité de traduction IA de l'application Zalo.
Dans le contexte de la mondialisation, la communication transfrontalière est devenue un élément essentiel de la vie. La fonction de traduction IA de Zalo permet de briser les barrières linguistiques pour 78 millions d'utilisateurs, en connectant les utilisateurs nationaux au monde entier , garantissant une communication fluide et facile. Cette fonctionnalité est efficace dans des domaines tels que le travail avec des étrangers, les affaires, la fabrication, le tourisme et les services hôteliers, etc. Ce n'est pas seulement un outil de traduction, mais ouvre également la porte à l'intégration et à un pont culturel entre le Vietnam et le monde. De plus, la fonction de traduction IA de Zalo est également choisie par de nombreux jeunes utilisateurs de la génération Z, contribuant ainsi à soutenir l'apprentissage et à se connecter avec des amis dans le monde entier.
On estime que 1,1 million d'utilisateurs utilisent régulièrement AI Translation chaque mois avec plus de 26 millions de demandes de traduction.
Instructions pour utiliser la fonction de traduction IA sur Zalo.
Selon le rapport « The Connected Consumer Q4/2024 » publié par Decision Lab, Zalo est la plateforme la plus appréciée des 3 groupes de générations d'utilisateurs : Gen X, Gen Y et Gen Z, en particulier des utilisateurs de Gen X - ceux nés entre 1965 et 1980. Les données d'enquête montrent que 68 % des utilisateurs de Gen X aiment utiliser Zalo pour la messagerie quotidienne, tandis que le taux d'utilisation de Facebook et Messenger pour la messagerie dans Gen X n'est que de 14 % et 12 %, respectivement.
L'intégration par Zalo de nombreuses fonctionnalités d'IA directement dans la plateforme pour près de 78 millions d'utilisateurs actifs mensuels confirme également la stratégie « AI-First » de la plateforme de messagerie la plus populaire au Vietnam à l'heure actuelle. On estime qu'environ 23 % des utilisateurs de Zalo utilisent des fonctionnalités liées à l'IA. En intégrant les fonctionnalités de l'IA dans la vie quotidienne des utilisateurs vietnamiens, Zalo démontre sa capacité à maîtriser la technologie et sa position de pionnier en matière d'investissement et de recherche au service des utilisateurs, de sorte que l'intelligence artificielle n'est pas un concept lointain mais a de nombreuses applications pratiques dans la vie.
Zalo brise les barrières linguistiques grâce à la fonction de traduction des messages sur l'application.
En particulier, avec la détermination du Parti et du Gouvernement à façonner et à promouvoir l'ère du développement technologique du pays, démontrée par la Résolution 57-NQ/TW (Percée dans le développement de la science, de la technologie, de l'innovation et de la transformation numérique nationale) et le Décret 147/2024/ND-CP (Création d'un espace de réseau social transparent, sûr et responsable), Zalo affirme activement son rôle de pionnier, apportant une contribution importante au mouvement « Éducation numérique pour tous » en particulier et à la promotion de la transformation numérique et de l'économie numérique au Vietnam en général.
https://www.vietnam.vn/fr/nen-tang-viet-xu-ly-26-trieu-yeu-cau-dich-thang-cho-1-1-trieu-nguoi
#metaglossia_mundus
Raymond Théberge estime que l'offre d'événements de recrutement en français ou bilingue de la GRC à l'extérieur du Québec est insuffisante.
"La GRC ne fait pas assez d'efforts de recrutement auprès des francophones en situation minoritaire, juge le commissaire Raymond Théberge
La Gendarmerie royale du Canada (GRC) se retrouve une nouvelle fois dans la mire du commissaire aux langues officielles. Raymond Théberge estime que son offre d'événements de recrutement en français ou en format bilingue à l'extérieur du Québec est insuffisante.
Le problème [...] est particulièrement grave parce qu’on sait déjà [...] que ça fait 10-20 ans que la GRC a de la misère à offrir certains de ses services en français hors Québec. La GRC
Gendarmerie royale du Canada
ne s’aide vraiment pas en refusant de recruter activement et de façon véritablement équivalente des francophones hors Québec, estime le spécialiste des droits linguistiques, Mark Power.
L’avocat franco-ontarien chez Juristes Power Law est estomaqué à la lecture du rapport d’enquête du Commissariat aux langues officielles que Radio-Canada lui a transmis pour l'analyser.
Même s’il ne connaît pas le plaignant, il estime que celui-ci a rendu un fier service au public en se tournant vers le commissaire.
Le plaignant reproche à l’institution fédérale qu'en date du 11 avril 2024, elle a organisé des évènements de recrutement à l’extérieur du Québec – notamment en Colombie-Britannique, en Alberta, en Saskatchewan, au Manitoba, à l’Île-du-Prince Édouard et en Nouvelle-Écosse – uniquement en anglais, et ce, même dans les bureaux désignés bilingues.
Il note également qu'à pareille date, en Ontario, des 14 évènements de recrutement affichés sur le site Web de la GRC pour la région de l’Ontario, un seul était offert en français.
Sa plainte a été jugée fondée par le Commissariat aux langues officielles.
La minorité francophone désavantagée
Dans sa réponse au commissaire, la GRC assure qu’elle fournit des efforts soutenus pour rendre ses processus accessibles aux francophones en organisant des activités sur demande ou en participant à des événements ponctuels et régionaux qui peuvent être offerts en français à l’occasion et au besoin.
Même si ceux-ci ne sont pas toujours affichés sur son site internet, comme les activités de réseautage, l'institution cite en exemple des séances de recrutement trimestrielles en français organisées par la division de la Colombie-Britannique ou, en Nouvelle-Écosse, des évènements de promotion des carrières dans des écoles francophones.
Une réponse insuffisante pour le commissaire.
