1. Introduction

1.1 The Legal Problem with Putting Twentieth Century Archives Online

Perhaps the single most important feature of the digital revolution is that for the first time since the Library of Alexandria, it is feasible to imagine constructing archives that hold all culture produced or distributed publicly…The scale of this potential archive is something we’ve never imagined before…but we are for the first time at a point where that dream is possible.1

Digital technologies now offer opportunities for searching and accessing archival2 cultural material on an immense scale. Governments, commercial enterprises and cultural institutions worldwide are testing the possibilities. But copyright has emerged as a major barrier, at least in relation to the creative output of the twentieth century. In part this is the unintended consequence of the Berne Convention3 system of international copyright protection – conceived in the pre-digital era. The Berne Convention is the foundation of modern copyright law and most countries are now signatories. It is based on two relevant key principles. The first is automatic protection – meaning that a copyright owner need not register his or her work4 or comply with any other formality for it to be protected by copyright.5 The second is absolute discretionary permission – meaning that any third party user wishing to reproduce the work requires the owner to ‘opt-in’ and licence the use, unless a particular exception or statutory licence applies.6

While the Berne framework has greatly assisted the development of copyright based creative industries around the world, these twin principles create difficulties for digital users of archival works which are still in copyright. As owners of copyright interests in the material are not required to be registered, they are not required to be easily located. As archival material is, by definition, not recently created, it is often the case that the primary use or market for which the material was originally licensed was pre-digital, and no advance permission was given for digital use. So, the owner, or owners, must be found to check whether or not they will permit the new digital use. Here is the difficulty. The copyright owner may since have gone out of business, changed addresses or business names, died or in the case of legal entities, been deregistered. When the owner cannot be found, or at least not easily found, the work is termed an “orphan work”. However, no precise definition is settled.7 As discussed below, there is debate over how difficult it must be to find an owner – and how diligent the search – before the work really is an orphan.

While this paper focuses on the issue of orphan works and how they are dealt with in relation to in-copyright, out of print books in the Google Books Settlement, it should be acknowledged that there is a related subset of copyright barriers around digital archive projects which are not strictly about orphan works. These include difficulties to do with getting a response from a copyright owner who is found; the transaction costs of individual negotiations; tracing ownership where copyright interests have been assigned, willed to estates, come under company administration or otherwise dealt with over time; and over the “tangle of uncertain copyrights”8 concerning digital rights in pre-digital works. In the recent US Copyright Office Inquiry into orphan works,9 a great many submissions identified problems which were not strictly about orphan works.10 Many instead concerned the kind of copyright barriers just described.11

Orphan works and these related copyright barriers apply to all kinds of media, but are particularly vexing in relation to recorded media such as television archives, which have multiple layers of underlying rights owners such as script writers, actors, footage owners and composers.12 These barriers can add significant cost, delay and risk to proposals to make archival copyright material digitally available. Lawrence Lessig notes the irony that, because of copyright barriers, it is now easier to search and access cultural material created in the nineteenth century:

How is it that we have created a world where researchers trying to understand the effects of media on nineteenth-century America will have an easier time than researchers trying to understand the effect of media on twentieth-century America?13

1.2 The Particular Problems of Large Scale Access Digitisation Projects

In large scale digital access projects these copyright barriers manifest in sizeable transaction costs and risks.14 The biggest of these by far is the Google Library Project, which to date has digitised some seven million books. There are many others in progress around the world, including other text-based projects;15 projects relating to film16 and broadcasts;17 and portals that collect and display all kinds of cultural material.18 As collective licensing of rights in archival material is usually at best patchy in the digital environment, many large scale projects get bogged down in searching for individual copyright owners, negotiating with those found, and assessing the risks where permissions cannot be obtained. The Berne system protects the rights of copyright owners to give consent to reproductions and other uses of their works, and where silence does not constitute consent. But for many users the resulting transaction costs create entry barriers to the development of new digital content access models.

