That Slate Article

That Slate article by Farhad Manjoo is insanely popular.  It drives me nuts because it's all lies and distortions.  I didn't want to give his article any more attention than it already has, but apparently that approach was not working.

Here then is a list of many of the factual problems in this article:

“Who says two spaces is wrong?” they wanted to know. ... Typographers, that's who. The people who study and design the typewritten word decided long ago that we should use one space, not two, between sentences.

Not true.  As a group, typographers have come to no such conclusion.  Some people who call themselves typographers make this claim, but many do not.  There's no sort of typographers' convention or union or club where they all get together and decide these things.  There's no document, no vote, no scholarly paper or journal.  There is only Farhad Manjoo telling you so.

James Felici, author of the The Complete Manual of Typography, points out that the early history of type is one of inconsistent spacing.

Let's look at what Mr. Felici actually said in that article he linked to.  For starters he said this: “I am not a type historian”.  OK, so not an authority on the subject.  He also says that the books he can find all seem to use wide sentence spacing.  That hardly sounds inconsistent.

In fact, if Mr. Felici were a type historian (and honestly he is somewhat, just not enough on this particular topic), he'd know that from about 1550 to 1900 was a period of incredible consistency in type.   Almost every book in the english language was printed with an em space between sentences.  And every typography manual in that period that describes sentence spacing says to use an em space (or more accurately an em quad).

Hundreds of years ago some typesetters would end sentences with a double space, others would use a single space, and a few renegades would use three or four spaces.

Mr. Felici didn't say this so I assume that we're back to Mr. Manjoo speaking from his posterior.  In fact as I said, the em space was incredibly standard.  Further, no typographer at that time counted spaces one, two, three.   They measured the distance in ems and used whatever spacing elements (thin space, 5-per-em, 3-per-em, en space, em quad, hair space) were needed to make things fit.  Anyone who knows anything about typography should know this.

Inconsistency reigned in all facets of written communication; there were few conventions regarding spelling, punctuation, character design, and ways to add emphasis to type.

Typographers had plenty of options for adding emphasis to type, and they used them.  In fact one could argue that they had more control over spacing for emphasis than we have today.  As for standards and conventions, some details (but not sentence spacing) were a mess early on.  But by the 18th and 19th centuries, those other details had been ironed out.  Much like today, there were certainly issues that were hotly debated.  Sentence spacing was never one of those issues though, so I'm not sure how any of this is relevant even if it were accurate.

But as typesetting became more widespread, its practitioners began to adopt best practices. Felici writes that typesetters in Europe began to settle on a single space around the early 20th century. America followed soon after.

Since it rests on a lie, this is a distortion.  It's not true that they “began to adopt best practices”.  Best practices changed after being standard for over three hundred years.  If he had said it that way, you might ask the obvious question “what changed?”.  What changed was industrialization, and widespread use of the Linotype, a machine that was severely limited in spacing capabilites compared to hand-set type.

I cover this in several other articles on this blog, but essentially we stopped using wide sentence spacing because it complicated the printing process (see these postings: “A River Runs Through It”, and “The Cost of Printing Errors on the Linotype”).

Every modern typographer agrees on the one-space rule.

This is an outright lie.  For one thing, he immediately mentions that the American Psychology Association thinks that two spaces are needed.  And then twists the story by saying that they prefer one space for published works, although I can find no evidence that this is the case.  He goes on to quote the Chicago Manual of Style.  That's interesting because on their website they say (in the first answer) “Some people, however—my colleagues included—prefer [two spaces]” Of course, Mr. Manjoo does not quote the Modern Language Association's answer, which says “As a practical matter, however, there is nothing wrong with using two spaces after concluding punctuation marks”.

As I said earlier, there's no organization of typographers in which such a thing could be decided.  There's not even a clear definition of “typographer” with many apparently having applied this label to themselves (which may be part of the problem).

Most ordinary people would know the one-space rule, too, if it weren't for a quirk of history. In the middle of the last century, a now-outmoded technology—the manual typewriter—invaded the American workplace. To accommodate that machine's shortcomings, everyone began to type wrong. And even though we no longer use typewriters, we all still type like we do.

(First a nitpick: “began to type wrong”?  Were they typing more correctly before typewriters existed?  There was no such thing as typing.)

There's a significant problem here.  The typewriter was invented in 1873 (ish).  And yet wide sentence spacing was standard practice in English for hundreds of years before that.  And for another 80 years after its invention.  Wide spacing has literally nothing to do with the typewriter.

If you don't believe me, you can just look at almost any old book before the invention of the typewriter.  Or you can listen to Mr. Felici, in the very same article that Mr. Manjoo references above, where he says “the use of double spaces (or other exaggerated spacing) after a period is a typographic convention with roots that far predate the typewriter.”

