If you have no choice, is it acceptable to translate a recommendation letter for yourself?

Thanks for asking this question.

In light of recent strong comments I have made on closely related issues elsewhere, let me say that I think that getting a letter translated from one language to another is an absolutely kosher academic practice. The classy thing to do would be to also include the original (e.g. how do you know that the readers will not speak that language?) and also indicate who the translator was.

However, the translator should not be the student. That is a problem because:

(i) It is an obvious conflict of interest.
(ii) Recommendation letters are often meant to be confidential, and this violates confidentiality.

If you absolutely cannot get anyone other than the student to translate the letter then you should clearly indicate "translated by the student" and expect to have your honesty applauded and the letter largely dismissed.

I must say that my heart opens up for a student who is living in a context where there is no qualified third party to translate a letter into English. I have been to academic departments in several non-anglophone countries and never encountered such a situation...but of course I have not been everywhere, nor to a random sampling, nor to any academic department in a "third world" country. That's a tough situation. Translating the letter yourself does not seem like the best answer.

Let me also say the following: if you are a non-anglophone student whose English skills are far superior to those of the faculty at your university [and assuming that you are applying to anglophone graduate programs, of course!], then you might try to cultivate relationships with anglophone professors elsewhere in the world. Twenty years ago that would have been preposterous advice, but due to the proliferation of mathematical interaction via the internet, it seems very viable today. For instance there is a small but positive number of students with whom I have had sufficiently substantial interaction on MathOverflow and (more often) math.SE so that I would be glad to write them a strong recommendation letter. If you are a math student, you can always try writing to any professor and having mathematical interactions with them. They are not obligated to respond (I certainly do not always respond...), but they often do (I often do...) especially if you show them something truly promising.

[At some point in the previous paragraph I forgot that I was supposed to be writing for a general academic audience rather than an academic mathematical audience. But since I am not sure how far my advice extends outside of mathematics, I will leave the m-word in.]

Among US students applying to US graduate programs, it is increasingly frequent for at least one of three recommendation letters to come from the director of a summer undergraduate research experience (REU) than a faculty member at the university. Such letters are not necessarily the most penetrating -- they read very similarly, perhaps because of the implicit motivation to paint one's summer research experience in a positive light -- but they often get the job done, i.e., they lead to admissions.

Let me also say that a letter of recommendation for graduate admission is not always the most important part of the application. If I get an application from a university that I have not heard of, and letters from faculty that I have not heard of and whose reputations I do not know, I can only take the letters so seriously no matter what they say. (And it is quite true that not everyone knows how to write a good "American-style recommendation letter". This does not necessarily get counted against the student; it just doesn't get counted for them.) If you are coming from an "obscure program" then your goal is to convince the readers of the applications that your training is equal to (or superior than!) the training that students in more familiar programs get. So it can be helpful to include very specific information about coursework: e.g. not just the title of the course and the course grade but the textbooks used. If you wrote a paper which does not make any research contribution but shows a solid understanding of graduate-level material, by all means include that as part of the application. Also be sure to take all the applicable standardized tests and do your best on them (and don't cheat on them!!).

I'd like to suggest instead of asking if you could self write the letter, send your advisor an email that says something like this. "Thank you so much for agreeing to write this letter for me. We have done a lot of great work together over a long period of time and I'd like to highlight some things that you may wish to talk about in your letter (of course feel free to choose not to use any of these examples if you so desire):

• Example 1
• Example 2
• ...

This will allow your letter writer to at least have some phrases he/she could say in relatively good English, but you aren't actually writing the letter. The advisor will likely edit them and add more phrases, but at least it is a good starting point. This way, no confidentiality is breached because the advisor can still ignore all of your examples and you have no idea whether he/she chose to do so or not. I also like Pete Clark's idea of sending in an untranslated letter if it is in a language that can be easily translated in most English speaking institutions.

No.

• It is a breach of confidentiality.
• If the recipient of the letter is not informed that the student translated it, then that is (at the least) unethical.
• If the recipient of the letter is informed that the student translated it, then I see no immediate breach of ethics. But it's no longer a confidential opinion, may have been subtly altered by the student, and likely won't carry much weight as a result.
• In either case, how can the recipient be sure that the student's translation is true and unbiased?

The proper way forward, as I see it, would be to get the professor to write the letter himself and then get the letter professionally translated. Enclose copies of both the original letter and the translation, declaring that the letter has been translated into English (by either a professional translator or another professor).