Here is an article that is almost entirely introduction, with a stubby little body at the end. Our excuse for it is that we know that some of our readers enjoy thinking about questions of language and typography. They enjoy thinking about them so much that they are willing to read a thousand words of blithering on the subject.

Are you ready for the introduction? Here it comes.

The computer keyboards we all use—whether on phones, on tablets, on laptops, or on desktop computers—are priceless historical documents. They preserve layers of history going back a century and a half to the Sholes and Glidden typewriter, which introduced the QWERTY keyboard. The staggered arrangement of keys, which line up in rows but not in columns, comes from the mechanical necessity of running a type lever from the key to the typebar.

Each subsequent advance has left its mark on our keyboards. The Remington 2 gave us the shift key, so we could have lower-case letters. Backspace keys came later, along with tabulators for typing in columns. The more expensive electric typewriters introduced a carriage-return key, which returned the carriage to the beginning of the next line.

Computers added characters useful in various computer activities—backslashes, brackets, backticks, tildes, and sundry alternate, option, command, function, and control keys. The Macintosh keyboard was the last important development of the computer keyboard, which means that our keyboard layouts were frozen in the 1980s. Touch-screen keyboards are a step backwards in many ways, often making use of the double shift that was popular on portable typewriters a hundred years ago but mostly extinct by the 1930s.

You will notice that at no point in this development from typewriter through computer did the keyboard go through a typesetting stage. This is an important observation: our keyboards were never meant for setting finished text. Typesetting was a separate discipline from writing. In fact there were separate lines of development in keyboards for typesetting, as for example the Linotype and Monotype keyboards. Like Homo neanderthalis, they were stockier, perhaps a little clumsier, and excellently adapted for their task; and like that unfortunate side branch of the human family, they were displaced by their more generalized cousins. We might say that the Macintosh keyboard, with its awkward but serviceable key combinations for proper typographic characters, retains a small percentage of Neanderthal DNA; but on the whole the typesetting keyboard is an extinct dead-end in keyboard evolution.

Thus we have been stuck with a series of clumsy compromises, which have led to a series of clumsy adaptations. For quotation marks and apostrophes, it saved a key—and all the mechanics that went with it—to invent a character that would sort of do, if you weren’t too fussy, for both opening and closing single quotes and apostrophes. Likewise with double quotes. Two hyphens would have to do for an em dash. We have kept those clumsy compromises on our computer keyboards of today.

But typeset text looks amateurishly awful with typewriter quotation marks and apostrophes and double hyphens for em dashes. So our software tries to guess which of the various possibilities we really mean when we press that apostrophe key. Is it really an apostrophe? If it’s at the beginning of a word, we’d better make it an opening quotation mark. That’ll make ‘em happy. This algorithmic substitution is usually called “smart quotes” or some similarly silly name.

And now, finally, we come to the little mystery of pedantry we mentioned at the beginning of the article.

For many years, LibreOffice Writer—the word processor on which Dr. Boli depends when he needs a full-scale word processor—installed by default with a curious inconsistency. It would correct typewriter quotation marks to proper printed quotation marks—but it would leave typewriter apostrophes alone. (This default was finally changed a couple of years ago.) Of course you could change this behavior in the preferences, but why was it ever set that way in the first place? Dr. Boli had always assumed it was an unaccountable oversight.

But the truth seems to be otherwise. It appears that, among the sort of people who write code for that kind of software, there is some kind of urban legend that a proper apostrophe is the typewriter apostrophe. The printed apostrophe is a closing single quotation mark. This is just the sort of pedantic distinction that easily gains traction among people whose livelihoods depend on pedantic distinctions—after all, using exactly the right character makes the difference between your program compiling and your compiler blowing a raspberry at you. The mere fact that it happens to be completely wrong does not prevent the pedantic distinction from taking firm root in the minds of the susceptible.

The subject came up when Dr. Boli was reading the reference manual for an academic writing tool called Madoko, which was written by a Microsoft engineer and is still kindly maintained on line by Microsoft, probably in a fit of forgetfulness, since it has not been seriously updated for five years. In the section on smart quotes, our engineer points out how much care he has taken to get the apostrophes right:

Madoko will only smart quote single quotes if the last quote is not directly followed by a letter. This often prevents wrong quotation with words like “can't”:

   ‘really, I can't do this’, he said. I can't and mustn't do this.

Yes—if he had not put that extra work into the processing of the single quotes, those apostrophes might have come out looking the same as a closing quotation mark.

Dr. Boli himself has long since given up on the idea of allowing software to decide how to distribute apostophes, quotation marks, dashes, and so on. He uses a keyboard layout with key combinations for all those characters, and after nearly thirty years of using the same layout, his fingers naturally find their way to those combinations.

But it pleases him to discover this strange little urban legend circulating through the coding community. He feels like an entomologist who has discovered a new species of small black insect. To him it is an object of delight and fascination, even if the rest of the world thinks it’s just a bug.


  1. D. Smolken says:

    I suspect (coming at this from a programmerly point of view) that the Microsoft engineer is close to being correct about the original reason, but not quite there. This may have originally been done to make it easier for programmers to keep track of whether the last quotation that was opened has been closed. This will avoid turning the next apostrophe into an opening quotation mark when it should be a closing one.

    In other words, I think programmers decades ago weren’t worried about ending up with ‘really, I can’t do this’ but with ‘really, I can’t do this‘.

    • Dr. Boli says:

      You may be right. However, from an algorithmic point of view, the simple question “Is there a space or carriage return before it?” produces more consistently correct results than this Microsoft engineer’s more complex assessment. If the answer is yes, it’s an opening single quotation mark. All other cases are apostrophes or closing single quotation marks. As we see with WordPress, it goes wrong if there’s an apostrophe at the beginning of a word, but it certainly works better than the Madoko algorithm.

Leave a Reply

Your email address will not be published. Required fields are marked *