It’s kind of a thing

June 20, 2002

The amazing thing is that I’ve read more than one good essay about punctuation in my life. I don’t know if the others are on-line (one was in a high school English textbook, if you can believe it), but this one is and it’s worth reading. Informative and witty: a combination I admire.

Of course, I can’t help but think of my own writing style while reading such an essay. I will admit that a giddy thrill accompanies my use of the semicolon, which probably arises from the mistaken impression that I’ve mastered use of an obscure tool, and no one will deny my love of the parenthetical aside. I can get away with that by claiming not to be a serious writer (and I’m not, particularly if you define “serious” as “paid”), but then I have to explain my use of courtesy titles and overcomplicated sentence structure.

Ah, forget it: I’m pretentious. I use my vocabulary and grammatical skills and arcane affectations to make me appear smarter. Or funnier. I forget which. (via Webword) #

Capital idea

Here at ZedneWeb, we (by which we mean “I”) pride ourselves on our unflinching presentation of useless ideas. Where reputable publications might ask “Why bother?”, we boldly forge ahead to tell you, our readers, More Than You Needed to Know.

This time around, we bravely tackle the oft-ignored topic of capitalization. When we write capital and lower-case letters, what are we really doing? What differences exist between “something” and “Something”? How come we have a special word for upper-case letters but not for lower-case letters?

Seriously, though (no, really!), this is actually something I’ve thought about from time to time. It was inspired by the endless conflict between the use of case-sensitive and case-insensitive naming in various bits of computer science. For example, domain names are case-insensitive, so “example.org” and “Example.Org” mean the same thing. URLs, on the other hand, are case-sensitive, so “http://example.org/bob” is different from “http://example.org/BOB”. (The scheme and domain name parts of the URL are defined to be case-insensitive, but not all software is aware of that.)

There are two arguments for case sensitivity. The first is that it’s much easier to implement. To compare two strings, all you need to do is check whether each character is the same. With a case-insensitive comparison, you have to check whether each character is the same or you could change the case of one of them in order to make them the same. With larger character sets like Unicode, this can require looking up values in enormous tables.

The second argument points to existing practices in other fields such as math, where a and A are generally considered to be different variables.

The argument for case-insensitivity is that it’s how people tend to think (outside of mathematical equations). When we’re reading or writing, we don’t ascribe any difference in meaning to “something” or “Something” or “SOMETHING” or even “SoMEtHiNg”. There is information in the use of capital letters, but it is connotative, not denotative, and it differs based on the context. I, for example, interpret text written in all-caps as shouting when it’s on a computer, but I’ll frequently write in all-caps on forms, because it’s easier to read.

In the case-insensitive world, “A” and “a” are two forms of the same letter, not different letters like “A” and “B”. At the same time, the difference between the upper- and lower-case forms is more important than the difference between the roman and italic forms (“A” and “A”) or the various weights (“A”, “A”, etc.).

The question I ask is this: does case apply to individual letters or to the word as a whole? One might assume the former, based on experience with word usage, printing practices, and the operation of the “shift” key on the keyboard, but consider the alternative for a moment. Most words are either written in upper-case, lower-case, or “title-case” (the first letter is upper-case and the remainder are lower-case). Thus: “EXAMPLE”, “example”, and “Example”. Each form is used in different places, but that’s pretty much all you’ll see in normal usage. (Small caps add two more forms, which I’ll approximate thusly: “EXAMPLE” and “EXAMPLE”. These can be considered stylistic variants of the title- and lower-case forms.)

“Example” has seven letters. If each character has case individually, there are 128 ways of writing it, yet only three of them are common.

What about mixed case words? Some names, such as “DePalma” and “McIntosh” are traditionally written with an internal capitalized letter, many programmers use mixed case to write readable variable names, such as “GraphicsWindowContext”, and certain companies squish words together to get their names, as in “FedEx” and “CompuServe”. Do these cases indicate a need for per-character capitalization? Perhaps, but consider the structure of these words. Instead of seeing them as single words with internal capital letters, imagine them as multiple title-cased words glued together. Thus “DePalma” is “De+Palma”, where the plus sign represents an invisible glue character, and the two words “De” and “Palma” are both title-cased.

The remaining examples of mixed-case words are unusual acronyms, such as “LotR”, and cases where capitalization is deliberately made weird, as when transcribing the dialog of Torgo, the lovable caretaker from Manos: The Hands of Fate. These are both clearly stylistic choices. LotR could be, and often is, written LOTR, and the odd capitalization of Torgo’s speech is intended to suggest his unusual manner of speaking.

Whether a word is written in upper-case, lower-case, or title-case depends a lot on the context. However, there are two sorts of words one encounters. Normal words, like “polish”, are written in lower-case except under certain circumstances like the beginning of a sentence. Other words, like “Polish”, are almost never written in lower-case. Each word has an intrinsic capitalization, which is combined with a visual style to create the final presentation. For example:

No styledid I see Donald?
NormalDid I see Donald?
TitleDid I See Donald?
Lower-casedid i see donald?

We could, therefore, replace the current system of having different codes for capital and lower-case letters with a system that marks words as being either normal or important and draws capital or lower-case letters depending on the word’s context and style.

On other hand, being able to do something doesn’t mean we should do it… #

(Hey, when we promise you More Than You Needed to Know, we deliver.)