The difficulties posed by orphan works have been widely recognised, and various legislatures have considered or implemented reforms to their copyright legislation. The particular problem of orphan works in large scale digital access projects was acknowledged by the US Copyright Office in its Inquiry into Orphan Works: Final Report (2006).19 However, the recommendations it adopted did not address the special needs of these projects, in particular the need to reduce transaction costs. Rather, it favoured the “diligent search” model, variations of which already exist in jurisdictions such as Canada and Japan.20 The diligent search model requires the would-be user to attempt to locate a copyright owner by carrying out certain prescribed or “reasonable” activities such as searching industry registers and publishing notices of intended use. Delay is inevitable – particularly if an application for use must be made to a regulatory body. Depending on the particular scheme, the reward for going through this process will either be a licence to use the material (if granted by the regulatory body), or some form of reduced legal risk if the owner subsequently emerges and objects to the use.

The “reduced risk” variant of the diligent search model was put forward in the subsequent draft legislation introduced to the US Congress.21 Broadly, the texts of these bills provide that a user who makes reasonable efforts to locate the owner of an orphan work in accordance with the relevant search criteria will be entitled to certain legal immunities against copyright remedies and legal costs if the owner later comes forward and objects to the use. This is certainly better than nothing and does go some way towards reducing legal risk for users. But it does little to reduce transaction costs in large scale digital access projects. As one commentator puts it, these bills “conceive of orphans as adoptable on a case-by-case basis only, not at scale, with all the in-depth investigation that the analogy to adoption suggests.”22

Even once the searches have been carried out, under the US bills protection from legal risk and further cost would not be absolute. The copyright owner who emerges and objects to the use is still entitled to seek damages unless “reasonable” compensation is negotiated. The potential for dispute under such a complex model is obvious.

Diligent search models may work reasonably well to balance the interests of owners and users in the case of one off “adoptions” where the user is well resourced with plenty of time, but they are clearly a compromise of interests which do not adequately address the needs of large scale users in the digital environment. Nor do they address the concerns of many copyright owners who perceive unfair compromise to the twin bedrocks of automatic protection and absolute discretionary permission, afforded to them under the general rules of the Berne copyright system. The US orphan works bills attracted strong criticism from some copyright owner groups23 and did not pass Congress. They now appear to be in limbo.24

2. The Google solution to orphan works

The US orphan works legislation was controversial enough. But as the bills were going through their public debate process, settlement of the Google Book Search litigation25 was being negotiated in private. The resulting Google Book Settlement26 is akin to dropping an atom bomb on the copyright system. It makes the US orphan works bills look positively coddling of copyright owners. Most significantly, the Google Book Settlement is attracting controversy over its use of the opt-out provisions of class action law. If the settlement is endorsed, the parties will be able to exploit a vast digital archive of books while sidestepping the copyright barriers associated with the Berne opt-in system. Nobody else “has ever been able to so dramatically flip the default position of copyright law.”27

This paper evaluates the orphan works provisions of the Google Books Settlement as they await judicial fairness hearings now set to commence on 7 October 2009.28 These provisions – if endorsed by the Court – will create, via class action rules, what one commentator has called a “magic device” for solving orphan works issues.29 However, this device will directly benefit only Google – and give it a huge competitive advantage in a field it already dominates. The scheme has been described by one commentator as “an elaborate scheme for the exploitation of orphan works.”30

I argue that whatever its particular flaws, the model set up by the Google Books Settlement is a useful template for thinking more generally about the needs of mass archive digitisation projects and, ultimately, the copyright balance in the digital environment. Legislatures who proactively solve these orphan works issues will boost their citizenry’s knowledge and foster local digital innovation. We should therefore think about whether Google Book Settlement-type schemes can be made more generally accessible in the public interest.

