Why you should LaTeX your dissertation; or, why you don’t have to write your dissertation in Word.

If given the choice, I choose not to use Microsoft Word. My experience with the program over many decades now has been that, despite its almost complete dominance of the word-processing sphere, Word is a program that does not handle long-form scholarly writing well. It is, like its kin in the MS Office Suite family, a program designed with businesspeople in mind: people writing letters and reports, making presentations, processing financial data, and communicating within a rather narrow band of kinds of documents. While this is all well and good, and probably covers a significant proportion of the electronic documents out there, it has not met my needs as a reliable program for typesetting long papers, dealing with footnotes and bibliographic information, handling images, captions, tables, and cross-references, and doing this all in a flexible, intelligent, and, not least of all, aesthetically pleasing manner.

Most importantly, I hate wrangling with software that I am forced to use not because it is designed for what I am doing or is the best fit for my needs, but because it enjoys an inexplicably huge market share, and, also presumably, because people in the humanities are not, by and large, writing their own software to meet their needs. I watched many colleagues in graduate school pulling their hair out towards the end of their Ph.D.s because Word would continually crash while they were putting the finishing touches on a 60-page chapter, or as they tried to generate a table of contents, or when they finally were attempting to put all their completed chapters into one long document. I myself had moved away from Word before I began my postgrad; all of this, plus my continual frustration with Word while working a job, convinced me that I would only use it to typeset my dissertation if I absolutely had to.

Fortunately, one of the best and most useful things I learned in college was a markup language called LaTeX. LaTeX and its variants were designed with the needs of mathematicians, scientists, and engineers in mind — TeX is particularly excellent at handling mathematical expressions, equations, tables, figures, and the like — and was pretty much the unofficial standard mode of typesetting at MIT. Every problem set I ever received during my undergraduate years had been formatted in TeX (or a variant thereof), and, like pretty much anyone with some exposure to the hard sciences, I quickly began to be able to recognize TeXed documents when I encountered them. (The default Computer Modern font is the first dead giveaway; although TeX is highly customizable in the fonts department; read on.)

And, like many MIT undergrads (at least in the nineties), I learned to use LaTeX to typeset my own documents: at first particularly term papers, because of the powerful bibliographic-handling powers of its counterpart, BibTeX, but also letters, meeting agendas and lists, short writing assignments, and, eventually, my resume. Like HTML, TeX is a markup language that gives instructions for how to format text; it’s not coding per se, but style-guiding. It is incredibly powerful, highly customizable, pretty easy to get the hang of in its most basic form, and unsurpassed in creating documents of beauty and simplicity.

Over the years, I have tried to convince other humanists who may never have encountered LaTeX that they should consider learning it; and I have rejoiced when I have run across others in history and allied fields who also use TeX to do their writing and typesetting. (Philologists, linguists, and people who deal with a lot of non-Roman alphabets tend to be familiar with LaTeX. And, of course, apostate scientists and engineers like me and a few of my colleagues.) This post is an attempt to enumerate as well as understand in greater depth my reasons for doing my own work in LaTeX, and to perhaps convince others that learning an unfamiliar markup language is worth their while. In the process, I also want to consider some of the benefits and challenges of using TeX typesetting in the humanities, and discuss why it is that more of my colleagues in history remain unfamiliar with it, despite its enormous strengths for anyone working on long manuscripts with lots of complex bibliographic information.

Why LaTeX?

There are a host of reasons — technical, aesthetic, psychological, educational, and practical — for choosing LaTeX as your typesetting environment. Here are my main justifications for why I have chosen and stuck with TeX, and why I think it’s worth considering, even if you’ve never heard of such a thing before.

LaTeX is Open Source.

Because LaTeX is an open-source means of typesetting, it is both free to you, and supported by a vibrant community of users who are constantly improving and adding to its functionality. You do not have to buy LaTeX, or any programs associated with it. This is particularly good news for graduate students: no more having to shell out tons of money to Microsoft just to write your seminar papers, for LaTeX is free. It works on all the major operating systems (Windows, MacOS, Unix/Linux), and, what is more, has very user-friendly standard installs of the TeX distribution for each of these environments. I run TeXLive on MacOS. If you’re working in Windows, you’ll probably want to get a perl interpreter (why perl does not come packaged with Windows is completely beyond me!); beyond that, a standard Windows install and package manager like MikTeX is all you’ll need.

The TeX user community at tug.org and beyond is an incredible resource as well. When you run into problems — and you will — you will almost always be able to solve them with a quick web search for your issue and a perusal of the bulletin boards and blogs devoted to LaTeX and fixing common problems. Chances are, someone else has run into the same problem you are having, and has posted how to solve it, or had their question answered by an expert on a message board. There is no LaTeX problem I have ever run into that I couldn’t solve in pretty short order by consulting the web. I have never even had to create a new posting about an issue — everything has always been solved before me, or I have found enough guidance out there to solve the problem myself.

LaTeX is designed for scholarly writing, and handles everything you can throw at it intelligently and with aplomb.

Whether you are producing a cover letter, a C.V., an article, a manual, a seminar paper, a dissertation, a book manuscript, or just about anything else you can think of, LaTeX has a package or a style for it. It is the original “there’s an app for that.” Need to change your style of page numbering midway through a document, or reset a counter for figures, notes, or pages? Easy. Want to play around with how all your headings are displayed? No problem. Want to product an incredibly long and complicated document with chapters, headings and subheadings, a title page, hundreds of footnotes, a bibliography, a list of archives consulted, a table of contents, and more? LaTeX was born to do it. As the woman who did the prechecks on my dissertation at the UW grad school said to me, “There’s nothing LaTeX can’t do.” “I agree!” I responded, knowing at that moment that everything was going to be just fine.

It is true: LaTeX has been able to handle everything I have thrown at it, and to do it in an elegant fashion, both technically and aesthetically. What is more, each time I have tried to do something new, I have learned something useful, enjoyed the process of figuring it out, and felt extremely proud of the results. Contrast this with how you tend to feel when you’re trying to get Word to stop forcing a page break when all the white space says it shouldn’t, or something else particularly frustrating that you’ve encountered in your tangles with that program. No contest there.

LaTeX allows you to write from your sources, and build highly modular documents.

When you’re working on a writing project that’s more than a couple of pages, you usually have bits and scraps that you want to hang on to, but that you’re not sure you want to appear in the final version. When you work in a word processor like Word, you usually have to cut and paste these onto a separate page at the end of the document, or into a new document altogether. This can be a huge pain in the butt, since it usually means scrolling around to different parts of your document, or switching back and forth between two documents, to find and grab the thoughts, ideas, notes, or other structuring information that you’re using as a writer, but which shouldn’t appear in the version the reader sees.

Separating editing and processing through a program like LaTeX means you can hang on to your work in the raw source, but have it not display in the typeset output. You can simply comment out the words by placing a “%” at the beginning of the line, sentence, paragraph, quotation, or note-to-self you don’t want to reader to see, and POOF! it’s gone, visible only to you as the source editor. This allows you to write in an incredibly modular way if you so desire, moving things around, commenting them out or putting them back in as your project evolves and as the argument demands. Everything is in one place: no more “excerpts” or “notes” files lying around, muddying the waters. It’s all in the source, ready to be brought to light.

This is particularly useful as a writing aid, especially when I am starting a new piece of writing. Often, I like to get all my quotes and sources in one place, and use them to build my argument. But, as I’m writing, I want to see how everything is displaying, and how long the finished document is, without those quotes and sources being included in the word or page count. If you input all your quotes and evidence, and comment it out, you can write around your sources and then add them in as you decide which ones are best for your argument. No more searching around for that quotation — it’s right there!

The ability to comment out parts of your source document is also quite useful when you’re playing with formatting. It means you can save one setting that you like on a commented-out line, replacing it with a line that calls for a different kind of formatting, but which you could always toggle-off and toggle back on if you want to switch back to the way it was before. Handy for the document preamble, especially if you write good comments about what each different setting will do.

The separation of editing and processing is also good if you want to collaborate on documents using a versioning control system. Google Docs can do this to some extent, but there’s nothing like a good old reliable CVS for really making sure that you’re not overwriting someone else’s work. Of course, the chances of finding someone in the humanities who knows what the heck I’m taking about here is so small that this point is pretty unimportant in the whole scheme of things. Still. It’s the truth.

BibTeX is the most powerful, customizable, and robust bibliographic management tool out there.

Probably the main reason I have stuck with LaTeX throughout graduate school, despite at least one professor’s complaints about not being able to “track changes” in a Word document to give me feedback, is not LaTeX itself, but its allied bibliographic management protocol, BibTeX. I have used many kinds of bibliographic software that are standard in academia — EndNote, RefWorks, and the honestly pretty awesome Zotero — but I have yet to find one that would make me switch from BibTeX. BibTeX is incredibly robust and flexible, handles cross-references and other complicated aspects of your database queries impeccably, and integrates seamlessly with LaTeX. It has at least one great GUI frontend, BibDesk, and the option of letting you really fool around in the raw DB text if you so desire. What should convince historians that LaTeX/BibTeX is for them is actually a slightly newer implementation of BibTeX called biblatex and its counterpart engine biber, particularly the historian style, which hews to Chicago/Turabian and — the thing I’ve not found in any other bibliography management system — handles archival sources absolutely perfectly. You can generate bibliographies that automatically separate unpublished and published sources, that generate a list of archival abbreviations for your reader, and that exclude certain types of sources you might not want to appear in your final list of references. Its power is unmatched in my experience, and it was completely critical in making my footnotes and bibliography a relative breeze when it came down to the wire.

