Rutgers - The State University of New Jersey
Charles Hedrick's Typography Pages
 

Fine Points of Typographical Usage

This page discussed certain fine points of typesetting. They are questions where I haven't seen any published rules. I've had to make decisions myself. I thought I should document what they are and why I've made them. If anyone knows of authoritative positions on these issues, please contact hedrick@nbcs.rutgers.edu.

Many of the issues discussed here would be settled by a "house style". I indicate my practice below. When combined with the main document, that defines my house style.

Text Figures

Use of text figures is a decision for the designer. It depends partly upon the font. In fonts with a small x height, you really want to use text figures. In fonts with large x height, you may choose not to bother. However my impression is that good typographers normally use text figures where they are available. I've seen a number of documents that use text figures but not small caps. I've never seen the reverse. (My practice is to use text figures where they are available, and to avoid fonts that don't have them.)

1 can sometimes be ambiguous. In certain fonts, the text 1 looks almost exactly like a small caps I. This can lead to problems. E.g. in one document Internet 2 was abbreviated I2. I was using small caps in this document for all acronyms. Thus the I was set in small caps and the 2 as a text figure. The result was almost indistinguishable from 12.

In documents where this occurs a lot, it is best to pick a font where the 1 looks different from the small caps I. In DTL Documenta and many other neohumanist fonts this is true. However in historical fonts the 1 looks like I. In such I font I would probably capitalize the I in I2, to avoid ambiguity. (My practice is to use fonts with a distinct text 1 for documents where this is an issue. For me that means DTL Documenta or a version of Janson Text where I have modified the design of the 1.)

In my paper on typography, I echo Bringhurst's characterization of text figures as "lowercase". I've seen some people take this farther than I would. What happens if you have to begin a sentence with a number? Do you "capitalize" the first letter by using a lining figure? NO! That's taking things too literally. If you use text figures, use them everywhere, except in contexts (e.g. titles) that are all caps. I think it's OK to begin a sentence with a text figure, though you may prefer to avoid the situation.

One solution is to avoid the situation. That is, try to reword sentences so that they don't begin with a number. But that's not always reasonable. Another solution is to spell out the number, e.g. ten rather than 10. Then you can capitalize it. I saw one example where Bringhurst started a sentence with a list of numbers: "Twenty-point, 22-point ..." Note that he spelled out the first number. However others think that such an inconsistency is a bad idea. I'm inclined to agree with this. I would almost certainly have used figures for the whole list. (A list of spelled-out numbers is harder to read, in my opinion.) My practice: If it's easy, word sentences so they don't start with numbers. But it's not a big deal.

Some authorities recommend that you spell out all numbers that aren't too big. In the Penguin house style, Tschichold said to spell out all numbers below 100. Others have said 20 or below. But Tschichold points out that there are exceptions, such as lists of numbers or other situations where the emphasis is on the numerical value (not his wording).

I haven't adopted a specific rule here, but I think my instinctive practice is basically Tschichold's, except that I normally write out only one-word numbers, i.e. numbers up to 20 and multiples of 10. I normally write out simple counts, e.g. "there are three different approaches to this problem". But I'm much less likely to write out meaurements or anything else that emphasizes the number, and I certainly wouldn't write out lists of numbers.

At any rate, if you need to start a sentence with 1776, I wouldn't spell out "seventeen hundred, seventy six" just to avoid questions about how to capitalize the figures. Readability outweighs rules. I see no problem beginning a sentence with a text figure.

As you'll see below, I believe the rule about capitalizing the first word only applies to normal English words. It doesn't apply to numbers, acronyms, logotypes, etc.

Small caps

There are several interesting issues with small caps.

Use of small caps is a decision for the designer. It depends partly upon the font. In fonts with a small x height, you really want to use small caps. Full caps are just too obtrusive. In fonts with large x height, you may choose not to. Typographers like small caps. But outside of books about typography, I seldom see small caps used in the text. It's more common for titles. I always use them where they are available. But competent designers may choose not to do so, at least when working with fonts that have a large x height. My practice is to use small caps consistently, and to choose fonts that have them available.

Now we come to the question of when to use them. Bringhurst's rules are sort of interesting. He starts by saying use them for all abbreviations and acronyms except (1) personal names, and (2) 2-letter geographical abbreviations. But he then notes that many people use small caps for 2-letter postal codes, and that in a document where there are lots of acronyms and abbreviations near each other, you're best to be consistent, and not treat 2-letter ones differently.