Dans son rapport d'enquête final daté de mars 2025 et dont Radio-Canada a obtenu copie, M. Théberge explique que les vérifications effectuées [...] ont révélé que l’offre d’évènements de recrutement en français ou en format bilingue hors Québec ne suffit pas, ce qui désavantage la minorité francophone et a une incidence négative sur ses chances d’emploi.
Les citoyens doivent pouvoir choisir l’une ou l’autre des langues officielles sans difficulté additionnelle causée en fonction de la langue choisie...
La GRC, un mauvais élève?
Ce n’est pas la première fois que la GRC se retrouve visée. De 2020 à 2025, l'institution fédérale a fait l'objet de 335 plaintes au Commissariat aux langues officielles.
...
Pour Me Power, une telle situation est inacceptable.
C'est la Gendarmerie royale du Canada et non la Gendarmerie royale du Canada anglais!
Une citation deMark Power, avocat chez Juristes Power Law
En science économique, il existe un principe voulant que l'offre crée sa propre demande. En droits linguistiques, au Canada, on parle d'offre active. La GRC gagnerait à faire beaucoup plus que de réaffecter des Québécois francophones et des Acadiens vers les autres régions du pays. Les gens sont là, il suffit de les trouver.
Radio-Canada a contacté la GRC qui n'avait pas répondu au moment de publier ce texte."
Benjamin Vachet
Publié à 9 h 00 UTC+1
https://ici.radio-canada.ca/nouvelle/2168691/grc-bilinguisme-plainte-francais-recrutement
#metaglossia_mundus
"Le 27 mai célèbre la diversité. Mais que signifie ce mot récent qui ne recouvre pas la même chose en anglais et en français?
La notion de diversité est née aux Etats-Unis dans un contexte historique précis...
Ce choix sous-entend que l'on est favorable à la défense de la diversité, qui est parfois devenue un argument quasi-marketing, expression d'une volonté de lutte pour l'égalité des chances et contre les discriminations.
Un mot traverse un océan
Aussi étrange que cela puisse paraître, le mot "diversité" est assez récent, dans la langue française. Il provient des Etats-Unis, où la notion de "diversity" s'est développée historiquement au moment où les politiques antidiscriminatoires se sont institutionnalisées, dans les années 1960.
Et notamment avec la mise en place de la politique d' "affirmative action" (mesures de "discrimination positive") après la promulgation du Civil Rights Act, en 1964.
En Europe aussi, la notion de diversité s'est répandue dans l'espace public avec la mise en place de normes juridiques. C'était dans les années 2000.
Mais le terme ne recouvre pas tout à fait la même réalité en français que dans le monde anglo-saxon.
"Diversity" vs. Donald Trump
Aux États-Unis, la diversité concerne les "minorités visibles", les différences ethniques et raciales en premier lieu. Donald Trump a fait de la "diversité" sa bête noire et il combat toute forme de tolérance à la différence dans les différents secteurs de la vie publique américaine, de l'armée aux universités.
Depuis son retour à la Maison blanche, Donald Trump a par exemple stoppé les enquêtes sur les usages excessifs de la force par la police envers les Afro-Américains, décidées à la suite du meurtre de George Floyd il y a cinq ans, et il est revenu aussi sur les initiatives d'embauche en faveur des minorités.
Il y a quelques jours, le président américain s'est félicité d'avoir "libéré" les forces armées de l'influence des théories sur le genre ou les inégalités raciales. Donald Trump a qualifié ces théories de "diversions" par rapport à la "mission essentielle [de l'armée] d'anéantir les ennemis de l'Amérique".
Multiplicité et différences
En français, le terme désigne plus largement la multiplicité des êtres humains, la variété de leur mode de vie, de leurs cultures, de leurs préférences individuelles.
C'est un mot qui prône l'inclusion de toutes et tous, sans distinction et surtout sans hiérarchie. En théorie, parler de diversité, la nommer, la reconnaître, c'est donc un premier pas pour accepter nos différences, qu'elles soient visibles ou non..."
(extrait de : Marie Duru-Bellat. La diversité : esquisse de critique sociologique. 2011. ffhal-00972952f)"
Sandrine Blanchard
27/05/202527 mai 2025
https://www.dw.com/fr/diversit%C3%A9-et-diversity/a-72690083
#metaglossia_mundus
"Advocates push for improved interpreter access
Ruby Pratka, Local Journalism Initiative reporter
editor@qctonline.com
Health care advocates are calling for improved access to interpreters in Quebec City hospitals after it emerged that an English-speaking immigrant mother was not offered interpretation services while hospitalized at the Centre hospitalier universitaire de Québec (CHU) for an emergency cesarean section last year.
For members of the English-speaking community, the right to receive health care services in English is enshrined in the Health Act, although only designated bilingual institutions systematically provide care in both languages; the CHU is not a designated institution. The sole designated institution in the region, Jeffery Hale Hospital, does not have a labour and delivery unit.
Patients across the province have access to an “interpreter bank” co-ordinated by Santé Québec, with interpreters in more than 100 languages, including English. However, evidence suggests that the bank is relatively little used by English speakers in the region – only 159 requests for English interpretation were made at the CHU from 2020-2025, compared to 4,456 for Spanish and nearly 1,400 for Swahili, according to Santé Québec. In the 2021 census, 10,130 people in the Quebec City area named English as their first official language spoken, while 7,850 said they spoke Spanish as a primary language and only 950 were Swahili speakers.
Anecdotal reports also suggest that health care providers don’t always ask patients whether they want an interpreter, leaving it up to the patient or their caregiver to request one.
“When you’re having contractions, you don’t have the headspace to ask for an interpreter,” said Marielle M’Bangha, co-ordinator of the Service de référence en périnatalité pour les femmes immigrantes de Québec (SRP- FIQ), which filed a complaint with the CHU on behalf of the anonymous patient, known as Mary. The SRPFIQ provides resources and support for immigrant women during and after pregnancy, including accompaniment for hospital visits. “Some people say it’s infantilizing [to suggest to a patient that they might need an interpreter] but it’s the other way around; the patient needs to understand.”
