Which software do learned professionals use - LaTeX or Google Docs or Word or any other software to write their books (especially engineering level scientific textbooks)?

Points to consider:

  • Could have diagrams, images, tables, graphs at quarter or half or complete page or in-between the text
  • Divided into chapters, should it be a different doc for each chapter in a Word/G.Doc like software?
  • Citation management should be easy
  • 14
    Some some people use all kids of obscure software, the vast majority use LaTeX. Jun 4 at 21:37
  • 15
    @DimitriVulis Outside of a few fields (including those overrepresented here) most people haven't even heard of LaTeX.
    – Bryan Krause
    Jun 4 at 21:49
  • 75
    @SachinMotwani: Writing complex equations is quite a task in LaTeX. Hmm, seriously? Not too long ago I had a job where I was required to regularly type equations in Microsoft Word - and it was really pain in the ass compared to LaTeX. Jun 4 at 22:28
  • 67
    Writing complex equations is quite a task in anything but LaTeX.
    – sleepy
    Jun 5 at 8:28
  • 7
    @KonradRudolph OP is asking about the people writing textbook content as far as I can tell, not professional typesetters.
    – Bryan Krause
    Jun 5 at 14:38

What publishers and authors choose or prefer for document processing really depends on the discipline.

Your profile says you are interested in "Electronics and Communication Engineering". For that kind of technical writing I suspect LaTeX is the system of choice, both for you and for your publisher.

It's easy to prepare large documents with several files. LaTeX manages citations, diagrams and images. As an author you focus on content. The publisher can then make format and design decisions that require no retyping.

The folks at TeX stackexchange will gladly help you.

  • 6
    Yes, the folks at TeX.Stackexchange are very helpful. And it’s a good thing they are, because beginners often need lots of help. I just looked at the first page of questions, and there are two asking how you change font size, and three asking about spacing between various types of items. These things sound simple, but they’re not.
    – bubba
    Jun 6 at 9:13
  • 20
    @bubba These things are (relatively) complicated because the entire philosophy behind LaTeX is different from MS Word. In LaTex you decide what kind of document you want and then LaTex decides what the appropriate fontsize is. So ideally you should have no need to change the font size yourself. Sometimes LaTeX makes bad choices but it general it makes better choices than 99% of people who are not professional typesetters.
    – quarague
    Jun 6 at 12:07
  • 2
    Yeah, I know the philosophy — just type the text and let TeX and the document rules control the formatting. It doesn’t work 100%. Take a look at the source for the TeX book or the LaTeX book. Both are full of little tweaks to improve appearance. Anyway, newbies think (rightly or wrongly) that they need to adjust formatting. So, either someone has to tell them how to do that (which is difficult), or has to teach them “the philosophy”. Either way, a lot of people would be in trouble if it weren’t for the generous folks at TeX.Stackexchange.
    – bubba
    Jun 7 at 12:06
  • 2
    The use of “styles” in MS Word follows the same approach — it doesn’t work 100% there, either.
    – bubba
    Jun 7 at 12:12
  • 2
    "As an author you focus on content." <- That's the epitome of wishful thinking my friend.
    – einpoklum
    Jun 7 at 20:03

I've written several books, maybe a bit "advanced", in mathematics, and I've used plain TeX in all (due to getting committed to plain, as opposed to La-, TeX, to be able to manage small formatting details that my publisher 25 years ago could not cope with...)

But, yes, some dialect of TeX.

For diagrams, I've used xypic, which is by now old-fashioned, but still adequate for my purposes.

By this year, it is a bit silly to use plain TeX, since competent publishers can do whatever reformatting on their own (rather than complaining that when they print out your draft, and hold it up to the light, the page numbers don't match up... sigh...) Plain TeX does obligate one to load fonts and stuff, which, yes, I figured out how to do ages ago... at at time when, also, the RAM in many computers was awkwardly small to include the actual font files for everything, as opposed to the font metric files... and that kind of lower-level stuff.

I gather that a slightly more modern, but maybe still several-years-old, graphical package for (La)Tex is "tikz"...

In any case, I'd wager that 99% of academic math people in the U.S. use some version of TeX for substantive writing. Yes, I can imagine that some might be coerced to use "Word"... but I'd tend to think that with current personal-computer capabilities, the end-product of a PDF document would be viewed as acceptable, regardless of the "engine" that produced it.

