When writing academic papers, I am really bad at improving what I have already written. I have heard that most of the time writing should be allotted to revisions. I know a few academics who are really good at keeping on revising until they are happy, but I simply can't do it. Knowing that a sentence/paragraph/section can be improved but not being able to do so is very frustrating.

My partial self-diagnosis:

  1. I refuse to make big changes, probably since it is a lot of work. (This sounds like I'm just procrastinating.)
  2. If write with collaborators (almost always), I do not want to change what they wrote or revised, unless it is obviously wrong. (This sounds like I lack confidence in my writing skills. Or I just don't want to upset my coauthors?)
  3. Before rewriting, I can't even re-read properly. I don't want to re-read the paper carefully and create a current copy of it in my head. I tend to skip parts. Even after I have re-read the manuscript, it is not always clear what the current state of the paper is.

I am sure I have many weaknesses that I am failing to verbalize in this question, but I'd like to hear what others did to train their rewriting skills. Also, I want to hear how you rewrite.

FYI, I am not a native speaker of English but I have seriously written only in English. My field is science and engineering.

  • 4
    +1 me too, and I am a native speaker. What has worked for me: Having someone to brutally revise early draft can help. On my own: reverse engineering an outline, physically cutting up a paper draft into pieces, and aligning pieces with the outline to reorganize while keeping many chunks of text intact, then rewriting to smooth transitions between the chunks.
    – Abe
    Feb 8, 2013 at 2:32

10 Answers 10


I recently posted a lengthy answer for a similar question a few days ago, the essence of which was to separate what you're going to say from how you're going to say it.

If you've done this for a paper, you can edit it focussing on writing style alone. This is a good way of avoiding the "big changes" you mentioned in your question: You will have made all these before actually formulating the text. As a consequence, you should also know precisely what it is you are trying to say in each paragraph.

Iteratively refining a text can get you stuck in dead ends, e.g. if you choose a certain formulation and then can't make it sound right. One thing I often do when I get suck with a paragraph or chunk of text that I don't know how to fix, is to just delete it and rewrite it.

If you get stuck on the specific formulations themselves, i.e. you don't know how to re-write a certain paragraph, you could try explaining it (remember that you know what you want to say, but not how to say it) out loud to an imaginary listener.

Reading a paragraph out loud is a good way of forcing yourself to re-read it. It's also a great way of checking if something sounds silly or is not really comprehensible.


If you're having trouble reading to yourself, you may want to try pairing-up with a colleague or co-author, and reading parts of the paper to each other. Granted, this may feel a bit awkward, but just look at it as editing the paper together. Working in pairs is known to improve motivation and productivity, and will basically force you to concentrate on the task at hand.

If you have problems concentrating in general, I can give you a few tips from my own experience:

  • Break down your editing into short bursts of at most an hour, and focus only on a part of the paper, e.g. the abstract, a figure, or any specific section.

  • The first few hours in the morning are the most productive for me. Try to find out where your own "best time" is.

  • Coffee. In certain cases, the caffeine can help you focus.

  • 8
    +1 for "delete and rewrite" and +1 for "read it out loud".
    – JeffE
    Feb 2, 2013 at 22:33
  • This is essentially what I am doing now. I spend a lot of time trying to iteratively improve, get frustrated, and then just delete and rewrite. Also, reading out loud most of the time, but somehow I can't focus on my own reading. If someone else reads it to me, I would find flaws in the flow much more easily.
    – Memming
    Feb 5, 2013 at 0:17
  • @Memming: Do you have problems concentrating in general? If so, most of the advice here will be moot, as the problem would not be specific to writing. In any case, I've added a few more suggestions above.
    – Pedro
    Feb 5, 2013 at 9:38
  • Good advices. Yes, working with a peer does improve my productivity. (Coffee also does help me focus, but I can't drink it for health reasons.) Thanks for the edits.
    – Memming
    Feb 5, 2013 at 12:46
  • One thing I found very helpful is to have the computer read the paper back to me. I am using a Mac and the text-to-speech functionality produces pretty good results (I prefer the "Alex" voice myself), except for mathematical formulae. This is probably not as good as reading to a colleague, but is much more convenient.
    – CC cat
    Jun 13, 2014 at 11:40

I think the more efficient way would be to start reading and focusing on how to write in the first place and then focus on editing your work. This approach is going to save you a lot of time.

Perhaps the fastest way is to get some professional help which often is free in academic institutions in the form of academic writing courses. If you get the opportunity through these classes to show your writing to a linguist you can gain a lot.

If that's not an option then some classics on writing are:

  • White, E. B., & Strunk, W. (1972). The elements of style. MacMillan.
  • Zinsser, W. (2006). On writing well: The classic guide to writing nonfiction. Harper Perennial.