“The mothers won’t always name their needs … and [the interpreter bank] depends on the availability of the personnel,” she added. “Once, we needed someone in Ukrainian, and that took a while.”
Access to English-speaking hospital staff is “very case by case,” said M’Bangha’s colleague, Hélène Lepage. “Those who speak English will do it, and they’ll do it happily. You can’t expect everyone to speak English, but it would be good to have someone on call for critical moments.”
Service “needs to be more widely known”
“The 24/7 emergency interpreter service is available throughout all departments of the CIUSSS de la Capitale-Nationale, and several programs regularly offer interpretation services to patients,” said CIUSSS spokesperson Mariane Lajoie in an email. “Sometimes, the patient or their representative may request interpretation services themselves, but when an appointment is scheduled and the staff member knows there is a language issue, an interpreter will be provided.” No CIUSSS representative was available for a follow-up interview at press time.
“The system failed this woman,” said Jennifer Johnson, executive director of the Community Health and Social Services Network (CHSSN), which advocates for health care access in English in the regions. “The resources and tools that should have been able to help her are in place … although [services] are supposed to be organized in advance, so I can see how that process could fail in a crisis situation.”
Johnson said the existence of the interpreter bank “needs to be more widely known among English speakers.” She cited a recent survey of English-speaking Quebecers which showed that nearly 30 per cent of English speakers in the Capitale-Nationale did not feel comfortable asking for help in English at health institutions. Forty-seven per cent had language assistance, although only one-fourth of those used a professional interpreter. One-third relied on friends and family, which Johnson said was a risky decision. “Friends and family don’t have medical training and may misinterpret something or omit an important detail. [Asking friends and family to interpret] is a very bad practice that people are resorting to because they don’t understand that interpreters are available to them.”
“If you ask for an interpreter, it’s the health institution’s responsibility to get one,” Johnson said. “If you don’t ask for [an interpreter], you most likely will not get one. You shouldn’t be afraid to ask for it, because it’s your health [on the line].”"
https://qcna.qc.ca/advocates-push-for-improved-interpreter-access/
#metaglossia_mundus
"Kim Hye-soon has been shortlisted for the International Prize for Literature awarded by the House of World Cultures (HKW) in Germany for her poetry collection "Autobiography of Death," published in German translation this February.
HKW announced Wednesday the six finalists for this year’s prize: Kim, Turkish German writer Dogan Akhanli, Canadian writer Sarah Bernstein, Ukrainian writer Anna Melikova, French writer Neige Sinno and American novelist Jesmyn Ward.
The award is jointly presented to both the author and the translator. Kim's work was co-translated from Korean by Park Soo and Uljana Wolf, who have been named as finalists alongside her.
"Autobiography of Death" was first published in Korea in 2016. The collection consists of 49 poems, inspired by the poet’s collapse at a subway station in 2015 and her reflections on collective tragedies such as the Sewol ferry disaster and the MERS outbreak.
The collection was translated into English by poet Choi Don-mee and won the prestigious Griffin Poetry Prize in Canada in 2019, making Kim the first Korean recipient of the award.
The award introduced the book as "a choir, each voice demanding a personal and therefore dignified death for itself. It celebrates both the fragile, enigmatic and unique inner world of each human being, and the connecting timelessness of cultural images, stories and worlds of thought."
Established in 2009, the International Prize for Literature honors an outstanding work of contemporary international literature and its first translation into German. Since 2023, poetry translations have also been eligible. The prize carries a total of 35,000 euros ($39,000) -- 20,000 euros for the author and 15,000 euros for the translator -- and will be awarded in the summer of 2025 during a literary festival hosted by HKW.
In 2017, Nobel Prize-winning author Han Kang was also a finalist for the award with the German translation of her novel "The Vegetarian."
German edition of "Autobiography of Death" (S. Fischer Verlage)"
May 29, 2025 - 17:14:17
By Hwang Dong-hee
hwangdh@heraldcorp.com
https://m.koreaherald.com/article/10498851
#metaglossia_mundus
"More sign language interpreters will be hired to support the 400,000 hearing-impaired people nationwide, says the Social Development and Human Security Ministry.
They can offer support when such people are receiving healthcare services, struggling at work, or receiving information or legal assistance, said Minister Varawut Silpa-archa.
Only 202 sign language interpreters are registered with the Department of Empowerment of Persons with Disabilities (DEP), he said, adding the number was too small to meet demand.
There are 432,495 hearing-impaired people, with many requiring assistance at schools and hospitals, he said. The DEP will hire more of them, this year and next.
The ministry also plans to employ more hearing-impaired people to work as interpreters so they can provide a true-to-context translation service and help those afflicted with listening difficulties improve their quality of life, Mr Varawut said. Two universities in the kingdom offer a sign language interpretation programme: Ratchasuda College, Mahidol University offers a bachelor's degree programme while Suan Dusit University provides a one-year certificate programme.
The latest initiative is being divided into three phases, including a 135-hour training course in sign language for social security officers as a short-term plan.
In the medium term, the DEP is expected to pilot the initiative by hiring independent interpreters as department officers. They are required to work at the 77 Provincial Disability Service Centres in Thailand, with an 18,000-baht monthly salary.
In this fiscal year, the subcommittee on sign language interpretation encouragement approved Suan Dusit University and the National Association of the Deaf in Thailand (NADT) as sign language interpretation assessment units, while the Deaf Thai Foundation will design an interpretation skill test, said Mr Varawut."
PUBLISHED : 2 JUN 2025 AT 06:28
https://www.bangkokpost.com/thailand/general/3039892/more-interpreters-needed
#metaglossia_mundus
Discover how to blend AI efficiency with NAATI-certified human expertise to deliver accessible, culturally appropriate translations that prioritise accessibility and ethical communication.
"As artificial intelligence (AI) becomes increasingly embedded in our daily lives, there’s a growing concern that these technologies are leaving many Australian communities behind.
