 Your new post is loading...
|
Scooped by
Charles Tiayon
February 3, 9:58 AM
|
WAXAL is an open-source speech database designed to support the development of voice-based artificial intelligence for African languages. "Google has collaborated with African universities and research institutions to launch WAXAL, an open-source speech database designed to support the development of voice-based artificial intelligence for African languages. African institutions, including Makerere University in Uganda, the University of Ghana, Digital Umuganda in Rwanda, and the African Institute for Mathematical Sciences (AIMS), participated in the data collection for this initiative. The dataset provides foundational data for 21 Sub-Saharan African languages, including Hausa, Luganda, Yoruba, and Acholi. WAXAL is designed to support the development of speech recognition systems, voice assistants, text-to-speech tools, and other voice-enabled applications across sectors such as education, healthcare, agriculture, and public services. “This dataset provides the critical foundation for students, researchers, and entrepreneurs to build technology on their own terms, in their own languages,” said Aisha Walcott-Bryantt, Head of Google Research Africa WAXAL’s launch comes amid growing efforts across Africa to develop language technologies that reflect local cultures and realities. In September 2025, the Nigerian government unveiled N-ATLAS, an open-source language model capable of recognising and transcribing spoken words and generating text, in Yoruba, Hausa, Igbo, and Nigerian-accented English. Similar initiatives are emerging in the private sector, where startups such as South Africa’s Lelapa AI are building tools like Vulavula, which offers speech recognition, translation, and sentiment analysis. By making this speech dataset openly accessible, WAXAL provides the fuel for a growing wave of homegrown efforts to bring African languages into the digital age. Although Sub-Saharan Africa is home to more than 2,000 languages, reports suggest that fewer than 5% of those languages have the resources needed for Natural Language Processing (NLP), which allows computers to understand and comprehend human language. This lack of representation in training datasets limits the effectiveness of speech recognition and text-to-speech systems for African users. Developed over three years with funding and technical support from Google, WAXAL addresses a major gap in global AI development. WAXAL provides speech data for 21 Sub-Saharan African languages, including Fulani (Fula), Hausa, Igbo, Ikposo (Kposo), Swahili, and Yoruba. The dataset contains more than 11,000 hours of speech drawn from nearly two million individual recordings. Under the project’s partnership model, contributing institutions retain ownership of the data they collected, while making it openly available to researchers and developers worldwide. “For AI to have a real impact in Africa, it must speak our languages and understand our contexts,” Joyce Nakatumba-Nabende, Senior Lecturer at Makerere University’s School of Computing and Information Technology, said. “The WAXAL dataset gives our researchers the high-quality data they need to build speech technologies that reflect our unique communities.”" Opeyemi Kareem 2nd Feb, 2026 https://techcabal.com/2026/02/02/google-joins-push-to-localise-ai-for-african-languages-with-speech-database/ #Metaglossia #metaglossia_mundus #métaglossie
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
"Ever since COVID-19, video conferencing apps have been a staple web tool in almost every working professional and student's life. Platforms like Google Meet, Microsoft Teams, Zoom, and more have constantly gained new bells and whistles since, with Google now looking to abolish language barriers with its latest Google Meet feature expansion.
Google Meet on computers supports real-time speech translation. Unlike translated captions, speech translation allows participants to speak in their native language, while the platform translates the speech in real-time into spoken sentences, essentially preserving the "flow of conversation by creating an audio translation dubbed over the original speech that mimics the speaker’s tone and speaking cadence."
Up until now, this feature has been limited to Google Meet on computers. Leaks recently suggested that the feature might make its way to the platform's mobile apps, and that's exactly what's happening now.
In a new Workspace updates post announcing speech translation's general availability for businesses, the tech giant also announced the feature's mobile expansion.
You can't try it out just yet
Speech translation will roll out to the Meet Android and iOS apps in the coming months.
In addition to the mobile expansion, the tech giant also indicated that it will make visual updates to the speech translation user interface. This applies to the feature's UI on computers. Additionally, Google will also make refinements to translation accuracy and nuance, which should apply to the feature on all surfaces.
It's worth noting that speech translation is limited to Business Standard and Plus, Enterprise Standard and Plus. Frontline Plus, Google AI Pro and Ultra customers. Those with a Google AI Ultra for Business add-on or a Google AI Pro for Education add-on can also access the feature."
Karandeep Singh Oberoi
Feb 4, 2026, 5:24 PM EST
https://www.androidpolice.com/google-meet-speech-translation-mobile-expansion/
#Metaglossia
#metaglossia_mundus
#métaglossie
"New study explores the value placed on support from family, friends, and society by those in various kinds of intercultural relationships.
Friends approving of romantic partners matters: research shows it's not just socially convenient but also linked to relationship quality and even physical stress responses. Gaining that approval can be even more important for couples who are crossing cultural boundaries, where relationships may attract scrutiny, disapproval, or misunderstanding.
Drawing on data from over 750 people in intercultural relationships, Hanieh Naeimi and team explore whose approval matters most for relationship quality. Writing in Social Psychological and Personality Science, they show that approval from friends, family, and society matter in different ways for different couples depending on cultural background and relationship stage.
Participants represented a wide range of backgrounds, with around half identifying as White and the rest as Black, Latin American, East Asian, South Asian, Middle Eastern, Native American, or multicultural. Their relationships varied widely in length, from under a year to several decades, and ranged all the way from just dating to committed relationships to marriage. To examine differences between couples, participants were grouped into three broad pairings: couples where both partners were White, couples with one White partner and one from a minority background, and couples where both were from minority backgrounds but with different cultural backgrounds.
All participants completed the same online survey. Firstly, they reported how much approval they felt from their family, friends, and society more generally using brief rating scales, before reporting on their relationship quality, including how satisfied and how committed they felt.
Feeling that friends approved of a relationship was the strongest predictor of both relationship satisfaction and commitment, with higher friend approval linked to higher satisfaction and stronger commitment for about 95% of couples. Family approval also mattered, but less significantly and consistently. It was particularly pertinent for Latin American and Middle Eastern participants and for couples in relationships of a year or less, likely because partners are still integrating each other into familial networks. Interestingly, approval from society at large had no measurable effect on either satisfaction or commitment.
Couples where both partners were from minority backgrounds benefitted the most from friend approval. Overall, the support of friends explained about 23% of the differences in relationship satisfaction and 26% of differences in commitment across all couples — suggesting that while family and societal approval sometimes help, feeling supported by friends is the most reliable predictor of a happy, committed intercultural relationship.
Because the data for this study was collected at a single point in time, we can't tell if social approval actually improves relationships, or if happier couples just feel more supported. Grouping couples into broad cultural categories may also mean more nuanced differences were missed.
Overall, though, the findings suggest that feeling validated and accepted by one's peers can strengthen a sense of satisfaction and commitment in a romantic relationship, helping couples grow even when family or societal approval is mixed.
Read the paper in full:
Naeimi, H., Muise, A., Di Bartolomeo, A., West, A., & Impett, E. A. (2025). With a Little Help From My Friends: Social Approval and Relationship Quality in Intercultural Romantic Relationships. Social Psychological and Personality Science. https://doi.org/10.1177/1948550625138944";
03 February 2026
By Emily Reynolds
https://www.bps.org.uk/research-digest/intercultural-relationships-whose-support-matters-most
#metaglossia
#metaglossia_mundus
#métaglossie
"Espresso Translations New York Earns ATC Membership and Elevates Translation Services NY Standards
PRESS RELEASE GlobeNewswire
Feb. 2, 2026, 05:49 PM
New York, NY, Feb. 02, 2026 (GLOBE NEWSWIRE) -- Certified translation demands are rising fast across New York and the USA, and clients are responding by choosing agencies backed by recognised industry standards. Espresso Translations New York has recently become an accredited member of the Association of Translation Companies (ATC), a milestone that strengthens its credibility for official and business-critical multilingual communication.
For clients comparing the top certified translation agencies in New York and the USA, the ATC serves as a key industry benchmark, listing vetted language service providers in its member directory. This makes ATC membership especially relevant for organisations and individuals who want added confidence that a translation agency operates with professional standards and quality controls.
That is exactly why Espresso Translations New York support organisations and individuals who require accurate multilingual work for legal, commercial, academic, and international use. These projects often involve court filings, immigration applications, academic transcripts and diplomas, business contracts, and regulated financial or compliance documents, where even small errors can create delays or complications. To meet that standard, the agency delivers translation solutions that protect meaning and context while keeping language clear, consistent, and reliable.
“Espresso Translations New York stands out by pairing recognised accreditation with practical language solutions for high-stakes needs,” said a spokesperson for this New York-based translation provider. “As a translation agency New York business community trusts, the company delivers certified translations, localisation, transcription, subtitling, editing, and multilingual support.”
This full-service capability helps clients manage multiple language requirements through one consistent provider, especially when accuracy and presentation quality must remain consistent across formats. Espresso Translation’s language solutions also support more than 150 languages, helping organisations expand internationally and assisting individuals with official documentation that requires precision and cultural sensitivity.
Quality standards remain a key differentiator in the USA translation market. The ATC member listing notes that Espresso Translations is ISO-certified and supported by a large network of linguists. This signals structured workflows and quality oversight for clients who want more than basic translation delivery. These credentials matter because certified translation work often supports decisions where errors can create delays, confusion, or compliance issues. Espresso Translations applies professional review processes that help reduce risks tied to inaccuracies, formatting inconsistencies, or wording that fails to match the intent of the original text.
As part of its full-suite translation services New York businesses count on, the company also provides professional transcription services for clients who need accurate spoken content converted into clear, usable text. This capability supports legal, academic, and commercial needs, ensuring interviews, meetings, recorded statements, and multilingual audio are documented accurately. Alongside ATC-recognised standards, the agency reinforces its position as a trusted provider for certified translation and multilingual projects.
Espresso Translations continues to raise expectations for certified translation services in New York and the USA by combining real-world delivery with recognised professional standards. With ATC membership strengthening credibility, the agency gives organisations and individuals added confidence that every project is handled with accuracy and accountability.
To learn more about Espresso Translations and request certified translation support from a translation agency New York clients trust, visit www.espressotranslations.com.
About Espresso Translations
Founded in 2018, Espresso Translations is a New York-based language services provider delivering certified translations, localisation, transcription, subtitling, and multilingual editing for businesses and individuals worldwide. The agency supports 150+ languages and applies structured quality processes to produce accurate, culturally sensitive communication."
https://markets.businessinsider.com/news/stocks/espresso-translations-new-york-earns-atc-membership-and-elevates-translation-services-ny-standards-1035775617
#Metaglossia
#metaglossia_mundus
#métaglossie
"The Army’s Western Hemisphere Institute for Security Cooperation marks one of the first military components reporting impacts from the platform.
The Pentagon recently tapped California-based tech company LILT to supply artificial intelligence-powered translation options to military forces worldwide.
According to a press release viewed by DefenseScoop before its public release Tuesday, the Defense Department’s Chief Digital and AI Office awarded LILT a flexible other transaction contract to expand its platform for military domain-specific vocabularies, after it was prototyped and proven via the Defense Innovation Unit.
“The platform enables rapid and accurate translation of text, video, and audio content into or out of English,” per the release. “This specialized, unique AI platform supports U.S. military missions that span from understanding foreign technical documentation and training materials to facilitating foreign partner exercises and supporting direct action missions.”
The Army’s Western Hemisphere Institute for Security Cooperation (WHINSEC) marks one of the first military components reporting impacts from the platform.
Advertisement
It historically would take personnel a full year to translate materials for the institute’s Command and General Staff Officer Course (CGSOC), which is delivered in Spanish and underpins instruction to students from a dozen Latin American and Caribbean partner nations. The course lasts 11 months.
In 2025, WHINSEC officials used LILT technology to translate the entire current-year CGSOC curriculum over the course of just a few weeks.