Here's what actually happened: people used wide sentence spacing on the typewriter, because that's what they'd been seeing in print, and doing in handwriting, for longer than anyone alive could remember.

(By the way, the typewriter does play a real role in this myth.  The two space typewriter habit caused a real problem with later Linotype machines.  I blogged about this in “Two Spaces - an Old Typists' Habit?”.)

Monospaced type gives you text that looks “loose” and uneven; there's a lot of white space between characters and words, so it's more difficult to spot the spaces between sentences immediately. Hence the adoption of the two-space rule—on a typewriter, an extra space after a sentence makes text easier to read.

As I already pointed out, this is not why they adopted the two space rule.  But I did hold back a little bit above, so that I could explain here that not only is he wrong about the adoption of the practice, he (and lots of other self-proclaimed typographers) have the story on the look of monospaced fonts 100% backwards.

Going from proportional to monospaced, most characters get wider.  But none more so than the space and period.  For proportional fonts, one space plus a period would be about the size of the letter “n” or maybe even a little smaller.  In a monospaced font, one space plus a period would be twice the size of the letter “n”.  The only sensible conclusion is that monospaced fonts needed less space than the print industry.

Don't believe me?  Look at history.  Remember when I said they used wide spacing on the typewriter to match print?  They matched it literally.  The print industry's em quad was about three times as large as most word spaces, so on the typewriter initially most people used THREE spaces between sentences.  But because of the monospaced font, this made gaps that appeared much larger than what people were used to seeing in proportional fonts, so eventually people fell back to two spaces.  [Update: 3/20/2014, I added a new posting that goes into more detail on this.]

History is clear that the typewriter demanded less space between sentences, not more.

Besides, the argument in favor of two spaces isn't any less arbitrary.

That's only true if you don't bother to offer the arguments in favor of two spaces.

Here's what two spaces gives us: an unambiguous full stop.  The period means more than just the end of a sentence, it also is used in numbers and initials and abbreviations, and for other reasons.  This can cause confusion, e.g. “Who's going?” “You and I. Smith also.”

Is that two people (You and Isabelle Smith) or three (you, me, and Smith)?

Of course since we read left-to-right, most of the issues of confusion aren't even technically ambiguous.  The grammar is obvious after you stop and analyze it.  But you shouldn't have to stop and analyze sentence structure, you should just read.  Wide sentence spacing avoids such issues.

As far as aesthetics, modern technology could give us a solution.  You could easily have a sentence spacing preference that adjusted the width of visible sentence spacing to the reader's aesthetic choice.  Except for one little catch.  Before a computer could do that (reliably) it'd have to know what a sentence is.  Which would be far easier with two spaces between sentences.  (Note that this blog implements this very feature.  The control is at the head of the right column).

So perhaps the best argument for TWO spaces between sentences is so that people like Mr. Manjoo could reliably and easily avoid looking at them.

Typing two spaces after a period is totally, completely, utterly, and inarguably wrong.

That's how his article opens.  If every single factual argument he offers is in fact not true, can we still believe his thesis?

Two Spaces - an Old Typists' Habit?

I've already thoroughly debunked the myth that wide sentence spacing comes from the typewriter.  [And even more thoroughly, in a more recent posting.]  But many myths contain a grain of truth.  One might guess that the typewriter was blamed for wide sentence spacing simply because the "habit" survived longer on the typewriter than it did in the print industry.  It turns out there's more to it than that.

If you don't want to read some of my other blog postings, let me bring you up to speed.  First came very wide sentence spacing, at least two or three hundred years of it, in almost every English language publication (and even hand-written letters) of the day.  Then, at almost the same time in history two machines were invented: the Linotype machine which revolutionized newspaper, magazine, and book printing; and the typewriter, which revolutionized professional correspondence and record keeping.  Initially both technologies continued to use wide sentence spacing.  But in time, various technical and cost-related issues with the Linotype lead to increasingly narrower sentence spacing.  By 1950 most professional publications no longer used extra space for sentences, but most typewritten communications still did.

That brings you up to speed on how we lost the extra space between sentences.  But it doesn't quite explain why there is so much vitriol now against extra space, particularly from editors who lead the charge in telling us that using two spaces is nothing but an old typists habit.

The Linotype saved time and money for a number of different reasons.  One of the more significant reasons was time saved in justifying text.  The Linotype introduced a little piece of hardware called a spaceband, a variable space that could be used between words.  As a line of text was being set, the Linotype operator could with a single lever resize all of the spacebands in the line to get automated and even justification.

Simulation of the operation of spacebands on the Linotype.  (In actual operation, this was all upside down.)

Initially, wide sentence spacing was maintained by adding an additional fixed space (usually an en space) next to the spaceband for sentences.  From a modern point of view one might wonder why they didn't simply "use two spaces", that is two spacebands between sentences.  This could have been because you'd end up with too much expansion between sentences, compared to the expansion between words, and it would have looked bad.  But there was actually a much more important reason.

