Internal:Public Policy/Orphan works/Notes

2006 Report on Orphan Works
Notes on it by Econterms
 * The 2006 proposal by the Copyright Office, makes a proposal about what to do about orphan works.
 * the proposal seems appropriate to me -- that the user must do a "diligent search", and if a copyright owner surfaces after a work has been classified as an "orphan" and used, that the copyright owner's recovery/penalty is capped.
 * It appears that the Copyright Office-drafted legislation in this report didn't become law, but did show up in House and Senate versions ; I haven't yet seen an explanation of why or how it didn't get done.
 * They may be holding the 2014 workshops to rebuild support for it or something like it.
 * Basically it says the user of an orphan work has to make "diligent" efforts to find the proper owner and license it. The Copyright Office does or would have materials online to help the person search and "diligent" would be defined partly by evidence that the user had used those materials.  Potential users who can't find the owner could use the material, labeling it as an orphan work.  Seems good to me.  Can we back that?   And/or, is the labeling requirement excessive for Wikimedia?
 * Orphan works are works whose copyright owners cannot be identified or cannot be located, and therefore permission cannot be obtained from them for using the works. Page 35 says the definition excludes works for which it is not known whether a copyright currently applies, e.g. because creation date or location or creator are unknown. That report refers to such situations as "problems determining copyright status. Wikimedia DC would definitely prefer that the legislation includes such cases, which are just like orphans in our practice. See our Oak Ridge example for a case like that. However, for clarity of communication, note that the Copyright Office report did not call those orphan works.
 * The report categorizes people who want to use orphan works. Wikimedians are mainly "large access" users, and also sometimes "enthusiast" users. (Relatedly, this means we can give up any particular orphan work -- we're not inscribing it permanently in stone or a movie shot. Instead we are high volume users who want it to be easy to manage orphan works en masse.  We're unlikely to pay for cases of infringement -- we'll take any such file down.)
 * Section IV.B (pp 44-59) discusses existing laws that relate to orphan works and give users some freedom. None of them seems to be generally relevant for Wikimedia so let's not comment in detail on them, but for background here's a sketch.  The "sections" seem to refer to the Copyright Act of 1976 and/or patches it put on previous laws.
 * Section 108(h) authorizes educational nonprofits to use copyrighted works for display purposes in the last 20 years of their copyright period ; this is part of the 1998 Sonny Bono 20-year extension to copyrights.
 * Section 115(b) is about compulsory licenses, which a copyright holder may have to give/authorize.
 * Section 504(c)(2) limits damages a reasonable good-faith user would have to pay after using a work that was not plainly copyrighted or a search for an owner came up empty. (The 2006 report says these damage-limits have not been controversial, and we like anyone want them to be incorporated into new/larger/easier orphan-works-use rules and systems, but it is not worth talking about much---we are hunting bigger game.)
 * Sections 203, 304(c), and 304(d) address "termination" of copyrights when there probably is a copyright but an owner wasn't found, and it includes the idea of a "reasonable investigation" which is a model for the "diligent search" idea in later proposals and which we can support.
 * Section 102(b) says an idea can be reused even when its expression is copyrighted in text, or computer programs, or images (pp 53-55).
 * Section 107 defines "fair use" (pp55-56).
 * Section 108 gives libraries and archives some permissions regarding preservation of a copyrighted work.
 * Section 117 gives some permissions regarding software, e.g. backups and alterations, usually associated with an abandoned copyright -- e.g., company went out of business) ; these cases are called "abandonware."

related notes and comments by econterms

 * Orphan works are not free content and cannot go onto Commons by Wikimedia's rules. Antony-22 says an orphan work classification is freer than a "fair use" classification, and therefore better for Wikimedia. We'll want an example of how and why at some point.  A file can be classified as both "orphan" and "fair use", gaining slight freedom-benefits from each of those.
 * Antony-22 says the copyright office did not include a "registration requirement" that would require registering the use of an orphan work, or the classification of the work as an orphan, in some kind of registry. They considered it, then did not include it.  Our clear preference is that a registration requirement is avoided. However, it would probably be good for there to be optional registries.  Wikimedia uploads are already searchable in a way that makes them registry-like.  Antony-22 points out that Wikimedia could also semi-automatically do searches of itself and other registries at the time of an upload.  Or, bots could review that afterward.  Wikidata services can help at some stage, since we'll have VIAF and other lists of authors and publishers.

