Speaking in Tongues
Science’s centuries-long hunt for a common language.
Michael D. Gordin. Scientific Babel: How Science Was Done Before and After Global English. University of Chicago Press, 2015. 424 pp. $30.
Latin thrived in Europe as a common language of natural science after the late medieval period. Only when the power of the Roman Catholic Church diminished during the Protestant Reformation did European savants gradually and consciously relinquish Latin as the language of science.
Elites everywhere began to favor vernacular languages, beginning with Martin Luther’s 1534 German translation of the Old and New Testament. In 1751 French savant Jean d’Alembert captured the trepidation over a looming Babel in science in his introduction to l’Encyclopédie. D’Alembert worried that a natural philosopher now would need to know seven or eight different languages and so waste a lifetime before ever beginning to learn the natural knowledge expressed in all the writings.
For d’Alembert the answer was straightforward: French should be made the universal scientific language. In practice early modern scholars were adept in several languages. As a rough rule, according to Michael Gordin’s Scientific Babel: How Science Was Done Before and After Global English, libraries in the Enlightenment period collected one-third in French, one-third in Latin, and the rest in the vernacular of the region. The Imperial Academy of Sciences in St. Petersburg had a substantial number of non-Russian members; in the late 18th century, Latin was the official language of the academy’s publications, German was the language of conversation, and French was the language of the academy’s minutes. To the west the chemist Torbern Bergman corresponded in his native Swedish, as well as English, German, Latin, and French.
Scientific Babel is a remarkably learned, absorbing, and witty book for all of us who think about how we use our own language, communicate with others in a foreign language, or read translations. Gordin’s book is of special interest, too, for readers of science and especially for those who know something of the chemical sciences. The title gives no clue that chemistry plays a main role in this history, but there is good reason to highlight the discipline. First, among the natural sciences chemistry is particularly notable for its dependence on the language of descriptive prose and systems of classification, as well as on symbols, images, and mathematical formulas. Second, during the course of the 19th and 20th centuries chemical science gradually became the largest of the sciences, even as it formed a bridge between the physical and biological sciences.
Gordin focuses on languages that appeared in scientific publications and in correspondence among scientists from the 18th through the 20th century, mostly in Europe and Russia. He distinguishes between language identity—the strongest language of the writer or speaker—and vehicular language—the one most suitable for communication.
By the mid-19th century, based on data from journals of science abstracts, what Gordin calls a “triumvirate” of languages had established itself in global scientific publications. From roughly 1880 to 1910 there was an almost equal partition of 30% each for English-, French-, and German-language articles and books. A graph showing global publication makes visible the increasing strength of German, until its peak in the early 1920s, and then of Russian, until its peak around 1970. English began its ascent in the 1880s and reached global domination by the end of the 20th century.
The Russian peak occurs during the Cold War, but Gordin also tells about an earlier important historical event in Russian science and the recognition of Russian as a major scientific language. Dmitri Mendeleev and Lothar Meyer both fought to be acknowledged as the first to publish on the law of periodicity of the properties of elements as functions of their atomic weights and what became known as the periodic table. Mendeleev’s claim for priority rested in his Russian-language publication that appeared slightly earlier than its translated version in a German chemistry journal. In the translation Mendeleev’s Russian word for “periodicity” was mistranslated in German as “phased” or “stepwise.”
For Meyer the publication in Russian had no status, and he complained that it was unreasonable to require German chemists to read reports in Slavic languages in addition to Germanic and Romance languages. Russian scientific achievements, however, particularly in chemistry, soon brought attention to the science being done in that country, but mastering the language was no easy matter.
The Mendeleev-Meyer story involves elements of grammar and vocabulary, with implications for scientific communication. Here and elsewhere in the book Gordin explains features of grammar, language structure, and vocabulary that cause learners difficulty in languages ranging from Latin, Russian, English, French, and German to constructed universal languages. With science literature expanding and patriotic nationalisms growing stronger toward the end of the 19th century, proposals proliferated for a new “neutral” universal language that would be easier to use than Latin. Among these were Volapük, Esperanto, Idiom Neutral, and Ido.
Esperanto and its revisionist rival, Ido, were among the favorite constructed languages for some scientists, including chemists Marcellin Berthelot and William Ramsay and mathematician Henri Poincaré. The original Esperanto had only 16 grammatical rules and a vocabulary of 900 roots drawn from English, French, German, Italian, Russian, and Spanish. The ending of a word indicates its grammatical role in speech: for example, o for a noun, a for an adjective. Ido, as a simplification of Esperanto, insists that a concept corresponds only to a single word so that there are no multiple meanings for any single word.
In the early 1900s the physical chemist Wilhelm Ostwald became a chairman of the Delegation for the Adoption of an International Auxiliary Language under the umbrella of the International Association of Academies, an organization for national scientific academies that numbered 22 members by the outbreak of World War I. By 1911 Ostwald favored Ido over Esperanto for use in science and later published a table of the chemical elements with new Ido names. He tried unsuccessfully to persuade the Dutch chemist J. H. van’t Hoff that the world’s premier journal of physical chemistry, Zeitschrift für physikalische Chemie, should publish abstracts in Ido. Following the outbreak of World War I, Ostwald abandoned Ido and encouraged instead the spread of German. That was not to be.
Gordin’s history recounts foreign reactions not only against Germans and Germany but also against the German language. This reaction was a consequence of two world wars, interrupted by a brief restoration of relations from the mid-1920s until the early 1930s, when Adolf Hitler came to power.