  • 9
    +1 for an answer based on actual experience rather than semi-informed speculation.
    – Dan Romik
    Jun 5 at 7:35
  • 1
    And if some publisher happens to want Word files, there are (per a quick search) plenty of ways to convert your LaTeX files to Word. Presumably the same might apply to plain TeX. (I don't know how well they work, having never needed to use one of them.)
    – jamesqf
    Jun 5 at 17:00
  • 13
    @jamesqf In my limited experience they work very badly, On the other hand, if a publisher is requiring MS word documents they deserve what they get :) Jun 5 at 18:56
  • @jamesqf I've done so once and I hope to never have to do it again. Jun 6 at 12:01
  • @jamesqf At work I occasionally need documents that are written in LaTeX but look like Word documents. It can be done but LaTeX produces tons of warnings.
    – quarague
    Jun 6 at 12:10

A few years ago I was the lead editor for a business engineering textbook. Right at the beginning of the project we had to clarify the question of which software to use. In this field, LaTeX is largely unknown because everyone uses Word. But since it was already clear that the book would have well over 1000 pages and Word has problems with such large documents, we decided on LaTeX. This was also due to the fact that the manuscript had to be a unified whole in order to reduce errors in the typesetting and thus the time needed for corrections. (This publisher had outsourced its editing to a country where non-native speakers worked on it).

The publisher provided us with an adapted template in which we worked. The finished manuscript in LaTeX was then converted internally by the publisher into a custom XML format, on the basis of which the typesetting was carried out.

We could also have submitted data in Word, PDF, HTML, RTF or other formats; this is all converted internally by the publisher. Don't worry about what format the publisher wants, but think about what you can work with best.

  • 2
    "This publisher had outsourced its editing to a country where non-native speakers worked on it" - wow. Just wow. I wouldn't go near a publisher like that. (Mind, I am a non-native English speaker myself).
    – Yuriy S
    Jun 6 at 13:05
  • 2
    @YuriyS I believe Springer, one of the main publishers in math, does that at least some of the time (I'm pretty sure they did this with one of my journal papers), but in any case a lot of publishers don't carefully correct English anyway (as evidenced by many books and papers I've read).
    – Kimball
    Jun 6 at 17:56
  • @Kimball I didn't want to reveal any names, but… yeah.
    – Johannes
    Jun 7 at 8:55
  • "LaTeX is largely unknown because everyone uses Word." <- Everyone in such fields uses Word because, and sorry for sounding conceited, they're being too crude and careless about typesetting to bother learning LaTeX; or - their publishers just take the MS-Word content and use something better.
    – einpoklum
    Jun 7 at 20:05
  • @einpoklum My impression is that the further away you get from mathematics/science and the closer you get to economics, the less interest there is in how things work and the more desire there is for pretence, for illusion without the work behind it. Microsoft is catering to exactly this target group. But these are prejudices, of course.
    – Johannes
    Jun 8 at 9:59

Publishers of books are normally pretty flexible about what they accept from authors. They don't want to put up barriers. It needs to be something that works for the author and that the production staff can work with. For textbooks there is usually a "copy editor" involved as well as a production staff. The publisher seldom (in my experience with three large publishers, though a bit dated) asks the author to produce final page proofs. Those are done by professionals.

The author is, then, required to proofread those pages and quickly provide any needed corrections. What happens in the background may be LaTeX or some older typesetting technology, though probably not hot lead anymore.

They might even provide professional help on preparation of images for the final copy.

A publisher may have preferences, but for books, probably not rigid requirements.


Over the past few decades, I’ve written hundreds of technical documents containing diagrams, tables, and mathematics, including a 400 page book about computer geometry.

Prior to starting the book, I used MS Word for everything. For the book, I switched back and forth between Word and LaTeX 3 or 4 times, and eventually settled on LaTeX. I still use MS Word or Google Docs for everything except books.

It took me a very long time to get the formatting set up the way I wanted in LaTeX. The memoir package was a big help — in the LaTeX world the solution to every problem is “there’s a package”. But your publisher might give you a document template, anyway, so you’ll have no choice about formatting.

