Tell the FCC to Protect Net Neutrality!Sign The Petition To Save The Open Internet
Questions and Considerations About Orphan Works
To be a bit simplistic, the NOI asks two questions: what should be done with the old efforts, and what about this whole mass digitization thing? It also notes some international developments, including some countries setting up collective licensing systems for orphan and out-of-print works.
In thinking about how we’d address those questions, I was concerned with a few different things. First, a number of recent cases have demonstrated that courts are willing to use fair use in new ways, like allowing reproductions of concert posters within a book detailing the Grateful Dead’s history, or allowing book scanning for indexing, or an e-reserves system. These various cases suggested that, certainly for some types of users and some types of uses, fair use could get them there without new legislation.
At the same time, these cases, and fair use generally, doesn’t necessarily fit all of the potential use cases for orphan works. Someone who uncovered a striking old poster and wanted to print it commercially would have a harder time making a fair use case than the Grateful Dead book publisher—they might be willing to pay the rightsholder but have no way of doing that, and would be left with the choice of either letting the awesomeness of the poster languish or go forward and, if the artist emerges later, face potentially far more in damages than they would have been happy to pay for a licensing fee.
This, in an oversimplified nutshell, is the archetypal orphan works problem, and it doesn’t disappear with some good fair use cases on the record. One way of alleviating this problem would be to reduce damages in exchange for an infringer showing that he’d made a diligent search for the copyright owner and came up empty-handed. For example, if the company that found the old poster could show that it tried diligently, and failed, to locate the copyright holder, it could go forward with its reprints and production, but if the copyright holder later emerged and sued them, they’d have to pay—but not the punishing damages that currently attend copyright infringement.
This damages-reduction-for-diligent-search idea was the core of the Copyright Office’s 2006 proposal, and one that we at PK liked. Over the course of two years and much political wrangling, though, it became freighted with more and more caveats and hoops for people to jump through, making it increasingly unlikely that, even if it passed, anyone would actually be able to use it. But a simple, easy-to-use version seemed like a good idea, and something that could complement fair use nicely in allowing more people to access orphan works.
There is a countervailing concern about passing such a law, though, in terms of whether or not it would undermine the case for fair use. If courts started to see a diligent search as a standard part of the process, they might start to think of it as a requirement for fair use—even though fair uses and the uses of orphan works aren’t always overlapping. For instance, someone can make fair use of an orphan work, just as they can make use of a known work, without any attempt to contact or find the author. Parodying the work of an unknown author is just as much fair use as parodying Roy Orbison, and it shouldn’t matter if I’m quoting a book whose author is known or unknown in news coverage about the book.
The Copyright Office also noted a handful of licensing schemes, and a number of groups and associations of copyright holders have at times suggested that they could form a sort of licensing body on behalf of unfound authors, able to grant permission to users or orphan works in exchange for a licensing fee. This idea was concerning for a number of reasons.
First of all, there's the possibility that it could undermine fair use. That’s because, in deciding whether or not a particular use is fair, one of the major things a court has to consider is the effect of the use on the potential market for the work. If the new use displaces sales of the old work, that weighs against the defendant in the overall scheme of things. Likewise, it hurts the defendant if the use displaces the ability of the copyright holder to license uses of the work—a more subtle form of harming the market for the original work, but still one that the law will recognize and place on the scales. But that shouldn't mean that just sticking a price tag on something can remove it form the bounds of fair use.
And it's even more problematic that the body setting those prices would have no way or representing the authors themselves. After all, the entire point of orphan works is that the authors can't be found in order to represent themselves. In other words, this licensing body would be in direct competition for licensing fees with the people who actually created the works in the first place. Add to this the fact that the licensor has very different motivations from the original author. Authors frequently create new works for money, sure, but they also create them for the purpose of self-expression, to build their reputation, and for other, more prosaic purposes of communication. Keep in mind that a snapshot and a letter home are both copyrighted—the incentives of the author are thus going to be different from the incentives of a licensing body that owes its existence to fees it can charge for the works it licenses out. This disconnect between the motivations of many authors and the motivations of the body that was supposed to represent them was one of the criticisms of the initial Google Book Search settlement. And that's just taking into account the structural biases that would accompany a good faith licensor, not including the potential for abuse and corruption that has occurred in other situations.
When it came to mass digitization questions, the program proposed within the original (and now defunct) Google Book Search settlement plan was impossible to ignore as a model. It was one of the most detailed and specific plans for using a mass-digitized body of works to grant access to copyrighted works. But it had a number of flaws regarding user privacy and potential technological restrictions, and, on a larger scale, stood ready to be a monopoly on orphan works access. In large part, this was because anyone else who tried to do what Google had been attempting would likely have just been sued into the ground with punishing statutory damages hanging over their heads.
