Scanning copyrighted works and making them searchable on the Internet sounds to many authors like it shouldn't be legal. But on October 10, a Federal Court in New York released a judgment holding that may be. That judgment raises significant issues for authors, publishers, and consumers, giving new scope to fair use of copyrighted materials in the USA.
The company doing the scanning, naturally, was Google. Google had entered into agreements with various universities including the University of Michigan (operator of the HathiTrust e-book pooling system), the University of Wisconsin, Indiana University, and Cornell University. Under these Agreements Google would be allowed to scan library books and make them accessible to be searched on the Web as well as providing digital versions of the book back to the universities (which would donate them to HathiTrust).
For works with known authors HathiTrust would have two benefits: full-text searching and access for people with certified print disabilities. For orphan works most of the books in HathiTrust would also be available for full text viewing online (Indiana U didn't allow this).
Various authors and authors' organizations opposed this plan, including the Authors Guild in the USA and similar organizations in other countries, as well as individual authors. The National Federation of the Blind, among others, got court permission to file a brief as an intervenor. Google's use of the books is not at issue here; there's a separate lawsuit about that. Here, these organizations were claiming that the HathiTrust system violated copyright. And, after applying the fair use tests under US copyright law (which I've written about before, links below), the judge disagreed. And in disagreeing, the judge made 3 holdings that open a door to what could become a very significant change for the e-book and audiobook markets.
1. Text-to-speech is transformative use. The first step in determining whether something is fair use is to see whether the use transforms the original work or just copies it. Here the court held that because scanning the books allowed for print-disabled scholars to have access to them through text-to-speech, that would pass the transformative use element of the fair use test. Finding that print-disabled customers aren't considered a significant market, the court determined that providing access to them wasn't the intended use of the original work (which was enjoyment and use by sighted persons). Technology that made these books accessible to such customers would be transformational, and therefore using that technology would be fair use.
This is a pretty big deal. Usually the transformation has to be something relating to the nature or character of the work. The ordinary example is parody or satire. Straight-up copying without any modification isn't something that courts tend to find to be a transformation. This is one of the first cases, if not the first case, where a court held that just changing the format of a work would transform it.
2. Copying the whole book can be permitted in certain cases. The third step in the fair use test is determining whether the amount of copying is reasonable in relation to the purpose for the use. Once the court found that text-to-speech was transformational, it would be very difficult for the court to hold that only partial copies would be permitted. As a result, the court held that HathiTrust was allowed to copy the entire book.
Although it follows logically from #1 and isn't the first time it has happened, this is also a big deal. Many people seem to think that there's a "10% rule" - there's nothing illegal about copying 10% or less of a work. This case is a very extreme example of why relying on rules like that is a bad idea.
3. Technology for print-disabled readers may help fill a mandate under the Americans with Disabilities Act. Under the ADA, institutions are required to help remedy discrimination against and difficulties met by people with disabilities such that they can compete on an equal basis with people not having those disabilities. Since one of the purposes of HathiTrust is to provide print-disabled customers with access to texts, HathiTrust is fulfilling that mission under the ADA by scanning the books and making them available in text-to-speech by using technology.
But does this mean that only universities and institutions are allowed to provide that technology? Or can the private sector develop technologies for this as well? The answer to this second question becomes very important.
In a world of e-books, that question becomes very timely. The original versions of the Kindle contained text-to-speech functions. In response to complaints and veiled threats of litigation, Amazon allowed publishers to choose whether or not to enable that function on a book-by-book basis. And the newest Kindles apparently don't offer the feature at all.
Does this mean an enterprising app developer could fill the gap, developing technology to allow print-disabled readers to convert their Kindle and other books to audiobooks?
If the answer to that question is yes, then that would raise a very interesting possibility. Under the Sony Betamax judgment, making copies of TV programs was found to be a legal activity because it facilitated time-shifting, and therefore Sony did nothing illegal by developing the Betamax. The movie studios complained that people were going to use the Betamax to do illegal copying of copyrighted content, but the studios lost that argument. (I'll talk a lot more about this judgment in a future background post, trust me!) So on that same reasoning if it's fair use for a print-disabled person to convert a book text-to-speech, it should also be fair use for someone to develop an app to facilitate that.
But an app like that would likely have a pretty significant impact on the audiobook industry. Publishers have sold audiobook rights separately from print. There's an industry of voiceover professionals who would lose their work if text-to-speech apps became commonplace. And authors would, of course, receive no royalties.
Of course, Kindle books are protected by DRM. Will that make a difference? Should it?
I suspect a court will be looking into that before long.
Legal Minimum post on fair use criteria:
Articles on Amazon phasing out text-to-speech on Kindle: