Last month, in this piece about “Stig of the Dump” and typefaces, I mentioned how, many years ago, I used to deliver training courses on desktop publishing applications. I enjoyed teaching people the differences between typefaces, “beyond the basics of why body text is usually set in a serif, rather than sans serif, font”. [If you need more clarification about what a serif is, and what it does, see this later piece, “More about serifs”.] Yesterday I started reading Susan Hill’s “Howards End is on the Landing”, in which she embarks (to quote from the back-of-book blurb) “on a year-long voyage through her books, in order to get to know her own collection again”.

On page 9 of my Profile Books paperback copy (“Typeset in Transitional by MacGuru Ltd”) she discusses fonts and mentions her preference for Garamond over Times Roman. She writes:

“But whatever the font, it must have a serif or I cannot read much of it. There is a long modern novel on one of the shelves of the Small Dark Den that I certainly should read. It came to me with recommendations from all sides, but I can’t read it because it is printed in a sans serif font, Arial probably, and I simply cannot force my eyes to take it in for more than a few lines. Why should that be? Do others agree with me? Do publishers think about these things? This one does.”

I wonder how many people answered her questions.

The answer to the question “Why should that be?” (regarding her inability to read more than a few lines of text in a sans serif font) is one that I discussed hundreds of times in training rooms, mostly in the 1990s. The simple reason is that the serifs help to draw your eye along the line, almost subliminally. They make the text more readable. Without the serifs to guide you your eyes are working harder to process each string of letters. Each letter in a sans serif font stands out more. That’s why sans serif fonts are most commonly used for headings and not for body text. The only novel printed without serifs that comes to mind (and which I have managed to read all the way through) is “The Curious Incident of the Dog in the Night Time”. I figured that the difficulty imposed on the reader by the font helps you to appreciate some of the dislocation and difficulties encountered by the book’s narrator Christopher, who has autism. For most people this would be subconscious or subliminal but if you take an interest in fonts you’ll notice it consciously, and that might also be enough to put you off reading the book.

In last month’s piece I also mentioned going through my bookshelves when I first learnt about fonts, checking the typesetting information (where it was given) on each book’s copyright page. The first book that I found that was printed in a sans serif font was “Ten years in an open-necked shirt” by John Cooper Clarke, but that’s a collection of poems, not a novel (and the typeface is rather too small for me these days). There’s a John Irving book in which whole chapters are printed in a sans serif font, to show that they’re excerpts from a work by one of the characters, rather than part of the main narrative. It’s either “The world according to Garp”, “The Hotel New Hampshire” or “A Prayer for Owen Meany”. I know that all of these Irving books, which I read in the 1990s, are still in the storage box I put them in sometime in 2012 and I won’t dig them out just yet to check. I do recall, though, that all of the words spoken by Owen Meany are in capitals, which is always less readable than sentence case.

This brings me back to readability. As I mentioned above, serifs make text more readable. So does black text on a white background (i.e. how all books are printed). So does the right balance between the lines of text and the space between the lines (the “leading”, which rhymes with bedding rather than weeding); paragraphs of double-spaced text are harder to read than paragraphs of single-spaced text. The usual combination is 20-25% extra leading, so 12-point text would be set with an extra 3 points of leading (and this would appear on the copyright pages as “12 / 15 pt”). The 2014 reissue of “Stig of the Dump” is “Set in 13.5 / 20.5 pt Sabon MT Std”, which is a little more space than I am used to, but it is much more readable than the 1971 edition that we started out trying to read.

The size of the font and the number of characters per line are also key considerations for readability. As a rule more than 80-90 characters per line will make text difficult to read. Once you get past 100 characters per line you are, effectively, reading in CinemaScope, moving your head from left to right to see the whole line. If you have ever had to read whole paragraphs of text printed in landscape rather than portrait orientation you’ll know what I mean. In such cases the text should have been printed in multiple columns, and then the maximum characters per line should be around 40 (“an alphabet and a half”).

If you’re a Londoner you might recall the London listings magazine City Limits, launched as an alternative to Time Out in the 1980s. I was loyal to it for years, until its inaccurate cinema listings sent me to two non-existent screenings in the same week. It was also at times nearly impossible to read. I recall one multi-paragraph article which I really wanted to read but couldn’t finish. It was printed in white text on a black background in capital letters. It bugged me that I couldn’t get through it but if I’d known then what I learnt in the 1990s I would have known that the fault lay with the typesetter and not with me.

I’ll finish with a cautionary note from my experiences in the training room. For years I had confidently announced that serifs draw your eye across the line and make text more readable, and most delegates accepted this. One afternoon, training a room full of scientists, I was challenged about my claims. Where was my evidence? And how much more readable is text with serifs compared to text without serifs? I had no evidence other than what I had read in printing and publishing guides, the “received wisdom” of generations of typesetters and publishers. And I had no quantifiable data about readability. I still don’t. At the time we had limited access to the newly emerging World Wide Web. There might well be something there to answer these questions but I haven’t found it yet. If I find anything authoritative I’ll post a link here, especially if it says something straightforward like “serif typefaces make text 25% more readable than sans serif typefaces, where all other aspects of the font are identical”. But if you share Susan Hill’s views you’ll know that serif typefaces are 100% more readable than sans serif typefaces.