Readwyn 0.1.0

Is Google Books legal? What the Authors Guild case decided

Search Google Books for a word in a copyrighted book whose publisher hasn't licensed it, and you get “Snippet view”: a few thin strips of the page with your word in them, and nothing more. That's legal in the United States. On 16 October 2015 the Second Circuit Court of Appeals held that Google's scanning of whole books without permission, its full-text search and its snippets were all fair use, and that handing digital copies to the partner libraries was lawful too. The authors asked the Supreme Court to hear the case, and it denied the petition on 18 April 2016, so the ruling stands.

How snippet view is built

The court's opinion describes the feature in more detail than Google's help pages do. A snippet is normally an eighth of a page: on a book with 24 lines a page, three lines. A search for one term shows at most three snippets, and always the same three, however often you repeat it and from however many computers. You can see more by searching for other words, but Google permanently withholds one snippet on every page and one whole page in every ten, which it calls “blacklisting”. The court put that at about 22% of a book's text. Snippet view is switched off entirely for dictionaries, cookbooks and books of short poems, where one snippet might be all a reader needs, and since 2005 a rights holder can remove a book from it with an online form.

Snippets are the third of four levels. Google's Library Project help page lists full view for books in the public domain, limited preview when a rights holder allows some pages, snippet view, and no preview for books Google knows of but hasn't scanned. You can download the full-view books; our post on getting public-domain scans from Google Books and HathiTrust shows how.

Why the court called it fair use

Judge Pierre Leval wrote for the panel. He found that the copying served a purpose the books themselves don't: it let anyone find out which of more than 20 million books used a term, and how often, without giving them the book. Google showed no advertising on the search results. The plaintiffs' lawyers had researchers run searches on the authors' books for weeks; they never reached 16% of any text, and what they got was scattered across the book in no order. The US Copyright Office's summary gives the gist: the court accepted that the search might cost some sales, but found the snippets too fragmentary to be a substitute for buying the book.

The ruling has limits. It covers snippet view only “as snippet view is presently constructed”, and Leval wrote that a way to read a coherent 16% of a book would raise a different question. It's a US court applying the US fair-use statute, section 107 of the Copyright Act. Which books get full view is a separate matter of public-domain status, and that differs by country: Google's help page asks you to request a review if you think a book is in the public domain where you live. The court also ruled on the libraries' copies. Through Google's download system the partner libraries had taken at least 2.7 million digital copies of their own volumes, and the bare chance that a library might misuse them didn't make Google liable.

From the 2005 lawsuit to the appeal

Google's Library Project began in 2004. By the time of the lawsuit its partners included Michigan, Harvard, Stanford, Oxford and the New York Public Library, which picked books from their shelves for Google to scan. The Authors Guild and individual authors sued on 20 September 2005. The next month five publishers (Simon & Schuster, McGraw-Hill, John Wiley & Sons, Penguin Group USA and Pearson Education) filed a separate suit in the same Manhattan court, asking a judge to stop Google from resuming its scanning on 1 November. In October 2008 the parties announced a settlement under which Google would pay $125 million and a new Book Rights Registry would pass on money from book sales, institutional subscriptions and advertising to authors and publishers.

Judge Denny Chin rejected the amended version on 22 March 2011. As James Grimmelmann's reading of the opinion points out, the deciding point was that it released Google from liability for future acts, beyond what a class action may settle. Chin also urged the parties to move from an opt-out scheme to an opt-in one, and The Guardian reported that some 6,500 authors had already opted out.

The case went back to the original question, and Chin granted Google summary judgment on 14 November 2013. The authors who appealed were Jim Bouton, Betty Miles and Joseph Goulden; their brief names the books at issue as Ball Four, The Trouble with Thirteen and The Superlawyers. The Guild itself stayed on the appeal, but the court noted that a 2014 ruling in its case against the HathiTrust Digital Library had already found it lacked standing to sue for its members' copyrights.