“LILT AI Translation has catapulted WHINSEC into the future of AI and its nexus with security cooperation and professional military education across the Western Hemisphere. We can now rapidly integrate international forces from countries that partner with the U.S. at all echelons: in the classrooms, during multinational exercises, and foreseeably in coalition combat outposts,” Army Col. Eldridge Singleton, 9th WHINSEC commandant, said in a statement. “That’s what we train for at CGSOC.”
The press release did not disclose the total value of the contract. A Pentagon spokesperson did not respond to DefenseScoop’s request for more information.
The announcement comes on the heels of DOD leaders launching a new AI Acceleration Strategy, which directed personnel to integrate the emerging technology into their daily operations."
Brandi Vincent
February 3, 2026
Brandi Vincent is a Senior Reporter at DefenseScoop, where she reports on disruptive technologies and associated policies impacting Pentagon and military personnel. Prior to joining SNG, she produced a documentary and worked as a journalist at Nextgov, Snapchat and NBC Network. Brandi grew up in Louisiana and received a master’s degree in journalism from the University of Maryland. She was named Best New Journalist at the 2024 Defence Media Awards."
https://defensescoop.com/2026/02/03/ai-enabled-translation-military-use-cdao-lilt-contract/
#Metaglossia
#metaglossia_mundus
#métaglossie
"Los chicos tienen un vocabulario que los adultos no entienden.
El nuevo diccionario adolescente: un traductor para entender qué quieren decir los chicos hoy
Miércoles 04 de Febrero de 2026, 07:08
“Dejá de tirar beef amigo”, “Es bait, no caigas”, “Amiga, devoraste con ese outfit, serviste face card”, “Subió una foto en blanco y negro sin poner descripción… está farmeando aura de misterioso”, “Cómo carreaste ayer en la fiesta, tenés alto rizz, estás en tu prime”, “No sé cómo no te diste cuenta, si ese pibe es una red flag andante”, “Tengo que hacerte un update del chisme porque te falta lore”, “Pegué un re glow up desde que entré a mi fitness era, siento que estoy llegando a mi peak”.
Estas son solo algunas de las palabras que integran el “nuevo diccionario” de los sub-18: una serie de códigos que los hermana y que, a su vez, los diferencia del mundo adulto que no puede llegar a comprender de qué hablan cuando los están escuchando.
“Hay dos cosas que yo siempre digo sobre el léxico joven. Primero, que la lengua cambia, es una constatación; no es tan violento el cambio, pero en el léxico es muy fácil incorporar una palabra nueva porque es el área de la lengua más susceptible a cambiar. Por otro lado, hay que recordar que a cierta edad hay una necesidad de diferenciarse identitariamente con los adultos o figuras de autoridad. Se crea un grupo de pertenencia que se une y se ata a la lengua”, opina Santiago Kalinowski, director del departamento de Investigaciones Lingüísticas y Filológicas de la Academia Argentina de Letras.
Algunas palabras clave
AURA: una especie de puntuación invisible de qué tan “cool”, respetado o imponente sos. Ejemplo: “Esto me da +1000 de aura”.
BAIT: viene del inglés y significa literalmente “cebo” o “carnada”. En el mundo de los adolescentes e Internet, se refiere a contenido creado específicamente para provocar una reacción, engañar a alguien o generar una pelea. Ejemplo: “Esto es bait, no caigas”.
BEEF: conflicto o pelea entre personas. Ejemplo: “Estos dos tienen un beef”.
BEIGE FLAG: significa “bandera beige” y es un término nuevo para describir algo raro o aburrido de una persona que no llega a ser malo (red flag) ni bueno (green flag).
BRAIN ROT: podredumbre cerebral por estar expuesto en demasía a contenido digital “basura”.
CARREAR: viene del inglés “to carry” (cargar o llevar a cuestas) y nació originalmente en el mundo de los videojuegos, en donde el “carry” es la persona que tiene todo el peso del equipo. Ejemplo: “La música estaba malísima, pero mi amigo empezó a bailar y carreó la noche”.
DELULU: es una abreviatura de la palabra inglesa delusional (delirante). Se usa para describir a alcés guien que tiene fantasías o expectativas poco realistas, especialmente en el amor o respecto a sus metas personales.
DEVORAR: se usa para decir que alguien hizo algo extremadamente bien, con mucha confianza o con un estilo impecable.
FACE CARD: se usa para hablar de la belleza o el atractivo en el rostro de alguien, a veces figura en oraciones en inglés como “face card never declines”, o en español “la face card de esa chica es increíble”.
FARMEAR: hacer lo mismo mil veces para ganar puntos o ítems y así subir de nivel o progresar. Se usa en el mundo de los videojuegos. Se puede combinar con “aura” y decir: “Estoy farmeando aura”.
FOMO: en inglés significa “fear of missing out”, o sea la ansiedad o miedo de quedarse afuera de algo. Ejemplo: “Literalmente solo vine por FOMO, estaba cansadísima”.
GLOW UP: es estar en tu mejor momento.
LORE: Es la historia de fondo de alguien o de algo, da más contexto o historia a una situación puntual. Ejemplo: “No te puedo explicar por qué es así, es que no conotodo su lore”.
PRIME: estar en tu mejor momento. Ejemplo: “Estoy en mi prime”. Es como “glow up”.
RED FLAG: traducido del inglés significa “bandera roja”. Se suele utilizar cuando en una relación amorosa hay señales que indican comportamientos que no nos gustan o que podrían considerarse tóxicos. Ejemplo: “No viste las red flags”, “Era una red flag andante”.
SERVIR: esta expresión se usa cuando alguien, especialmente una mujer, se luce haciendo algo. Ejemplo: “Ella sirvió rostro (face card) ayer”, “ella sirvió actuación”.
SIX SEVEN: los adolescentes lo usan como una jerga viral sin un significado fijo; sirve más como un código de identidad ligado a desconcertar a los adultos.
VIBE: la vibra que algo o alguien te genera. Ejemplo: “Este chico me da malas vibes”.
W MOMENTO: es común en el lenguaje de los videojuegos, transmisiones en vivo y redes sociales. La “w” viene de la palabra “win” (victoria). Un “W moment” es un momento de victoria o una situación donde alguien hizo algo genial.
Los orígenes
El origen de estas palabras es variado: algunas provienen del mundo gamer (como farmear aura, o carrear la partida), otras nacieron a partir de memes virales (six seven), otras están más asociadas a la comunidad LGBTQ+ (devorar, servir, slay) y otras a redes sociales como Tik Tok (Ok mañana). Es un vocabulario que viene en gran parte del inglés, el idioma predominante.
“Hay una jerga gamer y cada tanto una palabra que circula ahí pega un salto y se usa en otros ambientes, como en el streaming. La palabra “lore” nadie la explica en los mundos gamers, pero cuando uno ve streams y la usan, la explican. Es una palabra vieja en inglés que significa “saber”, entre culta y arcaica, que por la trayectoria sinuosa de los juegos de rol como el Minecraft empieza a circular como algo más popular, y trata de saltar a un ámbito más general”, analiza el experto.
Añade que se ve una trayectoria al ámbito de chimento, por ejemplo. “Hoy hablan de Luciano Castro y el lore que hay detrás”, ilustra Kalinowski. La farándula, entonces, es una de las áreas en donde palabras como lore -es decir, la historia de fondo o trama de un personaje- cobran relevancia. Pero no todas las palabras que usan los de la Generación Alfa (2010- 2024/5) -algunas las comparten con los más chicos de la Generación Z- están amplificadas por los medios.
La mayoría son indescifrables para los papás, los hermanos mayores, los abuelos y otros adultos. En este último tiempo surgieron reels de Instagram o videos de Tik Tok en donde los sub-18 ponen a sus padres frente al desafío viral de leer palabras de su generación sin tener idea de lo que están diciendo.
“Hay que advertir que lo que sucede en esos años de la adolescencia son rasgos del cronolecto -propios de la edad-, en donde los chicos buscan diferenciarse de las generaciones mayores. Lo que suele pasar es que a ciertas edades se establece un código para hablar con la gente de ese grupo y después la enorme mayoría de esas palabras se dejan de usar, no es que se incorporan al léxico general”, agrega Kalinowski.
Para el especialista, de hecho, son muy pocas las palabras surgidas del léxico adolescente que realmente “trascienden” y se logran incorporar de forma permanente al léxico general. El caso más emblemático es el del “re”, un prefijo devenido en adverbio de grado.
De las palabras gamer como lore saltamos a las que provienen de otros espacios del mundo digital. Por ejemplo “six seven”, la palabra del año 2025 según Dictionary.com que irrumpió en TikTok como uno de los memes más repetidos entre jóvenes y adolescentes, no significa nada puntual. Es una palabra que nació como un código y que tiene su fundamento en irritar a los adultos, quienes no entienden por qué se usa.
Es habitual, entonces, que los padres escuchen a sus hijos adolescentes decir six seven en toda clase de situaciones. A una pregunta cualquiera, como “qué hora es” o “qué es eso” los chicos responderán “six seven”.
Hay otras que son más curiosas, porque no se usan tanto de forma oral pero sí en comentarios, sobre todo en TikTok. “Ok mañana”, frase que aparece ante chistes que no funcionan, suele acompañarse de un corazón violeta y es uno de los ejemplos; “ella jura” es otro.
Hoy en día los chicos ya no dicen que alguien tiene “flow” cuando quieren decir que alguien emana una energía especial que no se puede comprar. Ahora dicen que esa persona tiene “aura”, un giro de tuerca más místico o “épico” si se quiere.
El aura -término asociado al mundo del anime, cómics de superhéroes y al deporte, pero con presencia en el mundo gamer ahora- se conforma de “puntos invisibles” de prestigio que una persona suma o resta a través de acciones o actitudes.
Los chicos dirán que alguien tiene “+1000 de aura”, por ejemplo, cuando hace algo calificado como épico. Los gamers irán un poco más allá y podrán decir que está “farmeando aura” (farmear, que viene de la palabra en inglés farming): así se refieren a alguien que hace determinadas cosas para acumular “puntaje” de respeto social.
“Siempre se identifica la tecnología en estos tipos de comportamientos, es una dinámica que ves repetitivamente, hoy ves ciertos formatos que dominan el momento (los reels de Instagram o shorts de YouTube) y si ese contenido se crea deslocalizadamente entonces ahí el propio medio favorece la transferencia del léxico más rápido. Sobre todo del español y del inglés, que se convirtió en la lengua más influyente después de la Segunda Guerra Mundial”, concluye Kalinowski. /La Radio 1029 /Clarin"
https://www.contextotucuman.com/nota/373167/el-nuevo-diccionario-adolescente-un-traductor-para-entender-que-quieren-decir-los-chicos-hoy.html
#Metaglossia
#metaglossia_mundus
#métaglossie
"Des progrès inégalés ont été réalisés ces dernières années pour rendre la Bible accessible dans toutes les langues. Elle l’est déjà pour quelque 6 milliards de personnes, et les efforts se poursuivent à un rythme rapide.
Des progrès remarquables : c’est le constat qui s’impose à l’heure du bilan. Il y a plus de deux siècles, des sociétés bibliques et leurs partenaires se sont fixé pour défi de traduire la Bible dans toutes les langues connues. Plus de 6 milliards de personnes, soit 6 personnes sur 8 dans le monde, ont aujourd’hui accès à l’intégralité de la Bible dans la langue qu’ils comprennent le mieux. Au total, la Bible a été traduite dans quelque 760 à 790 langues – les chiffres varient un peu selon les sources. En comptant les traductions partielles, plus de la moitié des 7 398 langues parlées dans le monde sont couvertes.
On observe une nette accélération de l’avancement des travaux depuis quelques années. Au 1er août 2025, il ne restait plus que 544 langues sur liste d'attente pour une traduction des Écritures. L’année précédente, il y en avait 985 et en 1999, elles étaient plus de 5 000 à ne faire l’objet d’aucun projet.
