The practice originally came from typesetting. If anything, it's more relevant and appropriate now than it was in the days when your font choices were limited to 12-point Pica and Elite.
That's not true. Go look at the us constitution, which is not typeset. There are bigger spaces after periods, see article I section 2, lines 3 and 4 for two examples.
That's a weird non-sequitur that has nothing to do with the topic, considering that the Constitution was handwritten, but it's still interesting to read about how things were done in that era:
> The practice [of two spaces after the period] originally came from typesetting
I'm just challenging that assertion, which is commonplace, but there's plenty of evidence to the contrary, many of which you can see with your own eyes.
Literally centuries of typesetters and
printers believed that a wider space was
necessary after a period, particularly
in the English-speaking world. It was
the standard since at least the time
that William Caslon created the first
English typeface in the early 1700s (and
part of a tradition that went back further),
and it was not seriously questioned among
English or American typesetters until the
1920s or so.
Extra space after each sentence was and is a good idea, whether you're talking about handwritten documents or typeset ones. Typewritten documents arguably benefit the least from the practice. Point being, when single-space advocates say that double-spacing is obsolete and unnecessary because we don't use typewriters anymore, they are just plain wrong.
Programs that do typesetting for you (Markdown, LaTeX), will ignore the number of spaces after the period in your "source" code and lay it out in the output with a single space.
I've gotten in the habit of single-space in variable-width fonts, and double-space in fixed-width fonts.
It tries to guess which periods indicate end-of-sentence (based on whether the next letter is capitalized) and will add additional horizontal space accordingly, unless you tell it not to by declaring your document uses “French Spacing”.
(Edit: and my memory may be serving me incorrectly here, but I recall 20 years ago having a support package that also detected double-space in the source as another indicator of end-of-sentence, but that was probably a non standard extension.)
> Splitting text into sentences is actually really hard.
It would be a lot less hard if we were able to unambiguously tell the computer whether a period ends a sentence. I wonder how we could possibly do that...
(Less sarcastically: a lot of Unix programs - Emacs, vim, troff, etc. - solve this problem by defaulting to double-spaced sentences, thus resolving the ambiguity. Kinda odd that this ain't the Tex approach; there's already precedent by using doubled-up ``backquotes'' and ``single quotes'' to produce “smart double quotes”, and lone `backquotes' and `single quotes' similarly become ‘smart single quotes’.)
I remember reading the TeXbook and learning that the space after a period is wider if there's a capital afterwards and the various ways of suppressing this behavior.
Thanks. I appreciate the attention to detail! FWIW, the distinction between the two is blurry given that most people I’ve known in academia never use TeX directly other then through LaTeX.
This is doubly nonsense. TeX works very hard to make good looking spacing. Markdown emits HTML, which has a completely separate rendering engine for spacing. And the only reason HTML looks wrong is that the programmers of rendering engines weren’t as good as Knuth, and van Rossum’s anti-intellectualism lobbied against semantically rich markup for sentences.