Wikimedia Deutschland statement on orphan works
Source: WMDE Statement on Orphan Works, Berlin, August 10, 2012

This statement was in response to a proposal, but the proposal was not really summarized. Notes and comments by Econterms (talk):
 * orphan works in their definition have been "handed over to an organization" -- so there's some difference from our definition
 * Regarding "diligent search" (the same term used in the U.S.): "It would . . . be desirable to specify the sources that are to be consulted in such a way as to create a specific, conclusive list. This would help to keep search costs to a minimum and would create the necessary legal certainty."   (I'm inclined to agree)
 * "It is also important that all the listed sources are, without exception, freely accessible within the meaning of the 8 Principles of Open Government Data (http://www.opengovdata.org/home/8principles). This is the only way to ensure the availability of low-cost, professional search services and to encourage the development of appropriate business models." (agreed)
 * "It is . . . crucial to establish open standards for the catalog and thus create the conditions necessary to simplify the development of appropriate business models and to boost competition within the digitization market." -- note here the idea that the use of orphan works is oriented toward business/competitive use ; not my focus
 * "It is also very much in the interests of publishing houses and collecting societies that their catalogs are included in the search. If they are, it will be possible to identify the rightholders of certain works and to conclude contracts on the digital use of those works."
 * "Wikimedia Deutschland e.V. believes that it is crucial that commercial partners are involved in searching for rightholders and in preparing works and making them available so that the necessary resources for the comprehensive projects exist. . . [and] the public sector struggles to handle large-scale digitization projects."
 * ". . . governments need to systematically rule out the “free-rider” problem. This refers to the danger that rightholders might wait until their work has been digitized and possibly “upgraded” (e.g. restored) before claiming back their rights and using the works themselves. To address this issue, countries need a regulation that would require rightholders to reimburse the cost of digitizing a work if they use it for commercial gain after it has been digitized."
 * "The current copyright situation means that much of the cultural material of the 20th century cannot be made publically accessible. This is not in the interests of users or rightholders. We therefore need a regulation that allows all those interested in doing so to make works accessible in a legally certain way that does not involve incalculable economic risks."
 * "we fear that the proposed regulation will not provide an adequate solution to the problem of orphan works, particularly as it is limited to only a few privileged players. . . it is unclear why the group of privileged institutions tasked with digitizing orphan works and making them publically accessible remains limited to such a small number." (this refers to aspects of a solution---designated organizations doing the digitizing---that does not seem to come up in the U.S. proposals)
 * "the catalog of provisions (Article 1(2) a-c) does not include photographs unless they have been published in books, magazines or journals. It can be said that the huge number of unpublished or uncataloged photographs for which the copyright and context of their origin can no longer, or only with great difficulty, be ascertained forms the main supply of orphan works. In the context of providing images for a free online encyclopedia such as Wikipedia, this material is an invaluable source for illustrating entries on topics like contemporary history. During the consultation process on the proposal for the directive, the interests of commercial photo agencies were apparently given priority over those of the general public. This is a major failing of the proposal." (the U.S. proposals do not seem to face this problem)
 * "the proposal for the directive still does not provide a genuine incentive for mass digitization. The text does not specify either the time period or the exact process involved in a “diligent search”. However, the search methods and parameters to be used must be defined in order to help prevent a serious underuse of cultural works in the future."
 * "it would be a good idea to encourage rightholders to make these works available under a free license. . . the ShareAlike clause contained in certain Creative Commons licenses can prevent “free riders” from acquiring works for their exclusive use."

More sources to review

 * Andrew Orlowski. The Register. 29 Apr 2013 UK.Gov passes Instagram Act: All your pics belong to everyone nowr
 * Materials from big Berkeley Law conference on orphan works. Wow.  Slides are there; they are reviewable.
 * After some more reading let's get a review from w:User:Levendowski who apparently helped organize the NYC edit-a-thon on intellectual property law and is a law student focusing on these topics.

Hearing notes
These are John's notes for the hearing on March 10–11. Most of these are viewpoints expressed by committee members, but I've interspersed a few notes to ourselves which are usually introduced with something like "We should...." On the first day I attended sessions I, II, IV, and V and skipped session III. On the second day I only was present for the audience participation sesson at the end of the day.