Le Lévitique en Banda-Linda
Comment le travail est-il organisé ? Basé dans la région de Lucerne, le Suisse Christoph Müller travaille depuis le début des années 2000 en tant que conseiller en traduction sur des projets en Afrique centrale. Il vient tout juste de terminer le contrôle d’une traduction du Lévitique en Banda-Linda, langue parlée dans la République Centrafricaine.
Interrogé par téléphone, il explique : « Le conseiller doit pouvoir justifier d’une formation biblique et de connaissances en linguistique. Après une formation continue plus ou moins longue, en fonction de son profil et de ses disponibilités, il commence à exercer sous la supervision d’un consultant plus expérimenté, puis se met à travailler de manière plus ou moins autonome. Il ne doit pas forcément parler la langue dans laquelle la traduction est effectuée, même si, à la longue, il finit naturellement par avoir quelques notions. »
Une approche collaborative
Le rôle du conseiller est d’épauler et de former l’équipe de traducteurs locaux. Ces derniers doivent, en principe, avoir au moins un niveau bac et des connaissances bibliques. « Il y a une cinquantaine d’années, il arrivait encore que des missionnaires effectuent des traductions dans une langue qui n’était pas la leur, mais ce modèle a été abandonné, poursuit Christoph Müller. De nos jours, les traducteurs sont recrutés dans la population locale. » Cette approche collaborative vise à valoriser les compétences des communautés concernées, pour un meilleur résultat final. Le contrôle est réalisé en présence des traducteurs et d’un locuteur natif qui n’a pas participé au projet ; il sert d’interprète au consultant chargé de la vérification du travail. Un examen supplémentaire a lieu au sein de la communauté.
Toutes ces traductions sont le fruit d’une intense collaboration entre différentes organisations d’inspiration chrétienne, comme la Société internationale de linguistique (SIL), pour qui travaille Christoph Müller, ou l’association à but non lucratif Wycliffe, à Bienne. Celle-ci participe à des projets de traduction dans plus de 100 langues. Pour cela, elle dispose de quelque quatre-vingts collaborateurs – dont de précieux consultants envoyés sur le terrain.
Des questions épineuses
Dans la pratique, les difficultés sont nombreuses. Que signifie tel mot dans son contexte historique ? Que voulait dire ce texte à l’époque ? S'agit-il d’une figure de style et si oui, que signifie-t-elle ? Prenons par exemple ce verset : « Priez pour que vous ne deviez pas fuir en hiver » (Marc 13 :18). Comment traduire le mot « hiver » pour des lecteurs qui vivent dans une région où l’on ne connaît pas cette saison ? Certains traducteurs ont trouvé une parade en employant l’expression « temps froid ».
L'inévitable évolution de la langue pose un autre problème. Par exemple, de nos jours, on ne peut raisonnablement plus écrire : « Nous nous mîmes à genoux et nous priâmes, nous partîmes et nous arrivâmes » (Actes des Apôtres, 21 :5-8). Les Sociétés bibliques doivent donc périodiquement actualiser leurs traductions pour préserver la lisibilité des contenus. Ces dernières années, elles ont publié 13 révisions de la Bible intégrale et du Nouveau Testament dans des langues majeures, notamment une Bible pour les 78,6 millions d’Indiens parlant le tamoul.
Transformation digitale
En 2024, date du dernier rapport annuel de l’Alliance biblique universelle (ABU), 74 premières traductions des Écritures, soit 16 Bibles intégrales et 16 Nouveaux Testaments, ont été publiées. Un nouveau cap a ainsi été franchi : 100 millions de personnes ont reçu pour la première fois au moins une partie des Écritures dans leur langue. Le Nouveau Testament, par exemple, est maintenant disponible dans plus de 1 700 langues.
Pour accélérer encore la cadence, les grandes sociétés bibliques ont effectué leur transition numérique dans les années 2000. Selon l’ABU, la création de la bibliothèque biblique numérique (DBL), en 2011, a généré « une dynamique sans précédent ». Actuellement, plus de 2000 textes bibliques dans 2250 langues différentes et quelque 13 000 vidéos dans 42 langues des signes sont archivés dans cette bibliothèque. Beaucoup d’entre elles sont disponibles gratuitement via l’application YouVersion.
Bibles adaptées
Quelque 370 nouveaux textes dans 191 nouvelles langues y ont été ajoutés en 2024, portant le nombre total de textes disponibles à plus de 3 780. En même temps, les efforts pour préserver la diversité linguistique se sont poursuivis. Ainsi, un Nouveau Testament a été publié pour la première fois dans une dizaine de langues minoritaires comme le same du Sud, avec 600 locuteurs dans les régions septentrionales de la Norvège et de la Suède, et l’enggano, parlé par moins de 900 individus en Indonésie.
Les personnes qui ont des difficultés de lecture ne sont pas oubliées. Par exemple, la Société biblique espagnole a publié en 2024 une Bible en langage simplifié, avec un vocabulaire réduit le vocabulaire à moins de 4 000 mots – soit nettement moins que les 16 000 à 18 000 mots contenus dans les versions traditionnelles. Ce projet a obtenu le label « Easy-to-Read » de l’organisation Inclusion Europe.
Braille et langues des signes
La traduction des Écritures pour les personnes sourdes et malentendantes est une tâche autrement plus complexe, car des notions comme la miséricorde n’ont pas d’équivalents dans les langues des signes, ce qui oblige les traducteurs à faire preuve de créativité. Parmi les derniers projets menés à terme : la traduction d’une partie du livre de Josué et de la première épître aux Corinthiens dans une dizaine de langues des signes différentes.
Une quantité de textes en braille et de fichiers audio ont également été déposés dans la DBL ces dernières années. En 2024, ce sont notamment de nouvelles traductions dans des langues indiennes en braille – dont l’ao naga, le saurashtra, le rabha, le biate et le mara – qui ont été rendues accessibles.
Un travail sans fin
Si l’objectif des Sociétés bibliques s’est clairement rapproché depuis quelques années, beaucoup de travail reste à faire. Environ 1,5 milliard de personnes dans le monde n’ont toujours pas accès à une Bible complète dans leur langue, et 129 millions ne disposent purement et simplement d’aucune traduction, même partielle.
Parmi les langues non traduites, beaucoup sont parlées par des groupes de locuteurs peu nombreux, de sorte qu’elles sont menacées d’extinction. « Le travail de traduction de la Bible ne sera jamais vraiment terminé, car les langues et la société évoluent constamment ; les révisions seront donc toujours nécessaires », conclut Judith Sawers, responsable communication pour la région Centrafrique au sein de l’organisation SIL."
https://www.reformes.ch/eglises/2026/02/le-nombre-de-traductions-de-la-bible-explose-bible-traduction
#Metaglossia
#metaglossia_mundus
#métaglossie
Une exposition de livres s'est tenue hier, mardi 2 février 2026, au siège de l'Association pour la Recherche et l'Éducation pour le Développement (ARED), à Dakar. Cette rencontre s'inscrit dans la perspective de l'édition 2026 du Prix Cheikh Hamad bin Khalifa Al Thani pour la traduction et la compréhension internationale, une prestigieuse distinction internationale organisée par la fondation qatarie du même nom.
...l'accent a été mis sur les traductions de l'anglais vers le pulaar et du pulaar vers l'anglais, un axe particulièrement stratégique pour la compétition internationale à venir.
"Les ouvrages présentés témoignent d'une production jugée «très riche» par les organisateurs. Ils couvrent des traductions du pulaar vers l'arabe et inversement, mais aussi des travaux scientifiques et techniques. Toutefois, l'accent a été mis sur les traductions de l'anglais vers le pulaar et du pulaar vers l'anglais, un axe particulièrement stratégique pour la compétition internationale à venir. La délibération finale est prévue vers la fin du mois de mars.
Restez informé des derniers gros titres sur WhatsApp | LinkedIn
Prenant la parole, le directeur général de l'ARED, Mamadou Amadou Ly, a expliqué l'objet de cette rencontre. Selon lui, l'exposition visait à rassembler chercheurs, écrivains et éditeurs travaillant en langue pulaar, notamment ceux engagés dans la traduction entre le pulaar et d'autres langues, en particulier l'arabe et l'anglais.
«Le pulaar a été retenu cette année comme langue internationale et doit concourir au même titre que des langues comme l'anglais. Il était important de montrer, en amont de la compétition, la richesse de notre production», a-t-il souligné.
Fondée en 1991, l'ARED est une organisation non gouvernementale spécialisée dans l'éducation. Elle oeuvre pour l'amélioration de la qualité de l'enseignement à travers l'utilisation des langues nationales, en s'appuyant sur la formation, l'édition, la recherche-action et les innovations pédagogiques. Cette exposition s'inscrit ainsi dans la continuité de sa mission de valorisation des langues africaines comme outils de savoir et de développement."
https://fr.allafrica.com/stories/202602030469.html
#Metaglossia
#metaglossia_mundus
#métaglossie
"Gros plan sur ToumAI une start-up marocaine qui travaille à intégrer les langues et dialectes africains dans les usages que l’on fait de l’intelligence artificielle. L'entreprise, qui compte une quinzaine d'employés, propose une application permettant de mieux communiquer entre clients et plateforme téléphonique de service.
Ce n'est pas toujours évident lorsque l'on parle hausa, bambara ou le darija d'utiliser des plateformes de service téléphonique pour gérer son compte bancaire, réserver un hôtel ou bien régler ses problèmes de connexion internet. Les serveurs téléphoniques ou les plateformes d'appel ne parlent pas toutes ces langues, loin s'en faut. Aussi, la start-up ToumAI a-t-elle mis au point une application qui permet, avec l'intelligence artificielle, de répondre aux demandes des clients dans différentes langues africaines.
Youcef Rahmani, l'un des trois fondateurs de ToumAI et de l'application HolistiCX : « On a, à peu près, huit langues aujourd'hui qui sont vraiment prêtes en production et déployées d'ailleurs sur des applications. Donc vraiment, notre objectif, c'est finalement de couvrir les langues les plus parlées. Dans un premier temps, le swahili, le lingala en Afrique de l'Est et en Afrique du Sud, le zoulou, le xhosa. il y a plusieurs langues, souvent mal considérées par les logiciels et l' IA et que nous, on souhaite couvrir. »
« L'IA ne comprend pas votre accent »
Une intelligence artificielle qui s'adapte à l'oral, aux vocables africains, mais également aux intonations, si importantes, à la bonne compréhension d'une langue afin de mieux répondre aux besoins de celui qui l'exprime. Un service qui comble, pour toute une clientèle, un risque de fracture numérique et linguistique, souligne Youcef Rahmani
« En fait, c'est exactement ça ! ToumAI est né d'une de cette frustration. C'est que pendant longtemps, l'intelligence artificielle a été développée pour quelques langues et quelques cultures. Beaucoup de personnes se sont retrouvées exclues sans même que l'on s'en aperçoive. Au début, on ne s'en rendait pas compte parce que l'IA était reléguée à des rôles un peu obscurs pour le commun des mortels. Mais, lorsque Chatgpt est sorti, je pense que c'est là où les gens se sont rendus compte de la puissance de l'IA. Et en fait, pendant très longtemps, Chatgpt ne parlait pas vraiment d'autres langues à part les langues principales. Quand une IA ne comprend pas votre accent ou votre façon de parler ou votre langue, ce n'est pas juste un problème technique, ça devient un problème d'accès aux services, parce que l'on va avoir des logiciels qui intègrent de l'IA un peu partout, mais sans justement prendre en compte les différences culturelles et linguistiques. Nous avons voulu partir de la voix parce que la voix c'est l'univers où tout le monde parle. Et même quand on ne lit pas ou même lorsque l'on ne maîtrise pas le digital, on peut s'exprimer et obtenir des choses en parlant.»
HolistiCX a séduit des entreprises bancaires comme le marocain Attijariwafa Bank, des opérateurs téléphoniques comme Orange ou Inoui, et certaines entreprises immobilières comme Héritage. Les clients de ces entreprises ne se sentent plus exclus des services à cause d'une inadaptation technologique à la langue. Une solution pertinente qu'a accompagné le fonds de soutien Digital Africa et Malek Lagha, cheffe de projet.
