SYSTEM 'BUGS'

Recently, Google's search page was adorned with one of its playful animated "doodles." A woman sits at the controls of a giant old-fashioned computer and asks it to calculate her age. After the computer returns an answer (107), a tiny moth flies from its output tray.

The doodle honors the birthday of the late Navy Rear Adm. Grace Hopper, a computer-science pioneer who helped to create the programming language Cobol. And the moth alludes to a famous anecdote involving Hopper's time working with the Mark II electromechanical computer at a naval weapons lab in 1947. (click below to read more)

When a dead moth was found gumming up the works of the Mark II, Hopper taped the offending insect to a page in her logbook with the annotation, "First actual case of bug being found." In computing lore, this real-life "bug" gave rise to the usage of the word to refer to a technical flaw, as well as to the verb "debug" for ferreting out software errors.

But while Adm. Hopper may have popularized "bug" in the early days of computing, she did not in fact coin the term. Its roots go all the way back to the technical innovations of a young Thomas Edison in the 1870s.

As Alexander Magoun and Paul Israel explain in an article earlier this year in the newspaper of the Institute of Electrical and Electronics Engineers, "bug" appeared in Edison's notebooks as early as 1876 to describe problems in his systems. "Awful lot of bugs still," read one notebook entry about a plan for incandescent lighting. He also developed what he called a "bug trap" to catch relay errors in his early telegraph system.

Within a couple of decades, Edison's usage of "bug" became common enough to enter dictionaries. "A fault in the working of a quadruplex telegraph or in any electrical apparatus," was how Funk and Wagnalls' Standard Dictionary of the English Language defined it in 1893.

But could the Edisonian "bug" have been forgotten about by the time Hopper labeled her moth? It doesn't seem likely. In fact, "bug" was used by researchers working at Harvard University on the Mark I, the precursor to the moth-ridden Mark II. A 1944 logbook entry read, "Ran test problem. Mr. Durfee from I.B.M. was here to help us find 'bugs.'" As for "debugging," the Oxford English Dictionary traces the term back to 1945.

So when Hopper called the discovery of the notorious moth the "first actual case" of a computer bug being found, she must have been humorously playing off a known usage in technical circles. While we celebrate the achievements of a pioneering woman in computing, let's not perpetuate an etymological story that's a little buggy.