Science, with its talk of theories, hypotheses, and frameworks, at times seems like a pretty abstract endeavour. But when the hypotheses have been tested and the theories have been hatched, scientists have to condense it all down into concrete human language. One language: English.

Estimates of the dominance of English in scientific writing vary, but all converge on the conclusion that English is the language that is overwhelmingly used to communicate science. Up-to-date studies are lacking, probably because the language’s dominance is so obvious that nobody bothers. A 2003 analysis found that, by 1995, 87.2% of journal publications in the natural sciences were in English. Likewise, English publications outweigh those in local languages in all eight countries selected for a different study, by a factor of over forty in the case of the Netherlands. China stands as a notable exception, but even here scientists are under strong pressure to publish in English-language journals.

The hegemony of English is a recent - and historically unique - phenomenon. The predecessors of the zealously anglicising Dutch scientists would have used a combination of languages. First, Latin was dominant, but not exclusive (Isaac Newton published the treatise Opticks in his native language). Some daring professors began to lecture in their own languages at the beginning of the 19th century, and by the end French, German, and English coexisted.

“Utopian thinkers have dreamt for centuries about a world language, and in a way the dream has come true”

Scientists had to at least read all three, although the primary language depended on the field. If you did biology, for example, you relied on German sources. As one author in Nature put it: “In the region of scientific medicine the Germans enjoy at the present time an undisputed pre-eminence. Their medical books have taken possession of the markets of the world.”

In the 20th century, two world wars struck German firmly off the list of scientific languages, while the American Century brought with it increased use of English. By 1970, English accounted for 70% of literature in the natural sciences. Most of the remaining science was published in Russian. But Russian use fell in the 1980s and by the end of the 20th century, science was thoroughly monolingual.

Utopian thinkers have dreamt for centuries about a world language, and in a way the dream has come true. Proponents of Esperanto and the like always maintained that the use of a single language would remove communication barriers, increase cultural understanding, and promote peace. In science, the use of English does make life easier for those who speak it. Knowledge of English alone unlocks the world’s scientific conversation, whether in papers, at conferences, or in the increasingly international environment of research universities.

“We need more languages, not fewer, for everyone to feel included in one of humanity’s most exciting endeavours”

The obvious problem is that most of the world’s population is not born in an English-speaking country. Life as an English-speaking student of natural sciences at Cambridge is hard. There is a lot to learn. There are many skills to acquire before the determined first-year student can become a paper-publishing academic. But one thing they do not have to be worried about is language, other than maybe some points about scientific writing style. On the other hand, an equally determined first-year student at some university in Eastern Europe, starting out on their journey into academia, also has to master a foreign language to a level high enough to be able to coherently express scientific arguments. Even for students who are privileged enough to have access to high quality English-language education from an early age, this is not an insignificant task. Unfortunately, the effect of local language on academic output is more or less untestable because it is impossible to control for more important factors, but it is interesting to speculate to what extent the supposed superiority of English-speaking universities such as Cambridge is simply due to their location.

There is a cultural imbalance in access to English as well as a geographical one. Latin was not ideal as an international language, but at least it was politically neutral. English is not politically neutral. It is the official language of the most powerful state on the planet, which is itself a former colony of a former empire that made a point of imposing English on the people it colonised. There are some who expand the political connotations of English into a thesis of ‘linguistic imperialism’, making it into a project to extend (American) political and economic power. Although there are no colonial administrators forcing scientists around the world to write in English, the use of other languages is strongly discouraged by the apparatus of international academia.

But the most sinister impact of World English in science is on the languages that it displaces. When scientists write about their work only in English-language journals and talk to colleagues only at English-language conferences, the need for a comprehensive scientific vocabulary in other languages disappears. Words used to describe scientific and technical concepts undergo ‘domain atrophy’, like an underused limb. Domain atrophy is the reason Finnish does not have ‘promoters’ and Bulgarian does not have ‘quarks’ (both words exist, but are loaned from English).

Scientific English is undergoing the opposite process, rapidly expanding its vocabulary to encompass new concepts as they are devised and discovered. Such a large body of literature written in the language has amassed that research will continue on its trajectory towards complete linguistic homogeneity for the foreseeable future. Despite this, language, such an important marker of identity, is ignored entirely in discussions about diversity in science. We need more languages, not fewer, for everyone to feel included in one of humanity’s most exciting endeavours.