First day
SESSION I - Need for legislation
 * What are the recent "fair use cases"? Google Books?  Libraries seem split on whether these are adequate.
 * Having legislation prevents court decisions that vary by jurisdiction
 * Concern about orphan works as an "excuse" for commercial entities to exploit them.
 * Suggestion that music should be excluded completely
 * We need to consider whether fair use is enough, what's the "value added" of using orphan works
 * They're asking much more technical questions than we prepared for! So far:
 * Have recent court decisions on fair use changed your position?
 * Would it affect your position if courts overturned Google Books and other recent cases?
 * Would it affect your position if courts took the approach of expanding fair use to include socially beneficial uses?
 * Many creators don't have the resources to litigate.
 * Copyright's long term and lack of formalities captures a lot of things that weren't intended to be copyrighted by their creators
 * Infringers often aren't found due to their obscurity, uncertainty is the only protection for the creator. (Doesn't apply to Wikimedia, though.)
 * A lot of people think that expanded fair use is good enough, though not perfect
 * Is limitation of remedies the only solution?
 * A technical discussion of a US trade agreement, but implies the possibility that a law could impose formalities requirements that take effect 50 years after publication?
 * Maybe orphan works legislation could let details could be done by Copyright Office's internal rule making process
 * Yay James! Yay Ezra Pound!
 * DMCA-like process for orphan works? Is it too regulatory-like despite being simple-seeming?
 * Copyright is much easier navigated by sophisticated players rather than individuals
 * Everybody has this very fast, succinct manner of talking. We need practice on this.  It's our first time and we're learning a lot.
 * More questions from CO: something called a "savings clause", and do international efforts affect us
 * We should debrief on what we can do better to prepare next time we do something like this.
 * Marketplace can move faster than Congress, can create new business models
 * A one-line law that authorizes judges to consider orphan works status in deciding damages
 * We should emphasize that our use is very public and it is easy for rightsholders to notice and request removal of their work if needed. Also, we make works available for reuse, but it is the reuser's burden to make sure their own use is legal

SESSION II - Criteria for diligent search
 * The first speaker mentioned James' comments!
 * Too much variety to have hard guidelines on diligent search
 * Comment from Prager? that libraries have specialized skills that people in "real life" don't have
 * Best practices should be made by practitioners, who are both creators and users
 * But fair use is a flexible standard, why not orphan works?
 * We should comment on this idea of flexible vs. rigid guidelines
 * Diligent search shouldn't stop with the owner, it should include the original creator. It's a human rights issue!?  Especially in cases of rights termination.
 * Fair-use-like procedure results in judge-made law
 * Family photos and letters have privacy issues
 * Users shouldn't "reverse engineer" a diligent search after the fact
 * We should emphasize that our contributors are both GLAM and individual
 * Support for copyright records modernization
 * EU directive on orphan works?
 * Let insurance companies develop the guidelines
 * EU directive directed at GLAM, not very useful in other contexts.

SESSION IV - Types of works, especially photographs
 * Some contras exclude photographs unless they're embedded in a book
 * We should make clear that photographs are the most important type of work we want. Also mention that this does not override privacy concerns and we already have guidelines about privacy. Also, we support differences between non-profit and commercial use.
 * Association of Medical Illustrators guy is concerned that someone may separate an image from the journal article it's in and post it online, and then it would be an orphan work to a downstream user
 * This reminds me of all those medical images the 1918 edition of Gray's Anatomy that we use. Maybe that's another good example.
 * Again: uploading images stripped of metadata creates "instant orphans"
 * Canadian system considered a "failed system" -- they collect fees but the creators never surface
 * Should there be different provisions for different types of works?
 * Three categories: out of business, have name but can't contact, have no name
 * Orphan works and mass digitization are different issues
 * "Democratic access" to allow public to make use of digital technology, but this only really applies to GLAM. Limitations on reuse?
 * Creations especially in visual arts have a hard enough time making a living, putting a burden on them and restricting their remedies is harmful to them. But previous Copyright Office report suggested a small claims court for orphan works
 * The public is the prime beneficiary of copyright law, not the owner, according to the Constitution/Sony case, but not all agree

SESSION V - Types of uses and users
 * Should commercial uses be included? Should an distinction be made between commercial and noncommercial uses?
 * Split rights, for example a film writer retains the right to make a novelization after selling it to a producer
 * Simply republishing them versus making derivative works
 * Orphan works are not benefiting society--works not currently available can be made available
 * Not all orphan works scenarios are covered by fair use
 * What if a user thinks they performed a diligent search, but they really didn't?
 * Nonprofit organizations can still use works commercially
 * Keep it accessible to individuals.
 * Limitation of remedies should not disincentivize organizations from doing the digitizations