2.1 The Google Book Search Project

Google has built its fortune on supplying the world’s most competitive free Internet search engine, supported by online advertising sales. Google’s corporate mission is to “organise the world’s information and make it universally accessible and useful.”31

Some time in the early 2000s Google began its Google Book Search project. At this stage it did so through its Partner Program, where publishers licensed to Google the right to digitise and index books for online searching by the public:

Our agreements for those books allow us to display the page containing the searched-for term and a few surrounding pages. We also show links to enable readers to buy the book and we may show advertising related to the content of the pages, from which the publisher can receive additional revenue.32

Reportedly there are around a million licensed books in the Google Partner Program.33

In 2004 Google announced an expansion to the project in the form of the Google Library Project. The cornerstone was a partnership with several prominent libraries including Harvard, Oxford and the New York Public Library. These participating libraries allowed Google to scan their vast collections of copyright and public domain books from around the world for inclusion in the Google Book search index. In exchange, their reciprocal benefits included receiving digital copies of the scanned books, and, presumably, gold-plated indemnities against the (subsequently realised) risk of being sued for copyright infringement in relation to the scanned copyright books. Out of copyright, or public domain, books were made available online in full. In-copyright books could not be fully accessed, but online viewers could see searchable portions or snippets to, as Google puts it, “help people decide whether to buy the book or look for it in the library.”34

The Google Library Books Project differed from the earlier version of Google Book Search because Google had permission to access physical copies of the books, but not copyright permission to scan and display snippets of those still in copyright. Google did not seek prior permission from the copyright owners of these in-copyright books. Rather, it claimed that permission was not needed because its activities were covered by the fair use doctrine of United States copyright law.35 Nevertheless, it stated that it would allow “opt-out” by rights owners who contacted it to request a book not be included, and would also continue to offer agreements to those publishers who came forward to become partners in the Book Search programme.

It is estimated that the total number of books now digitised by Google and included in its Book Search index is around seven million and counting. Of these, it is reported that one million are licensed, one million are public domain, and the remainder are in-copyright but allegedly out of print.36 Google argues that these out of print, in-copyright books are niche product which may potentially be revived by their inclusion in Google Book Search. Many of these, Google claims, are orphan works. In Google’s view: “[o]rphan works represent an untapped wealth of information that can and should be made accessible to the public.”37

2.2 The Google Books Litigation and the Google Books Settlement

In 2005 the American Authors Guild and the Association of American Publishers commenced United States class copyright infringement actions38 against Google. Google maintained its defence of fair use. The merits of the litigation have received abundant academic attention,39 and it is not proposed to speculate here about whether or not Google would have established its defence. On 28 October 2008, the parties announced that they had reached a provisional settlement – the Google Books Settlement – subject to the ratification of the court in fairness hearings. As noted above, those hearings are now set for 7 October 2009, in New York.

The Google Books Settlement is very long and complex, and a summary and FAQ are available online.40 Its notable elements, for the purposes of this paper, are as follows:

• Google will pay a compensatory cash amount to each copyright owner of a book41 scanned without permission before 5 May 2009;42

• Google will be permitted to continue digitising books for its Library Search project, and to make various uses of the scanned books including displaying snippets, selling advertising around the displays and selling institutional subscriptions, subject to a 63% revenue share split with copyright owners;43

• an independent Book Rights Registry, directed by author and publisher representatives, is to be established at Google’s expense. The Book Rights Registry will act as a new collective rights management organisation distributing compensation payments and revenue share to copyright owners, locating missing owners, resolving rights disputes and building the rights database;44

• participating libraries, researchers and new educational subscribers can have access to the entire repertoire on certain terms – some paid and some not;45

• Google and the Book Rights Registry may agree on new commercial models for exploiting the books, and

• the US public will gain greater access to books online by being able to view greater portions of texts than would have previously ever been possible.

The plaintiffs have succeeded in building in a high level of discretionary control for individual copyright owners who actively manage their rights. The rights granted to Google are non-exclusive and each copyright owner can by itself or collectively (via deals struck by the Book Rights Registry) simultaneously licence their copyright interests to third parties.46 Copyright owners can also remove books or exclude certain specific uses on an ongoing basis, including the exclusion of a book from any new exploitation model that Google agrees with the Book Rights Registry.47 The Settlement preserves flexibility for all parties to maximise their respective market opportunities while providing certainty in relation to the legal status of rights governed by the Settlement. In this way the parties have achieved a far more innovative solution than they could have if they pursued the matter to final resolution in court.

