British scientist Michael Faraday addresses the Royal Institution in London in 1855, in a painting by Alexander Blaikley. Photo: Rischgitz/Getty
For centuries, science was a multilingual affair, powered by French, German, English and other tongues. But since the early 1970s, English has become the undisputed lingua franca of scientific papers, conferences, and discourse.
Writes Axios' Kaveh Waddell:English-speaking countries now have a huge leg up in technical research, including the current rages — artificial intelligence and quantum computing.But, while English is highly unlikely to be dethroned, its advantages are eroding.This is due to an increasinglyhealthy research environment in China, the fast transmission of research papers across the internet, and AI-aided translation technology that's shrinking the language barrier.
Background: Speaking English isn’t optional for scientists, especially leading-edge computer science researchers.
Jacqueline Le Moigne, assistant technology chief at NASA Goddard Space Flight Center, did her graduate studies in her native France. But when she began researching AI and computer science, she found that even in France, English was the language of choice for the majority of journals and conferences."French researchers publish and collaborate in English," said Paul Indelicato, senior research advisor to the Conference of University Presidents in France. "Conferences and workshops, even [those] organized in France by French researchers, are in English."Tellingly, Indelicato responded in English to questions posed in French.
The dominance of English gives native speakers a huge advantage, says Michael Gordin, a Princeton professor who specializes in the role of language in technological advance.
A New Zealand math or physics student learns specifically those subjects. But a student from China learns those — plus English. "That is a competing drain on your attention," Gordin tells Axios.It was not always this way. English has been a scientific language since the 18th century, but only eclipsed other languages in the 1970s, says Gordin. English achieved absolute dominance in the 1990s, alongside the rise of the internet, and today more than 90% of all academic papers are published in English.There is variation between disciplines. When she studied pure math, Le Moigne read mostly French. Her PhD thesis, however, required German sources on medical optics.
Countries that build up English proficiency can benefit from the advantages the language brings.
Since 2015, an annual study from EF Education First, which sells English-learning programs, reports a relationship between a country’s level of English proficiency and scientific innovation.But no one suggests that native English speakers are better at doing science. Instead, a firm grasp of the language allows them to collaborate with other researchers across borders, communicate seamlessly in conferences, and submit papers to prestigious journals with the smoothest, most persuasive prose.
The big picture: An increasingly independent academic environment in China means non-Chinese speakers are missing out on important information.
Andrew Ng, a leading AI researcher at Stanford, said last year that this creates an asymmetry: “China has a fairly deep awareness of what’s happening in the English-speaking world, but the opposite is not true,” he told The Atlantic’s Sarah Zhang.Testing Ng’s asymmetry theory, Jeffrey Ding, a researcher at Oxford’s Future of Humanity Institute, found in March that important AI-related documents were being translated into Chinese within days of being published. By contrast, few China-watchers are translating Chinese AI developments into English.
What’s next: Machine translation is improving quickly. Though its work is still riddled with errors, Google Translate has been boosted by AI; meanwhile, researchers are taking early stabs at real-time translation, an extremely difficult task for humans and computers both.
Eventually, says Ding, "maybe the critical language of the future won't be English or Mandarin, but Python."