Perhaps then you should start focusing on editing and I recommend these for the start:

  • Cook, C. K. (1986). Line by line: How to edit your own writing. Houghton Mifflin Harcourt.
  • Ross-Larson, B. (1995). Edit yourself: A manual for everyone who works with words. WW Norton & Company

BUT personally the most important guide for me was the edits/comments that I got back from my supervisors, mentors and senior collaborators during the years. I checked their edits over and over again to systematically diagnose what was wrong with my writing and I think those edits/comments were the most helpful resource. I went as far as creating a corpora of literature relevant to my fields of research to know how exactly people write in my domains of interest in engineering and social science but well that's going a bit too far in the beginning.

Also one thing that I have noticed which makes a major impact on my editing is switching the edits on screen and on paper. I usually first do a round on screen. Then print and do it offline and then switch again! I don't know about others but in my case I tend to focus on completely different issues on when checking the scree or printed material and if I only do one I will miss a lot more.

2014 EDIT: One additional method for editing your work is listening to it. Text-to-speech software are really helpful here and you will pick up issues that you might normally neglect. There is something magical about listening to your writing which is totally different from reading it. I definitely recommend trying this as well. Obviously higher quality text-to-speech software that have more and better natural voices will enhance the experience...

  • +1 for Zinsser. His is still the book I reach for when I'm stuck on a writing project. Oct 25, 2013 at 21:00

My simple strategy is as follows:

  1. Draft your paper. Make it as complete as you can.
  2. Give yourself atleast a12 hour period in which you dont look at the paper, no matter how tempting it is.
  3. Use a read-back program that can read the paper back to you loudly. Mac has a free built-in program.
  4. Listen and note which sections need reworking. If you don't like what you hear, its likely to be poorly written.
  5. Revise as if you are writing the paper for someone with little knowledge of your field.

The best advice I have got while writing my PhD dissertation was to focus on the arguments in the drafting stage and on the details in the revision stage.

  • 3
    12h is not enough. 7days are more like it.. (;
    – Ran G.
    Feb 3, 2013 at 5:12

I think you're question is a VERY good one and one that impacts a lot of people. Because of that I upvoted it. However, the answer, which I think you at least have an idea of, is "suck it up and get to work."

Writing well takes a lot of time. Writing well includes planning, writing, reviewing, revising, reviewing, revising (ad nauseam - and I do mean nausea). If you look at great writing, you'll see it's often written by great writers. You should not think that they get it right their first time. When I (I consider myself an OK writer, not a great one) write for publication I usually write the piece and then end up editing it 10 times. In the end I spend much, much more time in the editing processing than in the initial writing process.

Writing takes time. Writing well takes more time. If you want to write well, you need to be willing to push through the discomfort and keep working on it. That said, don't try to do it all at once. Edit several times spread across several days (or weeks, if you have the time) - a fresh mind helps.


I would recommend using some versioning control system (like SVN or git) for your paper. These tools gives you the freedom to change what ever you want with you paper, and have all the history recorded. You could practice any kind of change, and still keep the ability to revert to older versions. Even better, you could merge good elements from old version into newer versions. By reviewing your history and seeing what you changed, you can learn what types of mistakes you tend to make, and you can work to avoid them in the future.

Note that some popular note-taking tools, such as Evernote or Simplenote, also keep track of previous versions of your notes, although it's more primitive than Git or SVN.

DropBox provide a (terribly) simplified notion of versioning control. The advantage, however, is that it works "out-of-the-box" - no learning curve, or fancy tools.

  • Good advice about version. Have seen it used in so many places. Feb 7, 2013 at 20:18
  • I have been using svn (and trying to move to git). But, I haven't been studying the changes carefully. BTW, do you have a favorite merge tool? I use vimdiff, but it's not too convenient for comparing/merging multiple versions at once.
    – Memming
    Feb 8, 2013 at 14:00
  • 1
    I use magit, but didn't do enough merges to be able to recommend it. However, in general, it is wonderful and makes the usage of git so much nicer!
    – Dror
    Feb 8, 2013 at 15:32

One great way I have found to improve my writing, is to reread it side by side with someone not directly in the project (significant other, friend, ...).

So this person would read it while you are here and will directly ask you questions on what you meant or tell you this and that is not clear. Also you can propose better phrasing together, wondering if such formulation is clearer.

Supervisor comments are often great, but they most probably won't have time for such detailed discussion over your text. Having a "live" discussion is really so much more insightful on how your text is received. If you do it with someone you have a good relationship with, it can even be fun and motivating!