For those copyright owners who do not wish to be included in the Settlement at all, class action rules provide that they must take steps to opt out by a cut-off date. These copyright owners will not be entitled to claim their compensatory cash payments, nor will they be entitled to any other benefits of the Settlement – the Settlement will not affect their rights or Google’s defences should they decide to sue separately.48 The opt-out date has now been extended to 4 September 2009.

If a copyright owner does not opt out by the cut off date, it will be deemed to be included in the Settlement. Google will be able to digitise and make the permitted uses of the book. The Book Rights Registry will hold any unclaimed monies on trust for five years from when it is reported by Google. If the owner does not come forward within that period, the monies will be rolled back into the revenue pool administered by the Registry, with most of it ultimately being redistributed to other owners.49

Opt-out systems are a common device in class actions to ensure the efficient administration of large cases. Individual members can opt out and pursue their own legal action, but if they do not, they will be bound by whatever the result is in the case brought on their behalf. But its use in this particular litigation is revolutionary – and controversial – for two reasons. First, the Google Books Settlement applies to an enormous class – the class of authors and publishers whose books were published before 5 January 2009, and whose work is protected under United States copyright law. Due to the reciprocal effect of the Berne international copyright treaty system, this means most books which are still in copyright around the world. As one commentator has pointed out, those directly represented by the plaintiff bodies who originally sued Google were “only a small fraction of authors and publishers”50 compared to those whose rights are now implicated by the Settlement. Notices have been published in newspapers around the world and author and publisher bodies in various jurisdictions are providing information to their members. But it seems the enormity of the Settlement is still sinking in around the world.51

Second, the Google Books Settlement does much more than the usual terms of a deed of settlement between class litigants. The Google Books Settlement includes a cash payment to the plaintiffs and a release of liability for past actions, but, as outlined above, it does much more. The Google Books Settlement allows Google to continue and expand the possibilities of its Library Project; sets up a newly minted collective rights organisation to administer rights and payments; and, most significantly, solves the orphan works problem without the need for any bills to pass or any general public inquiry to be held.

The implications of the Settlement for orphan works significantly change the existing opt-in position under copyright law. For Google, the Settlement will remove the risks of using orphan works in ways which may not fall within the fair use defence. It can use orphan works for free, without having to try and find the owners first, and keep its share of the revenue it makes from exploiting them. For orphan works holders, their rights to sue will die should they fail to become aware of their rights and opt-out of the Settlement by the cut-off date. If they fail to claim their share of revenue generated by Google within five years of it being reported to the Book Rights Registry, they also forfeit this revenue.52

It may be that many copyright owners, although not previously aware of the Settlement or directly represented by the plaintiffs, may now decide they are happy with the level of control they are granted by the opt out, exclusion and removal provisions of the Settlement, and with the opportunity to receive a share of the revenue Google makes out of books that may not otherwise have earned anything for years. The Settlement does not just reflect only Google’s interests as a user but the interests of large representative publisher and author bodies, and of the participating libraries. It is in this sense a market solution to the issue of orphan works.

On the other hand, it is also possible that the terms of the Settlement will fail to obtain court approval. The debate over whether the Settlement is fair is already ramping up in the public arena, and several motions for intervention have been lodged.53 While some interested parties have raised privacy, consumer protection, public access and other concerns, most of the controversy centres on the orphan works issue – with critics strongly objecting to Google’s new monopoly (by virtue of the class action opt out rules) on the legal use of orphan works.54 Some argue this raises anti-trust concerns55 and the US Department of Justice has recently commenced an anti-trust investigation.56 However, others object on the basis that it is not in the public interest for orphan works problems to be privately resolved when so far there is no general legislative solution.57 Others are concerned that monies are being redistributed away from orphan works’ owners (for the purposes of the Settlement, they are those who fail to emerge and make themselves known to the Book Rights Registry within 5 years).