« C'est très pertinent parce qu’ils ont besoin de collecter de la donnée et qu’aujourd’hui, les logiciels en IA n'arrivent pas à traduire des langues africaines. C'est un marché qui a beaucoup de potentiel. C'est de la donnée très précieuse et qui compte également pour beaucoup d'entreprises dans le monde. Aujourd'hui, ToumAI a conçu une IA pensée dès le départ pour des contextes complexes, capable de comprendre non seulement les mots, mais les intonations. Je pense que c'est pour ça que c'est pertinent pour des entreprises internationales. »
Pour le moment, la solution proposée par ToumAI concerne huit langues africaines, mais le projet futur est d'adapter une quarantaine d'autres langues à des solutions IA."
https://www.rfi.fr/fr/podcasts/l-afrique-en-marche/20260203-toumai-rend-les-langues-africaines-intelligibles-pour-l-intelligence-artificielle
#Metaglossia
#metaglossia_mundus
#métaglossie
"In undergraduate Classics, translation is an unforgiving exercise, demanding almost mathematical precision. I’ve spent excruciating hours poring over lexicons and grammar books, only to face reproof for neglecting the odd particle. When so many English versions of ancient texts already exist, not to mention digital translation resources, it’s easy to question why we bother.
Yet translation should be more than mechanic substitution. It demands that the translator acts as a conduit, conveying the intricacies of emotion, style, and intention, while negotiating the hurdles of linguistic complexity. It involves a degree of compromise, balancing fidelity to the original with creative interpretation. When a piece of literature is transposed into the idiom of a new age, a new culture, each adaptation becomes a radical re-reading, not a straightforward reproduction. Rather than representing the work as a historical artefact, mute and moribund on the page, the process of translation can shore up unmined meanings. In ancient languages, with a comparatively restricted vocabulary, each word is capable of being expressed in English in multiple ways, giving rise to vastly divergent interpretations. Word choice becomes a declaration of intent. As the translator Emily Wilson points out, the Odyssey’s opening line, which Fagles translates as “Sing to me of the man, Muse, the man of twists and turns”, could equally be rendered as “Tell me about a straying husband”, a very different framework for the same Greek words.
Things inevitably slip through the cracks; wordplay in particular demands more than a literal translation. For instance, Wilde’s The Importance of Being Earnest in French translation frequently becomes L’importance d’être Constant, replicating the pun by renaming the protagonist, yet losing out on the connotations of deceptiveness. Moreover, there are concepts so tethered to their specific language that they defy straightforward translation. How far the unfamiliar should be domesticated is a consequential choice – is it better to retain culturally specific allusions, or facilitate understanding through parallels or explanations? English translations of Elena Ferrante’s Neapolitan Quartet embed words of dialect, a deliberate choice to ensure that the work remains firmly rooted in its original context, with its particular local colour. The rhythms of each language, which determine much of literature’s emotional impact, are likewise impossible to reproduce exactly. Pushkin’s Eugene Onegin, for example, is widely regarded as untranslatable, owing to the intricacy of its rhyme scheme, and the unique musicality of the Russian. The best that can be achieved is adaptation: Deborah Smith in her translations of Han Kang – The Vegetarian (2015) and Human Acts (2016) – attempts to emulate the cadence of the Korean, its repetitions and underspecifications, resulting in a stark prose that enhances the tragedy.
The insistence on preserving the original essentially untampered with is futile; excessive hand-wringing over what is being lost in the process can only stunt creativity. Translation is, in a sense, a work of realignment – nothing can remain fixed. Since utter fidelity to the original source is impossible, the objective should be to create something that works in one’s own language, a discrete piece of art, so that the translator is effectively another writer of the same book. The boundaries of language are always permeable – a good translator is an unscrupulous gerrymanderer.
After all, translation is an inherently malleable concept, and does not necessarily signify replication of the source material. Language is not exclusively about designation, but the meanings hovering between statements, the conveyance of a mood, a perspective, an intention. There is no need, then, for translation to adhere to semantic, generic, or even formal boundaries. In this expansive spirit, Louis and Celia Zukofsky wrote homophonic translations of the poet Catullus, rendering not only the meaning but also the actual sounds of the Latin into English (miser Catulle becomes “Miss her, Catullus?”). Anne Carson went even further in her ‘translation’ of Catullus’ poem 101, an elegy for the death of his brother; Carson’s version constitutes a single long sheet of paper folded concertina style into a box entitled Nox, an epitaphic reflection on her own brother’s passing. How far then can we push the definition of translation? What’s to stop any response to a literary work being considered a translation – is Petersen’s Troy (2004), for example, a translation of the Iliad (despite it being a terrible film)?
The politics of translation are similarly complex. To translate a literary work into another language is, in a sense, to appropriate it from its original context for the enjoyment of another set of people. Taken further, a French translation of, say, an Arabic text could be viewed as an implicitly colonial act, while the ubiquity of English translations raises the spectre of global monolingualism. But surely this kind of engagement can be part of a dialogue, not an act of imperialistic plunder? Accessibility is the most fundamental objective of translation; widening the reach of a literary work is a conservationist practice, sustaining and invigorating its author’s voice, rather than an attenuation of its power. It is not sufficient for a translator to be merely a linguistic intermediary; the practice demands cultural proficiency and a profound understanding of the recipient language. The art of translation is one of bridging cultural divides, so that literature may resonate with readers worldwide. Such interaction eases the discomfort of translingual encounters and fosters cross-cultural understanding.
It is this notion of a participatory culture via translation that enriches the literary tradition – Goethe wrote that “every literature grows bored if it is not refreshed by foreign participation”. Translation does more than keep the original alive (although sometimes I wish we’d just let Latin die); it also vivifies the recipient language, traversing linguistic boundaries to provide access to unfamiliar cultures, concepts, and perspectives. The translator is literary critic, co-author, cultural ambassador, and, most importantly, close reader, engaging in a fundamentally creative practice. So perhaps it’s misguided to ask what gets lost in translation. The more pertinent question is what may be found."
Beatrix Arnold
1st February 2026
https://cherwell.org/2026/02/01/the-art-of-translation/
#Metaglossia
#metaglossia_mundus
#métaglossie
"Investment in interpreter readiness supports better patient experience and outcomes, patient safety, and equitable care delivery at scale.
SUNRISE, FL, UNITED STATES, February 2, 2026 /EINPresswire.com/ -- Equiti announced today that its Martti solution’s 18,000+ qualified interpreters completed over 800,000 hours of specialized healthcare training in 2025, reinforcing the company’s commitment to reliable, high-quality language access for health systems nationwide.
This milestone reflects a year of extraordinary investment in interpreter preparedness as Martti supported millions of clinical encounters across over 320 spoken languages and dialects with an average connect time of just 19.4 seconds.
This investment is reflected in Martti’s interpreter training standards: each interpreter completes 120 hours of healthcare-specific training, triple the industry standard.
“Interpreter training is foundational to safe, effective communication in healthcare,” said Kerry Moreno, Vice President of Operations and Language Services at Equiti. “Completing more than 800,000 hours of training in a single year represents our commitment to ensuring interpreters are both fluent in their languages and fully equipped for the intensive clinical, cultural, and compliance demands of healthcare environments.”
Martti’s interpreter training programs are designed specifically for healthcare delivery, emphasizing clinical terminology, care workflows, patient privacy, cultural sensitivity, and regulatory compliance.
In 2025, Martti set a new standard by increasing its already stringent training requirements by 50% – from 80 to 120 hours – three times the industry standard of 40 hours. These expanded training requirements help ensure interpreters can support complex clinical communication across departments, specialties, and patient populations.
This investment directly supports health system priorities, including:
- Medically qualified, highly trained healthcare interpreters equipped to support the clinical and cultural needs of patients
- Faster access to interpreters, reducing delays at triage, treatment, and discharge
- Interpretation that delivers a better patient experience, improved patient safety, and reduced risk
- Lower administrative burden for clinical teams working with interpreters trained in healthcare-specific workflows
- More consistent compliance, helping organizations remain audit-ready without added operational strain
The 800,000-hour training milestone was achieved following Equiti’s launch of the unified Martti platform, bringing together the best capabilities from Voyce and Martti to support healthcare organizations at scale. By aligning interpreter teams and standardizing operations, Equiti reinforced Martti as the trusted, healthcare-focused platform delivering high-quality interpretation, industry-leading language coverage, and fast connection times to providers and patients.
“Equal access to care starts with communication,” Moreno added. “By investing deeply in interpreter training, we’re helping health systems embed health equity into daily operations, reinforcing a dependable standard of care.”
To learn more about Martti’s interpreter training programs and language access solutions, visit www.martti.io."
https://www.einpresswire.com/article/886037036/equiti-s-martti-interpreters-complete-more-than-800-000-hours-of-specialized-healthcare-training-in-2025
#Metaglossia
#metaglossia_mundus
#métaglossie
"Écritures et mémoires : la littérature traduite comme archive transnationale de l'histoire
Numéro thématique de la Revue internationale de traduction moderne
Coordination : Ramila Demane Debbih (Université Constantine 1, Algérie) et Jean-Pierre Castellani (Université de Tours, France)
Date limite de soumission : 30 juin 2026
Publication prévue : décembre 2026
Langues acceptées : français, anglais, arabe, espagnol
—
Comment la littérature traduite produit-elle du savoir historique ? Cette question, au croisement de la traductologie, des études littéraires et de l'historiographie, demeure insuffisamment théorisée malgré son acuité contemporaine. Nombreux sont les lecteurs qui découvrent et comprennent l'histoire de pays ou d'événements lointains davantage par la fiction traduite que par les seuls travaux académiques. Un lecteur algérien contemporain qui n'a pas connu la guerre d'indépendance accède à la texture sensible de cette période par la lecture de Mammeri, Dib, Feraoun ou Kateb Yacine. De même, un lecteur maghrébin qui lit des romans ukrainiens traduits sur la guerre actuelle (Oksana Zaboujko, Serhiy Jadan) accède à une compréhension intime du conflit que les reportages journalistiques ne peuvent offrir.
Ce numéro thématique ambitionne d'interroger frontalement le statut épistémologique de la littérature traduite comme source et mode de production du savoir historique. Il part du postulat que la traduction d'œuvres littéraires à contenu historique ne constitue pas un simple transfert linguistique, mais un acte d'interprétation herméneutique qui reconfigure la connaissance du passé pour de nouveaux lectorats.
—
Argumentaire
La littérature traduite ne se contente pas de raconter l'histoire : elle produit une forme spécifique de connaissance historique que les discours historiographiques classiques, archives institutionnelles, monographies académiques, manuels scolaires, ne peuvent générer. À travers le roman historique, le témoignage littérarisé, l'autofiction mémorielle ou la poésie narrative, les œuvres traduites donnent accès à des dimensions de l'expérience historique (affective, corporelle, subjective, polyphonique, contradictoire), souvent absentes des récits officiels.
La fiction comme productrice de savoir historique
La fiction n'est pas un « véhicule » transparent qui transmettrait de l'information historique de manière ornementale. Elle constitue une forme de pensée qui articule autrement les rapports entre événement, subjectivité, temporalité et vérité. Là où l'historiographie académique privilégie la distance critique et l'objectivation documentaire, la littérature opère par incarnation narrative et singularisation des voix. Elle donne accès à ce que l'on pourrait nommer une vérité incarnée de l'histoire : non pas la vérité factuelle des dates et des données, mais la vérité phénoménologique de l'expérience vécue du passé.
Ainsi, Si c'est un homme de Primo Levi, Inyenzi ou les Cafards de Scholastique Mukasonga, Allah n'est pas obligé d'Ahmadou Kourouma ou Nocturne du Chili de Roberto Bolaño deviennent, en traduction, des références historiques mobilisées parfois davantage que les travaux académiques sur ces mêmes périodes.