Creating tables in LaTeX is ridiculously complicated, compared to MS Word. For diagrams, the TeX purists favor tools like Tikz and Asymptote, in which you essentially create a picture by writing code. I find this approach impossible, so I make pictures in drawing packages, or PowerPoint, or CAD systems, and include them in the LaTeX document as PDF. That works fine.

I think LaTeX is faster for simple in-line math, but for big complex equations, I find Word faster because I can see the equation emerging as I type it, so I make fewer mistakes.

In the end, I chose LaTeX because I very much like the appearance of the Computer Modern fonts, and getting those to work in MS Word was painful. But some publishers will insist on changing fonts, anyway.

In research-level mathematics and physics, most people use LaTeX for formal documents like books and papers. In elementary mathematics and science, most people use Word. In engineering, I’d guess it’s about 50:50 — in academia, LaTeX use is common, but in industry it’s not.

A lot of folks in the TeX/LaTeX community dislike MS Word or any other software that costs money. So, there’s quite a bit of anti-Word propaganda, some of which is out-dated folk-lore. The people in the Word community don’t have a corresponding dislike for LaTeX because most of them have never even heard of it. So, as you’re reading, you’ll probably find that Word critics are more common than LaTeX critics.

  • 2
    I agree tables (and matrices in equations) can be annoying with LaTeX. Using an online table generator or a LaTeX editor with that feature can help, however, but I still haven't it done as smoothly as in the WYSIWYG systems. E.g. Tikz and pgfplots certainly have steep learning curves. While the TeX enthusiast community may favor them, I think most people who just use TeX to write e.g. papers tend to generate graphics through some other system(s), so you're not alone in that choice.
    – Anyon
    Jun 5 at 15:00
  • As for the spreadsheeet to .tex issue, if not natively supported anyway / with a macro (e.g., the older calc2latex, excel2latex), tex.se, has some tradition to assist in the identification of solutions for this e.g., here.
    – Buttonwood
    Jun 5 at 15:34
  • 11
    It's not (at least for many of us) about the software costing money, since you can get LibreOffice (which I believe is an MS Word workalike) for free. It's just that we find writing code for things a lot simpler than e.g. trying to draw when you have no artistic ability, or trying to figure out what all the supposedly "intuitive" icons do.
    – jamesqf
    Jun 5 at 17:04
  • 1
    @JochenGlueck Yeah, that's fair. I originally meant to make a comment just about tables (which I think are handled nicely in Word), but added the mention about matrices at the end. So I didn't intend it as a statement that Word handles matrices nicely.
    – Anyon
    Jun 6 at 1:35
  • 1
    "for big complex equations, I find Word faster because I can see the equation emerging as I type it," That is exactly why I mentioned LyX in my comment to @JochenGlueck at the top. If I were to write my dissertation in LaTeX in a text editor (as I did for my masters and some papers before) I would have likely never finished it in time. I now use LyX to prepare my lecture notes. And I use LyX to prepare tables to papers prepared in some shared LaTeX environment. Doing them in plain LaTeX is a very terrible experience and leads to errors that cannot be debugged.
    – Vladimir F
    Jun 6 at 8:52

A little late, but I just ran across some actual data (as opposed to all the speculation and opinion in the other answers, including mine). There was a study of LaTeX usage by François Brischoux and Pierre Legagneux, published in The Scientist in 2009 (Volume 23, Issue 7, Page 24). I don’t have access to the original paper, but I suppose the following table, which was reproduced here is a good summary:

enter image description here

Things may have changed since 2009, of course.

Many of the other answers here are trying to tell the OP which authoring system is “best” in some sense. That’s not what he asked — he asked which ones are most commonly used. But, if anyone is interested, there’s some data about “best” in this article, published by Knauff and Nejasmic in 2014. In their tests, they found that using MS Word yields higher productivity than using LaTeX. YMMV, as always.

  • 1
    It's worth noting that the table caption only mentions papers, not textbooks. I expect the trends to remain more or less the same also for textbooks (as people tend to use what they're familiar with), but the numbers would likely change.
    – Anyon
    Jul 12 at 14:52

If you're working with a publisher, ask them. Chances are they accept virtually all major software. TeX is the obvious one for scientific text, but Word is also fine if you know how to use it. If you're self-publishing then it really doesn't matter, whatever works for you is fine.

