How do I make it possible for my supervisor to edit my thesis while using LaTeX? I submit a pdf document but they want to edit it.... how do I do this without switching to MS Word?

  • 32
    I'm not sure there is a good answer here, but the normal advice from those of us who do have mixed workflows is realism. If you have to provide Word output on a regular basis, particularly to your boss, then use Word and do a good job with it.
    – Joseph Wright
    Sep 4, 2012 at 8:41
  • 51
    A workaround: choose another supervisor... ;-) Sep 4, 2012 at 9:00
  • 90
    As a supervisor I would never edit a thesis: it's not my work, but the student's.
    – egreg
    Sep 4, 2012 at 9:08
  • 6
    @egreg We don't know in this case, but many people see PDFs as print-and-read only, and only know the Track Changes approach in Word for commenting at all in an electronic form.
    – Joseph Wright
    Sep 4, 2012 at 9:57
  • 33
    @JosephWright Imho egreg is right: I too do find the idea that a supervisor wants to edit a thesis horrific. Isn't the student claiming at the start that he/she wrote the thesis alone? A supervisor can get a print and make annotations. Or a pdf and make annotations. But he should not edit the document. A thesis is not a teamwork. Sep 4, 2012 at 10:41

10 Answers 10


I get pretty many people convinced to edit the PDF either with "PDFXchangeviewer", which you can download for free (but it isn't open source), or using "PDF Annotator", which is used widely in the academic area.

With both it is easy and can even be fun to annotate texts. And there is one large advantage to word & co: You can see at the first glance, whether something was added.

To reassure my readers that their respective annotations have been included, I either use the perlscript latexdiff to display the changes, or I use a version control system like git, and e.g. latexdiff-git (I'll provide the link later).

So there are two steps:

  1. Get your supervisor to annotate the PDF;
  2. Reassure him/her that the comments are not lost.

So if you are in the position to negotiate, argue to give it a try.

EDIT: According to the comment below, the supervisor seems to accept writing into a PDF. Regarding the second step, I found it very useful to have a version control system like git, which took some hours (no more) to learn the very low level it needs to control a book. The main advantage is that you can compare any version with the present version. About the necessary steps see here: https://tex.stackexchange.com/a/44092/4736

  • 1
    Good advice. I helped them install a pdf reader which allows easy annotation with a range of tools and they are now happy. I'm very lucky to have a supervisor who is willing to read at least part of my thesis... Let me check 'latexdiff'
    – HCAI
    Sep 5, 2012 at 22:32

You need to talk to your supervisor to find out what he is willing to do. Then you need to modify your work flow to achieve that. If he is willing to read/edit raw/uncompiled LaTeX in MS Word, then that is easy. Just import your LaTeX files into MS Word.

If he is okay with the occasional bit of LaTeX markup, but not too much, then maybe you need to do citations and cross referencing by hand, but leave things like \section. This means having a really good preamble is important and thinking about each piece of markup you use.

If he wants to see the finished product complete with formating in MS Word, then it is probably best to use MS Word.

I tend to find that if I use good citation keys and keep the markup to a minimum people are willing to ignore the little bit of LaTeX. If they cannot ignore the LaTeX, then I either switch to Word or ask them to markup a hard copy.


This will likely get downvoted, but I think the best answer is to use Word. Making it so that your supervisor may feel frustrated dealing with your thesis is a very bad idea. Be glad your supervisor is willing to edit it, rather than just making notes on a hard copy that you are going to have to put into the file yourself. In my experience that is unusual.

I think if you do something that frustrates your supervisor and is likely going to be a continuing issue you are asking for trouble.

  • 1
    Sadly I agree, while I'd rather use Latex for most things the fact is that while it's better in theory for collaboration the speedbump of learning how is too much for most people. Jul 12, 2013 at 19:02

If you are familiar with Emacs you might want to consider using Org mode. The built in exporters make it easy to produce both Latex and Open Office documents (which you can convert to Word) from the same plain text source file.

I've used this approach successfully (albeit for shorter documents than a thesis) when I needed to produce files for both printing and electronic submission.

For reference the export section of the org mode manual is http://orgmode.org/manual/Exporting.html and there was a discussion about how to write a thesis using Org mode on the Org mailing list here.


Always are bad news that your supervisor want edit your thesis. This probably mean that the manuscript need so extensive rewriting that he is unable to correct with handwritten comments, or at least he consider easier write partially your thesis that explain what must be changed. The supervisor annotations can be useful first for general indications and finally to correct typographical errors, but not to rewrite entire paragraphs or change the meaning of your arguments or the document structure.

Therefore the first measure to avoid send a Word documents is take care of the contents and the clarity with which they are exposed, and left to the supervisor to be worried only of typing mistakes (let some, so that he can feel useful).

If this is not enough, submit the .tex file to people using Word (or a similar WYSIWYG) it's just a bad idea. Probably even convert to plain text (detex foo.tex foo.txt or pdftotext foo.pdf foo.txt) is the wrong way. There are people that simply not know how to handle a txt file. Even if they are able to open with Word, they will be scared by the plain format. Imagine the face of this people if the contents also include \emph{some} LaTeX code.