La traduction comme reconfiguration herméneutique
La traduction ne transmet pas cette connaissance de manière neutre. Le traducteur opère des choix (lexicaux, syntaxiques, paratextuels), qui transforment l'accessibilité, l'intelligibilité et la résonance émotionnelle du texte source. Traduire un roman historique implique de négocier constamment avec l'altérité culturelle et temporelle : faut-il expliciter les références historiques opaques ? Naturaliser les concepts intraduisibles ? Conserver l'étrangeté linguistique pour préserver l'altérité historique ?
Le paratexte, notes du traducteur, préfaces, glossaires, devient alors un dispositif pédagogique essentiel qui transforme la traduction en médiation didactique. Le traducteur devient ainsi une sorte d'historien parallèle qui accompagne le texte littéraire d'un appareil critique informel.
Archives transnationales et géopolitique des mémoires
En circulant d'une aire linguistique à une autre, la littérature traduite crée des archives transnationales de mémoires plurielles. Ces archives sont symboliques : elles consistent en la sédimentation progressive, dans l'imaginaire collectif d'une communauté linguistique donnée, de récits historiques venus d'ailleurs.
Toutefois, cette circulation est profondément asymétrique. Certaines histoires voyagent massivement (génocides européens, guerres mondiales), tandis que d'autres restent confinées linguistiquement (guerres coloniales africaines, conflits post-soviétiques). Les langues-pivots (français, anglais, espagnol) jouent un rôle central : une œuvre traduite vers l'anglais acquiert une visibilité mondiale, tandis qu'une œuvre disponible uniquement en arabe, swahili ou portugais demeure largement inaccessible.
Cette asymétrie pose une question éthique et politique : quelle est la responsabilité des traducteurs, éditeurs et institutions académiques dans la circulation équitable des mémoires historiques ? Peut-on penser une forme de « justice traductive » ? Comment favoriser les traductions Sud-Sud ?
Axes thématiques
Les contributions pourront s'inscrire dans l'un des axes suivants (liste non exhaustive) :
Axe 1 : Épistémologie de la fiction historique traduite
Le roman historique comme source de connaissance : quelles vérités la fiction transmet-elle ?
Statut du témoignage littérarisé traduit : document ou création ?
Légitimité du savoir narratif face au savoir historiographique
La traduction comme interprétation herméneutique de l'événement historique
Temporalités narratives et mémoire en traduction
Axe 2 : La traduction comme construction d'archives transnationales
Quand la fiction traduite devient référence historique pour des lecteurs étrangers
Patrimonialisation mémorielle et canonisation par la traduction
Littératures « mineures » et invisibilisation de certaines histoires
Retraduction et réactualisation du savoir historique
Études de réception comparée
Axe 3 : Stratégies traductives et médiation du savoir historique
Le paratexte comme dispositif pédagogique (notes, préfaces, glossaires)
Traduire les référents historico-culturels : expliciter, naturaliser ou conserver l'étrangeté ?
Intraduisibilité des concepts historiques et création lexicale
Le traducteur comme médiateur culturel et historique
Éthique de la traduction : jusqu'où contextualiser ?
Axe 4 : Géopolitiques et idéologies de la traduction littéraire historique
Asymétries dans la circulation des mémoires : quelles histoires voyagent ? (Sud-Sud, Sud-Nord, Nord-Sud)
Pourquoi certaines histoires sont sur-traduites et d'autres sous-traduites ?
Le rôle des éditeurs, collections et institutions dans la sélection des œuvres
Censure, autocensure et manipulation en traduction
Peut-on penser une « justice traductive » en matière de mémoire historique ?
Traduction et décolonisation des savoirs
Corpus et approches
Les contributions pourront porter sur tout corpus littéraire traduit à dimension historique (roman, nouvelle, autofiction, témoignage littérarisé, poésie narrative), sans restriction géographique ou temporelle. Les approches comparatistes, les études de cas et les analyses de réception sont particulièrement encouragées.
Exemples de corpus possibles (non limitatifs) :
Littératures maghrébines traduites (guerre d'indépendance, décennie noire, migrations)
Littératures des génocides et des conflits armés en traduction
Romans postcoloniaux et leur circulation transnationale
Littératures des dictatures (Amérique latine, Europe de l'Est, monde arabe)
Œuvres contemporaines sur les conflits actuels (Ukraine, Syrie, Palestine, etc.).
—
Modalités de soumission
Langues acceptées : français, anglais, arabe, espagnol
Format des articles :
Longueur : 25 000 à 40 000 signes (espaces compris), notes et bibliographie incluses
Format : Times New Roman 12, interligne 1,5
Normes bibliographiques : APA 7e édition ou MLA (au choix, avec cohérence interne)
Accompagner l'article d'un résumé en français et en anglais (150-200 mots chacun) + 5 mots-clés dans chaque langue
Notice biobibliographique de l'auteur (100 mots maximum)
Procédure :
Évaluation en double aveugle par des pairs internationaux
Notification d'acceptation/révision : 8 semaines après soumission
Calendrier
Diffusion de l'appel : 1er février 2026
Date limite de soumission : 30 juin 2026
Notification aux auteurs : 15 septembre 2026
Réception des versions finales : 31 octobre 2026
Publication : décembre 2026.
—
Bibliographie indicative
Berman, Antoine (1984). L'épreuve de l'étranger : Culture et traduction dans l'Allemagne romantique. Paris : Gallimard.
Berman, Antoine (1999). La traduction et la lettre, ou l'auberge du lointain. Paris : Seuil.
Casanova, Pascale (1999). La République mondiale des lettres. Paris : Seuil.
Casanova, Pascale (2015). La langue mondiale : Traduction et domination. Paris : Seuil.
De Certeau, Michel (1975). L'écriture de l'histoire. Paris : Gallimard.
Lavocat, Françoise (2016). Fait et fiction : Pour une frontière. Paris : Seuil.
Lukács, Georg (1965). Le roman historique. Paris : Payot.
Meschonnic, Henri (1999). Poétique du traduire. Lagrasse : Verdier.
Nora, Pierre (dir.) (1984-1992). Les lieux de mémoire, 3 volumes. Paris : Gallimard.
Ricœur, Paul (2000). La mémoire, l'histoire, l'oubli. Paris : Seuil.
Sapiro, Gisèle (dir.) (2008). Translatio : Le marché de la traduction en France à l'heure de la mondialisation. Paris : CNRS Éditions.
Venuti, Lawrence (2008). The Translator's Invisibility: A History of Translation, 2e édition. London/New York : Routledge.
White, Hayden (2017). L'histoire s'écrit : Essais, recensions, interviews. Paris : Éditions de la Sorbonne.
Coordination scientifique
Ramila Demane Debbih
Maître de conférences en Littérature française
Université Constantine 1, Frères Mentouri
Laboratoire Langues et Traduction
Algérie
Jean-Pierre Castellani
Professeur émérite
Université de Tours
France
Contact et soumission :
Email : dds.ramila@gmail.com / jeanpierrecastellani@hotmail.com
Plateforme de soumission : https://asjp.cerist.dz/en/PresentationRevue/571
Note aux contributeurs :
Cette bibliographie est indicative et non exhaustive. Les auteurs sont encouragés à mobiliser d'autres références pertinentes pour leur sujet spécifique."
https://www.fabula.org/actualites/132493/ecritures-et-memoires-la-litterature-traduite-comme-archive-transnationale-de-l-histoire.html
#Metaglossia
#metaglossia_mundus
#métaglossie
""...This innovative product empowers developers to integrate real-time voice transcription and translation capabilities into their applications, significantly enhancing multilingual support for businesses.
The DeepL Voice API allows businesses to stream audio and receive transcriptions in the source language, along with translations into up to five target languages. The API provides a seamless experience for users, ensuring that language barriers do not hinder effective communication.
DeepL Voice API will be widely available for customers with spoken communication at their core, with contact centres and business process outsourcing (BPOs) providers being the earliest adopters of this solution.
Transforming Multilingual Support
The DeepL Voice API turns language support from a staffing problem many contact centers face, into an easy-to-use solution that fits well with current systems. By adding real-time transcription and translation to how agents work, supervisors can handle issues better, and agents can assist customers in different languages without needing to pass them on to a colleague or revert to written communication to allow for translation.
On the operational side, the Voice API provides clear transcripts and translations that help with quality checks and training of customer service teams. This allows for quicker reviews, fairer evaluations across different locations, and clearer feedback on agent performance and gaps in knowledge. By minimizing issues caused by language barriers, like longer calls, repeated contacts, and expensive misunderstandings, the DeepL Voice API changes the overall experience for the end user.
...
Real time translation helps teams maintain service levels during nights, weekends, and holidays, when fewer specialized language agents are available.
Two way understanding, not just text on screen
Agents can follow the conversation through live translated audio, alongside on screen transcription and translation, so they can respond naturally and confidently in the moment...
The launch also includes a six week early access program for voice-to-voice capabilities, set to run from mid-February. This feature will allow agents to hear translated audio while communicating with customers in their preferred languages in real-time, further streamlining the customer experience.
Availability
The DeepL Voice API is available to all DeepL API Pro customers starting February 2. Interested businesses can get started by accessing the DeepL API documentation or contacting sales for Voice API access.
For more information on DeepL Voice API, including supported languages, please visit the website.
SOURCE DeepL"
News provided by
DeepL
Feb 02, 2026, 06:00 ET
https://lnkd.in/eFZxBCRm
#Metaglossia
#metaglossia_mundus
#métaglossie
"AI Translation Triumphs Over Human Translators in Korean Literary Contest
12 of 16 English professors preferred ChatGPT's version of Joseon-era poem, sparking debate on AI's role in cultural translation
By Park Jin-seong
Published 2026.02.02. 00:57
Updated 2026.02.02. 13:59
Artificial intelligence (AI) and humans have engaged in a competition to translate Korean literary works into English. Who emerged victorious?
Recently, the Literature Translation Institute of Korea under the Ministry of Culture, Sports and Tourism conducted a blind test involving 16 domestic English literature professors. The test compared an English version translated by a professional translator and one translated by ChatGPT for the Joseon-era poet Jang Yu’s poem “Be Cautious When Alone (Shindokjam),” which is set to be exported to English-speaking regions. Without revealing which translation was done by whom, the professors were shown the original Korean text and the two translations and asked which was better. The results showed that 12 professors chose the ChatGPT translation, two selected the human translation, and two declared “undecidable.”
Graphics by Baek Hyeong-seon
◇AI Translation Wins… “Exceeding the Threshold of Language Learning”
Professors who favored the AI translation praised ChatGPT’s deep understanding of Korean history and culture, as well as its effective preservation of the original’s rhythm and style. For example, in the passage “Above, the sky / Below, the earth / If you think I don’t know what I’ve done / Who are you trying to deceive?”, the human translator rendered “sky” as “Sky,” while ChatGPT translated it as “Heaven.” Considering the author was a Confucian scholar, “Heaven,” which incorporates the concept of a deity, was deemed more appropriate than the physical space “Sky.” Other evaluations included, “The parallelism of the original text was well-expressed in English literary terms” and “The concise word count preserves the original’s essence.”
Professors who preferred the human translation noted that it had “fewer non-grammatical sentences” and “a more natural title translation.” A professor who declared the translations undecidable stated, “The difference was not significant enough to insist on which translation is better if one of them is AI-generated,” adding, “It should be noted that AI has advanced to the point where it is difficult to distinguish from human translation.”
Experts assess that large language models (LLMs) like ChatGPT, designed to understand context well, inherently excel in translation and have reached a mature stage due to accumulated learning. Choi Byung-ho, a research professor at Korea University’s Human-Inspired AI Research Center, stated, “At least in Korean-to-English translation, it has reached a level where it can replace human translators,” adding, “It should be considered that all publicly available web data has already been learned, and the volume of learning has surpassed the threshold.”
The test was conducted by the office of Democratic Party of Korea lawmaker Min Hyung-bae, a member of the National Assembly’s Culture, Sports, and Tourism Committee. Min stated, “AI is already an irreversible reality,” adding, “While utilizing AI’s efficiency, it is time to contemplate the cultural context and ethics unique to humans.”
