If you were writing a dissertation in the present, what would you use?

I’m torn by all the options.
I do have guidelines that I will need to follow, but I can do that without using the Uni’s Word or LaTeX templates.

I have ruled out Word.


Scrivener with no doubts.


Scrivener. (Padding to 20 char).


Perhaps you are torn by the options to manage the content in a coherent way versus to assemble the content in an efficient way. The former can be done independently from the latter.

For managing (on macOS):

  • Curio
  • Obsidian
  • DevonThink

For assembling:

  • LaTeX. Caveat – I am strongly biased from nearly four decades using it.
  • Scrivener. While it certainly appears to offer significantly greater benefits to manage (organize) content compared with LaTeX or Word, the overhead required to get a publication-ready standard document (through export through Pandoc through … ) turns me off. Caveat – I’ve tried it once but never took to it, perhaps because my LaTeX preambles generally include page settings as well as LaTeX commands + environments that go well beyond what I could figure how to cobble together with or into a markdown to LaTeX translator without causing myself more agony than just doing it in LaTeX in the first place (see above about my longer experience already with LaTeX directly).
  • Not Word. At all. Caveat – My graduate students seem to have found a successful way to make it work for them regardless my beliefs that it is simply a bloated note-taking app.



I wrote my dissertation in Scrivener and used the option to compile MMD to LaTeX to produce the final document. It worked beautifully for me.


This is what it did the last time I used it. 22 page document and it decided to just drop out some pages. Not using it again. My advisor and I will just have to work out some workflow.

I don’t think of it as a bloated note-taking app, but a desktop publishing app where presentation is baked into content (along with failure and unpredictability).


Having tried Scrivener to write my thesis, it wasn’t as easy as LaTeX. Sure, there’s some code knowledge needed for LaTeX, but I found that quicker and easier to pickup the basics than Scrivener, especially to play nicely with references.

Teamed up with Git as a versioning tool and backup, plain text as tex files was really handy. I was on the Beta testing for Texpad and that is still one of the best LaTeX editors I’ve used, and makes it easier to use.

1 Like

This seems to be the “use an RPN HP calculator versus TI calculator” level discussion from many decades ago. I have to wonder why RPN never appears in tools such as GeoGebra or Mathematica or Maple or …

In any case, I’ve expanded my caveat on Scrivener. I don’t doubt that some folks find using Scrivener to be a beautiful experience. I just could never see properly to invest the time to master the markdown → LaTeX step to appreciate the added benefit in the content management side.


1 Like

You my friend have a lot more gracious good will toward Word than I do.


1 Like

One of the challenges with Scrivener is in the feedback cycle. I’m curious how others have dealt with that How to send drafts to advisors, then get feedback and incorporate into existing work? One option is to compile into Word, send, receive Word file with comments, just add word file into Scrivener and view in a pane on its own? When I was dealing with a single feedback cycle in Scrivener I just copied the comments from Google Docs into Scrivener as scriv comments, but I wouldn’t want to do that with a diss that’s having more feedback cycles from multiple folks (maybe–that’s how it works in my field).

1 Like

Good point. Feedback cycle.

I am old school. It came back to me in red ink on hard copy. I modified the source manually accordingly.

LaTeX is perhaps the least friendly place to do modifications for feedback. Translating hand-written comments on a PDF back to the source LaTeX file is … cumbersome. So, when you plan to use LaTeX, plan your document workflow to allow for cumbersome editing processes. Store the document in chunks (e.g. one file per sub-section) and use \input liberally.

A compromise might be to use LaTeX → Word, send the Word file for electronic editing, and then pull out the electronic edits from Word back to the LaTeX source file. Ugh.

Editing back from other people’s comments is just not easy with LaTeX.

The lesson is, with LaTeX, you have to write a perfect dissertation the first time or be prepared to spend time adding in the feedback comments. With Word, you can easily add feed back comments (they are automatically in place simply to “accept or reject”). You just have to hope that the edits do not change the numbering or positioning points for sections, sub-sections, equations, figures, or anything else that is not simply a string of words.


LaTeX but then I am so old fashioned, really I think that is probably all it is about.
I remember a friend over a decade ago coming to me in despair as one of the dissertations he supervised was being done on a single Word document: not even one document per chapter, which would have been sensible. As always the candidate sailed thru of course with no mishaps.

1 Like

Word is more capable than this, and can reference and number things.