LaTeX is a thing of beauty, and produces things of beauty.

Using LaTeX is an aesthetic experience on two levels. It is an elegant program in its design, implementation, and functionality; and the documents it produces are themselves extremely aesthetically pleasing. If you care about fonts, kerning, and proper text handling, LaTeX is definitely for you. But LaTeX is also for you if you like your software to work reliably and intelligently. It is not bloated or over-engineered. It is infinitely customizable with packages, styles, fonts, and other add-ons. It works beautifully; and when it doesn’t work, you as a user are able to fix it.

The difference again stems from the separation of editing and processing. Word is processing your text on the fly as you edit: it can’t see the whole you imagine, and it is making its adjustments based on the immediate surroundings, and what it can imagine you want to do is limited by that (hence its often hideous output). Because in LaTeX you edit your text, then process it, the compiler has a look at everything before it decides how it’s going to lay everything out. As a result, its “instincts” about how to format text, images, references, notes, and everything else are usually more correct (or better designed) than Word’s, which are situational, rather than holistic.

Why not Word?

My reasons for sticking to LaTeX have almost as much to do with my dissatisfactions with Word as they do the joys of working to master LaTeX. This is not to say that I will never use Word, or that I don’t know how to do all the things I’ve described above using Word — Word can do them, and will, albeit in an incredibly glitchy and frustrating way which drives me bonkers — it’s just that, if given the choice, I prefer to work in a typesetting environment that makes me feel like I am mastering something useful and well designed, rather than banging my head against a wall repeatedly for little reward. Does this sound familiar to you? It doesn’t have to be that way.

You are not a cubicle drone: you have a choice in how you typeset your work, and you can choose to use software that was designed for you, rather than without your needs in mind.

My main complaint about Word is that it’s a program designed almost entirely for business people: writers of reports, memoranda, letters, and other short-form documents that do not require the kind of complicated typesetting that scholarly writing consistently demands. And when you work in Word, you feel this constantly: this program was not designed for me and my needs. It makes you feel like you’re at war with your software, constantly trying to force it to do things they way you want it to, and constantly being stymied by its endless layers of supposedly helpful settings that put abstraction barrier after abstraction barrier between you and your work. It is technically true that Word can do all the stuff we as historians want it to do; but I ask you, does it do all of this smoothly, uncomplainingly, and with minimal frustration to you? If your experience has been anything like mine, the answer is a resounding “no.” Oh, you want that text to flow around this image in a sensible manner? Oh, you’d like me to position this page break in a place that doesn’t create enormous amounts of white space? Oh, you’d like to play with your styles and formatting without screwing up the whole document? Sorry, that will cost you several hours of annoyance. Please, read on.

When you work in LaTeX, you are working in an environment that was designed with the needs of scholars in mind. Once you get the hang of TeX, you will understand just how poor a fit Word is for the kind of work you do. You will realize that you can do more, more easily, more robustly, more powerfully, and with a smaller kilobyte footprint than you could possibly do in Word. And when things break, you’ll learn how to fix them.

Wrangling with LaTeX is more rewarding than wrangling with Word.

Okay, I’ve been painting TeX as this masterpiece, when the reality is that nothing is perfect: you will definitely experience frustrations with LaTeX, just as you will with any software. (I promise you that at some point you will have a serious breakdown over margins or document layout, something you think about almost never in Word.) But I want to suggest that your experiences wrangling with LaTeX to get something to display the way you want it to will be rewarding to you in ways that they can never be in Word. (And I promise you also that you will solve your margin problem, possibly with something as simple as a quick web search or a helpful message-board posting or, more involvedly, learning a bit about a new package, and you will feel so good about having done so that you will take yourself out for ice cream thereafter and try to explain to your friends why you’re so happy. Finally, I promise you that they will think you are insane.)

As I’ve been writing this post, wanting to make this argument about the comparative rewards of troubleshooting in TeX versus Word, I have been asking myself, is this really true? Don’t you learn things when you figure out how to do something in Word too? Aren’t those tricks useful later on? To be sure, the more you use any software or tool, the more familiar you are with its workings, and the better you get at bending them to your will as a user. You asymptotically approach mastery. This happens in Word, just as it does in LaTeX.

The difference is something I haven’t been able to articulate well thus far, partly because I don’t want to simply fall back on an argument that is in form equivalent to “it’s better to know how your car works at least on some level than to treat it as a black box.” (I believe this, but I’m not sure it’s sufficient, and I don’t think everyone feels this way.) But it has something to do with the level at which you’re interacting with the program: in the case of Word, the surface, where what is possible is limited by what the software designers imagined you might want to do, and is therefore premised on assumptions about who users are, which I think is pretty clear (businesspeople, admin assistants, writers of reports and memos). When you work in LaTeX, you are working more on the inside of things, and what you can change and modify to meet your needs is enlarged. More options are open to you because you participate in the design of the program, in a way.