◇The Stronghold of ‘World Literature Collections’ Cracks… Absurd Mistranslations Like “Kingbatne” Also Emerge
As AI’s translation capabilities have improved, publishers previously considered the stronghold of major publishing houses like Minumsa and Munhakdongne are now emerging to produce “world literature collections.” These publishers are using AI to translate works by masters whose copyright protection periods (70 years after the author’s death) have expired, without incurring costs.
Translation examples in the Odyssey /Courtesy of Park Jin-seong
A publisher that has been releasing science and technology academic books for over 30 years recently became a topic of discussion in the publishing industry. From October of last year, it released 12 books, including “The Little Prince” and “The Metamorphosis,” in just three months. This publisher, without professional translators, used Gemini for translation and had human editors review the content. The issue arose with the translation of Homer’s epic “The Odyssey,” which included, “Futile conversation is useless. Alppano? (None of my business),” a translation that sparked controversy. Newly coined terms like “Kingbatne! (I’m pissed!)” and “Seubuljae (self-inflicted disaster)” also appeared in the classical work’s translation. The publisher’s representative explained, “We couldn’t afford to produce books if we had to pay translation fees, so we used AI translation. We intentionally left the newly coined terms as they are, thinking they might be necessary for fun intergenerational communication.”
◇“Human Translators Will Utilize AI”
Overseas, business models where humans and AI collaborate have emerged. The UK-based GlobeScribe started a service last summer offering to translate a book for 100 dollars (approximately 145,000 Korean won). Considering that the minimum translation fee for a novel of around 1,000 manuscript pages in South Korea ranges from 3 million to 4 million Korean won, this is extremely affordable. The publisher uses AI to translate most of the text, with human translators refining parts with high literary value or complexity. UK translators are protesting. Ian Giles, chairman of the UK Society of Authors' (SoA) Translators Association, stated in an interview with The Guardian, “It is completely wrong to claim that AI can match or even surpass the delicate work of human translators by replacing authors.”
Domestic translator No Seung-young stated, “Given that AI learns and uses translation sentences refined by humans, most existing translated works will likely be sufficient with AI in the near future,” adding, “However, the importance of proofreading to prevent poor translations from AI has increased.”"
https://www.chosun.com/english/travel-food-en/2026/02/02/TTXCFMS2MJEINIZANZ5WVZB54Q/
"...While political discourse dominates headlines,... something different taking place. “People are independent. They can think for themselves, and they can have their own opinions,” he said. “In these small but growing communities, this is kind of a global village.”
Over time, Jerusalang has become more than a social gathering; it has developed into a network. Through the events, people have found jobs, apartments, travel companions, and romantic partners. One couple, Oland recalled, made a pact to keep attending events together, no matter what the outcome of their relationship was.
Participants hail from all over – Iran, Turkey, China, Korea, and beyond. Conversations often turn to heritage and upbringing, uncovering unexpected stories and highlighting how varied people’s experiences are.
Running Jerusalang has also shaped Oland himself. Drawing on his background in languages and business, he has learned more about marketing, community management, and social media strategy. “I don’t think there was an event where I didn’t learn something new about someone or about something,” he said. Most of all, he has learned how to build and sustain a community – and what people are looking for when they come to his events.
Jerusalang has begun expanding beyond its original format, hosting various events such as a multilingual open mic night where participants sang and played music from their cultures, which included performances in Cantonese, Chinese, Italian, and Arabic. “Other cities deserve this community as well,” he said. He also believes it is especially important for Israelis to learn more Arabic. In a city often defined by divisions, Jerusalang provides a bridging space where interaction is not forced but is casual, human, and authentic."
ByBATSHEVA SHULMAN
FEBRUARY 1, 2026 12:18
Updated: FEBRUARY 1, 2026 12:24
https://www.jpost.com/israel-news/article-884851
#Metaglossia : Gaza on my mind! Where are you?🤔🤔🤔😓😓😓😓
#metaglossia_mundus
#métaglossie
"Recently, it was announced that the Cambridge Dictionary has added around 6,000 new words, including Parasocial, Broligrachy, Delulu, Skibidi, Slop, Memify, Tradwife, Work-Wife, Work-Spouse, Gen Z, and Gen Alpha, among others. There has been heightened enthusiasm in sociological circles that the word “Parasocial” has been selected as Word of the Year 2025 by Cambridge. The term was first used by sociologists, Donald Horton and Richard Wohl, in 1956 to describe how viewers developed one-sided relationships with television personalities.
The factors that influence inclusion of a word in a dictionary include the popularity of the word in digital culture, new technological terminology, opening in new knowledge domains, changes in social setups, transformation of work relationships and even the pandemic. Here, the question arises, “Who decides what counts as a ‘real’ word to form part of a dictionary? How are the words shortlisted and selected?”
Colin McIntosh, Lexical Programme Manager, Cambridge Dictionary, clarifies, “It’s not every day you get to see words like Skibidi and Delulu make their way into the Dictionary. We only add words where we think they’ll have staying power. Internet culture is changing the English language, and the effect is fascinating to observe and capture in the Dictionary.”The words are included based on criteria broadly encompassing frequency of their use; consistency of their meanings, significant enough spread and commonality of definitions across regions.
The criteria of selection, though, seem quite straightforward, but have complexity from a sociological angle: How to decode “staying power” and “significant enough spread” of a word while shortlisting them for dictionaries and ascertaining the neutrality in the selection process.
Qarshiboyeva, M. and Abduraxmanova (2024) have underlined the problems, complexities and challenges of English lexicography. As English is a popular language in both the West and the East, words are coined in abundance in both regions. However, there is a chance of bias in giving more space and accommodation to the words germinated in the West due to a colonial superiority syndrome. The bias of considering the Indian subcontinent as less important continuously haunts the professional ethics of lexicography.
Here, the model of Little Tradition and Great Tradition conceptualised by Robert Redfield comes to be handy and relevant. Redfield puts forward that the culture of folk themes, oral traditions, dialects, and local deities forms the Little Tradition. Whereas, on the other hand, the culture professed by priests and hierarchy of religious leaders covering a legitimate form of all reflective, systematic and textually elaborated rituals/ epics may be considered as the Great Tradition.
Further, while shortlisting words, the possibility of analogically considering the West as a Great Tradition cannot be ruled out. Indian words like Juggad took comparatively a very long time to make their space in dictionaries in comparison to even slurs and slang from Western English. The colonial supremacy syndrome definitely needs to be counterweighed.
The connotation of the same word may be different across different regions. There are chances to overlook the definitions of the word from the “little traditions” or less significant regions. For example, the recently added word “work-wife” means “a woman with whom someone has a close, but not romantic, relationship at work, in which the two people help and trust each other in the same way that a married couple does.” If someone thinks of using this word in Indian work culture, even in a corporate office situated in India, then it may be considered too offensive, as the word wife /spouse cannot be separated from a romantic relationship in the Indian context.
The digital space is the new catchment area for the selection of words. Though the digital space is rich enough from a lexicographical angle but there also exists algorithmic biases which may creep into the selection of words which have lesser acceptance in the general public but have algorithmic abundance. For example, the word “Skibidi” initially started in videos on the YouTube platform as Skibidi toilet. The word, as accepted in the dictionary, means a word that can have different meanings, such as “cool” or “bad”, or can be used with no real meaning as a joke. TikTok and other platforms promoted their use as an online slang, and they could get their acceptance much more easily. Whereas than Indian terms, e.g. “Timepass” and “Dadagiri”, which were popular enough in the Indian subcontinent but had a lesser digital footprint, struggled for a comparatively longer period to find a space finally in the Oxford dictionary in the year 2017 only..."
Vijay Pal is an independent researcher with interests in the sociology of everyday life, education and gender issues."
https://doingsociology.org/2026/02/02/lexicography-through-a-sociological-lens-vijay-pal/
#Metaglossia
#metaglossia_mundus
#métaglossie
Develop AI-powered education tools in local languages, starting with Twi, Ewe, Dagbani, and Hausa.
"Ghana, Google ink AI education deal
Jan 30, 2026, 2:23pm GMT+1
Ghana is partnering with Google to develop AI-powered education tools in local languages, starting with Twi, Ewe, Dagbani, and Hausa. The program will later be scaled up to cover all 12 approved Ghanaian languages.
Of the 7,000 languages used worldwide today, education instruction is limited to 351, leaving many students with huge learning gaps. A shift to bilingual and multilingual education in Africa — which is home to around 2,000 languages — rather than teaching exclusively in the colonial language, usually English or French, has already underscored the benefits of children learning in their mother tongue, the UN said.
Ghana’s education minister emphasized the importance of including Hausa — spoken by around 22 million people across West Africa — in the AI initiative. “Language accessibility determines who can benefit from digital transformation,” TechAfrica News founder Akim Benamara wrote in a column last year.
— Paige Bruton"
Ghana, Google ink AI education deal | Semafor https://share.google/ZjUopsgMQ4Y7yVyWe
#Metaglossia
#metaglossia_mundus
#métaglossie
"L'IA complique-t-elle la tâche des universités ?
Un de mes collègues, professeur de traduction dans une université vietnamienne, a soupiré : « Chaque fois qu'on assigne une tâche aux étudiants, ils se contentent de coller la question dans ChatGPT et de soumettre le résultat. »
La semaine dernière, lors d'une réunion autour d'un café, un de mes collègues, professeur de traduction dans une université vietnamienne, a soupiré : « J'adorais enseigner la traduction. Mais maintenant, à chaque fois qu'on me donne un exercice, les étudiants se contentent de copier-coller la question dans ChatGPT et de soumettre le résultat. Je n'arrive presque plus à leur donner de retour. »
Paradoxe : les connaissances et les compétences ne sont plus rares, pourtant les frais de scolarité augmentent.
Ce phénomène n'est pas propre à l'enseignement de la traduction. Il reflète le malaise croissant qui règne dans le secteur de l'éducation face à la perturbation, par des outils d'IA comme ChatGPT et DeepL, de nombreux modèles de formation, notamment dans les universités, qui existent depuis des décennies.
Pendant des siècles, les universités ont fonctionné selon un principe très simple : le savoir et les compétences étaient rares. Pour acquérir savoir et compétences, les étudiants devaient payer des frais de scolarité, assister aux cours, lire des livres, réaliser des travaux et, finalement, obtenir un diplôme.
Un diplôme constitue à la fois une source de connaissances et une certification de compétences sur le marché du travail.
Mais aujourd'hui, l'IA peut expliquer, synthétiser, traduire et écrire en quelques secondes à un coût quasi nul. Paradoxalement, alors que les connaissances et les compétences ne sont plus rares et deviennent moins chères, les frais de scolarité universitaires augmentent sans cesse.
Le marché du travail réagit plus vite que les universités. Au Royaume-Uni, le nombre d'emplois disponibles pour les jeunes diplômés a chuté d'environ 33 % au cours de l'année écoulée, soit son niveau le plus bas depuis sept ans, principalement en raison de l'automatisation des postes de début de carrière par les entreprises grâce à l'intelligence artificielle. (Niveau d'emploi pour les personnes ayant peu ou pas d'expérience) et réduction des coûts.
Aux États-Unis, plus de 27 États ont supprimé ou réduit les exigences en matière de diplôme universitaire pour un large éventail de postes dans la fonction publique, dans le but d'élargir le vivier de talents et de remédier aux pénuries de main-d'œuvre ainsi qu'à « l'inflation des diplômes » (la tendance à exiger des qualifications scolaires plus élevées pour des emplois qui auparavant ne nécessitaient que des qualifications moindres).
Les entreprises réévaluent la main-d'œuvre à mesure que l'IA remplace de plus en plus les tâches répétitives et liées au codage qui étaient autrefois le domaine des jeunes diplômés.
Au Vietnam, la transformation induite par l'IA est manifeste dans le service client et le marketing, les chatbots et les outils d'IA remplaçant progressivement des rôles fondamentaux.