That’s one of the annoying things. It has lots of capabilities, then drops pages when you’re near a deadline. Thank goodness for git and Arq.

I started writing the article (shown above from Word) in Scrivener. It took a lot of time to get things working so I could compile to Word and LaTeX (pandocomatic and all that business).
When we entered the feedback cycle, that all broke down, so I just finished in Word, and submitted in Word, which is slightly embarrassing.

I’m kind of leaning toward LaTeX or markdown->LaTeX. Like Frank Costanza and tinsel, I find all the LaTeX markup very distracting. This might be better for a diss, but the resume I just wrote is a soup of

\cvline{2019}{Magnificent thing I did}
\cvlineitem{George P. Burdell, \textbf{look at me I helped}, Even Moore People. Paper name. Journal. }

It’s just hard to ignore it.
But as I said, paragraphs are probably better.

There have been a couple of attempts to remedy this that I can recall, LyX, and Manuscripts, neither of which are useable, and may have been abandoned.

Silly but … You’re not supposed to read the LaTeX markup while you work on the content. It’s supposed to be like overlooking the tarps and ladders in the room while you paint your walls. Only better. Because all you need do is learn to hit “command-T” (compile) when you want the markup to go away (try that while you paint your walls and see if that makes the tarps and ladders go away :upside_down_face: ).

Write your resume in Word. Cleanly formatting some content over two to three pages when that content has equations or no need for self-numbering is what Word does well. If you do so, also learn how to use styles. Don’t kludge stuff (e.g. putting extra carriage returns between headings and content when heading styles are designed to control this directly).

In any case, if you do write your CV in LaTeX, start with vanilla LaTeX article style and modify it sparingly with focused reasons. IMO, a well-designed CV can be generated simply using by a proper choice of basic setting packages (titlesec comes to mind here) and environments in the document preamble.





\mi{} BS & Somewhere & Someday \\
\mi{} PhD & ...


In the end, I might say that you are taking LaTeX too far for the easy cases (with your CV) and yet trying to avoid learning the direct power in LaTeX for what will be the harder cases (your dissertation) that LaTeX is designed to handle efficiently.

Being as passionate about LaTeX as I am, I’m glad to go off-line on discussions about it if you are interested.


1 Like

Thanks, might take you up on that.

I used moderncv for my resume. I like the look of it, and it looks more “industry/modern” than my former article-based LaTeX cv.

Yeah, as I said:

It’s not even a matter or reading it though, just looking at it overall it’s a mixture of markup and content, with nothing but curly braces to separate it.

Maybe Texpad has a theme and a can colorize it or lower the contrast on the markup.

I’ve been using LaTeX off and on for 10 years. Maybe when I have another 30 under my belt all that markup will just fade away and \textbf{important} will just look like important.

It really depends. If you’re doing anything scientific, mathematical, or technical with formulas/equations/mathematics/tables of data, I’d highly recommend LaTeX, while it’s got a learning curve, the fact you were provided with a template implies your institution cares about the presentation and formatting of the dissertation—and gives you a great starting point.

I would suggest that writing it in something that doesn’t support a bibliographical aid of some kind is going to be troublesome—something you definitely want to keep in mind.

Personally, if it were not scientific/technical (as above), I’d likely go with Scrivener. But give LaTeX a try. It’s not WYSIWYG, but with something like Overleaf it’s not that bad!


I have to agree with Rosemary. If you sit on the “other side” you want to avoid spending hours trying to teach the writing tools or apps to them, Word is often not such a bad solution. Many of my doctoral students (medicine) don’t even use a Mac (can you imagine?). And adding comments or corrections (and following up after multiple versions) is not so bad in Word.


I use Word. Using styles it’s absolutely fine. I use it in conjunction with Zotero which manages all my referencing within the Word document.

Word + OneDrive provides automatic versioning and saving.


I tried Markdown to LaTeX. I wrote my last major document for work using Markdown and converted to Word using Pandoc and found it was would have been easier to just write in LaTeX.

For me, the benefits of LaTeX came in with the referencing of everything, as well as the benefits of the acronym package. Tables as well were more straight forward that Word as well once you get used to it (or use a tool - I used to use Gnumeric to design and complete the tables, and that is able to export as LaTeX).

Another thought is to consult with your colleagues at the University … they will form a large part of your support network–through good times and bad–and see what they tried, are using, not using, and their recommendations.