IN “THE LIBRARY OF BABEL”, a story by Jorge Borges, a man loses himself in a gargantuan repository of every possible book in the universe. Google Books is not quite that vast, but it is big. Since 2004, Google has teamed with libraries to scan over 20m titles—including many that are out of print—and put them on the web for anyone to access. Users cannot read entire books, unless they are in the public domain. But unlike the sad hero in Borges’s dystopian tale, who never locates the catalogue to the collection, Google Books browsers can search for specific phrases and, without paying, read snippets of countless tomes.

A decade ago, a group of authors sued Google, claiming the service cut into their copyrights. After years of legal machinations, a federal district court ruled in favour of the internet giant in 2013. The plaintiffs—including Jim Bouton, author of “Ball Four”, and Betty Miles, who wrote “The Trouble with Thirteen”—appealed to the Second Circuit Court in New York and on October 16th, they were rebuffed again.

How can a company get away with digitising millions of books without the authors’ consent and showing them to the world? In his ruling, Judge Pierre Leval explains that copyright law gives “potential creators” the exclusive right to copy their own work in order to expand everybody’s “access to knowledge”. It’s not all about enriching authors. The “ultimate, primary intended beneficiary”, he writes, “is the public”.

Reproducing too many of somebody else’s words for profit and without adding value to the work generally constitutes a copyright infringement. But if a reasonable portion is spun into parody or criticism, or if the work is put to a “transformative purpose”, it counts as permissible “fair use” under the Copyright Act of 1976. That’s just what Judge Leval says Google does by uploading millions of books and rendering them searchable by the masses. The “purpose of Google’s copying of the original copyrighted books”, the ruling reads, “is to make available significant information about those books, permitting a searcher to identify those that contain a word or term of interest.” Another tool enables users of Google “to learn the frequency of usage of selected words in the aggregate corpus of published books in different historical periods.” There is no question, the court concludes, that these functions are “quintessentially transformative”.

The ruling has its limits: Google only supplies three eighth-of-a-page snippets for each book. And publishers and authors are free to opt out of the snippet-showing altogether. (You won’t find a word of the first “Harry Potter”, for example, or of “Fifty Shades of Grey”.) But the ruling gives people without subscriptions to expensive online databases instant access to a universe of knowledge. The decision is a boon to researchers and students who will retain quick access to millions of books they may otherwise never have had the chance to read.