I'm not a native speaker. However, I'm doing my PhD in an English country and I read Academic papers which are in English. How can I improve my English from those papers? Any ideas? Although sometimes those papers are not written by native English speakers and may still have problems.
The old adage is: "the language of science is bad English" (and at least in computer science it is 100% true). That being said, I have known PhD students who have failed mainly because of their lacking English language skills – they had ideas, but they could not communicate them to anybody, and they could not write them down in a paper for the life of them. You need decent English to write any paper, and you need pretty good English to write papers at top venues.
You ask how to improve your language skills from reading papers. I would say, you can't do that effectively at all. Clearly, reading anything written in English will help you to some extent, but if you feel like you need to improve your English, taking one or more (good) courses in technical English will help you more than reading through a few dozen papers.
However, there is one language-related skill that reading papers for is really useful – if you are setting out to write the first or one of your first papers of your own, nothing beats analyzing existing papers (at the same conference, in the same journal) to find out how papers in these venues are usually structured. This gets you started much faster than drafting something only for your advisor to tear it apart.
While I understand that there's a big difference between scientific and day-to-day language, I actually think you need both to succeed in academia.
While you do need to excel in research (e.g. publish papers, communicate ideas), only the "top level" will be done in pure scientific and formal language: writing papers and scientific presentations. A lot comes from communicating your ideas, either in formal and informal settings. The "level" of English varies from setting to setting: at some point, you would not want to host a prominent researcher in your country, and take him to lunch only to realize you do not know how to translate the menu.
That said, I think you need to improve your English overall, and the best way to do it is to expose yourself to the language as much as you find comfortable:
- if you usually read (non-scientific literature), switch from your native language to English
- when watching movies/TV shows, don't go for the dubbed versions: go for the original language, first with the subtitles in your language, then in English, and then no subtitles at all
- try and find yourself in social situations where English is the main spoken language. Listen to people and talk to them. I saw a big change in a lot of people in just a few months from this.
- take a course, possibly in technical English
- specific terminology and style of writing will come in time, with immersion in your field, but it is okay to use a dictionary for rare terms
- in addition to just using a dictionary (translations are not always stellar), use Thesaurus to find synonyms and antonyms. It is also useful when you know the correct word, but you feel your writing is too repetitive
- have some kind of spell checker activated when you type (not only your papers, but also your posts, chats and e-mails), use it whenever it tells you that you misspelled the word, and try to remember your mistakes
- start using an English-English dictionary, as often as you can when you come across an unknown word. For me, just typing "define:whatever" in Google works just fine.
In addition to other good observations... : the idiosyncrasies of English are not easily codified in "rules", so considerable familiarity with many good examples of standard use, to allow faithful mimicry, (in addition to more formal study and hearing-experience) is surely helpful.
That is, rather than "composing" in a vacuum (worst of all translating into English...), it is very convenient to be able to recall already-vetted phrases and wording-choices and simply re-use them.
This is especially true with regard to "articles" "a", "an", "the", or their absence, and related seemingly-innocent modifiers. Also, verb tenses.
Thus: imitation from good models.
The exercise is to learn English by writing. Take a few paragraphs from Academic paper. Print the text and then write it by hand. Check if you have any errors? No errors? Then take another piece of paper and hide the side of printed text so that you can't see two letters at the end of each line. Write the whole text again. Check for errors? No errors? Move the sheet further and repeat. Have an error? Start from the beginning. Write the text looking at a printed one. Then add a sheet of paper on a side and write again and re-check for errors. No errors - move paper further. If you experience a mistake you need to start from very beginning! Eventually you should be able to write the text by looking at one letter of each line only and then the whole text by memory without any mistakes. Congrats! Do the same for second time next day. Once you complete second text, re-write first one too. Check for mistakes in both. Got a mistake? Start from text 1! Now you do that for 10 text and you will be shocked!
Another thing you need to do is to practice writing sentences. Take any phrase that has a new structure to you (e.g. from the academic paper) and write it down. Then write again. Repeat this 10 times. Then change one word using a thesaurus. And write 10 times again. Change the same word again and write re-phrased sentence 10 times. Repeat this task for each word. This practice will allow to learn the limited range of rules that are used in English sentences.
Surely you may check English grammar from time to time if you really want to know why something is like that, however don't spend too much time doing it. You will become bored and overloaded with information very soon. Happy learning!