Could have diagrams, images, tables, graphs at quarter or half or complete page or in-between the text

If you're working with a publisher, this matters only if you need to refer to other pages while you are writing the manuscript. Otherwise it is irrelevant, do not worry about it. That's because by the time the manuscript's been reformatted it will not look like what you submitted and all the time you spent on this will be wasted. If you're self-publishing then whatever floats your boat.

Divided into chapters, should it be a different doc for each chapter in a Word/G.Doc like software?

Again if you're working with a publisher, this is irrelevant. If you're self-publishing then whatever floats your boat.

  • 1
    What do you mean by "scientific text" here? I'd be surprised if TeX is the obvious choice for authors of biology textbooks, for example.
    – Anyon
    Jun 5 at 15:05
  • @Anyon IDK, it looks like a lot of Biology journals, at least, are providing LaTeX templates. Jun 7 at 19:18
  • It looks like LaTeX usage in biology was essentially zero in 2009 (see my answer below).
    – bubba
    Jul 12 at 12:41

I sometimes help edit high-school mathematics textbooks, which have most of the requirements you mention (not citation management) and are published by an academic press that also does university textbooks, so I expect the process will be similar enough.

This kind of publishing is a group project. Authors generally aren't responsible for their own layout, just for providing the words/equations. In my experience that's usually done in MS Word, using either Mathtype or Word's equation editor; I can only remember one time when an author worked in LaTeX. (NB most of the authors I work with are high-school teachers; I suspect there'd be more LaTeX users among university academics.)

Other editors/etc. then format that material for house style, create diagrams according to the author's sketches, and lay it all out to generate a tidy PDF, usually one per chapter.

Breaking it up into chapters not only keeps file sizes manageable, but also helps with workflow - since there will be several different people working on the same book, you don't want to have each person go through the whole book before the next person in the sequence can start.


You had all kind of reasons to use LaTeX in other answers. The extra few you may want to consider:


  • repeatability: you can compile a LaTeX document several times and when you change a list you will not suddenly have all your headers change font (yes, I had that)
  • stability: the more text you type in LaTeX, the longer the compilation becomes - but you get something at the end. My good friend stopped his thesis on "(...) this equati" and then he could not neither add nor remove text.


  • unsung suffering when you want to change the default formatting. You really need to dive into the language.
  • you edit plain text so you do not have the usual hints on headers, bold text etc. (some IDEs can probably handle that)
  • collaborating is a real pain. No matter what, the "tracking" in Word is much better

CON that is actually a PRO

  • between what you type and what you see there is a compilation session so it is not immediate. And this is a good thing because you know you have no influence anyway so you concentrate on your text.

EXCEPT for equations, where anything more complicated is beautiful but requires a second CPU in your head to follow the \int{\frac{a\frac{\times}{... where am I...?

  • 12
    I would strongly disagree with putting collaboration as a con for LaTeX. The pain ASCII LaTeX source works very well with the revision control systems used for software development (e.g git, mercurial or Subversion), and the related diff tools. It helps if you follow the practice that each sentence starts a new line in your LaTeX source. This doesn't affect the final appearance, but means insertions and deletions leave clean diffs. Jun 6 at 0:25
  • 5
    @StephenC.Steel: I am a developer and use git on a daily basis. It is wonderful when developing with others developers. Having someone not used to it collaborate through diffs, merge requests, etc is simply not an option. This is light-years from a Word tracking that is simple and very intuitive. I can explain to my father by phone how to use it, something impossible with git & co.
    – WoJ
    Jun 6 at 7:47
  • 1
    @StephenC.Steel: it is even more obvious with online editors such as Google Docs where the collaboration is completely seamless.
    – WoJ
    Jun 6 at 7:50
  • 2
    @StephenC.Steel. Interesting typo. You probably meant “the plain ASCII LaTeX source”, but pain is often present, also.
    – bubba
    Jun 6 at 9:27
  • 1
    @jamesqf: yes, what I meant is that the fact that you do not have a WYSIWYG editor helps to understand, after some time, that it is not that useful to peek at what you wrote in its final form because it is going to be good anyway (or that you do not have much control in it). Except for equations :)
    – WoJ
    Jun 6 at 16:50

Not the answer you're looking for? Browse other questions tagged or ask your own question.