Of course, it is a great burden to put on someone else, it can take hours, so you need to be ready to return the favour or find another way to make up for it.

  1. Focus first of all on the content: your message -- the points you want to get across. Do not focus on the language -- the way you communicate the message.

    This is the most important guideline, IMO. Clear ideas will lead you to clear organization and clear language. Unclear ideas will not lead you anywhere useful. Do not start by worrying about the language.

  2. Especially if English is not your first language, write short, simple sentences. Very short. Very simple. Later, if appropriate, you can always combine them.

  3. The secret to writing is...reading. To start with, reread and rewrite your notes about the message, before trying to write the text that conveys the message (see #1).

    After you've written your message, read, reread, and rereread what you've written. Each time you read it, you will naturally improve (rewrite) it. When you read it, try to erase any knowledge of it beforehand - read it like your intended audience would read it.

  4. Repeat #3. Repeat it again. Reread to write better. (It will also help you read better.)

All of the above apply to editing, as well as to writing. If you are editing the work of someone else, then you are interested, first, in understanding that writer's message. If the message is not clear then forget about improving the wording and provide the feedback that you do not understand what the message is (and perhaps help by pointing to language that confuses you).

If the message is clear to you, then move on to how it is conveyed. If you read carefully it will be clear to you what is not as clear as it should be, what might be missing, and what might be extra (unnecessary). You will naturally discover problems of order of presentation. Just read and reread, carefully, and you will be fine.


When you need to rework bits of, or the whole, manuscript it might be a good idea to alienate yourself from the manuscript. One thing I noticed is that it's difficult to look at a piece of text I have worked on for weeks, with fresh eyes. Hard to be creative in formulating things when you are stuck in a particular thought pattern.

It does wonders to put that aside and go deal with something else entirely. If you can forget the existence of the manuscript, even better. When you take a look at it again some days after, you might be able to see awkward formulations, hanging sentences or unclear paragraphs. During an academic writing class I took a couple of years ago they had this fantastic quote, "the author is dead!"

While the full context of the quote is somewhat deeper and besides the point here, it should suffice to say that seen from the eyes of the reader, the author is long dead and gone... You, as the author, need to be aware of that and at least try to objectify yourself from the text in order to be able to see it as a reader might, and see the potential weaknesses.

On a different note, another thing that might be a good exercise in revising or reformulating scientific text, is to attempt to rewrite some existing text. For instance take a published article (could be yours or someone else's) and rewrite/summarise/revise it. Then submit both texts to a plagiarising-checking tool (there are some free ones online), to see if you can minimise the similarities in between the texts while keeping the message as intact as possible.


I also used to resist making big changes. At some point, I realized that this resistance was coming from a fear of having to delete sentences that my brain has already created and now feels emotionally attached to them as my precious pieces of writing. (Even if objectively they aren't such good sentences anyway!) Subconsciously, I would think: if I delete them now, who knows if I ever come up with the same idea on how to put something in words? The strategy that helped me is really a psychological strategy to address that fear:

Don't delete anything permanently. Cut a sentence that does not fit in your current text and paste it into your "sentences bank". I have a Notion page for this purpose called "My writing snippets", which collects all those phrases, sentences and even whole paragraphs that didn't make it into any manuscript yet. But it can even be a .txt file on your desktop. This way you don't dread permanently deleting your prose. And, with your next writing project, or when you're stuck in your writing, you can look over your sentences bank to see if there is anything that you can utilize. I've observed that even if I never use a sentence again, saving it calms my brain that it's still out there, just in case I need it.

This helped me move on with applying the most impactful changes, e.g. ones that improve the flow of the story, and require moving or removing many sentences or whole paragraphs.

In addition, I do recommend removing stuff from the sentences bank once it has been applied in a manuscript to avoid self-plagiarism.


I have a lot of problems with my writing (although I am a native English speaker), and was made to do a writing course by my PhD panel.

I did this course, Writing in the Sciences, on coursera, and I can thoroughly recommend it. It took a few hours of my time over around 8 weeks, but you could easily pick and choose which bits you wanted to do if you wish. There are assignments to complete, which can lead to a certificate if that is something that interests you, but you don't have to do them. I actually did all the assignments as it helped me practice writing and editing my own work. I also got to practice editing other participant's work. Although the instructor is from the medical sciences field (as are most of her examples), everything she teaches is applicable to other sciences. I am from the atmospheric sciences field myself.

I learnt some really good tips on how to go about the writing process itself (i.e. how much time to spend on each step in the process), but the majority of the course is about how to edit writing (either your own or someone else's) to make it more exciting and interesting to read. Many of the tips given already are included in the course.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .