Why Can’t I Digitize My (Institution’s) Library?

By David Hansen, J.D., Scholarly Communications Intern

On Tuesday Judge Denny Chin set a deadline of mid-September for Google, the Authors Guild, and the AAP to work out a settlement for Google Books. The lawsuit, filed in 2005, seems to have been going on forever, and I wonder what, in the meantime, libraries can do to move forward. After looking at my own (personal) digital library, I wonder how the same principles regarding digitization might apply to institutional libraries.

Over the weekend I joined Google Music, a service that uploads my collection of music and stores it . . . somewhere. Somewhere in Google’s cloud. With it, I can access my entire collection of music from any computer. It’s great.

What is not great is my internet connection. I’ve had the service for about a week, and at this point only about half of my music collection is uploaded. Uploading large amounts of data understandably takes time, and since Google Music “store[s] a unique copy of Your Music on your behalf,” each and every file has to be transferred.  Uploading these copies is generally considered “space-shifting,” which is something that Google –and the courts—have concluded is lawful “personal use.”

Apparently there are other approaches to what Google Music does. Ars Technica has published this article outlining the legal positions of Google Music, Amazon Cloud Player, Apple’s iCloud, and MP3Tunes. All three services provide online streamed copies of user’s music collections. Apple does so with licenses from the record labels.

Google Music and Amazon Cloud Player both seem to operate as a “digital locker,” making unique copies of the user’s own files. They presumably rely on time-shifting cases that make users’ actions lawful, and on the Cartoon Network v. Cablevision case (discussed at length in the Ars article) which held that Cablevision would not be directly liable for “publicly performing” the works in question, although it provided a DVR service that allowed users to record and retransmit their own unique copies of previously transmitted shows. The court in Cartoon Network placed some emphasis on the fact that each user only had access to their own personal and unique copies of the recorded shows.

MP3Tunes acts in a similar way, but with two differences: First,  MP3Tunes will delete redundant copies when more than one user uploads identical files. This de-duplication process, while obviously more efficient than the Google and Amazon services, may conflict with the Cartoon Network case because each user accesses one centralized copy of their song, rather than multiple users accessing multiple ‘unique’ copies of the their own recordings. The second major difference is that MP3Tunes is currently being sued by EMI. Most of the suit focuses on the safe harbor provisions of the DMCA, and whether MP3Tunes can be held liable directly, notwithstanding the Cartoon Network case cited above, for “publicly performing” the works in question.  But another major issue is whether space-shifting to the cloud is a permissible fair use.

For libraries that want to make digital copies of their print collections—i.e., space shifting—there are some limited exceptions in the law that permit copying for preservation (section 108 of the Copyright act).  There is a need, however, to provide more complete digital access to the entire campus community beyond that which is contemplated by section 108. The University of Michigan (along with Florida, Illinois, and Wisconsin) has recently announced that it will be making available to campus users copies of orphan works, held jointly by the University of Michigan and HathiTrust, based on an assertion of fair use and its own risk analysis. The fair use argument relies on the idea that only works in each respective library’s print collection will be made available online to their users through the HathiTrust; one print copy, one digital access. No one is gaining access to books they don’t already own—just different, electronic access to those already in the print collection. The parallel to the ‘digital locker’ analogy that supports Google Music is strong, and the fair use argument for Michigan is bolstered even more by the fact that it isn’t in it for the money (as Google is).

This fair use assertion makes an end-run around section 108, but looking at the fair use factors, it is still appealing. Even more so for Michigan because a large part of the scanned corpus of the HathiTrust comes from Michigan, so for many books it would also be able to make the argument that the digital copies are not just practically the same books that are in its collection, but that they are identical copies of UM books, meeting some of the concerns of the Cartoon Network court.  Other libraries have less to rely on in that respect, as fewer (or none) of their physical copies were scanned for inclusion in the database. But the fact that Michigan and these other libraries are only making orphan works available means that even if the fair use analysis is slightly off, there is still almost no chance anyone will be sued. The orphan works identification process that Michigan has used (detailed here) employs a more than reasonably diligent search for copyright owners, and leaves little chance that there are any rights holders available or willing to bring an infringement suit.

Risk notwithstanding, though, I wonder, what’s wrong with a library digitizing its entire collection (not just orphan works) under the space-shift theory?  If the library takes those books out of circulation (perhaps in high-density storage) and limits online access to one user at a time (essentially, recreating the limitations of a physical visit to the library), the fair use analysis is still very much in the library’s favor. Google, in its amicus brief in support of MP3Tunes, makes the point well:

“[j]ust as the Supreme Court has held that ‘time-shifting’—recording television broadcasts for later viewing—is a lawful fair use, Sony Corp. of America v. Universal City Studios, 464 U.S. 417, 455 (1984), so too is ‘space-shifting’ lawfully acquired music onto digital music players or cloud-based equivalents, Recording Indus. Assoc. of Am. v. Diamond Multimedia Sys., 180 F.3d 1072, 1079 (9th Cir. 1999). A contrary holding would treat tens of millions of iPod owners who lawfully acquire their media as no better than those who misuse new technologies to pirate music and movies. “

Should space-shifting books be any different? These cases, admittedly, deal with space- and time-shifting for personal uses, and not for uses of educational institutions. That distinction may be critical in the end.  But shouldn’t uses for “teaching . . .  scholarship, or research” — which are specifically called out in the section of the copyright act that codifies fair use — carry at least as much weight as “personal use,” which has no mention anywhere in the act?