Not everyone agrees with these criticisms. Some argue that Google deserves any advantage it has gained over orphan works via the Settlement by taking the enormous risk and investment of resources involved.58 Others argue that the percentage of books which are really orphan works is far lower than some suggest, and that the effect of the scheme will be to lower that number even further by encouraging owners to come forward and claim their share of cash and revenue share under the Settlement: “[a]s the registry starts sending out royalty checks, books will exit the orphanage in a rush.”59 It is also pointed out that once these works cease to be orphans, the Settlement will not prohibit licensing of these digital rights to third parties – either collectively via the Book Rights Registry or on an individual transaction basis – as Google’s licensed rights will be non-exclusive.60 The debate will continue in court.

The Google litigation is merely one manifestation of the continuing tension between “old” content creation models, which (broadly speaking) favour the strengthening of copyright owner rights and the opt-in system; and “new” digital enterprises which aggregate or facilitate distribution of content found online – seeking to expand user rights in the copyright material of others. The tension can be summed up by the contrast between Google’s view of itself as an “archivist”61 and of the competing view that such search engines are “parasites or tech tapeworms in the intestines of the internet.”62 The results of the fairness hearings, and the body count of copyright owners who decide to opt out of the Settlement, will be indicative of which view may prevail in practice, and ultimately how successful this model of compromise between content and search might be. In the meantime, it is interesting to consider how the Google scheme could be applied as a public model to orphan works in mass digitisations of other cultural material.

2.3 Possible Applications of the Google Books Settlement Principles

In his recent article in the New York Review of Books, author and Harvard librarian Robert Darnton suggested that Google has privatised an opportunity to create a digital Library of Alexandria which, in retrospect, should have been a public initiative:

Looking back over the course of digitisation from the 1990s, we can now see that we missed a great opportunity… We could have created a National Digital Library – the twentieth century equivalent of the Library of Alexandria. It is too late now. Not only have we failed to realize that possibility, but, even worse, we are allowing a question of public policy – the control of access to information – to be determined by private lawsuit.63

But perhaps it is not too late. The Google scheme inherently relies on regulatory intervention – in this case via the class litigation rules – to underpin it. The copyright system itself is a legal intervention in the market, which is in a constant feedback loop responding to new technologies and the uses they permit. The policy initiatives on orphan works so far have been slow to address the needs of the digital environment, but they can still catch up.

A basic starting point for considering a public licensing scheme is collective licensing. Collective licensing is far more efficient from a user’s point of view than individual licensing, as only one licence need be negotiated or registered for, standard fees or tariffs apply, and no upfront clearances need be carried out. The Book Rights Registry will be a new collective licensing body, albeit one which, unusually, gains its authority by court order. Some collective licence mechanisms already deal with the issue of orphan works in various ways.

Most relevant to current purposes are statutory, or compulsory, licences and, in particular, “universal” legal licences (e.g. those permitting use of all copyright material falling within the class of rights even where the copyright owner does not authorise use).64 This type of universal scheme naturally provides legal use of orphan works included in the class and type of use covered by the scheme, with no need for the user to make inquiries or even realise that a particular right is orphaned before the use occurs. Attempts to locate copyright owners whose works are included in such schemes are undertaken by the appointed collective management organisation administering the scheme after reviewing records of reported usage. As with the Google Books Settlement, unclaimed revenue is typically held on trust for a certain period while attempts are made to locate the owner. Once a rights-holder is found, it is invited to become a member to receive payment.

A related model is extended collective management schemes, or extended collective licensing (as is found in the Nordic countries).65 These schemes legally extend, via statutory provision, the representation of collective management organisations to deal with a class of works to copyright owners who are not members of the organisation, subject to opt-out by notification. As one commentator has noted, this type of scheme is “actually quite close to the scheme proposed for the [Google Books] settlement.”66

Both these models of collective rights management provide a low transaction cost model for solving orphan works problems by removing the need to find the owner upfront. The model is “user uses, owner registers to be paid.” However, the Berne convention requires that exceptions to the absolute discretionary rights of copyright owners to be limited to “special cases”,67 and this is a potential obstacle to a large scale paradigm shift to Google-style opt-out schemes in the digital environment.68