Our Comments: the Framework
Keeping all of these various considerations in mind, we tried to come up with a particular perspective and framework for looking at orphan works. First of all, it was important to recognize that there's no one solution for orphan works, because "the orphan works problem" has a wide variety of different sources. Just about any kind of work can become orphaned, and there are as many different kinds of potential orphan work users as there are users overall. Trying to say that there is one solution, to the exclusion of others, would be like calling chemotherapy "the cure for cancer," and scrapping radiation therapy, surgery, and all other forms of treatment (to say nothing of prevention, but that's an analogy for another blog post). So it's important to make sure that none of the proposed solutions will prevent others from working.
Another thing to keep in mind is why we're trying to do any of this in the first place. The entire point of copyright law is, according to the Constitution, "to promote the progress of science and the useful arts." In other words, the point is to spread knowledge and learning. Orphan works staying in the dark, unable to be accessed and seen by the public, can only frustrate that goal. So providing access to works has to be a primary concern.
That's not to ignore an important secondary concern, which is the other half of how copyright law is supposed to work: in order to meet the ultimate goal of promoting knowledge and learning, copyright law secures rights to authors, to incentivize them to create. Thus, as an important secondary concern, we want to see creators compensated.
It's the combination of these two goals that make the idea of a licensing entity for orphan works seem so frustrating. On the one hand, you have an entity that is increasing the barriers for access to the works by placing a price on them. This in itself isn't a bad thing generally—it's one of the standard ways we deal with commercial copyrighted works. But it falls apart because the tradeoff—cost to the public versus gain to the author—when you look at the reward side of the equation. In a situation where some other body can license the use of orphan works, the works' authors get nothing. In that case, neither the primary nor the secondary goals of copyright are being met.
Two other considerations should be taken into account when evaluating orphan works solutions: certainty and ease of use. As to certainty, the issue is that a lot of orphan works remain unused and inaccessible simply because, with no way to contact the author, and a highly variable range of potential damages, a user is denied not only the ability to get permission, but also the ability to estimate what a settlement might even cost should an author come forward later. Giving the user either the knowledge that she won't be sued or a known ceiling on cost will allow her to move forward.
Ease of use should be fairly self-explanatory—if an orphan works system is too cumbersome, or requires an expensive legal team to figure out—for any of the involved parties—it's likely no one will bother using it, and we end up just where we were before.
These considerations allow us to see what should and shouldn't be a part of an orphan works plan, and the basic outlines of the Copyright Office's 2006 proposal seem to get us there. Limiting damages gets certainty for the user, and allows access while preserving some remuneration for the author. Keeping the requirements clear and simple should allow a wide variety of good-faith users to use orphan works and grant people access to them.
Our Comments: Mass Digitization
Mass digitization, though, presents a separate series of issues. First of all, not all mass digitization projects are going to involve orphan works; many of the works will have known, findable authors, and many others will be in the public domain. But large-enough efforts to digitize and catalog works are inevitably going to scoop in orphans as well.
In discussing mass digitization, it's also important to separate out the different uses that are being made of a work. In the example of a book-scanning project, for instance, it'd be a fair use for someone to scan in a series of copyrighted books and hold their contents in a database for the purposes of making a text-searchable index, or conducting linguistic analysis on them. It would be much more likely an infringement if they took that same digital database and stuck it all on the web for everyone to see.
Fortunately, that's not generally the plan with such programs. They've been focused at universities, libraries (and university libraries), to do things like catalog the works, study them, and make electronic versions accessible to students and faculty with print disabilities. Many of them are making public domain works available to the general public, or to all of their patrons. For known in-copyright works, they can negotiate with rightsholders for licenses if they want to make those works more available in a digital form, or if those talks don't work, they can just leave them dark and buy paper copies for the stacks.
But how would orphan works be made available? If they're not in-print, they can't be placed on the physical shelves, and there's no one to negotiate licenses with. So again, the digitizer is left with potential liability. Given that the number of works at issue could be in the thousands, or hundreds of thousands, conducting a diligent search for the author of each might be prohibitively time-consuming and expensive. On the other hand, allowing a program to make a work available just based on an assumption that a work is orphaned could lead to needless infringement of known authors' works.
In addition to this dilemma, there are other problems with digitized libraries. They can create privacy problems—knowing someone's reading habits can tell you a lot about them—and in a digitized version, that information can be ever more fine-grained, including not just what they're reading, but when, and how quickly, and in conjunction with what else. Digital works are also subject to DRM—software meant to protect copyrights, but that is often used in ways that harm users.
These specific problems, coupled with the larger problem of accessing orphan works, can actually be used to cancel each other out somewhat. One plan could be to offer mass digitizers a limit on damages if they infringe copyright by making orphan works available to the public, but only in exchange for refraining from bad behavior, such as spying on users, installing harmful software, or acting anticompetitively.
There's a lot to be done on orphan works—to the point where it's unlikely that the problem can ever be completely solved. However, a number of good steps can be taken to alleviate the problem, and it can only be a good thing that the Copyright Office is taking up the problem again.