Alors que de nombreux programmes universitaires enseignent encore des compétences manuelles comme la rédaction de contenu ou la gestion de communautés, les entreprises ont rapidement remplacé les stagiaires et les nouveaux employés par des systèmes d'IA, privilégiant l'embauche de personnes capables d'utiliser l'IA pour améliorer les performances.
Cependant, tous les types de connaissances et de compétences ne se déprécient pas au même rythme. Les domaines qui peuvent être standardisés et rationalisés, tels que le droit, la comptabilité, l'administration, l'ingénierie des opérations et la traduction, sont les plus durement touchés.
J'en ai fait l'expérience directe, tout comme nombre de mes collègues du secteur de la traduction. J'ai perdu beaucoup de clients internationaux qui me confiaient la traduction de contrats et de documents types, car l'IA prend désormais en charge ces tâches plus rapidement et à moindre coût.
Mais j'ai encore d'autres projets, comme la relecture de traductions de Chat GPT, la recherche et l'analyse de groupes de patients pour tester des questionnaires de santé traduits par l'IA, la comparaison des réponses entre les groupes et l'adaptation du langage aux différents contextes culturels.
Ce sont des emplois qui requièrent du jugement, de l'expérience et de l'empathie – des qualités que l'IA, du moins pour l'instant, ne peut pas remplacer.
Un ami architecte a vécu une expérience similaire. Les logiciels et l'IA peuvent certes faciliter la réalisation de plans standard. Mais lorsqu'un projet doit concilier les besoins des personnes, le paysage, la culture, le budget et les contraintes légales, le rôle de l'architecte devient crucial. Aucun algorithme ne peut « comprendre » les individus et le contexte comme un professionnel expérimenté.
L'IA se rapproche de plus en plus de l'humain.
Ces histoires révèlent une tendance de plus en plus nette : l’IA remplace efficacement les tâches répétitives et standardisées ; mais plus elle se rapproche des humains, du contexte, des émotions, de l’éthique et de la responsabilité sociale, plus le rôle des humains devient irremplaçable.
Et c’est là que l’histoire ne tourne plus uniquement autour de la traduction ou de l’architecture, mais aborde directement une institution centrale de la société du savoir : l’université.
Si même l'IA peut obtenir d'excellents résultats à un examen, alors continuer à enseigner et à évaluer de la même manière ne fait que dévaloriser l'université. La valeur des universités aujourd'hui ne réside plus principalement dans la transmission du savoir, mais dans le développement de l'esprit critique, du jugement et des capacités intellectuelles des étudiants.
Cependant, la réalité au Vietnam montre que, même si ce n'est pas le cas pour tous, de nombreux programmes enseignent et évaluent encore de manière traditionnelle : prise de notes, apprentissage par cœur, réalisation de devoirs selon un modèle, tests basés sur des « exemples de réponses ».
Dans le contexte de l'IA, cette méthode pédagogique révèle clairement ses limites. Un rapport de groupe peut être rédigé en une soirée grâce à l'IA ; une présentation peut être créée en quelques minutes ; même les arguments et les preuves peuvent être « préparés pour vous ». Si l'évaluation se limite à la capacité de reproduire un contenu, alors plus les apprenants disposent de technologies, moins ils sont amenés à réfléchir par eux-mêmes.
Bien sûr, il y a aussi eu des évolutions positives. Dans certains programmes avancés, les étudiants doivent analyser les résultats de l'IA, comparer les points de vue, défendre des arguments contre des contre-arguments, travailler sur des projets concrets et assumer la responsabilité de leurs choix.
Ces établissements scolaires font figure de pionniers en intégrant l'IA dans l'enseignement, en organisant des formations pour les enseignants spécialisés et en élaborant des programmes d'études axés sur la maîtrise des outils d'IA. Toutefois, ces approches demeurent dispersées, variant d'un enseignant ou d'un établissement à l'autre, et ne constituent pas encore une orientation systémique cohérente.
La question cruciale n’est pas de savoir si l’IA va « freiner » les universités, mais plutôt : les universités vietnamiennes évoluent-elles suffisamment vite pour passer de l’enseignement des connaissances au développement de la pensée et du caractère humains – l’IA étant un outil puissant pour soutenir les apprenants et les travailleurs ?
Dr. Pham Hoa Hiep
Source : https://tuoitre.vn/ai-co-dang-lam-kho-dai-hoc-20251231112540395.htm";
Báo Tuổi Trẻ
30/01/2026
https://www.vietnam.vn/fr/ai-co-dang-lam-kho-dai-hoc
#Metaglossia
#metaglossia_mundus
#métaglossie
"Santo Domingo. - Legal translators or court interpreters, Alberto Gómez and Luis Manuel Pérez Guzmán denounced that the Attorney General's Office has established new measures that hinder the dynamism of the processes and the ability to respond in a timely manner to deliver documents to users. In this regard, they call on the Attorney General of the Republic, Yeni Berenice Reynoso, to take action, reiterating that our interest is for this measure to be reversed so that legal translators and court interpreters can provide a quality service to the population. They indicated that the new measures are to present the passport or identity document of the actors or of the people who appear in the documents, which has generated enormous delays.
You can also read: Deligne guarantees fair payment for expropriations after protest in Los Alcarrizos
"This measure is not covered by any law that supports it. The role of the legalizations department of the Attorney General's Office is to certify signatures of both legal translators and notaries; that is what they are authorized to do." "We wanted to raise our voice on behalf of the thousands of users who daily seek these services, as well as on behalf of all legal translators who are being affected by the implementation of this measure, which was abruptly and without prior notice put into effect two weeks ago," said Alberto Gómez, who is the director of Tralega Traducciones Legales. Luis Manuel Pérez Guzmán, director of Dragoman Traducciones reiterated "this measure has caused much uproar and displeasure in the population that requires their legalized documents on time." Alberto Gómez and Luis Manuel Pérez Guzmán, who represent the main legal translation companies in the Dominican Republic, maintain that this joint effort has the mission of helping the population that requires these services for different reasons, for example, legalization of translations of birth certificates, marriage certificates, death certificates, legal documents for studies, among others. “In addition to offering a quality service, in each request, we guide clients to resolve their procedures effectively. We do our work out of vocation, aware that the most gratifying honor a human being can have is to serve at the right time,” pointed out Alberto Gómez and Luis Manuel Pérez Guzmán, who have more than 20 years of professional experience."
Paola Castillo
January 29, 2026
https://deultimominuto.net/en/uncategorized/legal-translators-or-court-interpreters-denounce-absurd-measures-in-the-legalization-of-documents-and-call-on-the-attorney-yeni-berenice-reynoso/
#Metaglossia
#metaglossia_mundus
#métaglossie
"Judiciary clarifies interpreter deployment at Opuwo court
The Office of the Judiciary has clarified how indigenous language interpreters are deployed at courts, following concerns over the availability of Otjiherero and Damara/Nama interpretation services at the Opuwo Magistrate’s Court.
Judiciary spokesperson Vikitoria Hango tells Nampa in a media response that the current structure of the judiciary presents challenges in ensuring that all indigenous languages are represented at every court simultaneously.
She explains that the allocation of interpreters is guided by geographical location and the predominant languages spoken in each region.
“The distribution of interpreters is determined by the geographical location of courts and the dominant language in that area. When specific language interpreters are required, a systematic process is followed to source them in advance,” Hango says.
She adds that the Office of the Judiciary has not identified any backlog of court cases attributable solely to a shortage of interpreters, noting that mechanisms are in place to ensure proceedings remain fair and accessible.
Language access remains a key component of the justice system, particularly in rural areas where multiple indigenous languages are widely spoken.
The Opuwo Magistrate’s Court serves communities across the Kunene region."
By Namibia Press Agency
29 January 2026
https://www.namibian.com.na/judiciary-clarifies-interpreter-deployment-at-opuwo-court/
#Metaglossia
#metaglossia_mundus
#métaglossie
This two-volume, 1,600-page dictionary is the result of more than fifty years of collaboration among Native speakers, educators and linguists and contains more than 19,000 entries. The dictionary also includes a grammatical sketch of the language, a pronunciation key and guides to noun forms and verb conjugations. Designed by Michael Alpert, director emeritus of the UMaine Press, this edition of the dictionary features a user-friendly layout of entries with examples of everyday usage for each word.
"The University of Maine Press published the second edition of “Peskotomuhkati-Wolastoqey Latuwewakon: Passamaquoddy-Maliseet Dictionary” by David A. Francis, Robert M. Leavitt and Margaret Apt.
This two-volume, 1,600-page dictionary is the result of more than fifty years of collaboration among Native speakers, educators and linguists and contains more than 19,000 entries. The dictionary also includes a grammatical sketch of the language, a pronunciation key and guides to noun forms and verb conjugations. Designed by Michael Alpert, director emeritus of the UMaine Press, this edition of the dictionary features a user-friendly layout of entries with examples of everyday usage for each word.
Passamaquoddy and Maliseet communities in eastern and northern Maine will be hosting book launches in the coming weeks, and other events are being planned for the spring at UMaine and its regional campus, the University of Maine at Machias. Goose Lane Editions of New Brunswick is distributing the dictionaries in Canada.
To place orders, contact Betsy Rose, betsy.rose@maine.edu. More UMaine Press publications are available at umaine.edu/umpress/."
https://umaine.edu/news/2026/01/new-edition-of-passamaquoddy-maliseet-dictionary-published-by-umaine-press/
#Metaglossia
#metaglossia_mundus
#métaglossie
"One out of every two people worldwide speak one or more languages. With over 7,000 languages spoken globally, bilingualism is a skill that will forever be useful.
In the North Penn School District alone, there are around 80 languages spoken among students and families, and this year, the district has decided to award those who can speak multiple languages “proficiently” with the PA Seal of Biliteracy (PASB).
If a student, by the time of high school graduation, is proficient in both English and one or more additional language, they are eligible for the seal. This includes ELs (English Learners) and those who speak another language at home.
“[Through this award, students can receive] credit for being bilingual in their native language. Bengali, Gujarati, or Arabic are commonly spoken languages in our district population [that we don’t offer courses for here at North Penn,” Curriculum Supervisor Rachel Earley said.
The first step in getting the seal is making sure all the prerequisites are met. Students must meet all the requirements needed to graduate high school, as well as prove proficiency in English and their second language. This must be done via grades in English and world language courses, as well as one additional document of proof each. The secondary piece of proof can be attained through various “pathways.”
English proficiency can be proven through a passing score on a state assessment, an AP exam, or a test of equal status, or, alternatively, through a portfolio that “meets the criteria for listening, speaking, reading, and writing at the intermediate or higher English proficiency level,” which will be reviewed and approved by the school’s Seal of Biliteracy committee.
Additional language proficiency can be proven through school transcripts from another country (including Puerto Rico as a territory) proving three or more years of instruction averaging a B or higher, a passing grade on an approved language assessment, or a portfolio that matches the same criteria as above.
To view more details and all criteria necessary for the seal, click here.
In a world where the word “fluency” is used extremely liberally, certification of bilingualism can provide security for those who see it, such as employers and universities.
According to the Pennsylvania State Modern Language Association (PSMLA), this award “provides employers and universities with a method of identifying people who are bilingual, recognizes students with 21st century skills, and encourages the study of other languages and cultures.”
And while this award does look good on job applications, and might even be the last deciding factor over a competing candidate, students can also receive language credits in college depending on the school they are considering.
Those on the aforementioned Seal of Biliteracy committee are set on encouraging applications for the PASB. This being the inaugural year for the award at North Penn, they look forward to seeing many more students receive it in the future."
https://www.knightcrier.org/student-life/2026/01/30/pa-seal-of-biliteracy-recognizes-multilingual-students-at-north-penn/
#Metaglossia
#metaglossia_mundus
#métaglossie
"A series of events will be held throughout the year to commemorate this historic initiative. The ‘Languages 250 at Trinity (1776-2026)’ programme will showcase the important work of the School of Languages, Literatures and Cultural Studies and promote the cause of modern language education on the island of Ireland.