General notes for the first day
 * This was our first time doing something like this and we learned a lot. We were given a difficult job being that we got the very first session and we didn't know what to expect with the format.  Also it was 9AM on a Monday immediately after Daylight Saving Time started.
 * They asked a lot of specific questions about how recent developments have changed the panel members' views. They talked a lot about the Google Books ruling that what they were doing was fair use and no one was entitled to any compensation.  Some of the libraries said that this was good enough and if the ruling is upheld they don't feel orphan works reform is necessary, but not all of them took this viewpoint.  They also asked about recent orphan works reform in the EU and a long-standing system in Canada.
 * They mentioned further workshops in other cities, including New York and (I think) SF. There might be an opportunity for WMNYC and WMF to send representatives to these.
 * I made an audience comment at the end of the fourth session, on the scope of types of works. Earlier in the session the representative for the Association of Medical Illustrators and a few others mentioned fear that someone would take their images and post them on Twitter or the like without any metadata, thus creating "instant orphans."  I stated that most of our images come from either PD sources or are user-contributed, giving the 1918 edition of Gray's Anatomy as an example of the former.  I said that our interest is in using works for our educational mission, that come from the wide gap between 1923 and the last ~10 years when freely licensed works started being made, and that we don't want random images that someone improperly posted to Tumblr.
 * James and I discussed that it's a good idea to give further statements at the audience participation session late tomorrow. I think a good topic to emphasize is that we already make very limited use of fair use works under strict restrictions (no free equivalent, rationale needed, must be low resolution), and the expansion of fair use from the Google case won't expand the types of work we can use, while orphan works reform will as long as it is not overly burdensome to our diverse range of individual volunteers on the Internet.

Second day

 * I gave the following statement:


 * We are grateful for the opportunity to participate in these roundtables and give the unique viewpoint of the Wikimedia movement. I have two points I want to discuss based on the discussions we've heard here.


 * First, I want to emphasize that we want any reform to minimize burdens on our volunteers who contribute to our projects, who range from high school students to preeminent scientists, to people from foreign countries which include developing countries. The amount of burden in any reform will affect how we are able to use orphan works in the pursuit of our educational mission.  Any reform that imposes large burdens such as an escrow fee, or that does not limit or remove monetary remedies for non-profit educational use, would likely prevent us from being able to use them.


 * Second, We've heard a lot of discussion on fair use as a vehicle for orphan works reform, and I want to comment on that. Since our goal is provide material that is freely available, the Wikimedia projects make very limited use of fair use. In fact, our internal criteria for non-free content are much stricter than the law would allow for. We only use works under fair use if no freely licensed equivalent could possibly be created.  For example, we generally do not host non-free photographs of living persons or of existing buildings, because a free equivalent could be created by a volunteer taking a photograph themselves.  Thus dealing with orphan works solely by expanding fair use is unlikely to allow us to benefit from these works.  Thank you.


 * The organizers noted that they don't want us to just reiterate our spoken comments in our written statement, but want us to respond to what others said.
 * I happened to sit next to Lisa Shaftel from the Graphic Artists Guild, who had been part of the last two panels the day before. We had a brief conversation where she asked me why Wikimedia wants to use orphan works instead of paying creators to expand our content.  She referred to my example from the previous day of the 1918 Gray's Anatomy, saying that there are plenty of later works that we could use if we paid the rightsholders.  She seemed a little less upset when I told her that WMF has started to make some grants available.
 * More broadly, I think we should be clear in our written statement about the specific types of orphan works we’re interested in, which I think aren't the ones that the creator’s groups are opposing reform over. We mostly want to use photographs and documents in historical archives, I think, rather than anything anyone's trying to commercially exploit or make their living off of.  We should make it clear that we're not like Twitter or Tumblr, we don't republish copyrighted works flippantly and we strongly discourage our contributors from doing so.  (In fact, if you go the Wikipedia upload form you have to choose whether it's a free work, a fair use, or something you found somewhere on the Internet.  If you click the last of these, you get a very scoldy message telling you not to upload it.)
 * I also had a brief conversation with another panelist, Jeff Sedlik from the PLUS Coalition, which is essentially a free image search that gives you ownership information and metadata. He said he'd unsuccessfully tried to contact the Foundation about integrating this system into our upload procedures. I said that I thought it was more of a community issue and I’d figure out where on Commons he should start a discussion.