My dissertation is going to be a thesis by publication. I am going to publish a couple of papers and put those into one document with a couple of leading words (another 20 pages or so) and this is going to be my thesis. I feel that the monograph I wrote for my master thesis (which was pretty extensive) was actually pretty helpful as a reference now (I am continuing the research on the same topic). Thus I would like to know: Should I write a monograph for myself? Obviously I am not going to publish that, so I won't have all the work with writing it nicely (or at least more nicely than I personally need). My mentor (who is not my doctor father) told me this is too much work and I should rather use that time to publish more. What do you think? I am researching of A.I. and Machine Learning, thus I will have a lot of math and a lot of programming. The monograph I would write would be about the math, not the programming (that should document itself).

The comments seem to misunderstand a couple of things:

  1. My mentor is not my doctor father. My mentor has not say in anything and no involvement - he is just there to give me tips.
  2. There is no question what my thesis will be. It will be a thesis by publication. The monograph would just be for myself.
  3. I cannot "publish a paper based on the thesis": My thesis is going to consist out of multiple (about 6) publications.
  • 1
    Didn’t you write things down nicely in your paper?
    – Wrzlprmft
    Commented Nov 15, 2016 at 13:23
  • Really, this is your call, especially, if your mentor essentially supports a quick solution in the way of dissertation by publication (some don't, they want a book). The monograph is, of course, a learning experience, which may be just the thing you want. But nobody can really answer this question. Writing a monograph is a different animal than a paper; its length imposes totally different way of organising the writing than a paper. Commented Nov 15, 2016 at 13:27
  • 4
    "this is too much work and I should rather use that time to publish more": I agree. Rather, try to see if you can publish a paper based on the thesis. Commented Nov 15, 2016 at 13:27
  • @Wrzlprmft: No. Paper are written in such a way that they should be understood by themselves. Thus I would simply everything that is not needed for this particular paper. Thus I leave out generalizing stuff. E.g. parameter in formulas that are zero parameter for this particular paper will be left out. E.g. the formula can be weighted with a parameter, but in this paper the weight is always 1, so in this paper we leave the parameter out.
    – Make42
    Commented Nov 15, 2016 at 13:27
  • Well, I'm writing a thesis by publication too, and it has basically two parts before all the papers, the first beeing an introduction to the topic which is about 30 pages long in my work. Wouldn't it be possible to fit such information there.
    – user64845
    Commented Nov 15, 2016 at 14:53

1 Answer 1


This question was asked a long time ago, but remains on the unanswered queue, so I'll try to answer.

In math and computer science, generally articles matter more than books, for junior researchers trying to establish themselves. So, the usual advice is to focus time on writing short, high-quality articles that can be published quickly, rather than monographs, until after tenure. Doing a "thesis by publication" is ideal. I essentially did that and was able to put two strong papers on arXiv before I even defended my thesis. Peers of mine who wrote the monograph first, and later tried to extract the articles lost a year on that task, after getting their PhD.

Many authors write a book after a series of papers, to tell a complete story and write things down in a cleaner way, which became clear to them as they published those individual papers.

That said, it's good to check all details carefully even if some of those checking steps are omitted from the articles. And, it's wise to be well-organized and keep notes regarding what's true more generally, nice ways you've found to think about/explain things, etc. That way, when you are ready to write a book, you will have a large collection of notes to draw from. But in terms of the optimal allocation of your time, it's best not to spend time polishing those notes when you could be getting publications. Lastly, realize that the story will become clearer to you as you write those individual papers, shepherd them through the publication pipeline, and give talks on them. So it's more sensible to write a monograph/book after you've done all that learning and honing of your own understanding.

  • 1
    Many years later: I agree with this. Instead of writing a monography which looks and sounds good enough to be published, one should just write for oneself in notes. Those notes can be much shorter and still contain all information. Ideas are not a linear text one's mind, but more like a non-linear map. Writing for publication requires much more time than writing for oneself; these types should not be conflated. Writing for publication also requires a lot of "marketing work"; e.g., making images extra pretty. In my case, the dissertation is a mixture of the published papers and new content.
    – Make42
    Commented Apr 11 at 11:59

You must log in to answer this question.

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