Ruminations on type

Fri 02 Nov 2007 12:41 PM

Carefully deployed fonts and typefaces can add clarity and precision to a manuscript, but it makes is unclear what to do when presenting the same material in lecture.

In forall x, I differentiate bits of the object language from metavariables by writing the former in roman letters and the latter in script letters. As it's typeset in the book, the 'x' in the title of the book is a metavariable. This week in Intro Logic I lectured on the definitions of satisfaction and truth. Although I explain the difference between object language and metalanguage elements, I have a hard time differentiating my letters when I write on the chalk board. My block As are distinct from my script As, but other letters tend to be ambiguous. Of course, there is no difference at all when talk out loud about As and Bs.

In a recent paper, I distinguish objects from concepts by writing the former normally and the latter in all caps. The title ('What SPECIES can teach us about THEORY') plays on this convention. Tomorrow I am going to present the paper at the Creighton Club, and there is really no way to mark this distinction in speech.

These conventions are fairly standard, and I used them because I think they help to clarify somewhat subtle differences. They work because they are not really essential. If someone reads forall x and skips the explanation of why some expressions are in Chancery and others are in Computer Modern, I don't think the difference is distracting. It underscores a distinction, but it can be transparent to readers who are attending to the content.

Contrivances like these become problematic when they are distracting or when they are the only indication of crucial distinctions. To take a non-scholarly example, many webcomics typeset dialogue in different fonts for different characters. Some authors pick fonts that are more decorative then readable. Some use it as an excuse not to connect speech bubbles clearly to speakers. If a reader can only figure out the dialogue by looking at the font, then the typsetting is no longer an effective way of emphasizing certain information; it has become the sole means of conveying that information.

Similar points might be made about italics in scholarly papers. They can be effective for emphasis, but they are not substitute for actually saying something. Depending on the typesetting and the quality of the copy, a reader might not even notice careful and clever italics. Continental italics, obtained by extended letter s p a c i n g, are especially easy to miss.