Probably a better (better, not good) solution is convert the TeX file or to HTML, RTF or ODF formats with some of the available tools (pandoc, etc). This is a task relatively trivial for LyX users. Hopefully is this way you can export tables and images as well. Although these formats are supported by Word, it could better save as .doc with Word (or OpenOffice, Abiword, etc.) and enable Word version control before submitting (or ask your supervisor to mark the comments in some form). If there are problems exporting the document, send a plain text saved as Word text. Finally, when you receive the corrected manuscript, pass the changes to the original TeX file.

  • 3
    If you enable track changes in the word document before sending it to your supervisor you don't have to worry about if they understand it. If their technical level is low enough, also default the view to Final (vs Final: Show Markup) to avoid confusing them with Word's change tracking markup while still recording all their edits. Alternately Word 2010 has a decent multi-pane diff view for .doc files. I think it was added in W2007. Sep 4, 2012 at 15:23
  • @DanNeeely, thanks for the clarification. I really did not want to give any tips on Word. In Windows period I used reluctantly only when it was the undisputed fact standard (aka WordPerfect extinction-level event) but soon I switched to Linux where I've managed between antiword and StartOffice or successors for interact with Word users.
    – Fran
    Sep 4, 2012 at 20:04

They should be able to directly markup / comment on the file using Adobe Reader XI which no longer requires that a file be explicitly enabled for this feature. If they are still using an older version:

  • Make a .pdf as normal
  • Find someone who has a newer Professional copy of Adobe Acrobat (you could use the ``free'' 30-day trial)
  • Open the .pdf and enable Commenting and Measuring:

Adobe Acrobat Professional X

  • Inform them that they may view the .pdf and mark it up as desired using Adobe Reader 6 or later.

If that won't work, use pandoc or some other tool to convert the LaTeX to Word and send it (or find another supervisor).

One final alternative which in theory should work would be to use LyX to import the document, then convince your supervisor to install it and use it instead of Word --- it even includes a track changes feature which is presumably why Word is considered appropriate.

  • The Professional version of Acrobat is useless since Adobe Reader already offers comment features.
    – pluton
    Jul 12, 2013 at 17:53
  • 1
    It's not useless if one needs to enable the Commenting feature for the older versions of Adobe Reader which require it --- it's Adobe Reader XI which freely enables such markup w/o a .pdf being especially enabled.
    – WillAdams
    Jul 12, 2013 at 18:28

You can convert from latex to Word directly using pandoc.


I have a similar issue with a customer at the moment. They wanted project reports in MS-Word, so they can add comments and suggest modifications using the Track Changes feature and I'd much rather use LaTeX (for all the usual reasons, plus being able to edit in Emacs). At the end of the day I settled for a commercial solution that converts PDF to .doc and does a quite decent job of it. I then merge back whatever changes are needed into my TeX sources. It's a bit more laborious, but it keeps both them and myself happy.

I'd rather not give publicity to the commercial software in question here, but it's rather trivial to google it up.

Good luck!

  • Yes I have ended up doing something similar too... However my pdfs often won't convert to .doc due to some artefact which I have yet to identify...
    – HCAI
    Jul 16, 2013 at 9:31

Latex file source is plain ascii and it should not pose any difficulty for your Prof to read contents. As for styling, ask his opinion first and try to implement accordingly in Latex and impress him with an output better than his MS Word could do!

  • 13
    Sorry for being rude, but in the reality, it does not work this way. People either want to read raw code (and then they use LaTeX) or they don't (and then they want you to sumbit in Word). As well, I cannot imagine that I tell my boss: "Here is the TeX source, everything you need is in there."
    – yo'
    Sep 4, 2012 at 9:02
  • 5
    @tohecz I'd tell my supervisor: "This is my work; I'll gladly implement any suggestion you're going to make."
    – egreg
    Sep 4, 2012 at 9:11
  • 7
    @egreg The point here is that the supervisor has no idea how to make suggestions without using Word. I simply print the file and use a pencil to correct something, but many other people want to use Word for corrections. (The pencil method got out-of-fashion today, I guess).
    – Marco
    Sep 4, 2012 at 10:06
  • 5
    @Marco There is a difference between "making suggestions" and "correcting". Should the supervisor be allowed to rewrite a paragraph? To correct an equation? To correct the spelling? To add a citation? If the supervisor wants to write down suggestions he can use word. But he should use another document and not touch the original thesis. Sep 4, 2012 at 11:04
  • @UlrikeFischer IMO it really comes down to what is done with the edits/markup. A blanket accept of everything without looking at it is bad; but if the edits are in the form of "you misspelled X", or "you need a comma here"; making the correction directly to the original will take less time for the reviewer to do and not require any integration afterwards. "X is confusing, if what you mean is Y to clarify it I suggest changes A, B, and C" is often clearer if done as an in place edit. If it was confusing enough that the reviewer misunderstood you writing, a rewrite is probably needed anyway. Sep 4, 2012 at 17:16

Thanks everyone for the useful discussion above. I'm in the same position, with committee members who aren't familiar with LaTeX. My plan is to write the draft chapters for committee review in Word, then spend a few days at the very end shifting everything over to LaTeX for typesetting.

That will lose some of the advantages of writing in LaTeX, like being able to structure things with invisible comments, but it's probably the appropriate way for me to accommodate my committee members' experience with different sorts of software and get the feedback I will need to get this thesis to the defence stage.

You must log in to answer this question.

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