Senior readers will recall that in North America (USA/Canada) it was required to show a reading knowledge of another language besides English in order to complete the academic requirements for a PhD in basic sciences. It is no longer necessary now in physics and chemistry but other top notch universities like Harvard still have this requirement for a PhD in mathematics. For chemists studying electrochemistry, Russian was recommended, but for organic chemistry German was beneficial. Undoubtedly, German had the largest compendium of handbooks, factual information etc in pure sciences. Gmelin and Beilstein Handbuch were one of the largest databases in chemistry and they still are in the form of Reaxys (A database in Elsevier) since the mid 1700s.

I was writing an educational article encouraging students to explore foreign language publications in chemistry with the help of online tools. A reviewer asked if there is any knowledge difference between those who knew foreign languages and those who did not. This is a very subjective idea.

Do those that have to learn a foreign language for their PhD, especially in sciences, really benefit from being able to read foreign language paper/book/ database/patents?

(I am especially interested in German, French or Russian language literature for research in math, physics or chemistry.)

  • 2
    I am closing this question because it does not ask for general information, but is a survey. Thus there is no base to vote on answers as all answers are equally valid. For example, you can ask for is one of the following: 1) a list of arguments for and against learning a non-English language (where the answers can feature anecdotes, but cannot only be based on it), 2) surveys of how often scientists need non-English languages, 3) statistics of how often non-English papers are cited nowadays. — Please edit your question accordingly, trying not to invalidate the existing answer. – Wrzlprmft Sep 2 '19 at 19:59
  • 2
    The problem with your question is not a lack of specificity but that it’s a survey. – Wrzlprmft Sep 3 '19 at 7:14
  • Delete this question please, as it seems that "publishing" a question is far more difficult than publishing in a good journal. – M. Farooq Sep 3 '19 at 13:01
  • 1
    I edited this post to make it not a survey and am reopening – ff524 Sep 3 '19 at 15:07
  • 4
    Related: academia.stackexchange.com/q/118092/40589 – Dan Romik Sep 3 '19 at 15:45

Readers, who learned foreign languages for their PhD, do you remember any case where you benefitted from reading a foreign language paper/book/...

I'm a (pure) mathematician in the US, and I've literally benefited hundreds of times from being able to read mathematical French. Not only are there great papers in my area that are being published in French today, but a huge proportion of my field's seminal papers were written in French during the twentieth century. (I know that for people working in other areas of pure math, German plays a similar role.) In fact, I find it very hard to believe that someone would be able to have a successful research career in various areas of pure math without being able to read papers in a language other than English. Certainly such a person would be at a huge disadvantage.

  • 4
    In contrast, foreign language knowledge is entirely pointless in applied math. Nobody has published anything of importance in a foreign language in 20 years, and the fact that I speak German and French has not come in handy once in my 20-year career. – Wolfgang Bangerth Sep 2 '19 at 22:46
  • 4
    @WolfgangBangerth My experience is the opposite of yours. I regularly find important insights in papers that more than 200 (sic) years old and written in Latin. – JeffE Sep 3 '19 at 18:06
  • 1
    @WolfgangBangerth: Please, please make your comment into an answer!!! – user112604 Sep 3 '19 at 20:06
  • 1
    @user112604: I don't really have anything to add to that one sentence. My area is computational math. There was some work in French and German up to the 1980s, but few of it is (i) still relevant or at least (ii) has been re-packaged into English language textbooks. – Wolfgang Bangerth Sep 3 '19 at 20:31
  • 1
    @WolfgangBangerth: It is still very useful and should become an answer. – user112604 Sep 3 '19 at 20:51

Strongly seconding @Ben Linowitz' remarks: "even nowadays" many papers important to me appear in languages other than English, especially French, but also German. Decades ago, perhaps a majority of significant sources of interest to me were not in English, with few equivalents available in English, so sufficiently fluency in both French and German (and, at that time, if I'd known, I'd have studied Russian also). Fortunately, I had (by chance) studied French and German in high school and undergrad, so this was fine.

For that matter, prior to WWII the U.S. was certainly not dominant in mathematics, whether "pure" or "applied", so sources in that time would not have been in English. Whether or not currently papers are mostly in English, it would make me very uneasy to be unable to directly understand older sources. Indeed, when I want to see first-hand origins of contemporary things in pre-1800 sources (or early 1800's) I am happy to have also studied Latin in high school.

It would make me very uneasy to have all that stuff be incomprehensible to me. To ignore it (pretending that everything has been replaced by English-language stuff) seems to me parallel to accidentally acting as though anything not available on-line, or that hasn't been digitized, has ceased to exist. The world is just 20 years old?

So, even though the grad program in math at my university has reduced non-English language requirements, and my eventually drop them entirely, I do encourage my students to not ignore non-English sources. Part of the point is that "Google translate" is often helpful enough so that one can make sense of (small) non-English sources... rather than ignore them. Perhaps use of machine translation software as an "assist" is the best contemporary version of "sufficient fluency".

  • 3
    acting as though anything not available on-line, or that hasn't been digitized, has ceased to exist. The world is just 20 years old? — More accurately: The world did not exist between about 1940 and about 1995. Most older mathematical literature that can be digitized has been digitized and is available on the web, from places like the German National Library. – JeffE Sep 3 '19 at 18:08

Reading German, French or Russian is a nearly useless skill for physics research. Only very old or obscure research is not published in English.

Physicists do need to be able to work with colleagues with a range of language backgrounds. Usually that is not reading. Speaking Mandarin is an increasingly useful skill.

  • "Speaking Mandarin is an increasingly useful skill." True, either Harvard or Stanford graduate program allows Chinese as one of the foreign languages for math PhD besides French, German and Russian. – M. Farooq Sep 4 '19 at 1:32

This answer will be a bit different. Back when (back, way back) I needed to have a reading knowledge of any two of French, German, and Russian. I picked the first two. There were interesting current works in French (Bourbaki), older works in German (mostly translated). Russian was an outlier because of the cold war and the split between Eastern and Western scholarship at the time. Lots of things were being done independently in the (then) Soviet Union from what was going on in the US and the rest of Europe.

But I never needed any of what I learned. Bourbaki was interesting, but the knowledge was generally available in English. I didn't need historical works and the 20th century German stuff was mostly available. But recently, I wanted a translation of a mid century Russian paper and can't get it done. Machine translation for math is terrible - really terrible.

Today, just about every educated person in the world speaks and writes in English. I've done a lot of collaborations and conferences in Europe and the language is always English.

But, I'll still suggest the studying a "foreign" language is good for a math student since it trains the mind to think in a somewhat different way than studying math itself does. In fact, I think that it does researcher a lot of good to have a certain mental agility outside the strict confines of their field. A lot of mathematicians use music, actually, to provide something like that. But language study is also a "mental stretcher". I'll note that continental and Asian scholars get this skill when they learn English. It is good, likewise, just for the mental training, for native born English speakers to learn some other language.

At one time Latin was the language of scholarship. Today it is English.

  • You raised an important point which I am highlighting as well. What to do with machine translations when you have equations in the text. Of course chemical texts are not that full of equations, like a math paper, but machine translation messes up the sentence if an equation or an odd Greek symbol is embedded. Google translate is a disaster in such cases. – M. Farooq Sep 4 '19 at 1:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.