The School of Languages, Literatures and Cultural Studies last night kickstarted a year of celebrations to mark 250 years of modern languages in Trinity.
Trinity College Dublin was the first university in these islands to introduce the study of modern continental languages. In 1776 Trinity appointed professors in French, German, Spanish and Italian. They were the first university Chairs in modern languages and, in the case of the Chairs in French and German, the oldest continuous Chairs in these languages in the world.
The establishment was spearheaded by the newly appointed Provost John Hely Hutchinson, motivated by his desire to equip college graduates with language skills for foreign travel and cultural engagement, opening them up to the exciting possibilities of Enlightenment Europe.
Last night’s opening ceremony in Trinity Long Room Hub featured an address by Pat Cox, President, Jean Monnet Foundation and former President of the European Parliament.
Pat Cox is pictured above on left with Provost Dr Linda Doyle and Thomas Byrne, T.D, Minister of State for European Affairs
Pat Cox said: “We are marking one of the world’s and Ireland’s most enduring academic traditions, the institutionalisation of modern language teaching in French, German, Italian and Spanish established through professorships here at Trinity College Dublin 250 years ago.
“Beyond history, the marketplace, career opportunity, and technological aids, fluency in a foreign language builds a self-enriching bridge to new cultures, literature, cinema, networks and friendships, offering a different prism through which to view the world. A shared language confers a sense both of connection and belonging even in the most trying of contexts.”
Provost of Trinity College Dublin Dr Linda Doyle responded:
“Trinity is immensely proud of its 250-year-old tradition in modern language education and remains committed to the promotion of modern languages to create a more open, tolerant, and culturally diverse world.”
Starting in January, a series of events will be held throughout the year to commemorate this historic initiative. The ‘Languages 250 at Trinity (1776-2026)’ programme will showcase the important work of the School of Languages, Literatures and Cultural Studies and promote the cause of modern language education on the island of Ireland.
Highlights include a conference on ‘Translation in Ireland: Past and Future’ in May and in October a conference on ‘Modern Languages in Irish Life’. Also in October, the Hely-Hutchinson memorial lectures will be delivered by Professor Yves Citton, Université Paris 8 (Vincennes-Saint Denis) and author Ulrike Draesner.
Michael Cronin, Chair of French 1776 (above), said:
“The celebrations in 2026 represent a unique opportunity to honour Trinity College’s pioneering role in modern language education, and to build on this legacy to further deepen and strengthen Ireland’s European connections.”
Mary Cosgrove, Professor of German 1776, said:
“The origins of the Chairs of Modern Languages some 250 years ago was a matter not just of institutional, but of national and international significance: their establishment put Trinity, and Ireland, on the map as a pioneer of Modern Languages education at the highest level. In this history, Trinity emerges as a radical, modernising force.
“The advent of the Chairs was furthermore in line with a revolutionary spirit and progressive sensibility as proclaimed in the American Declaration of Independence of 1776: the Chairs’ emergence indicates the start of a democratising shift, in the university context, from the ancient to the modern, from the study of classical languages to the learning of vernacular and foreign languages.”
Timeline: Chairs of Modern Languages in Trinity:
4 August 1774: Newly appointed Provost John Hely Hutchinson outlines his proposal to the Board of Senior Fellows that ‘it would be highly useful to have Professors of the modern Languages established in the college’
August 1775: R. Antonio Vieyra Transtagano arrives in Trinity to teach Spanish and Italian (Vieyra is the author of a Portuguese-English dictionary and a Portuguese grammar)
September 1775: Anthony D’Esca arrives in Trinity to teach French and German (Voltaire scholar, born in Berlin)
29 October 1776: Letter from King George III confirming appointment of two professors and granting sum of £200 per annum from the royal estates in Ireland to be divided equally between them ‘and to commence from the 29th day of September last.’
15 January 1777: Vieyra and D’Esca granted honorary degrees of LL.D.
26 January 1786: Rev. Frances Bessonnet appointed Professor of French (due to death of D’Esca in January 1785). Lt. Col. James Philip Hamilton appointed Professor of German.
1790-1801: Francis Amyot: Professor of French and German
1802-1841: Charles Willomier – Professor of French and German
1824: Evasio Radice – Professor of Italian and Spanish
1842-1866: Ignatius George Abeltshauser – Professor of French and German
1849: Basilio Angeli – Professor of Italian and Spanish
1866-1907: Albert Maximilian Selss – Professor of German
1862: Augusto Cesare Marani – Professor of Italian and Spanish
1869: Robert Atkinson – Professor of Romance Languages (French, Italian and Spanish combined)
1907: Maurice Alfred Gerothwohl – Professor of Romance Languages
1907: Robert Alan Williams – Professor of German
1909: Thomas B. Rudmose-Brown – Professor of Romance Languages"
Posted on: 30 January 2026
https://www.tcd.ie/news_events/top-stories/featured/celebrations-to-mark-250-years-of-modern-languages-at-trinity/
#Metaglossia
#metaglossia_mundus
#métaglossie
"The path towards a nationwide language resurgence is winding, but schemes like The Queen’s College Translation Exchange (QTE) show the way forward.
School children, primarily in Key Stage 2 and 3, attend sessions run by Oxford undergraduate students who pick the texts. Previously they’ve been tasked with translating graphic novels, poems and comics.
The emphasis lies in the process of creative translation, rather than “functional” translation.
In 2025, 22,000 students aged 11-18 took part in the college’s Anthea Bell Translation Prize and the 2025-2026 competition will run from Monday until 27 March.
Dr Charlotte Ryland, the founding director of The Queen’s College Translation Exchange, said: “We’re trying to present an alternative narrative to the doom and gloom.
“Doing linguistic work in the service of something creative and purposeful, which is what we think is missing from school curriculums.”
The prize has proved extremely popular since it’s 2018 origins and in 2025, QTE were able to add Russian language as their sixth language strand included in the translation prize.
Ryland said: “It’s a Russian language prize rather than Russian literature. So there’s texts from writers who are grew up in and lived in countries before the Soviet Union and that’s a really important part of how it is presented.”
Oxford University Professor of Brazilian Literature and Culture Claire Williams added: “Language teaching should be embedded in education, as it used to be.
“The earlier children start learning another language, the more they appreciate the nuances between languages, never mind not having the inhibitions that self-conscious teenagers do.”
In the 2023/24 academic year, less than half (46%) of eligible students in England opted for a language GCSE.
That is compared to 98% of upper secondary students in the EU who study a minimum of one foreign language, according to a recent University of Cambridge study.
QTE also values the languages spoken outside of the school gates, too.
Ryland said: “It’s so important. If we’re going to raise the profile of languages and language learning across the country, we have to start out there with the languages that are in our classrooms rooms, in our homes, and validate those.”
Although the government’s pledge to once again join Erasmus+ in 2027 is a step towards re-building bridges, modern foreign language provision remains in critical condition across the education spectrum in the UK.
A host of universities in the East Midlands are facing modern language course closures.
At the University of Nottingham, all undergraduate modern language degrees will be suspended under the proposed Phase 2 of Future Nottingham, which was announced on 6 November.
They would be the only Russell Group university not to offer modern languages. This unpopular decision comes as a result of the drastic reduction in applicants.
Moreover, the University of Leicester has also proposed closing their languages department and will not accept students from 2026/27 onwards.
Williams said: “I felt shock and disappointment at the short-sightedness of closures and the lack of investment in language-learning.”
Nottingham’s 2023/24 cycle only saw 42 applications to study BA Hons French, in comparison to Oxford University’s 331 applications for the same cycle.
At Oxford University, French still receives the highest number of applicants with 331 in 2024, with Spanish in second, having 226 applicants. However both these numbers have consistently dropped from 2021.
Total 2021 MML applicants (not including joint honours degree with subjects such as English) came to 1,017 in 2021 but fell by 18% by 2024 with 831 applicants, according to Oxford’s own faculty data.
Italian saw the greatest decline with an 84% decrease in applicants from 2021 to 2024.
With such a numerical disparity, the issue of a two-tiered university system manifests as a legitimate concern.
Williams said: “I’m really worried that this might happen, unless we start teaching languages at primary school, or capitalising on the multilingualism of primary schools.”
The QTE has made steps to address this regional disproportionality at secondary school level.
The translation exchange is currently working with six schools in the North West of England to pilot their Think like a Linguist programme.
Students in Blackpool and Rochdale are experiencing the programme that introduces children who live in more monoculture environments to language learning.
Even more worryingly, out of the students who have continued with languages at A-Level and receive university offers, even fewer actually accept to study them at a higher level.
According to the University of Nottingham’s own data, only less than 10% of students holding offers to study Hispanic Studies accepted in 2023/24 – four out of 42 students.
Williams said: “Perhaps more candidates will apply to Oxford. Hopefully they don’t decide that Modern Languages is not a ‘useful’ degree course.
“My students have learned valuable transferable skills: formulating a persuasive argument, intercultural criticism, close reading, confidence in expressing their ideas in several languages, time management, translation, to name a few.
“Knowing another language is cool, a superpower, not something to be ashamed of.”"
https://www.swlondoner.co.uk/life/30012026-translating-comics-an-oxford-college-leading-language-revival-in-the-uk
#Metaglossia
#metaglossia_mundus
#métaglossie
|
"Google has collaborated with African universities and research institutions to launch WAXAL, an open-source speech database designed to support the development of voice-based artificial intelligence for African languages.
African institutions, including Makerere University in Uganda, the University of Ghana, Digital Umuganda in Rwanda, and the African Institute for Mathematical Sciences (AIMS), participated in the data collection for this initiative. The dataset provides foundational data for 21 Sub-Saharan African languages, including Hausa, Luganda, Yoruba, and Acholi.
WAXAL is designed to support the development of speech recognition systems, voice assistants, text-to-speech tools, and other voice-enabled applications across sectors such as education, healthcare, agriculture, and public services.
“This dataset provides the critical foundation for students, researchers, and entrepreneurs to build technology on their own terms, in their own languages,” said Aisha Walcott-Bryantt, Head of Google Research Africa
WAXAL’s launch comes amid growing efforts across Africa to develop language technologies that reflect local cultures and realities.
In September 2025, the Nigerian government unveiled N-ATLAS, an open-source language model capable of recognising and transcribing spoken words and generating text, in Yoruba, Hausa, Igbo, and Nigerian-accented English.
Similar initiatives are emerging in the private sector, where startups such as South Africa’s Lelapa AI are building tools like Vulavula, which offers speech recognition, translation, and sentiment analysis.
By making this speech dataset openly accessible, WAXAL provides the fuel for a growing wave of homegrown efforts to bring African languages into the digital age.
Although Sub-Saharan Africa is home to more than 2,000 languages, reports suggest that fewer than 5% of those languages have the resources needed for Natural Language Processing (NLP), which allows computers to understand and comprehend human language. This lack of representation in training datasets limits the effectiveness of speech recognition and text-to-speech systems for African users.
Developed over three years with funding and technical support from Google, WAXAL addresses a major gap in global AI development.
WAXAL provides speech data for 21 Sub-Saharan African languages, including Fulani (Fula), Hausa, Igbo, Ikposo (Kposo), Swahili, and Yoruba. The dataset contains more than 11,000 hours of speech drawn from nearly two million individual recordings.
Under the project’s partnership model, contributing institutions retain ownership of the data they collected, while making it openly available to researchers and developers worldwide.
“For AI to have a real impact in Africa, it must speak our languages and understand our contexts,” Joyce Nakatumba-Nabende, Senior Lecturer at Makerere University’s School of Computing and Information Technology, said.
“The WAXAL dataset gives our researchers the high-quality data they need to build speech technologies that reflect our unique communities.”"
Opeyemi Kareem
2nd Feb, 2026
https://techcabal.com/2026/02/02/google-joins-push-to-localise-ai-for-african-languages-with-speech-database/
#Metaglossia
#metaglossia_mundus
#métaglossie