I conclude that in documents with lots of abbreviations and acronyms, these rules reduce to using small caps all the time, except for personal names.

Not everyone use these rules. Some publications appear to use small caps only for fairly long abbreviations, such as 5 character or longer, or 3 characters or longer. This (like other questions discussed in this document) is commonly defined by a "house style."

I typically deal with technical documents that have lots of acronyms. Thus I normally use small caps for all abbreviations and acronyms, no matter what their length is. I guess I'd use full caps if they represent personal names, though I don't think that has come up. (I assume by this Bringhurst means initials.)

Many fonts have normal small caps but not italic small caps. So what do you do if you have something in italics, and an acronym occurs in the middle, e.g. "I like RUCS." If I don't have italic small caps, I'd use normal small caps, "I like RUCS," rather than leaving it in uppercase, "I like RUCS." While Bringhurst doesn't give a rule, there are several examples in "Elements of Typographical Style" where he does this.

What do you do if a sentence starts with an acronym, e.g. "RUCS is great." If you're using small caps, I would set RUCS in small caps. Some people are bothered because this seems to violate the rule about capitalizing the first word.

To understand my answer, you need to realize that small caps are used in two completely different ways.

Let's go back to the question of a sentence that starts with an acronym. The problem with using a full cap for the first letter is that by mixing case, it makes it look like the small caps are being used as lowercase. E.g. in "RUCS" if you put R in full caps and the rest in small caps, RUCS, it looks like a word or name Rucs, not an acronymn. So my rule is:

If small caps are being used as capitals (e.g. in acronyms or abbreviations), don't change them when you put them at the beginning of a sentence.

If small caps are being used as lowercase, then use normal rules for capitalization. Turn the first letter into full caps at the start of a sentence or elsewhere that capitalization is expected.

Oddly spelled things

More and more companies are using special spelling. What do you do if a sentence starts with iDrive.com? Do you capitalize the i? I say you do not. My basic rule is that you never want to do anything that would confuse the reader. Apparently iDrive.com cares how their name is capitalized. If you turn it into IDrive.com, the reader may be confused about proper capitalization.

I claim that you only capitalize the first letter of a sentence when it starts with a regular English word. For regular words, the reader knows the spelling. He will not be confused when you turn the first letter into a capital letter, because he knows how the word normally goes. But for anything abnormal, the reader won't know the proper capitalization. So if you change the first letter, you're going to confuse him. Hence iDrive.com, www.rutgers.edu, or http://www.rutgers.edu should be left as is if they start a sentence.

In fact I argue that things like iDrive are no longer words. They're closer to being graphics. As such they are not subject to the normal rules that apply to words.

This brings up another issue: Some people suggest letterspacing things that aren't real words. The claim is that with a URL or a hostname, people don't read them like a word. Rather, they have to look at and remember the exact letters. Letterspacing makes it easier to see all the individual letters. If you take this approach, you'll probably letterspace the same things that you wouldn't capitalize at the beginning of a sentence. I would also consider setting items like this in a contrasting font, or possibly bold. I typically set hostnames and URL's in a sans font. This emphasizes the fact that they aren't normal text.

One final problem with these odd objects: How do you use small caps? In one document the term "RUNet 2000" appeared throughout. I used small caps for other acronyms. After trying various possibilities, I decided to leave RUNet alone, i.e. to leave it as full caps and lowercase. Turning the RUN into small caps, RUNet 2000, produced a *very* odd look.

Dashes

In my paper, I repeat Bringhurst's position. When you are separating phrases -- like this -- use either an en dash with normal word spaces around it or an em dash set flush with the words around it. He, and I, prefer the en dash. (I use -- to denote an en dash when typing in ASCII or other contexts that don't have it. -- would never be permitted in a printed document.)

Some people have strong preferences here. However I looked through my books on typography and found just about an even mix.

The goal is to produce a visible separation, but one that doesn't disrupt the even appearance of the page. The choice may well depend upon the particular font, although it would normally be defined by the "house style."

BACK TO TOP

For more information, contact hedrick@rutgers.edu
Last updated: Wednesday, 03-Dec-2003 01:07:20 EST
© 2003 Charles Hedrick. All rights reserved.

 

Search Rutgers