I am about to start planning my bachelor's thesis in computational linguistics, and I'm looking into how I can make the process as smooth as possible. So far, I've found that using some sort of version control system for keeping track of one's changes seems like a good idea. I've also started collecting papers using Zotero for easily exporting to BibLaTeX and keeping notes.

But those are only tools. What I'm most uncertain about is the actual writing process, especially since my native language isn't English. Is there a collection of common mistakes I could benefit from reading, or other every day tips and tricks? Anything, really!

  • Version control systems are good. I recommend Mercurial. LaTeX is also good. To improve your writing, the best thing is to practice. Aside for your research writing, you could do worse than post questions and answers to online forums. The Stack Exchange sites for English and Writers are quite helpful. Dec 21, 2013 at 19:17
  • 1
    @FaheemMitha any particular reason for recommending Mercurial instead of git, svn et al.? Just curious.
    – user7112
    Dec 22, 2013 at 8:06
  • @dgraziotin: Well, this is a bit OT, but distributed version control is technically superior to centralized in all ways that matter. I use Mercurial and like it. I think (as do others) that it is has a better user interface and is more "friendly" than git, though of course git is preferable to svn. These topics are discussed to death on the net, so you could take a look at comparisons. Dec 22, 2013 at 8:19
  • @FaheemMitha I am using all of them regularly (software engineer). That is why I was curious!Let's end the OT here, or a flamewar might start :-) Thanks!
    – user7112
    Dec 22, 2013 at 8:28

2 Answers 2


This is a pretty broad question, but if you are looking for a general guide on how to write there have been a number of books suggested to me.

Writing for Computer Science (Zobel) This is a decent book on explaining the principles on why and how you should write computer science articles. It's very computer science specific, with a bit more emphasis toward databases and operating systems, but I think it's applicable to most technical fields (of which computational linguistics should fall into).

I like this one because it's a general guide to writing that outlines the process and the structure.

Elements of Style (Strunk and White) A commonly-recommended book that a lot of people swear by but I found it a bit too low-level for my liking and actually had a hard time finishing it. A professor described it as "too prescriptive".

Bugs in Writing (Dupre) This is an okay book as a reference manual for improving your prose, but isn't a good guide for figuring out the writing process. There's good advice in this book (and it's actually rather fun to flip through just to learn about various style points) but can be difficult to use as an actual guide.

Style: Toward Clarity and Grace (Williams) I've gotten through a few parts of this book and it gives very specific advice on how to proceed on improving your writing, especially explaining specifically how to introduce ideas in paragraphs, how to flow them together to create longer manuscripts, and how to restructure and correct text so the main ideas come through. As I surmised, it is in fact very analytical and detailed and does describe the rationales behind its recommendations. I think this book is a great resource for those who have some words written, who know they have to make their writing "flow better" or "more clear" but aren't really sure how to do it. It appears to be an especially valuable book for the "second pass" through anything, where you have written down your initial ideas in part-writing, part bullets, and are trying to convert it into a real readable piece of work. (Previously, I wrote: This book was JUST recommended to me and I haven't read it, but it is apparently very analytical and detailed, and outlines the rationales behind why one should write in the way it discusses. In fact, I think there a jab at the Elements of Style in this book where the author laments that "Be clear" is GREAT advice. We all KNOW that. The real problem is, "How is it that I can actually implement 'be clear' in my writing?" However, I haven't gotten far enough to comment on it yet.)

So there's a little about what to read. As for what you should know? Well, that's a more difficult question :)

  • Thank you! I have been reading through half of Writing for Computer Science and it certainly seems like a great book to keep close for reference in the writing process.
    – user10157
    Dec 23, 2013 at 21:44

Best thing you can do is have a well defined topic. Since you don't mention that in your question, I don't know if you have a research topic/question yet. In my experience, that's by far the most crucial thing in having a thesis go smoothly. The student who know what they want to work on usually have everything fall into place.

You must log in to answer this question.