Extreme font neepery is one of those hobbies that I’ve never gotten into. I mean, I have my preferences: I like serif fonts over sans serif fonts and I think anyone over the age of fifteen who uses Comic Sans for anything should question their life choices, but I am not someone who reads the little “this book was typeset in…” note in the back of books and arguments about Georgia vs. Palatino vs. Times New Roman leave me looking for some better form of entertainment.
But, it turns out those subtle differences in curlicues and kerning may be much more important than you’d think.
Back in 2012, Errol Morris, in a column in the New York Times, did an experiment. He presented a quiz titled, “are you an optimist, or a pessimist?” in which he presented some cheery, reassuring text about the likelihood of an asteroid hitting the earth and then asked people whether or not they thought it was true, and how convinced they were that it was true (or not true). People saw that reassuring text in one of six fonts: three serif (Baskerville, Computer Modern, and Georgia), two sans-serif (Helvetica, Trebuchet), and one ridiculous (and sans-serif) (Comic Sans).
It really shouldn’t surprise anyone that no one believes anything written in Comic Sans. If you Google for “most hated font in the world,” Google will instantly pull up an article about Comic Sans. There is an entire website dedicated to hating Comic Sans. There’s a website dedicated to helping Comic Sans addicts overcome their tendency to use it inappropriately. There is a Ban Comic Sans tumblr that is mostly a collection of wildly inappropriate uses of Comic Sans. If you want something to be taken seriously, of course you shouldn’t use Comic Sans. (Or Papyrus, for that matter.)
Anyway. According to Morris’s results, which he had analyzed in detail by David Dunning, a psychology professor from Cornell, people were markedly less likely to agree with the statement if it was made in Comic Sans. Though, when looking at active disagreement, Comic Sans was more in the middle of the pack; the “winner” there was actually Georgia.
It’s worth noting that although they had a substantial sample size (45,000 responses) it’s a thoroughly non-random sample: this was a web survey done via the New York Times website. Who reads that site? Well, they are “educated, affluent, and influential,” according to the page explaining the audience to potential advertisers. This was not a “study” per se, but an experiment done by a guy with an online column. The results were still pretty striking, though.
Far and away the most trustworthy font was Baskerville, a serif font that dates back to 1757. Michael Bierut, who recently reprinted Morris’s essay as a handsome paperback, said that in his mind, “Baskerville speaks with a calm, confidence-inspiring English accent, sort of like Colin Firth.” The difference between Baskerville and Computer Modern was fairly small, but it’s worth noting that not only was Comic Sans distrusted, the other sans serif fonts were less trusted as well. Helvetica is supposed to be a respectable font, the screen equivalent of a grey business suit. No one really notices it, no one really objects to it. The most serious charge is that it’s “insufferably bland.”
My (unscientific) theory is that web sites for many years have primarily used sans serif fonts. As a technical writer in the late 1990s I remember arguments about whether serif fonts were actually easier to read, even on-screen, and in fact the Georgia font (which was considered generally less reliable in Morris’s experiment) was a serif font designed for ease of on-screen reading. Printed material, by contrast, tends to use serif fonts. I think people have been conditioned over the decades to view sans serif as less reliable because it’s associated with the Internet, where you routinely see claims made that are not true. I think this association (print = reliable, web = unreliable) is probably even stronger among Times readers. (And, I will note, the New York Times website uses primarily serif fonts, mimicking the look of its print editions — it even uses them on its mobile site.)
Anyway. The important thing to know here: if you’re trying to persuade people that your facts are right, you should use Baskerville, especially if they’re New York Times readers (or seem like the sort of people who ought to be New York Times readers). If you don’t want to use Baskerville, use some other conservative, old-fashioned serif font with a long pedigree. And for the love of God do not use Comic Sans, but surely you didn’t need me to tell you that, did you?