While AI promises to bridge communication gaps, it often fails to account for the cultural and literacy nuances that make language meaningful. This is where culturally appropriate and accessible translations become essential in the Australian context.
Australia’s Multicultural Reality
According to the 2021 Australian Census, 27.6% of the population was born overseas, and 22.8% speak a language other than English at home. The top five languages spoken at home (excluding English) are Mandarin, Arabic, Vietnamese, Cantonese, and Punjabi. This linguistic diversity reflects a deeply multicultural society—one that requires communication strategies that go beyond simple translation.
The Digital Divide in AI Language Models
Recent research from Stanford University highlights a stark reality: most large language models (LLMs) are trained predominantly on English or other high-resource languages, leaving speakers of low-resource languages at a disadvantage. These communities not only face reduced access to accurate AI-generated content but also risk being misrepresented or excluded altogether from important communications.
For example, while English speakers benefit from AI tools that understand idioms, tone, and context, speakers of languages like Swahili or Vietnamese often receive generic or inaccurate translations. This isn’t just a technical issue—it’s a matter of equity. As AI becomes a gateway to education, healthcare, and economic opportunity, the lack of culturally sensitive translation can deepen existing inequalities.
Why Cultural Context and Accessibility Matter
Language is more than words—it’s a reflection of culture, values, and lived experience. A literal translation might convey the basic meaning of a sentence, but without cultural context, it can miss the mark entirely. This is especially critical in sectors like healthcare, legal services, and government communication, where misunderstandings can have serious consequences.
But cultural appropriateness is only part of the equation. Accessibility is equally vital. AI-generated translations typically do not account for the literacy levels of the target audience. For example, a machine might produce a grammatically correct translation in Arabic or Simplified Chinese, but if the community reading it has an average literacy level equivalent to Year 7, the message may still be inaccessible—even if technically accurate.
The 2021 Progress in International Reading Literacy Study (PIRLS) found that while Australia’s Year 4 reading scores have improved overall, there is a widening gap between high and low achievers. This disparity is even more pronounced among students from non-English-speaking backgrounds and those in regional or remote areas. These findings underscore the importance of tailoring communication to meet the literacy needs of diverse communities.
This is where professional translators play a crucial role. They don’t just translate words—they adapt tone, structure, and vocabulary to ensure the message is understood by the intended audience. At Sylaba Translations, we prioritise this human-centred approach to ensure that every translation is both culturally appropriate and accessible.
Our Approach: AI + Human Expertise
At Sylaba, we’re big proponents of combining AI with human expertise. While we leverage AI tools for efficiency, we rely on NAATI-certified translators to refine and perfect the output. This ensures the final translation meets the highest standards of sensitivity and clarity.
We also advocate for plain language writing before feeding content into AI systems. Simplifying the source material not only improves translation accuracy but also makes the original message more inclusive for everyone.
Our approach is multi-layered: we start with clear, accessible writing, use AI for initial drafts, and then have professional translators review everything. This ensures the end product is not only accurate but also culturally relevant and easy to understand.
Looking Ahead: A Call for Ethical AI
As we move toward a more connected and automated world, the need for ethical, culturally aware AI is more urgent than ever. Policymakers, developers, and content creators must work together to ensure that AI systems are trained on diverse, high-quality data and evaluated for cultural sensitivity and accessibility.
The future of communication isn’t just about speed or scale—it’s about understanding. And that starts with recognising the value of every voice, in every language, and every culture."
BY ADMIN
JUNE 2, 2025
https://www.techguide.com.au/news/computers-news/why-culturally-appropriate-translations-are-the-next-frontier-in-ai-powered-communication/
#metaglossia_mundus
"Paraphrasis is a podcast dedicated to the art and practice of literary translation, brought to you by a team of graduate students in the Department of Comparative Literature at Harvard."
https://www.paraphrasispodcast.com/podcast
#metaglossia_mundus
Metaphor translation has been a matter of concern in Translation Studies because its interlinguistic transfer can be impeded by cross-cultural and crosslinguistic differences. Since the inception of Conceptual Metaphor Theory (CMT), which focuses on the conceptual structure of metaphorical language, a range of studies have emerged to investigate metaphor translation from a cognitive perspective, presenting an eclectic mix of research questions and methodologies. This paper is targeted at illustrating what the cognitive approach has offered to Translation Studies by providing a critical overview of recent research in metaphor translation from a cognitive perspective. It is pointed out that cognitive theory can get to the heart of metaphor, an essential cognitive device for meaning-making, as well as translation, a cognitive activity. Illustrations from the literature show that a cognitive approach can account for in-depth conceptual transfer in the analysis of product- and process-oriented metaphor translation. The cognitive approach also provides important insights into translation as cross-cultural communication by offering a redefinition of culture. Within this context, the paper provides multilingual illustrations while paying special attention to translation between culturally-distant languages, e.g. English-Chinese and French-Chinese translation. Lastly, it is argued that there is potential in combining cognitive theory with translation theories such as Descriptive Translation Studies and the Interpretive Theory of Translation.
Wenjie Hong, Caroline Rossi. The Cognitive Turn in Metaphor Translation Studies: A Critical Overview. Journal of Translation Studies, 2021, 5 (2), pp.83-115. ⟨hal-03342406⟩
#metaglossia_mundus
A certain amount of political baggage had to be shaken off before such a feat could be realised – not just in the right words, but in the right spirit.
"Waiting for Godot has been translated into Afrikaans: what took so long
At last, the most infamous latecomer in all of literature has arrived – not in the flesh, but in South Africa’s Afrikaans language. Irish playwright Samuel Beckett’s best-known drama, Waiting for Godot, now also lives as Ons Wag vir Godot.
Published and staged in 2024, the translation was inspired by the official centenary of Afrikaans in 2025.
As a Beckett scholar, I think it’s worth asking why Afrikaans is so late on the scene – and why it matters.