A variation on this idea is offered by Darnton. With hindsight, he suggests, a public version of the Google Books project could have been built by a coalition of libraries, supported with action by Congress, who, “could have done the job at a feasible cost and designed it in a manner that would put the public interest first.”69

He suggests briefly that the idea of a “rental based on the amount of use of a database” could have been one possible way the scheme could have worked.70 Such a scheme could still work if the details could be satisfactorily worked out. After all, the Google Settlement is a non-exclusive scheme. Authors and publishers remaining in the Settlement can still licence their books to other digital access schemes if they wish, or switch off uses they wish to licence exclusively elsewhere. There is no reason why a publicly accessible solution cannot now be found – nor why such schemes could not be applied outside the medium of print.

Other commentators have since suggested that perhaps the terms of the Settlement itself could be opened up to allow other would be digital users to obtain the benefit of the class action indemnity in addition to Google, or to have Google grant some kind of compulsory licence to them.71 But this would arguably stretch the class action rules to breaking point,72 and in any event, would likely be resisted by Google who has taken very significant risks in becoming the defendant in major copyright infringement litigation to get to the point it has. It is likely that other digital users will, unfortunately, have to take their own litigation risks or wait for orphan works legislation before they can safely use works of uncertain ownership in their projects.73

Finally, in reviewing how the Google Books Settlement principles could be applied to public models, it is worth considering Google’s own submission to the US Orphan Works Inquiry.74 This submission – made in 2005 when the Google Library Project was at the height of its controversy – is interesting for two reasons. First, because of the philosophy Google applies to orphan works, and secondly, for its suggestions on how legislative change might have been introduced to solve the orphan works issue without recourse to the class settlement which has since overtaken the issue.

Google’s philosophy on orphan works is to query the presumption that owners who cannot be contacted to give permission to use their works do not want any use of their work to be made. Google calls this a “conservative assumption mandated by the current state of copyright law”,75 and argues that the class of orphan works owners must include at least some who do not care if others use their works. Public policy demands, in Google’s view, that the presumption of non-use be altered so that orphan works can be removed from “purgatory” for the benefit of human knowledge.

Google’s suggested solution was that a register system be used to determine which owners “signal a lack of interest in enforcing their copyrights by not updating the records of their ownership.”76 If an owner did not update their ownership records, then, Google argued, certain uses of the work – to be determined in the legislation – should be permitted without risk. Google also suggested that any such register should be capable of automated search – a request which foreshadows the future of peer-to-peer machine licensing in mass digital projects. However, there is doubt over whether such a registration requirement would fall foul of the Berne “no formality condition to protect” principle.77 Arguably, it depends on how the registration requirement is framed. Some submissions to the US Orphan Works Inquiry argued that any form of opt-out registration scheme would infringe this Berne requirement.78 However, at least one commentator argues that not every form of opt-out scheme requiring an owner to register its objection to a use would infringe – for example the Nordic extended collective licensing schemes, discussed above, which operate on an opt-out principle.79 This issue bears further inquiry.

All of the above suggestions are only theoretical at this stage, and will attract concerns that will need to be worked through. The Google Settlement is tailored to the particular interests and concerns of the parties to the litigation at hand and the public so far has had no direct role in shaping it. In contrast, any kind of new scheme developed to address orphan works in digital projects will need to satisfy a number of criteria. First, it will need to fill a specific perceived need in relation to use of the kind of material in question. It will need broad industry support. It will need to address policy concerns about consumer protection and public access on equitable terms. It must comply with international treaty obligations. It must ensure that incentives for creators remain strong. It must be flexible enough to permit new developments in licensing models which enable the industry or national economy in question to maintain international competitiveness. And last, but not least, it must retain the key benefits of low transaction costs and legal certainty which large scale digital users require.

This is no easy task, but it is not too late to watch and learn as the Google Book Settlement makes its grand attempt to reform copyright law by court action. And as Google has pointed out, “fortunately there isn’t an either-or choice between the legislation and the settlement.”80 Google supports new attempts to introduce orphan works legislation and says that passing such legislation “will remain one of Google’s priorities.”

