Press "Enter" to skip to content

How to change García into Garcman

Brief by Central Staff

Typography – October 1999 – Colorado Central Magazine

How to use a computer to change García into Garcma

We never heard of anyone named “Garcma,” either, nor of a place called “San Josi.” So we don’t have any good excuse for the errors we made last month in publishing Jeanne Englert’s review of Brujas, Builtos & Brasas: Tales of Witchcraft and the Supernatural in the Pecos Valley by Nasario García (rather than Garcma).

Our abject apologies to all involved.

There’s no excuse, but we do have an explanation. Somewhere on its electronic trip from Jeanne’s machine to ours, one of the computer programs involved didn’t like accents. Since such glitches are common enough, we should have noticed — but didn’t.

Machines ignorant of foreign accents, however, are an old problem, and why we have canyon instead of cañon and pinyon instead of piñon. Modern computers simply take a more novel but less understandable approach when — as it turns out — they don’t know everything.

The ASCII (American Standard Code for Information Interchange, and usually pronounced “ass-key”) character set uses only 7 bits, and thus has a limit of 27 (128) distinct characters.

That’s more than enough to accommodate the regular upper- and lower-case alphabets, as well as numerals, punctuation, and the like.

But it’s not enough for the accented characters like the í in García or the é in San José. Desktop computers work around this by using 8 bits to represent characters, giving 256 (28) possibilities.

The problem is that there’s not one standard like ASCII for those extra characters, but several. And generally, if a program doesn’t know what to do with an accented character, it converts it from 8-bit to 7-bit, thus turning García into Garcman.

We’re not sure where it happened in the many programs that text goes through (Jeanne’s computer to the Internet through a variety of computers to our computers), but we should have caught it before it went to print.

We did fix it on our website, anyway, where there’s yet another way to represent such characters: í is written as í in the Hypertext Markup Language used for constructing World Wide Web pages.