Godot in many tongues
First written in French, En attendant Godot was published in 1952 and debuted on stage the next year.
The action involves two tramps, Vladimir and Estragon, who have a series of absurd conversations and encounters as they wait for a man called Godot who never arrives. Beckett would self-translate the drama into English in 1954, calling it “a tragicomedy in two acts”.
Since then, translations of the play have exploded. By 1969 – the year of Beckett’s Nobel Prize for Literature – Waiting for Godot could already be read in dozens of languages, including Albanian, Marathi, and even Icelandic.
Samuel Beckett and South Africa
Beckett’s connections with South Africa are surprisingly varied. As a young man, he unsuccessfully applied for a lectureship at the University of Cape Town. His 1951 novel, Molloy, was translated from French into English with the help of a South African student, Patrick Bowles. And in 1968, Beckett made a donation to the then-banned resistance party, the African National Congress, in the form of a manuscript for auction.
This gesture was unprecedented for the Irish writer, who was wary of political causes. Yet not only did Beckett feel strongly enough about apartheid’s injustices to make this donation, he also refused to let anyone perform his plays before South Africa’s racially segregated audiences.
Read more: The case of the acclaimed South African novel that 'borrows' from Samuel Beckett
Already in 1963 Beckett had signed the petition Playwrights Against Apartheid. He would continue to refuse performance rights in South Africa until 1980, when the Baxter Theatre was allowed to stage Waiting for Godot with a racially integrated cast.
Nevertheless, unauthorised Godots materialised before this. Athol Fugard, the South African playwright whose own dramas were influenced by Beckett, directed one of the earliest South African productions in 1962. Featuring an all-black cast, it testified to the play’s political charge, which Fugard emphasised:
Vladimir and Estragon … were at Sharpeville or the first in at Auschwitz.
It’s reasonable to think that Beckett would have supported this protest performance. But he would probably have denounced the first and unofficial Afrikaans version, Afspraak met Godot, translated by Suseth Brits and performed in 1970 at the Potchefstroom University College (now North-West University) behind closed doors.
For different reasons, Beckett would also have frowned on the substantial “borrowings” in Afrikaans novelist Willem Anker’s 2014 novel, Buys.
Domesticating a European classic
Fully sanctioned by Beckett’s estate and beautifully translated (from the French and English) by now-retired professor of French at the University of the Free State Naòmi Morgan, Ons Wag vir Godot arrives at a different moment altogether.
The translation retains the gallows humour of the original while adding local flavour. For instance, where Vladimir originally names the Eiffel Tower as a picturesque site to commit suicide, his Afrikaans counterpart nominates Van Stadensbrug, a bridge over a ravine in the Eastern Cape. The slave-like Lucky once entertained his master with European dances: “the farandole, the fling, the brawl, the jig, the fandango”. These now become a South African mix: “volkspele, die riel, die pantsula, selfs die horrelpyp” (folk games, riel dance, pantsula dance, a hiding).
In translation-speak, Ons Wag vir Godot is therefore fully “domesticated”: the play’s universality comes through even though – and perhaps even more so because – it’s anchored in a particular place and time.
This struck me when I attended the play’s limited-run production, expertly directed by Dion van Niekerk, at the 2024 Vrystaat Kunstefees (Free State arts festival). Its set managed to thread together subtle South African roadside details: a toppled rubbish bin, pylons on the horizon, a (broken) picnic bench.
In the text itself, we encounter familiar place names, sayings and cultural clues. Consider how Beckett’s abstract phrase “the essential doesn’t change” is grounded in African mythology: “Jakkals verander van hare, maar nie van streke nie” (The leopard doesn’t change its spots). Then there’s the charming touch of the dog in Vladimir’s song snatching “’n stukkie wors” (a piece of sausage particular to South Africa) rather than a measly “bone”.
Godot and the Afrikaans canon
Ons Wag vir Godot achieves its most profound tribute to Beckett and Afrikaans through its intertextual richness. Both the French and English originals are highly allusive texts: they invoke other works of literature to increase their range of meaning and subtlety. Morgan is attuned to this subtlety and to the parallels to be found in Afrikaans literature. There are references to works by canonical Afrikaans writers like Eugène Marais, Totius and C.J. Langenhoven, each adding its own resonance.
Yet the dilemma any translator faces is not so much in bringing in the new, but in striking a balance with the old. Consider the judicious swapping of a line from Percy Bysshe Shelley for a line from C. Louis Leipoldt.
In the English version, Estragon looks up forlornly at the moon and half-quotes the English Romantic poet: “Pale for weariness … Of climbing heaven and staring on the likes of us.” In the Afrikaans, he gives us a fragment from the wistful poem, Die Moormansgat: “ek kyk na die lig van die volle silwermaan” (I behold the light of the full silver moon). At face value, this lacks the detached, woeful quality of Shelley’s line. But read in the context of Leipoldt’s poem, it is every bit as poignant.
The virtue of waiting
“Vladimir would agree,” Morgan concludes in the preface to her translation, “that a century is a decent amount of time to hone a language for the translation of one of the best-known dramas in world literature”.
And indeed, the riches of the Afrikaans language are on display in this sensitive, witty and allusive rendering of Beckett’s European classic. But it’s also true that a certain amount of political baggage had to be shaken off before such a feat could be realised – not just in the right words, but in the right spirit. Of course, if Beckett’s play teaches us anything, it’s the virtue of waiting."
Published: May 29, 2025 4.28pm SAST
Rick de Villiers, University of the Free State
https://theconversation.com/waiting-for-godot-has-been-translated-into-afrikaans-what-took-so-long-257345
#metaglossia_mundus
As English gains ground in Algeria, questions emerge about identity, access, and the cost of sidelining French in education, work, and public life.
"The shift from French to English reflects deeper struggles over postcolonial identity, global integration, and the politics of language...
Several factors are driving this shift...
