On the Measure of a Line

Anything but the simplest of documents is an argument about attention.

Jost Hochuli

Ask anyone to improve a document and they will change the font. It is the most visible lever and the least effective one. The reader who abandoned your report on page three did not abandon it because Calibri is thin; they abandoned it because each line ran a hundred and ten characters wide and their eye could not find the start of the next one.

This is not a matter of taste. The return sweep — the movement from the end of one line to the beginning of the next — is a motor task, and like every motor task it has an error rate that rises with distance. Past about seventy-five characters the rate becomes high enough that a reader begins re-reading lines without noticing. What they notice is fatigue, and fatigue gets attributed to the content.

The range, and why it is a range

The conventional figure is forty-five to seventy-five characters for a single column, with sixty-six given as the ideal. The width of the range is the interesting part. It is wide because the constraint is not really about characters at all; it is about the number of fixations the eye makes per line, and that depends on the reader, the difficulty of the prose, and the size of the type relative to the viewing distance.

Which is to say: the number is a heuristic standing in for a measurement nobody takes. That is fine. Heuristics that are right within a factor of 1.5 are worth more than measurements nobody takes.

Typography has one plain duty before it and that is to convey information in writing. No argument or consideration can absolve typography from this duty.

Emil Ruder, Typographie, 1967

Ruder is writing about ornament, but the sentence applies with more force to layout than to decoration. An over-wide column is not an aesthetic misjudgement; it is a failure to convey.

What the measure costs

Fixing the measure on A4 means giving up horizontal space, and the objection is always the same: the page then has a large empty right margin and looks unfinished. Two answers. The first is that a margin is not empty — it is the reason the eye can rest at the end of a line, and every printed book you own has one.

The second answer is better: put something in it. Sidenotes cost nothing to implement, need no page-break arithmetic, and let a reader take an aside without leaving the sentence they are in. A footnote asks the reader to travel to the foot of the page and back. A sidenote asks them to glance right.

The final column is the only one that matters, and it is almost constant. The typefaces differ in every respect except the one under discussion — which is the whole argument, arrived at from the other direction.

An objection

The obvious rejoinder is that nobody reads printed documents any more, and that a PDF is read on a screen at whatever width the window happens to be. This is true of most PDFs and false of the ones that matter. A contract, a letter, a report going to a board, a thesis, an offer of employment: these are printed, or they are read at full page width on a large display, which amounts to the same constraint.

And the documents that are genuinely read in a reflowing window should not have been PDFs. That is a separate essay.

The number is a heuristic standing in for a measurement nobody takes.

Conclusion

Set the column first. Sixty-six characters, or something within ten of it. Then choose leading — roughly 1.4 to 1.5 for a serif at reading size, more as the measure grows. Then, and only then, argue about the typeface, and notice how much less the argument matters than it did before.

References

1. Bringhurst, Robert. The Elements of Typographic Style. 4th ed. Hartley & Marks, 2012.
2. Hochuli, Jost. Detail in Typography. Hyphen Press, 2008.
3. Ruder, Emil. Typographie: A Manual of Design. Niggli, 1967.
4. Tschichold, Jan. The Form of the Book. Hartley & Marks, 1991.