3. Conclusion

Legislators are yet to find a satisfactory solution to the problem of orphan works in mass digitisation projects. Meanwhile, Google is developing its own private scheme. If the parties to the Google Book Settlement succeed in getting this scheme judicially sanctioned and winning over a critical majority of copyright owners, they will have created, with the court’s assistance, something akin to a new statutory licence for digital use of archival copyright material. They will neatly solve the orphan works issues which plague digital archive projects. They will also unlock revenue and public access – at least to a greater extent than previously possible – to such material. On these grounds alone they will have succeeded where many have failed. As one supporter of the Settlement observed: “[o]nce in a while something big is so suddenly achieved to be truly revolutionary in scope. The Google book project surely fits in this rare category.”81

The Google Books Settlement model inspires us to think of how its benefits could be made accessible on a public model capable of application in many jurisdictions. Copyright owners, digital users and governments around the world will be watching this new phase of the Google Library Project with interest.

---

* Senior Lawyer, Special Broadcasting Service (SBS) and board member, Audio-Visual Copyright Society Ltd (Screenrights). This paper was originally presented at the Unlocking IP Conference, “National and Global Dimensions of the Public Domain”, University of New South Wales, Sydney, Australia 17 April 2009. My thanks to Jan Constantine of the Authors Guild, Ishtar Vij of Google, Graham Greenleaf and the anonymous reviewers for their useful comments. Thanks also to Francis Johns, Natalie Buck and Michael Easton for help with research. The opinions expressed in this paper are my personal views and not those of SBS or Screenrights.

1 L Lessig, Free Culture (2004) available at http://www.free-culture.cc/, at ch 9.

2 In this paper “archival” is used loosely to refer to works for which the primary use has expired or which are no longer in wide circulation in their original form. Examples would include books which are out of print, films no longer in distribution or computer games made for obsolete formats.

3 Berne Convention for the Protection of Literary and Artistic Works (Paris, 1886 as revised) (hereafter Berne). The relevant principles of Berne have been backed by World Trade Organization dispute procedures since the introduction of Article 9(1) of the Agreement on Trade-Related Aspects of International Property Rights (Marrakesh, 15 Apr 1994) (hereafter TRIPS).

4 For convenience, in this paper I use the term “work” as shorthand for all material protected by copyright under international treaty law as reflected in Berne, TRIPS and associated treaties concerning copyright.

5 Berne, above note 3, article 5(2).

6 Ibid, articles 9(1), 9(2).

7 US Copyright Office, Inquiry into Orphan Works: Final Report (Jan 2006) (hereafter US Orphan Works Inquiry) available at www.copyright.gov/orphan, at 21. (accessed 31 July 2009).

8 M Lemley, “An Antitrust Assessment of the Google Book Search Settlement” (2009) available at http://ssrn.com/abstract=1387582 (accessed 31 July 2009).

9 US Orphan Works Inquiry, above note 7.

10 Ibid, at 40.

11 See, for example, the submission of UCLA film and Television Archive (25 March 2005) available at http://www.copyright.gov.au/orphan/comments/index.html (accessed 31 July 2009). Para 2(c) refers to the problem of disputed or unclear rights in films, using the example of the Paramount film “Eight Girls in a Boat.” Other parts of the submission deal with “true” orphan works issues.

12 For a discussion of the difficulties of clearing public broadcaster archives for online reuse, see S McCausland, “Getting Public Broadcaster Archives Online: Orphan Works and Other Copyright Challenges of Clearing Old (but In-copyright) Cultural Material for Digital Use” (2009) 14 Media Arts Law Review, 140.

13 L Lessig, above note 1, ch 13.

14 E Fraser, “Antitrust and the Google Books Settlement: the Problem of Simultaneity” (draft paper, 23 June 2009) available at http://ssrn.com/abstract=1417722 1-26, at 8, commenting that the transaction costs would be “unimaginable” if Google had undertaken licensing of all books included in its Library Project.

15 See, for example, G St Clair, “The Million Book Project in Relation to Google”, (2008) 47 Journal of Library Administration, 157-163.