Finally, if a library can digitize its own library and make it available to patrons, can that library pool its digital holdings with other libraries, so that there is no needless duplication of digital copies? Storing these works in digital format is not cheap, and while my meager 20GB music collection has taken a half a week to upload to GoogleMusic’s “cloud”, the creation and duplication of millions of volumes of digital volumes is a monumental and inefficient task.  Such a restriction, as amici in the MP3Tunes case have argued, would be incredibly burdensome to both digital libraries and users in general.

8 thoughts on “Why Can’t I Digitize My (Institution’s) Library?”

  1. Nice thoughtful article Kevin.

    One issue I have is that this article incorrectly states that Google stores files differently than MP3tunes. Ars Technica got it wrong if they originally claimed this. Every major net company and every major web service (email, web hosting, file storage) ALL use deduplication technology behind the scenes. It’s been standard practice for a decade.

    The claim that Google or Amazon is somehow storing files differently than MP3tunes is not true. Deduplication is standard throughout the industry. In our case against MP3tunes we demonstrated that EMI and the courts own servers use deduplication.

    — MR

    CEO, http://MP3tunes
    CEO, http://DAR.fm

  2. You can do all the things you wish to do and more.
    Consider with the ‘cloud’ services you mention the copyright violations that they facilitate in their current form. Without accusing you of illegal filesharing there is no verification that you own the files you are uploading. Without that verification any use of those files to stand behind others is counterfeiting and giving value to counterfeits which devalues all copyrights. What has been called piracy is really counterfeiting. We would like to get your take on our position in the depth of personal communications. Let us know if you are up to it.

  3. Great article, David. Freshest thinking yet.

    The final piece of the puzzle is how do you verify “owners who lawfully acquire their media?”. The authors of the amicus probably saw that as a truism … but it is not, of course. Just like it was not a truism that Kazaa users were simply owners sharing media that they lawfully acquired with their friends (i.e. like their family home movies).

    (We proposed a system just like the one you are talking about back in 2009 in our Amicus Brief in the Google Book case. It’s based on verification, registration and immobilization methods used in the stock market and employs cloud technology. It’s novel and non-obvious enough to be patentable, we believe. We call it The Digital Content Exchange)

    1. For libraries, the verification is a non-issue. I don’t think anyone really suggests that libraries are sitting around with large deposits of illegally acquired books on their shelves. For consumers, congress hasn’t done a great job when it changes focus from copyrighted content to the actual technology of the content (e.g., Audio Home Recording Act, Digital Millennium Copyright Act), and I would suggest that a mandatory registration scheme might run into 1st Amendment hurdles. That being said, a voluntary system might do a lot to make personal collections more liquid, while avoiding liability.

      In terms of ‘upstream’ adoption of such a system, I don’t think most manufacturers or service providers have much incentive to do so. Sony (Betamax) says pretty clearly “the sale of copying equipment [or service?], like the sale of other articles of commerce, does not constitute contributory infringement if the product is widely used for legitimate, unobjectionable purposes. Indeed, it need merely be capable of substantial noninfringing uses.”

      1. The idea of having one central pool of resources for all libraries seems like a good method for curtailing IT costs. Regarding this quote:

        “Finally, if a library can digitize its own library and make it available to patrons, can that library pool its digital holdings with other libraries, so that there is no needless duplication of digital copies?”

        Libraries may not be “sitting around with illegally acquired books on their shelves,” but who would verify that each library then accessing the “pooled” resources in fact owns a physical version of the item and has indeed placed the item in storage. I believe we would ultimately be dependent on presuming all individuals at institutions will behave professionally, but would that suffice for publishers who may balk at this idea and demand process/documentation that ensures only appropriate individuals are accessing (primarily, meeting the above conditions)?

        1. This is a good point, though I think the verification would be pretty simple. Most, if not all, academic libraries have their catalogs available online where anyone can view them. Since institutional user-access would most likely flow through those catalogs, anyone with internet access can check print holdings against those that link to a digital copy in, for example, HathiTrust or any other such repository that centralizes access.

          In terms of libraries behaving “professionally,” I don’t think there is much to be concerned about. Librarians can be infamously timid when it comes to pushing the bounds of copyright, so much so that it makes searching for case law on library-copyright law issues very frustrating; they are never sued, so there are few precedential cases that deal with fair use in academic libraries (Georgia State and AIME v. UCLA are two new, notable exceptions to the that statement).

Comments are closed.