English is largely seen as a pragmatic tool for global communication, rather than a cultural imposition. This perception distances English from the emotionally and politically charged connotations that continue to surround the French language, especially among younger Algerians who seek to redefine their identity in a postcolonial world.
Yet the pivot to English is far from straightforward. The transition raises profound structural questions, particularly for Algeria’s intellectual and academic elite. Professor Abderrezak Dourari, a linguist and specialist in translatology, warns of the risks of abrupt linguistic substitution on a national scale:
“Even if it is a sovereign decision, replacing one language with another reconfigures the structure and activity of national elites. Scholars and researchers who work in French or Arabic risk being rendered voiceless for a generation.”
The challenge lies not only in learning a new language but in reshaping the cognitive and professional frameworks that come with it.
“One cannot simply switch from French to English across all levels of discourse. The language used by educators is more conceptual and nuanced than that of learners,” Dourari explains, “for professors and researchers, forging a new linguistic habitus will take years, causing performance losses and disrupting the continuity of knowledge and social contribution.”
A multilingual future
This tension between global ambition and internal readiness is at the heart of Algeria’s language dilemma. While English may promise greater access to international networks, research collaboration, and economic mobility, the sudden displacement of French could destabilize long-established academic and professional ecosystems.
The impact of these shifts extends beyond classrooms. In business, media, and even everyday conversation, the competition between French and English reflects deeper questions about Algeria’s postcolonial identity and its orientation in the world. Public discourse on national language policy is gaining visibility, often framed around whether Algeria should pursue full integration into Anglophone global circuits or maintain its Francophone heritage as a cultural and intellectual anchor...
This ongoing transformation reflects broader global patterns where languages do not merely coexist but often compete in shaping access, power, and belonging. For Algeria, the challenge lies in managing this shift with foresight and inclusiveness, ensuring that the embrace of English does not come at the cost of continuity, equity, or cultural depth..."
By Nourredine Bessadi, researcher, independent consultant, and translator
2 June 2025
https://globalvoices.org/2025/06/02/algerias-shift-to-english-is-about-more-than-just-language/
#metaglossia_mundus
New rules in Quebec require French to be the dominant language on storefront and product packaging, sparking concerns about costs, compliance and fines.
"French must be “markedly predominant” on store signs
Under the new rules, French must take up twice as much space as other languages on store signs and commercial advertising. That means stores with English names, like Canadian Tire, Best Buy and Second Cup will have to include generic terms or descriptions in French on their storefronts that take up two-thirds of the space devoted to text.
Michel Rochette, Quebec president of the Retail Council of Canada, said businesses must ensure they’re complying with municipal bylaws and landlords’ requirements when changing their storefronts, which can be time-consuming.
“They want to comply with the rules. It’s not a question of willingness,” he said. “It’s a question of capacity and authorization.”
But last week, French-language Minister Jean-Francois Roberge said many companies have already updated their signs, and pointed specifically to food retailer Bulk Barn as being largely in compliance. “It’s possible to do it,” he said...
Labels on product packaging must already be translated into French.
But there’s an exception for trademarks in other languages, which don’t have to be translated.
The latest regulations take aim at generic terms sometimes included in trademarks, like “lavender and shea butter” hand soap, for example. Those descriptive terms will now have to be translated as well...
Smaller businesses must register with language watchdog
Quebec already requires businesses with 50 or more employees to undergo a “francization” process to ensure French is the dominant language in the workplace. That requirement is now being extended to companies with 25 to 49 employees.
The businesses must register with Quebec’s language watchdog, the Office quebecois de la langue francaise. Francois Vincent, Quebec vice president of the Canadian Federation of Independent Business, said the majority of small businesses in the province already operate in French...
Fines can reach $30,000 a day
Businesses in violation of the new rules can be fined $3,000 to $30,000 per day for a first infraction, and up to $90,000 per day for a third offence...
However, Roberge has said the language office won’t be looking to slap immediate fines on non-compliant businesses, as long as they’re taking steps to fix the problem.
Rochette and Vincent issued an open letter last week asking for an extension from the Quebec government. They say the province had promised companies would have three years to adapt to the new rules. Bill 96 became law in 2022, but the final version of these regulations was only published in June 2024.
The government confirmed Friday that the regulations would take effect on June 1, as planned. “In one year, there’s time to do a lot of things,” Roberge said earlier in the week..."
The Canadian Press
Maura Forrest
Jun 01, 2025
https://nationalpost.com/news/canada/quebecs-tougher-language-laws-begin-for-signage-packaging-and-workplaces
#metaglossia_mundus
In a striking demonstration of real-time augmented reality, Rokid captivated international guests at a recent event in Hangzhou
"...This immersive experience highlighted Rokid’s cutting-edge on-lens translation feature, turning the glasses into a live communication bridge between cultures. As one French delegate greeted the host city, his words were automatically rendered in Chinese on the receiving end:
“ Merci beaucoup, Monsieur le député-maire. Je suis très honoré de visiter votre ville. ”
“Thank you very much, Mr. Mayor. I’m very honored to visit your city. ”
The Mayor responded and remarked in Chinese:
“Yuhang, c’est une entreprise créatrice qui a accueilli plusieurs milliers d’entreprises technologiques. Ce produit est l’une de nos entreprises créatrices qui a juste été développée.”
“Yuhang is an innovation hub that has hosted thousands of technology companies. This product is one of the innovative creations, which was just recently developed.”
The delegate echoed the sentiment, impressed by the system’s real-time performance:
“C’est très impressionnant et ça fonctionne plutôt bien. Je suis très impressionné par le système. J’espère que lors de la prochaine conférence générale, nous pourrons tous en profiter pour pouvoir communiquer tous ensemble.”
“It’s very impressive, and it works quite well. I’m very impressed by the system. I hope we can all use it to communicate at the next general conference.”
