CAPTCHA

In 2000, computer scientists at Carnegie Mellon — Luis von Ahn, Manuel Blum, Nicholas Hopper, and John Langford — needed a name for their new test. The test asked users to read distorted text that computers could not decipher, proving the user was human. They devised the acronym CAPTCHA: Completely Automated Public Turing test to tell Computers and Humans Apart. The backronym was a stretch, but it worked because it sounded like capture, and capture is exactly what the test does: it catches bots.

Capture descends from Latin captura, from capere, to seize or take. The Romans used capere constantly. A captivus was someone seized in war. A capito was a big-headed fish, easy to grab. The word branched into hundreds of descendants: captain (one who takes command), capable (able to take hold), recipe (literally 'take!' — the imperative form), and accept (to take toward oneself). The Latin root cap- is one of the most fertile in the entire language.

Von Ahn, a Guatemalan-born mathematician, later realized that the millions of hours humans spent solving CAPTCHAs could be redirected to useful work. In 2007, he launched reCAPTCHA, which used distorted text from old books that optical character recognition had failed to read. Every time you solved a reCAPTCHA, you were digitizing a fragment of a scanned book. Google acquired the project in 2009 and used it to digitize the entire archive of the New York Times and millions of books for Google Books.

Modern CAPTCHAs no longer use distorted text. They ask you to identify traffic lights, crosswalks, and bicycles in photographs. The test has evolved, but the name persists, carrying its hidden Latin root. Every CAPTCHA is still an act of capere: the system seizes a moment of your attention and uses it to confirm you are not a machine.