You weren't allowed to put two spacebands next to each other.

It was physically possible.  But as any well-trained Linotype operator of that era new, this mistake could at best end up with a bad linecasting, with extra fins of metal sticking out where the two spacebands touched.  You might also end up squirting out a bit of hot metal during the casting.  And perhaps worst of all, it was possible to damage the spacebands by putting them next to each other.

This in itself almost seems like the smoking gun.  But Linotype operators were typesetters, not typists.  They still didn't have the "one space or two" mindset of a typist, but rather "how much space".  The Linotype had several different fixed spacing elements that could be combined much as with traditional hand-assembled type, and the typesetter would never place two spacebands together when more space was needed.  If wide sentence spacing was called for, an experienced operator would have no confusion with "two spaces" or putting two spacebands together.

The Linotype used a very different keyboard from a typewriter.  The metal lever on the left was used to place spacebands in the line of text.

This was the case for the first seventy years or so of Linotype operation.  But in the 1950s and 1960s another technology began to become popular, the teletypesetter perforator.  This was not exactly new technology, as the Monotype used something similar since the late 1800s.  But it was relatively new to the Linotype.  And it was introduced to set type faster, and more cheaply.

The teletypesetter separated the typesetting task from the actual running of the Linotype, by storing the typesetting information on perforated tape.  Another machine could then be attached to the Linotype, and replay these typesetting tapes to produce the castings of the lines of text.

One reason this saved time and money was that the Linotype wasn't limited by the speed of the operator.  One Linotype could be run continuously, setting the instructions produced by two or three operators simultaneously.  But another big advantage was that teletypesetters eventually replaced the traditional Linotype keyboard with a modified typewriter keyboard.  The intent was that typists, with a little training, could be used to set type, further saving money by using relatively unskilled labor.

A teletypesetter perforator, with a typewriter keyboard.  The gauges for justification are visible in the upper right.

As the typist typed, the teletypestter punched codes into a paper tape to identify the letters being entered.  When the typist pressed the spacebar, an instruction was added to the tape to use a spaceband.  The teletypesetter also had two gauges to guide the typist in basic typesetting.  They showed the operator the minimum and maximum width of the line, assuming the spacebands in the line were at their minimum or maximum widths.  Once a line of text was long enough for what ever width column was needed, the typist could tell from the gauges.  They'd press the carriage return, the appropriate instructions for a new line would go onto the perforated tape, and the gauges would reset to zero.

Of course, the problem with spacebands still existed, and we now see a very real problem with the typists' habit of pressing the spacebar twice between sentences.  If the typist made this mistake, at best, the error would have to be found and the line of text retyped.  At worst, the mistake might result in damage and repairs and downtime on the Linotype machine.  One can imagine an editor's blood boiling the umpteenth time they had to hear about a delay on the Linotype because a teletypesetter operator had made the dreaded mistake of pressing the spacebar twice.

“By far the biggest item of cost of operation is damage to matrices and spacebands, which to a very great extent is traceable to ignorance on the part of perforator operators of the limitations and performance of the Linotype.” –British Provincial Press Productivity Team, 1952

It's my belief that this is the real source of the myth of the typists' habit.  The two-space habit was a very real problem on a teletypesetter.  It also explains the editors' hatred of the habit, a hatred so strong and deep that it would be passed down to editors today.

The next step in technology was the introduction of computerized justification.  It was no longer necessary for the teletypesetter operators to justify the text.  They could make paper tape instructions with nothing but the plain text.  That tape was fed into a computer reader, which would calculate the justification and line breaks, and produce a new tape with all the needed Linotype instructions.  The software would never place two spacebands together.  This is the source behind the claim that this software would eliminate runs of multiple spaces.  This is true only where spaces really means spacebands.  Other spacing elements could still be combined for additional space, although nearly a century of cost-cutting had mostly eliminated the practice.

This became particularly ironic when the teletypesetter and justification software were transitioned from the Linotype to phototypesetting.  Phototypesetting equipment still had a variable space, and they still called it a spaceband, but the actual spaceband hardware itself no longer existed.  At this point the space and the spaceband were pretty much the same thing.  Even though the original technological reason for eliminating multiple spacebands had disappeared completely, the solution to the problem remained in place.

Modern software still has a notion of a variable space, although nobody calls them spacebands.  Software can also solve many of the other problems that lead to elimination of wide sentence spacing.  Rivers of whitespace can easily be avoided by spacing between letters, and adjusting where words wrap and break.  Yet despite all the power and possibilities of modern technology we are still overrun by worshippers of the Linotype voodoo cult, yelling at us about extra spaces for reasons they don't even remember.

I suppose eliminating spaces has become just another old habit that's too hard to break.