Practical Typography: You’re Doing it Wrong

A graphic designer friend sent along a link to a very handy introduction to typography, Practical Typography by Matthew Butterick. The entire guide is freely available and the site itself illustrates many of the principles laid out by the author.

Possibly the most useful part of the guide for academics is the page on research papers. Butterick recommends a few big changes to the typical double-spaced, 12-point Times New Roman* with 1″ margins we’ve come to know and love(?). Instead, Butterick suggests bigger margins (and thus smaller lines), a bit less than 1.5 spacing, and a slightly smaller font. The amount of text on the page actually goes up a bit, but the added white space makes the whole thing more readable. Most of his changes will look familiar if you’re used to LaTeX – LaTeX’s article class uses most of these principles by default. Tl;dr: Bad. Good.

I’m still working my way through the whole guide, but so far I’ve found it very useful and accessible, and with just the right amount of snark.** Also recommended to anyone trying to convince a co-author to use a single space between sentences.

* Butterick’s Times New Roman (TNR) hatred is interesting. Part of his dislike of the font comes from it being optimized for printing small characters on bad paper, as it was originally used. Modern variants are a bit thicker and thus look better when printed at the larger sizes that are typical. But part of Butterick’s dislike simply comes from TNR serving as a signal of typographic apathy. Here I disagree (though without the aid of any actual training in typography). Because TNR is both ubiquitous and otherwise inoffensive, it serves nicely as an unmarked category, suitable for when you really don’t have much need to call attention to your font choice – or better yet, when you explicitly want your font choice to go unnoticed. It’s also installed by default on every word processing device available. So, for example, while I might love for my undergrads to write in Butterick’s delightful TNR-alternative Equity, it’s much easier to make them write in TNR. Also, as Butterick elsewhere notes, documents meant to be shared for collaboration necessarily require fonts that are system fonts. TNR is both a default system font for Macs and PCs and everyone’s used to it. That said, I might consider changing my working papers to something a bit nicer in the future. Put differently, much of this advice only makes sense when you are about to share your work with someone you hope is going to consume it, rather than someone obligated to read it or actively contributing to it.

** E.g. “Many system fonts are not very good. This is less of a problem on the Mac. But some of the Windows system fonts are among the most awful on the planet. I won’t name names, but my least favorite rhymes with Barial.”



  1. Fr.

     /  November 6, 2013

    My technique for two years has been to compose essay templates on Google Docs and export them to docx/odt/rtf formats for students to use for their homework. It works pretty okay: around 75% of the homework I grade now comes in better layout (e.g. paragraph spacing ≥ 1.15em).

    The fonts remain an issue. I long for Helvetica, so I actually use Arial in the templates. Many students are too disturbed by the idea of typing in sans and switch the paragraphs to TNR or Times. The only solution would be to coerce all of them to work with Google Docs, where you get a reasonable choice of online fonts as well as PDF export capability and more.

    (I’m actually already planning to coerce all 2nd semester students into using Google Docs.)

  2. About that whole “always one space, never two” thing… (I personally prefer two spaces, but find the intensity of the discussion rather absurd.)

    • Thanks! The intensity is a bit silly, but I admit I’ve been sucked into it too many times, and saved most recently by LaTeX’s habit of simply ignoring the second space. To Butterick’s credit, I think, he doesn’t invoke the (mythical) typewriter history of the two space alternative to justify abandoning it, but relies heavily on the current consensus of professional typographers. So, he embraces the conventional nature of the debate, but argues that the current convention is for one space.