Thirteen years ago, a deadly strain of avian flu known as H5N1 was tearing through Asia's bird populations. In January 2004, Chinese scientists reported that pigs too had become infected with the virus—an alarming development, since pigs are susceptible to human viruses and could potentially act as a "mixing vessel" that would allow the virus to jump to humans. "Urgent attention should be paid to the pandemic preparedness of these two subtypes of influenza," the scientists wrote in their study.
Yet at the time, little attention was paid outside of China—because the study was published only in Chinese, in a small Chinese journal of veterinary medicine.
It wasn't until August of that year that the World Health Organization and the United Nations learned of the study's results and rushed to have it translated. Those scientists and policy makers ran headlong into one of science's biggest unsolved dilemmas: language. A new study in the journal PLOS Biology sheds light on how widespread the gulf can be between English-language science and any-other-language science, and how that gap can lead to situations like the avian flu case, or worse.
"Native English speakers tend to assume that all important information is in English," says Tatsuya Amano, a zoology researcher at the University of Cambridge and lead author on this study. Amano, a native of Japan who has lived in Cambridge for five years, has encountered this bias in his own work as a zoologist; publishing in English was essential for him to further his career, he says. At the same time, he has seen studies that have been overlooked by global reviews, presumably because they were only published in Japanese.
Yet particularly when it comes to work about biodiversity and conservation, Amano says, much of the most important data is collected and published by researchers in the countries where exotic or endangered species live—not just the United States or England. This can lead to oversights of important statistics or critical breakthroughs by international organizations, or even scientists unnecessarily duplicating research that has already been done. Speaking for himself and his collaborators, he says: "We think ignoring non-English papers can cause biases in your understanding."
His study offers concrete examples of the consequences of science’s English bias. For instance, the latest population data on the fairy pitta, a bird species found in several Asian countries and classified as vulnerable, was not included in the latest assessment by the International Union for the Conservation of Nature. The reason, again, was that the paper was only published in Chinese.
For the new study, Amano's team looked at the entire body of research available on Google Scholar about biodiversity and conservation, starting in the year 2014. Searching with keywords in 16 languages, the researchers found a total of more than 75,000 scientific papers. Of those papers, more than 35 percent were in languages other than English, with Spanish, Portuguese and Chinese topping the list.
Even for people who try not to ignore research published in non-English languages, Amano says, difficulties exist. More than half of the non-English papers observed in this study had no English title, abstract or keywords, making them all but invisible to most scientists doing database searches in English. "I think this issue is actually much larger than many people think," Amano says.
This problem is a two-way street Not only does the larger scientific community miss out on research published in non-English languages. But the dominance of English as science's lingua franca makes it more difficult for researchers and policy makers speaking non-English languages to take advantage of science that might help them. For example, of 24 conservation directors in Spain surveyed by Amano and his team, 13 said that a language barrier made their jobs more difficult by limiting their access to information on conservation.
It's also worrisome that English has become so prestigious for scientists that many non-English speakers avoid publishing research in their own languages, Amano says. For context, Dutch scientists publish more than 40 papers in English for every 1 article in Dutch, according to a 2012 analysis by the publication Research Trends. The desire to publish in respected English journals is even prompting journals in some countries to decrease or cease publishing in their local languages.
Federico Kukso, a MIT Knight Science Journalism fellow who has reported on science in Spanish and English for more than 15 years, says the bias extends beyond how scientists view studies; it also manifests in what science the media chooses to focus on. The Argentina native has previously written about how English-language media tends to ignore the work of Latin American scientists, and especially when these scientists collaborate with American or British scientists.
The hegemony of English-language science—and science journalism—has led to the elevation of the work of British and American scientists above that of other nations, Kukso says. He gives an example from earlier this year, when an accomplished Argentinian paleontologist named Sebastián Apesteguía helped discover a new species of dinosaur. Most English-language media didn’t even mention him, instead focusing on his American collaborators.
"They don't cover the scientific breakthrough of scientists in Latin America, Asia, Africa, until someone dares to translate it," Kukso says of English-language science journalists. "It's as if non-English science doesn't exist at all."
Amano thinks that journals and scientific academies working to include international voices is one of the best solutions to this language gap. He suggests that all major efforts to compile reviews of research include speakers of a variety of languages so that important work isn't overlooked. He also suggests that journals and authors should be pushed to translate summaries of their work into several languages so that it's more easily found by people worldwide. Amano and his collaborators translated a summary of their work into Spanish, Chinese, Portuguese, French and Japanese.
Scott Montgomery, a geologist at the University of Washington, agrees that this is an important issue that needs solving. However, when it comes to methodology, Montgomery, who has written extensively on science communication and participated in early peer review process of Amano's paper, thinks that the study "lacks real substance beyond adding to a literature of complaint that has emerged over the past 20 years."
The authors took little effort to differentiate between research that was peer-reviewed and research that wasn’t in their Google Scholar searches, Montgomery says, making it hard to quantify how much serious research is published in non-English languages. He adds that the authors ignore the historical context of this problem. Just a few decades ago, scientific communication was much harder because there was no dominant lingua franca to publish and share research in, he says.
"There were more language barriers, and they were thicker and higher," Montgomery says.
While the rise of English as a global and scientific second language does handicap some scientists in other countries, it has also been instrumental in greasing the wheels of communication, he argues. Montgomery is also skeptical of the proposals of Amano and his collaborators to ramp up translation for scientific research. "Scientific translation—which I did part-time for 10 years—is not cheap or fast, and machine translation is a very long way from doing the job, if it ever will," he says.
Scientists in all fields would benefit from learning another language, Montgomery says—including native English speakers. But he believes that the best solution to science's language barrier is encouraging scientists worldwide to study English. This may seem unfair to say as a native speaker, he concedes, but as English continues to spread and thrive worldwide, he says it is increasingly necessary. "It is a difficult process, with a rough justice to it," Montgomery says. "But it is profound, human and repeatedly proven."
Montgomery and Amano agree one at least one thing: Ignoring language barriers in science is dangerous. "Someone needs to seriously start tackling this issue," Amano says.