In the United States, German was spoken and taught in many schools before World War I. But by 1919, 22 states had banned the speaking of German. That same year the U.S. Supreme Court ruled the ban unconstitutional. Later, eminent German exiles from Hitler’s Reich published mostly in English after arriving in other countries, making German journals less relevant. Nazification of German journals also decreased their relevance as Nazi political ideology dictated the inclusion or exclusion of scientific authors and their research. In neutral Sweden, a country previously known to be Germanophile, English replaced German after World War II as the first foreign language taught to children. The division of the German nation into eastern and western sectors also worked against the use of German as an international language; English became a prime language in the West, as did Russian in the East.
In Russia science and technology took off after World War II, just as in the United States. In 1948 more than a third of all non-English technical data appeared in Russian. By 1958, among 50 language sources used to collate data for the periodical Chemical Abstracts, Russian abstracts numbered 17% of the total (with English at 50.5%), which was greater than German (10%) and French (6%) combined. In 1970, 23% of the chemical literature was published in Russian. The last chapters of Gordin’s book examine U.S. and Soviet attempts to cope with translating and reading this huge body of scientific literature in the context of Cold War rivalry. There were two approaches: machine translation and direct human translation.
On the machine front linguist Léon Dostert at Georgetown University directed the first attempt, in which the IBM 701 electronic computer was used in 1954 to translate Russian into English. The project was eventually funded by the CIA, along with the National Science Foundation and U.S. military. In the Soviet Union, Aleksei Liapunov, a leading figure in cybernetics, discovered a scientific abstract about Dostert’s project and launched a competing English-to-Russian project, which by 1964 made the Soviet Union the leader in the field of machine translation.
Machine translation aided an already existing Soviet program at the Soviet Academy of Sciences to collate, translate, and publish scientific and technical information from around the world. By 1966 this institute was reputed to be the largest scientific information service in the world, with a permanent staff of 2,500 assisted by 22,000 language and disciplinary specialists who annually produced more than 700,000 abstracts translated into Russian.
Machine translation was not the only way to go, either in the Soviet Union or the United States. Soon after World War II the American Earl Maxwell Coleman saw a market for scientific translation. He employed freelance translators, some of them physics graduate students who had studied Russian. In 1970 Coleman’s company became Plenum Publishing, which produced 72 Russian scientific journals in English translation. If one counts American cover-to-cover translations of Soviet journals into English, then 81% of physics literature appeared in English in 1969. Between 1980 and 1996 English jumped from 74.6% of all entries in Chemical Abstracts to 90.7%.
What is astonishing in Gordin’s account of the taming of scientific Babel is the emergence of English as a global scientific language. This turn of events was partly, or perhaps even largely, because, as Gordin says, a scientist in Pakistan or Italy or Brazil had to follow both American and Soviet science after World War II and eventually could follow both with English as the default language. The electronic revolution and digitization also favored English in the earliest program languages and computerized databases because of their development in mostly English-language countries. In 1981 the prestigious and indispensable print publication Beilsteins Handbuch der organischen Chemie switched to English, becoming an English-language electronic search engine in 1994. Since 1881 the Handbuch had provided the names, structures, properties, and reactions of all known organic compounds, numbering some 9 million by the early 2000s.
Gordin’s history of the movement toward a single language for scientific literature raises challenging questions. Did it have to be this way? And how long will the dominance of English last? Has the movement to English been partly a matter of numbers, of native English speakers spread throughout the globe from Canada to Australia to India? This is not likely since, according to Gordin, Spanish has never figured among the statistically significant languages in the production of scientific knowledge, unlike Chinese and Arabic.
Gordin’s account implies that we might want to investigate, too, the connections between vehicular language and the history of regional and global trade. Some languages of trade, like the original lingua franca used in medieval Mediterranean ports, are mainly spoken, with lingua franca itself a mixture of Italian, French, Spanish, Greek, and Arabic. In contrast the Greek-language variant Koine was a language of commerce and learning in Hellenistic and Roman antiquity from the 4th century BCE through the early Middle Ages.
The story of Latin’s decline, the ascendance of vernacular languages in science, and the recent dominance of English together give rise to the philosophical question of whether monolinguism is “good” for science. Many mathematicians still pay a great deal of attention to French-language mathematical journals. Is mathematics more creative in French? The American linguist Benjamin Lee Whorf suggested in 1940 that observers and experimentalists respond to physical evidence in similar ways only if their linguistic backgrounds are similar or can be calibrated. Is what or how we think, as well as how we communicate, dependent on the language we use? Gordin is agnostic in this debate, but the historian and philosopher of science Thomas Kuhn acknowledged Whorf’s influence in his own development of the notion of “paradigms”—scientific worldviews that are incommensurable with one another because of the way each differently structures how we see the world.
Or is scientific knowledge not language dependent? Might science itself be a language? Gordin ends this history of scientific language with the example of H. Beam Piper, a science-fiction writer, and Carl Sagan and Iosif Shklovskii, astrophysicists in the Search for Extra-Terrestrial Intelligence project (SETI). Both the fiction writer and the scientists suggested that an image of the periodic properties of the naturally occurring elements might constitute a metalanguage accessible to all intelligent life in the universe. Gordin regards such a proposition as unlikely, but this last example using the periodic table nicely ties together his engrossing history of language in science.