(A great place to start if you're new here is my Wide Spacing Roadmap.)

Thursday, December 13, 2012

A River Runs Through It

[One of my primary goals here is to develop a set of articles that explains the real history of how we lost wide sentence spacing.  Eventually I'll lay the entire case out in some permanent documents, but in the mean time I'll be making little bits of my case in individual blog postings.  This is just one piece of that case, the first of many to come.]

One of the significant technological factors that influenced the aesthetics of sentence spacing was the invention of the line-casting machine in the 1880s, and in particular the Mergenthaler Company's Linotype machine.  With only a little training, an operator of the Linotype could be four or more times faster at setting type than the best traditional compositors doing it by hand.  In only a few years, the Linotype began reshaping the industry.

One of the Linotypes' advantages was that lines could be automatically justified.  This was accomplished because the standard word space was not simply a blank spot in the mold for the line.  It was a wedge called a "spaceband" that could be used to push words apart as needed.  So if all the spaces in a line were set with spacebands, then as the line became near full, the machine would spread the spacebands out until the line was justified.

But there was an aesthetic downside to this massive increase in productivity.  The Linotype had a very bad habit of creating "rivers" in the output, especially in narrower columns of text.

A river is a big area of white space that draws the eye.  It's caused when spaces on several lines in a row all line up, and the larger the spaces are the worse the problem will be.  It's considered very poor form in typography to create these rivers.  Traditional compositors could alleviate rivers by adding space between letters, which would would keep the word spaces from becoming overly large, and would decrease the chance that they would line up with spaces in lines above or below.  They could even add more space between letters in one part of the line so that the spaces in adjacent lines were pushed  away from each other.

The Linotype however could not do this automatically.  It was only possible to add spaces between letters by hand.  That by itself would slow things down to lose any advantage of the Linotype, but of course it would actually be worse than this.  A linotype compositor could only set one line at a time, while a traditional compositor using a composing stick could set several lines at once.  In setting type by hand, rivers could be eliminated before proofing, but on the Linotype, you'd have no idea how the rivers were forming until after you'd already cast the line, and at least one, if not several more.  Letter spacing on the Linotype would have been slower than setting type by hand, and would not have been worth it.
Simulations of  Linotype typesetting methods on the left and traditional methods on the right.  The rivers on the left are much more apparent, even though the text and line breaks are identical.

The solution initially was just to live with the rivers.  Over time, narrower fonts were favored.  Not only did they save paper, they put more words (and hence more spaces) into each line, so the spaces wouldn't have to expand as much for justification.  As narrower fonts became more popular, narrow word-spacing also became more popular both to match the narrow fonts, and to reduce the rivers.  The Mergenthaler corporation pushed these changes forward as aesthetic improvements in general, but it seems to me the primary goal would be to improve the aesthetics of Linotype's automated output.

Even with these changes, the Linotype still created rivers and holes with much greater frequency than handset type.  The next thing to go was wide sentence spacing, and again, it was Mergenthaler who was leading the charge for this new trend.

I'd like to do a bit more research into the history of "rivers".  There's almost no references to rivers in the 19th century (but I do see references to "pigeon-holes" going back as far as 1824 so far).  On the other hand, in one reference in 1895, William Morris notes the excessive appearance of rivers as a reason why he founded his press (in 1891).  It's not clear if the Linotype was common enough at that point to have this kind of influence, although it's possible.  The first commercial machine was in use in 1886, and there were 1100 machines in operation in the United States by 1894.  It's also just as possible the he was simply talking about sloppy compositing in general.  It seems likely that the same financial pressures that made the Linotype a success also put pressure on traditional compositors to cut corners.  And as with every profession, there are always those that just suck at their jobs.


  1. A linotype operator worth his sault would not have set the lines on the left. Your demonstration is a false example.

    1. The simulated example is precisely what I remember from daily reading of newspapers in the past. I would therefore not describe the example as "false".
      The strong river effect as described is apparently a normal result of the Linotype technical system. Now I'm sure it can and should have been mitigated by it's operators. But that would have probably come at time and effort, or cost.
      Perhaps that is why it was prevalent in newspapers anyhow.