To double space or not? Farhad Manjoo at Slate is against it, most vehemently, and Norm Geras agrees. Moe Lane means to go on, as does Meghan McArdle.

Hmph. It’s just Yet Another Example of bastardizing an originally flawed, but essentially decent, system in the name of Art.

Double spacing originated in the days of typewriters as a convenience for typesetters. The period (“full stop” in BritSpeak) is such an insignificant little dot that it can easily be missed in a sea of monospaced Courier font, especially when (as usual with older typewriters and/or careless typists) the page also had a goodly number of spots and blemishes created by bits of congealed ribbon ink flung up by the flying typebars. Double spacing provided a bit of emphasis, allowing the poor fellow rowing the Linotype machine to detect the ends and beginnings of sentences without stopping to inspect the manuscript with a magnifier. Properly typeset print, even today, puts an em space between sentences and en spaces (smaller) between words, then adjusts their sizes proportionately as part of justifying the line.

Fast forward to Dr. Berners-Lee, whose original specifications for presentation of material on the World Wide Web were based on his awareness that wildly varying displays were in use. His intent was not that it be pretty, but that it be more or less equally comprehensible on almost any screen. One of the resulting decrees is a property of HTML that has frustrated people who want to make things look nice ever since: All white space is presented as a single space, except paragraph breaks, which are two lines.

What, the SOB couldn’t even define a <tab> for paragraph indent? Sadly, no. The good Doctor was a geek, and geeks are always prepared to abandon common practice and substitute their own judgement. Printers, beginning with Gutenberg and continuing since, had worked out the system: en spaces for words, em spaces for sentences, indents to flag paragraphs, a nice (as in “carefully considered”) adaptation of best practice in written manuscripts, and a nice (as in “it works well”) balance between readability and a certain pleasing symmetry. Berners-Lee made many decisions that seem to me shortsighted or simply wrong — it still pisses me off that he didn’t include a simple line-drawing facility, making the simplest two-dot scattergram into multiple KB (or MB!) of bandwidth-hogging image — but the clunky, blocky paragraphing is arguably the ugliest.

It was most problematic to graphic designers, whose attempt to migrate from the wide open spaces of magazine and advertising layout to the more-confining practices and presentations of the Internet was frustrated by the limitations of HTML. Some of the features and facilities could be perverted into forcing the production of attractive layouts; wide (mis-)use of the <frame> and <table> tags and copious instances of &nbsp; in the source could, and did, approximate the flexibility they were used to, but many of the remnant features, particularly Mosaic’s stubborn insistence on re-scaling and re-arranging things to fit different screens, frustrated their attempts to replicate their print-based efforts. Dammit, that’s hideous! It doesn’t look anything like my layout! Well, no, it doesn’t — but a monitor screen is not a magazine page. It might be argued — and I did, elsewhen — that the designers should have made more attempts to conform to the medium rather than trying to force new pixels into old paradigms, but if there is a more stubborn class of people than artists in full musth I have not yet encountered them.

The solution (?) was technology, in the shape of wide penetration of many-pixeled monitors, and Cascading Style Sheets. CSS allows close control of layout and design, at the expense of greater bandwidth. Web pages can now have multiple columns, designers can now specify typefaces and graphics and expect them to be presented properly, and (best of all) the classic system of paragraphing, complete with indents and single line spacing, is possible. CSS permits its own perversions, of course; there are a number of Web pages and blogs I no longer visit because careless or pugnaciously specified page widths or complex layouts “break” my screen. In the end, though, and modulo bandwidth availability for the remaining poor souls on dialup, CSS has been a worthwhile advance.

One of the features of CSS is abandonment — finally! — of Berners-Lee’s diktat that all whitespace reduces to a single space. People whose aesthetic, developed over decades of typewriter-pounding, demands two spaces after the full stop can now have their preferences honored. To my eye the result isn’t attractive when incorporated into fully-justified text, because when the spaces are stretched to make the line fit the gap is larger than results from the em space between sentences in book typography. Of course, if I think graphic designers should have accommodated themselves to video screens rather than forcing the result, I certainly have no standing to demand that the conventions of print typography should rule. If that’s what Meghan, Moe, and others want, they can and should get it.

Modest proposal: treat the system that converts what the author typed as a siliconized implementation of the Linotype operator — which it is — and what the author types as “manuscript format”, which it is. CSS should recognize the sequence period-space-space as a sentence end, and substitute period-em space. That’s what the typesetter did in days of yore, and to my mind the result is prettier. Authors whose thumbs hit the space bar twice after a period by reflex are accommodated, Gutenberg’s decision that sentences need to be set off with a bit of extra space is likewise adhered to, and nobody’s cold dead hands need be pried off the long thin bit at the bottom of the keyboard. How about it, W3C?