Back in July, Audible announced a new feature called “Captions,” which uses automated transcription to produce read-along text that appears on a smartphone screen while an audiobook as playing. The company pitched this very much as an “educational” resource, pledging to offer free access to students and schools as well as providing the option to turn on the captions for paying Audible subscribers. The roll out was planned for the beginning of the 2019 school year, on 10 September.
Right from the start, publishers made their opposition to this move clear, believing that it would seriously harm sales of ebooks to have Audible make the text of a book available with the audio purchase, as well as being a violation of existing contracts for different book formats and editions (i.e., Audible have the audio rights and can therefore sell an audiobook, but not the text rights). The company maintains that because there is no ability to scroll through the entire book — the transcript appears on the screen phrase by phrase as it is read — there’s no conflict with existing digital editions. There is also the option to offer translations within the Captions feature, apparently, which opens up a whole other front of problems, since rights for books tend to be sold territory by territory and may well be owned by different publishers.
Some publishers immediately called the feature an “unauthorized and brazen infringement of the rights of authors and publishers,” and the Association of American Publishers together with Chronicle Books, Hachette, HarperCollins, Macmillan, Penguin Random House, Scholastic and Simon & Schuster filed a suit in New York on 23 August to prevent their titles from being included in the Captions feature without their explicit permission.
Although Audible continues to deny that the feature violates any agreement, the company did file a stipulation on 28 August agreeing to exclude the works of the publishers involved in the legal action from the feature until the whole lawsuit is resolved. The Captions feature will go live next week for Amazon-owned and public domain titles, however.
This isn’t the first time that Amazon and book publishers have done battle over audio. In 2009, Amazon had to backtrack on a proposed “text to speech” option on the Kindle 2 that would have meant a reader could listen to a machine-generated reading of any book without also purchasing the audiobook. In that instance, publishers won the right to disable the feature on their titles. A decade on, it’ll be fascinating to see if the courts come to a similar conclusion for the opposite situation: audio-becoming-text instead of text-becoming-audio.
Part of the reason this whole situation seems really relevant to our interests is because I suspect that we’re not very far off a similar moment in podcasting, given the encroaching prevalence of automated transcription within podcasting apps and platforms. Apple and Google are already doing it (to varying extents), and one would guess that Spotify won’t be far behind. This under-the-hood feature is, for now, being sold to podcasters and listeners alike as a tool to enhance podcast search and episode discovery.
But should these transcripts ever become publicly visible, it’s not hard to imagine a situation where a publisher would rather a listener visit their own website for an accurate transcript (perhaps enhanced with advertising or calls for support) than experience it via a machine-generated version on a third-party platform. In that instance, who owns the text version of a podcast’s audio? It’s a murkier scenario than with book publishing, where all aspects of a work tend to be defined by the author’s contracts.
For another instructive and partially related example, we can look to song lyrics. The lyrics and annotation website Genius recently won a dispute with Google over its work being surfaced as search results without credit or link, or coming via a service that pays the originator for the use of the lyrics. It’s yet another of those brave new digital world problems — who owns the text version of what you hear? I’m betting a podcast test case will be coming soon to a courtroom near you.