Many in the room were impressed by the spontaneous nature of the exchange. Conversations traditionally requiring human translators unfolded organically, with glasses as the bridge. The demonstration underscored the potential of Rokid’s real-time subtitle feature for international business and travel, education, and diplomacy. By eliminating communication barriers, the Rokid Glasses represent more than a technological leap. They mark a shift in how we connect as global citizens. This breakthrough moment in Hangzhou proved Rokid’s mission to “Leave Nobody Behind”, including language, access, and understanding."
https://www.laweekly.com/seamless-on-lens-translation-impresses-international-guests-at-rokid-event-in-hangzhou/
#metaglossia_mundus
"University of Washington team builds multi-speaker, spatial audio translation system
28/05/2025
A team from the University of Washington (UW) has developed a prototype translation system that tackles a challenge facing many public space translation tools: understanding and distinguishing between multiple speakers in real-world, often noisy environments.
According to a report from the University of Washington, the new system, dubbed Spatial Speech Translation, uses off-the-shelf noise-cancelling headphones equipped with microphones to isolate, translate and relay speech from multiple people in a space. Unlike conventional translation technologies that focus on a single speaker, the UW system preserves the direction and characteristics of each speaker’s voice and follows them as they move.
The project was presented at the ACM CHI Conference on Human Factors in Computing Systems in Yokohama, Japan, on April 30 and the research team’s proof-of-concept code is now publicly available for others to build upon.
The system emerged from the frustrations of lead author Tuochao Chen, a UW doctoral student, who struggled to understand a museum tour in Mexico due to noisy conditions. “Other translation tech is built on the assumption that only one person is speaking,” said senior author Shyam Gollakota, a professor in the Paul G. Allen School of Computer Science & Engineering at UW, in an article on the university’s website. “But in the real world, you can’t have just one robotic voice talking for multiple people in a room. For the first time, we’ve preserved the sound of each person’s voice and the direction it’s coming from.”
The technology introduces three main innovations. First, it uses a radar-like algorithm to scan a space in 360 degrees, instantly detecting how many people are speaking. Then, it processes translations while maintaining each speaker’s voice quality and volume. Lastly, it dynamically adapts as users move, ensuring spatial audio cues remain accurate.
Rather than relying on cloud processing, which can raise privacy concerns around voice data, the system runs locally on devices powered by Apple’s M2 chip, such as laptops and the Apple Vision Pro.
The team tested the system in ten real-world settings, indoors and out. In user trials involving 29 participants, the prototype was preferred over baseline models that lacked spatial tracking. Another study showed that users found a 3 to 4 second translation delay more acceptable than shorter latencies, which tended to produce more errors.
While the system currently supports common speech in Spanish, German and French, the researchers note that existing translation models can be expanded to cover up to 100 languages. However, specialised terminology and jargon remain outside the tool’s current capabilities.
“This is a step toward breaking down the language barriers between cultures,” said Chen in the original University of Washington article. “So if I’m walking down the street in Mexico, even though I don’t speak Spanish, I can translate all the people’s voices and know who said what.”
The research was supported by a Moore Inventor Fellow award and the UW CoMotion Innovation Gap Fund. Co-authors include Qirui Wang, a research intern at HydroX AI and former UW undergraduate, and Runlin He, a UW doctoral student."
https://www.inavateonthenet.net/news/article/university-of-washington-team-builds-multispeaker-spatial-audio-translation-system
#metaglossia_mundus
"Ottawa rushes to build its own AI translator as government use of free tools soars
Demand for official translation services is falling despite a rise in content creation. As public servants turn to free online tools, the government is racing to build its own AI translator.
By David Reevely and Murad Hemmadi
A stop sign in both English and French in Kingston, Ontario. The federal government employs 1,300 people to translate a range of documents.
May 28, 2025
OTTAWA and TORONTO — The federal government’s translation bureau is rushing to devise an AI-based tool for public servants after being spooked by the use of free services on potentially sensitive materials.
“Industry is moving at lightning speed and the bureau needs to accelerate the cadence to transform its services,” says a presentation deck produced by the bureau last December, which The Logic obtained through an access-to-information request.
Talking Points
The federal government plans to roll out a single, AI-based tool for all public servants. Some departments have already built their own tools, while staff are also using insecure free online products on potentially sensitive documents.
Demand within the government for the translation bureau’s services is falling, but workers face pressure to handle larger workloads with the use of AI, according to their union
Starting in June, the federal department that houses the government translation service, Public Services and Procurement Canada (PSPC), is planning a trial run of an in-house tool it’s calling PSPC Translate, departmental spokesperson Jeremy Link told The Logic by email.
If PSPC Translate works for staff in the department of about 19,000 people, the tool will be used more widely. It’s the first “lighthouse project” under the federal government’s artificial intelligence strategy, meant to help the public service learn how to build and implement new AI tools and create something that can be scaled across government.
Although this effort long pre-dates Prime Minister Mark Carney’s run for office, he’s pinning a lot on PSPC Translate and projects like it, promising to use artificial intelligence to make the government more productive so the Liberals can spend less on operating expenses.
Translation is one area where public servants are already using AI to transform their own work—if not always for the better.
The 1,300 workers in the bureau translate between English and French, but also into Indigenous languages, foreign languages and sign languages when the government, or contractors doing government work, need it.
The unit reported a 17 per cent decline in demand for its services from federal departments and agencies in the 2023–24 fiscal year, the deck says, even though “all indicators are showing that content creation is on the rise.”
The reason, the bureau concluded, was that people were going elsewhere. Many public servants are using “free internet tools,” scattering government data all over the place, including onto servers outside Canada. Some departments are creating their own tools, wasting money through duplication and complicating the work of public servants who use different tools in different departments.
While demand for the bureau’s services is dropping, its human translators face growing workloads because it’s outsourcing less to freelancers, according to Antoine Hersberger, vice-president for the Canadian Association of Professional Employees unit that represents the workers. The union claims the agency’s five-year business plan would cut a quarter of staff via attrition.
“We’re pushed to do higher and higher volumes, especially using AI tools to force us to work faster and [at] lower quality,” Hersberger said. AI tools are quicker for some translation tasks, but humans still have to check outputs. And, Hersberger explained, public servants using the bureau’s self-service tool can also ask for a human review of their AI translation, adding to the translators’ workloads.
Many free services like Google Translate are powered by large language models, the same technology behind ChatGPT. The bureau’s translators currently use commercially available software from DeepL and TradooIT, which employs an older but more accurate form of AI called neural machine translation.
But both kinds of tools, and even careless use of a dictionary, can generate errors. The deck includes images of bad translations, such as a sign at what appears to be a security screening site telling people what to do with keys, coins, pocket knives and smartphones. People reading English are told to put these items “in a bin.” French-speakers are told to Placez ces objets dans une poubelle—“put these objects in the garbage.”
In another example, a road sign in English tells cyclists to ride single file. The French mistranslation says to ride un seul fichier, using a word, fichier, that means a bundle of documents, not a line of people; à la file is what the writer wanted.
(A web search for “single file” turns up the Parisian-French idiom en file indienne as the favoured translation, which any Canadian public servant would avoid. Seul fichier is the next suggestion.)
The translation bureau has been dealing with such quirks its whole existence and has an “extensive repository” of bilingual texts that it has used to train AI models to create “accurate translations,” PSPC’s Link wrote.
In the absence of a centrally run translation tool, some departments have created their own. Justice Canada has been using its own AI-based translator, called JUSTranslate, since October 2023, according to the documents The Logic obtained. It uses Microsoft’s Azure cloud-computing platform and can translate text from multiple languages into English or French.
One of the Justice Department’s needs is secrecy: pasting documents subject to solicitor-client privilege into public cloud-based translation services isn’t allowed. JUSTranslate (like PSPC Translate) is approved for what the government calls “Protected B” material. That designation means that if it were released, the material “could cause serious injury to an individual, organization or government.”
Another issue is formatting. Not for official legal publications—those still go through human translators—but for materials like presentation decks and even written submissions to courts and tribunals. Those ought to look more or less the same in both English and French, but substituting text slide by slide is a pain and a time-sink.
In its first year the Justice Department has used JUSTranslate on about 20,000 documents, including in tests and experiments, the documents say, working out to more than 57 million words. Some of that work would otherwise have been done by the translation bureau, the department acknowledged.
For ordinary Canadians who use free AI translation tools, the money the federal government spends on the bureau may seem like a waste, Hersberger acknowledged. But clients who switch to AI tools “usually run back to human translators after a few months” after finding “massive mistakes or omissions in their documents,” he claimed.
As an officially bilingual country, Canada “needs to ensure the same level of quality for both languages,” Hersberger said, and “the technological tools are just not enough right now to do that.” Instead of shrinking the bureau, he called for the government to provide funding so that translators can experiment with new technologies like AI while maintaining the quality of their work.
Longer-term plans for PSPC Translate include specialized translation tools for technical fields, adding Indigenous languages, voice-to-text transcribers, and eventually AI-assisted live interpretation, according to the documents obtained by The Logic."
https://thelogic.co/news/exclusive/ottawa-rushes-to-build-its-own-ai-translator-as-government-use-of-free-tools-soars/
#metaglossia_mundus
"Journal of Translation 21(1) (2025)
Journal of Translation 21(1)
The latest issue of the Journal of Translation, JOT 21.1 (2025), is currently being published on a rolling basis. At the end of the year once the issue is complete, the Editor’s Foreword will appear here and this file will be replaced with a single, full-issue PDF. For more information and to view articles already published in this issue, please visit https://www.sil.org/resources/publications/jot/21.1.
Amateur English-Kiswahili Interpretation in Tanzanian Pentecostal Churches: Challenges and Strategies
Mlundi, Simon
Interpreters apply different strategies to overcome challenges during the interpretation process. This study examines the frequency of strategies selected and utilised by amateur church interpreters in Tanzania in overcoming challenges of translation between English and Kiswahili. The analysis reveals that these interpreters use strategies such as expansion, compression, and skipping, as well as approximation, message abandonment, and incomplete sentences. More practice and self-training are recommended for amateur interpreters to use interpreting strategies more effectively.
The Great Taboo: Omission and Addition in the Mongolian Translation of “Genghis Khan and the Making of the Modern World”
Dorjsuren, Suvdanchuluun
In literary translation, alterations such as omission and addition are often made to adapt the source text to the target language. But how frequently do translators make such changes, and what motivates these alterations? This paper explores these questions by applying both qualitative and quantitative methods to compare Weatherford’s Genghis Khan and the Making of the Modern World with its Mongolian translation by Ganbold Gonchig. The theoretical framework of this study posits that translators modify the source text to accommodate linguistic, pragmatic, cultural, and ideological..."
https://www.sil.org/resources/publications/jot/21.1
#metaglossia_mundus
"Oman launches sign language interpretation for social media posts
31 MAY 2025
Muscat – In a move to promote inclusivity and accessibility, the Consumer Protection Authority (CPA) has launched a new service to interpret its posts on the social media platform X into sign language.
The service, the first of its kind in the sultanate, ensures that vital consumer awareness messages, alerts, and campaign updates reach members of the deaf and hard-of-hearing community.
The selected video posts will feature a sign language interpreter, making the content accessible to a wider audience.
“This initiative reflects our commitment to serve all segments of society and to promote equal access to information,” a CPA official said. “Sign language is a vital communication tool, and through this effort, we aim to bridge the gap and empower every consumer.”
The CPA plans to expand the service to cover a broader range of topics and interactive content in the coming months.
The initiative also aligns with Oman’s national strategy to improve digital accessibility and uphold the rights of people with disabilities."
https://www.muscatdaily.com/2025/05/31/consumer-authority-of-oman-launches-sign-language-interpretation-for-social-media-posts/
#metaglossia_mundus
|
Scoop.it!
