Through music, online learning, and AI, old dialects are finding new life and helping younger audiences reconnect with identity and tradition.

At the 2026 Tibetan New Year Gala, 26-year-old singer Mamcu stepped onto the stage. As the lights dimmed and came back up, a lively Afro-style beat filled the auditorium, set against the traditional Tibetan patterns of her costume.

She opened with a line in Tibetan: "Droma has finally grown into the person she once admired."

The song, Sa Sa Sa, which tells the story of Droma, a Tibetan girl finding her voice — a journey that closely mirrors Mamcu's own — has captivated listeners since its release in October 2025, amassing tens of millions of streams. Nearly 90 percent of its audience is aged 20 to 30.

"I was amazed that so many young people enjoy it, even without understanding the Tibetan lyrics," Mamcu said, reflecting on the song's ability to transcend language and cultural boundaries.

READ MORE: Expanding career paths for Tibetan youth

Yet behind this cross-cultural appeal lies a more urgent reality.

According to the China's Endangered Languages series, published by the Commercial Press, China is home to more than 130 languages. Yet 68 have fewer than 10,000 speakers, 48 fewer than 5,000, and 25 fewer than 1,000. Some are spoken by only a handful of people and are on the brink of extinction.

For Mamcu, these figures are more than statistics — they are a lived experience.

"Many young people in cities no longer speak their hometown dialects," she said."But for me, my native language is the truest way to express who I am. It's part of me that I will never lose."

Born into the Jiarong Tibetan community in Aba Tibetan and Qiang autonomous prefecture, Sichuan province, Mamcu grew up surrounded by ethnic music in her mother tongue.

"In Tibetan communities, girls are often expected to be gentle and stay in the background. But the women in my family have always chosen their own paths. From my grandmother to my mother, they each carved out a place in music," she said.

Following in their footsteps, Mamcu began shaping her own career. At 18, she collaborated with a rapper and performed a hook in her native language for the first time.

"That experience planted a seed — to let the Tibetan language grow and thrive through music," she said.

Today, Mamcu is not only a celebrated singer but also a cultural bridge, drawing audiences across China toward the Tibetan language and traditions.

She shares tutorial videos online, breaking down Tibetan lyrics with phonetic guides. Many fans have started learning alongside her.

"Music should not be boring," she said."Blending the Tibetan language with different styles is an exciting experiment in itself."

Reviving lost voices

Mamcu's efforts reflect a broader linguistic reality.

A similar story is unfolding in Suzhou, East China's Jiangsu province, where 38-year-old Hu Shuning has been confronting the gradual disappearance of her local dialect.

After returning from her studies in Spain in 2014, Hu discovered that her 11-year-old cousin could neither speak nor understand the Suzhou dialect.

The realization was striking: even in a family that valued dialect education, fluency was slipping away.

"At that moment, I could see it disappearing before my eyes," she recalled.

Determined to do something about it, Hu began taking classes to refine her pronunciation and earned a teaching certificate in early 2015.

She soon discovered that most existing learning materials were "too academic for ordinary learners". So, drawing on her background as a Spanish teacher, Hu developed her own materials — including free textbooks, exercises, and audio recordings — which she has shared online over the past decade.

What surprised her most was that many of her learners came from outside Suzhou.

"That gave me a real sense of fulfillment," she said. "It showed me that my language is needed."

Building on this momentum, Hu partnered with a local bookstore in 2025 to launch night school classes.

Four sessions have been held so far, attracting participants — mostly aged 20 to 40 — who are eager to reconnect with a fading cultural heritage.

As she worked to promote the dialect, Hu often found herself asking how it could be revitalized. She noted that, in the past, many cultural forms — novels, pingtan (a traditional storytelling and singing art form), and Kunqu Opera — had flourished in it, but its artistic vitality is now fading as fewer people speak it.

"The Suzhou dialect, with its seven tones and subtle tonal shifts, poses a particular challenge for creative expression," Hu explained.

Encouragingly, Hu has seen younger creators begin to experiment with the language. One of her online students, a content creator on the Chinese video platform Bilibili, inputs Suzhou dialect phonetics into a virtual singer, allowing songs to be performed in the local tongue.

"A language can only stay alive through continuous creation," Hu said.

Old words, new tools

Beyond teaching, Hu is also exploring how technology might help preserve the dialect. In recent years, she has collaborated with iFlytek, a leading Chinese company in AI and speech technology, to develop AI-generated Suzhou dialect speech.

"Teaching a machine to speak the dialect means annotating every sentence," Hu explained. "The tonal variations make the process even more complex."

Although the product has yet to reach the market, similar efforts are already underway elsewhere. One notable example is Wang Mubin, a project manager at Guangdong Qunyu Interactive Technology, based in Chenghai district of Shantou, South China's Guangdong province.

Wang and his team developed an AI-powered toy named Dingguagua, designed to recognize the Chaoshan dialect. Modeled after the local lion-headed goose mascot, its name comes from a Chaoshan phrase meaning "excellent".

"Chenghai is known as a hub for toy manufacturing," Wang said. "With this toy, we aimed to create something that could both translate the Chaoshan dialect and help children learn it, while also strengthening their sense of cultural identity."

The project grew out of a common challenge faced by many local families: older generations often speak only the dialect, while children are raised using Mandarin, making communication increasingly difficult.

"The Chaoshan dialect has no standardized writing system and depends entirely on oral transmission. As fewer people use it today, it could quickly become endangered without technological support," Wang said.

After more than a year of work, the team collected over 2 million voice samples across different age groups and everyday scenarios. The AI-powered toy can now recognize Chaoshan speech with about 90 percent accuracy — surpassing many comparable products — and comes preloaded with 40 children's songs in the dialect.

ALSO READ: Threading the past forward

"I hope AI can inspire young creators to produce content in the dialect," Wang said."These materials can then be shared and appreciated, giving the language a real chance to thrive."

The emergence of projects like this reflects a broader shift in attitudes toward dialect preservation in China.

Hu recalled that more than a decade ago, her efforts to promote the Suzhou dialect were often dismissed as unnecessary.

"People would ask, 'Why protect it? Isn't everyone still speaking it?'" she said.

Today, however, a growing number of learners recognize that dialects are in decline and that their cultural space is shrinking.

"For young people, awareness is only the first step. They need to understand that dialects matter — not as something optional, but as something worth holding on to," Hu said.

Meng Shuyan contributed to this story.

Contact the writers at mengwenjie@i21st.cn