Here, as I promised in the debate on serifs, are some comments on Gerard Unger’s book While You’re Reading.
First impressions: as you would expect from a type designer, it is beautifully designed. It is set using two of Unger’s fonts, Capitolium and BigVesta; nice big typesize, very clear, not too many characters in a line. But I found it hard to read, and it took me a little while to figure out why.
If you read a book on trees, or Renaissance painting, you would expect illustrations, which would add to your understanding of the text. What kind of illustrations would you expect in a book on typography? That’s right, examples of typefaces. The best such examples are samples of text, of course. But, unlike trees or Renaissance paintings, typeset text competes for the attention of your reading brain with the book you are reading. A couple of times I turned a page and continued reading the verso page, and it was a few moments before I realised it was unconnected with what went before. I suppose this is an occupational hazard of a book about typography; but something boring like putting the examples in frames might have helped.
This illustrates a point Unger makes early on: the permanence of letterforms, and the very limited scope that a type designer has for change. He gives a sample of type from Nicolas Jensen in Venice more than five hundred years ago; it is legible without any strain.
Anyway, I was reading it partly to see if it shed light on the readability of sans serif typefaces. So first, definitions.
My copy of Chambers’ dictionary scarcely distinguishes between the two terms:
- clear enough to be deciphered; easy to read; readable (rare).
- legible; easy to read; interesting without being of highest quality.
However, Unger, quoting Walter Tracy, says,
Whereas Dutch has only one word (leesbaarheid), English has two: legibility and readability. Legibility refers to the ease with which letters can be distinguished from one another: whether, for example, there is a sufficient difference between a capital I and a lower-case l. According to Tracy, readability is a broader term that refers to comfort: if you can read a newspaper for a long stretch at a time, it is readable. At the same time, both these levels can be regarded together as being part of legibility, and readability then refers to the way the writer uses his language and makes his text easy to follow and understand.
Unger quotes older typographers such as Jan Tschichold and Stanley Morison who laid down rules based on their own good sense rather than (non-existent) research. These two felt that the best body type doesn’t draw attention to itself (though Tschichold had earlier taken a different view which he later repudiated). Indeed Morison is best known for the widely-used Times New Roman.
But what does research show? Unger describes in detail the physical aspects of reading: the way we become lost in a book, the jumping eye movements (saccades), the look-ahead preprocessing. But not much about what goes on in the brain, because not much is known.
There is also remarkably little on the question of readability of different faces. What little research that has been done shows no clear differences, and also is vague about the difference between reading from a page or a screen, or different kinds of reading (books, newspapers, road signs, telephone directories, etc.) He says,
The question of which basic kind of type is the more legible has never been definitively settled and continues to exercise minds to this day. At the same time it is a debate in which the waters have been muddied because it has been conducted more often with emotional than with rational arguments.
The one thing that does seem to have an effect is that people read more effectively a typeface that they are used to reading.
For example, he tells the story Daniele mentioned, of the Dutch newspaper that changed from serif to sans serif:
In 1986, Trouw switched from Ionic (1926) to Frutiger (1976), making it a beautiful but rare phenomenon: a newspaper set entirely in sans serif type. Twelve years later the constant stream of complaints could no longer be ignored: its typeface was costing the paper readers. Trouw‘s designer, Erik Terlouw, decided to make another switch, this time to Swift (1985) – a seriffed type [designed by Unger] – and readership rose again. However, this rise cannot be attributed solely to the change in typeface, since at the same time changes were made to the layout, and the paper’s editorial direction changed …
How hard it is to run a controlled experiment!
But since most of the body text we read is in seriffed fonts, it is surely best to stick to these in our own writing. Beside me on the sofa as I write is Metro, one of the most typographically diverse newspapers. It is hard to discern a rule, but it seems that they use sans serif (in various weights) for headlines and short (one-paragraph) news items, and serifs for more substantial articles.
Let me finish with a couple of gems.
In a short text, combining text and images by using letters made up of animal shapes is not a problem, but try it with a long text and you soon find it doesn’t work: you keep trying to see the images as well as read the text.
… Kurt Schwitters, presenting one of his “theses of typography” in issue 11 of the magazine Merz (1924): “Never do things the way someone else did them before you. You could also say: always do things differently from the way others do them.” For practical purposes you could hardly have a more dogmatic rejection of all dogma.