I’m still not satisfied with this answer, and I hope commenters will help me sort out what I am trying to get at here; but I will say this: When I get something to work in Word, I feel frustration at the existence of the problem, anger and my inability to find out what is causing it, exhaustion with my inability to solve it elegantly, boredom at my attempts to work around it, and relief when it’s finally sorted. When I get something to work in LaTeX, I feel confusion, a desire to know what’s happening, interest and a desire to solve the problem, enlarged knowledge once I’ve figured it out, and pride in my resulting handiwork. Two very different experiences. I’ll leave it at that for now.

Why not LaTeX?

To be fair, there are plenty of arguments for not journeying too far down the rabbit hole. The first and most important one is publication: I have not thus far run across a major journal in my field that proclaims on its web site that it will accept manuscripts in TeX. I find this a bit sad, since many journals in the sciences expect to receive documents in LaTeX, and provide TeX stylesheets or templates for those submitting manuscripts, which I think would make the whole process easier for everyone, not least of all the editors and typesetters at those journals. But I have the sinking feeling that most history journals would be a bit befuddled, or annoyed, if I tried to get them to accept my TeX source. There are some okay LaTeX-to-RTF converters, but it is true that this is one major problem that stands in the way of wider adoption of TeX outside the sciences. The DOC’s dominance in the humanities remains unchallenged, and individuals are not going to do it. This is a big structural roadblock that should be an enormous caveat for people considering following my advice in the above paragraphs, and I won’t try to deny its importance.

Alongside the issue of publication is that of collaboration: as I am embarking on a couple of article projects with colleagues, I am realizing that a lot of time with Word is in my future. That’s okay — articles are short, and Word can handle them pretty well — but there’s a nerdy part of me that wishes I could just collaborate with people using a versioning system and the raw TeX source. That is a dream that will certainly remain unfulfilled.

There are other obvious reasons not to move to LaTeX as well: having to learn a markup language that most people in the field have never encountered, working in a text-based environment rather than a GUI (although people who like to work WYSIWYG should consider LyX), having a steep learning curve before you become comfortable with the vagaries of LaTeX, and having people complain when you don’t send them DOCs to track changes on. But, despite these cons, I remain convinced that, for scholars working on large complex manuscripts, LaTeX really is the gold standard.

I harbor no illusions about this, though: the pressures to hew to Word are enormous, and the rewards of doing work in LaTeX are largely personal rather than social. No one in the humanities will pat you on the back for TeXing your dissertation, and there aren’t any prizes to be had for “best TeX source” or “most beautiful manuscript.” (It’s too bad — that’s an award I could really compete for in my field.) Your labors will be largely invisible to your colleagues, and only your friends in the sciences or from back in college will look at the finished product as a masterpiece of markup as well as of scholarship. You will have to be content with the knowledge that you really did a good job, and the satisfaction that comes from that job well done. (And you may impress some scientists down the road at some point. If you tell them, they will be impressed.)

I also will say that I do not think that TeX is good for everything. But what it is good for — typesetting complex and/or lengthy manuscripts, dealing with references, notes, and counters, and other allied challenges of word processing — it really is best for, I think. Even just knowing such things are possible and out there should be enough to get any scholar who works in Word to ask herself whether she really is working in the best possible way.

The rewards of making something.

For me, in the end, the issue comes down to temperament and disposition. I enjoy the challenge of solving a problem when I know that doing so will teach me something that I will use in the future. I also enjoy feeling like the products of my labors are just that, not solely in terms of their intellectual content, but in terms of how they are presented, formatted, and laid out. I want to produce something I can be proud of as both a work of scholarship and a work of the printed word. LaTeX allows me to do that all the time; with Word, those rewards are harder-won.

If you do not garner any satisfaction from making something from scratch, from tinkering and getting something right, from learning a bit about what’s going on inside the black box of your word-processing, LaTeX is probably not for you. But if you like to do things right, and elegantly, and beautifully; if it matters to you how your document is produced; if you take a certain amount of perhaps unspoken pride in the process of crafting your document alongside your words; then you should consider learning LaTeX. It’s one of the most valuable skills I have ever taught myself, and it involves me in a continuous process of learning and developing those skills while I practice my scholarship. Doing the two together is enormously rewarding for me, and I hope others find reasons to do so as well. (I know that I am not alone here!) I hope that this post at the very least opens more people’s eyes to the options that are available to them, so that they do not assume that they have to do everything in Word if they’d prefer not to. Perhaps most importantly, the more humanists who use LaTeX, the more and better tools there will be for us to do our work well.