16 See, for example, the European Film Gateway, a Consortium of European Film Archive Bodies http://www.europeanfilmgateway.eu/ (accessed 31 July 2009).

17 Such as those undertaken by the British Broadcasting Corporation (BBC) and Japan Broadcasting Corporation (NHK), each of which have recently provided online public access to 1,000 hours of archival programming. For a review of these projects and the copyright clearance issues they faced, see S McCausland, above note 12.

18 One of the largest multi-media portals is the non-profit Internet Archive http://www.archive.org/index.php (accessed 31 July 2009).

19 This category of project was separately considered in the US Orphan Works Inquiry, above note 7, at 43.

20 For a review of these various legislative schemes see Ian McDonald, “Some Thoughts on Orphan Works” (2006) 24 Copyright Reporter, 180-198.

21 Orphan Works Act of 2008 (HR 5889) and Shawn Bentley Orphan Works Act of 2008 (S. 2913). The bills can be found at www.copyright.gov/orphan (accessed 31 July 2009).

22 G Harper, “Mass Digitisation”, blog post (31 May 2008) available at http://blogs.tdl.org/digitize/2008/05/31/google-book-search-2 (accessed 31 July 2009).

23 The Bill was opposed by visual artists and photographers, the MPAA and other groups concerned at the potential effect on unauthorised use of their works, particularly in the context of the internet where photographs, images and film clips are easily orphaned or pirated. See http://www.myspace.com/opposeorphanworks and http://orphanworks.blogspot.com/ (accessed 31 July 2009).

24 See http://www.govtrack.us/congress/billtext.xpd?bill=s110-2913 (accessed 31 July 2009). The bills passed the Senate but did not pass by the end of the 110th session of Congress, and are therefore “dead.” They could possibly be reintroduced in a new session.

25 Authors Guild, Inc. v Google, Inc., No. 1:2005cv08136 (S.D.N.Y. filed Sept. 20, 2005); McGraw-Hill Cos. v Google, Inc., No. 1:2005cv08881 (S.D.N.Y. filed Oct. 19, 2005).

26 Settlement Agreement, The Authors Guild, Inc., Association of American Publishers, Inc., et al. v Google Inc. No. 1:2005cv08136 (S.D.N.Y. Oct. 28, 2008) available at http://www.googlebooksettlement.com/intl/en/Settlement-Agreement.pdf (accessed 31 July 2009).

27 Fraser, above note 14, at 8.

28 The original fairness hearing date was 11 June 2009, however this was vacated due to the extension of the opt-out date for the class action from 5 May 2009 to 4 September 2009: http://www.googlebooksettlement.com/help/bin/answer.py?answer=118704&hl=en#extension_explanation (accessed 31 July 2009).

29 J Grimmelman, “How to Fix the Google Book Search Settlement” (2009) 12 Journal of Internet Law, 10-20, at 14.

30 Brewster Kahle, “It’s All About the Orphans”, blog post (23 February 2009) available at http://www.resourceshelf.com/2009/02/25/brewster-kahle-it’s-all-about-the-orphans (accessed 31 July 2009).

31 Google corporate mission statement available at http://www.google.com/corporate (accessed 31 July 2009).

32 Google Inc, Response to Notice of Inquiry Regarding Orphan Works, submission dated 25 March 2005 to US Orphan Works Inquiry (hereafter Google Orphan Works Submission) available at www.copyright.gov/orphan (accessed 31 July 2009), at 2.

33 J Perez, “Google Book Settlement, Business Trumps Ideals”, PC World (30 October 2008) available at http://www.cio.com/article/458176 (accessed 31 July 2009).

34 Google Orphan Works Submission, above note 32 at 2.

35 17 USC § 107. See further Ruth Allen, “Google Library: Why all the Fuss?” (2005) 23 Copyright Reporter, 105-111.

36 R Darnton, “Google and the Future of Books” (2009) 56 New York Review of Books, available at http://www.nybooks.com/articles/22281.

37 Google Orphan Works Submission, above note 32, at 2.

38 Above note 25.

39 See eg Allen, above note 35, M Williams, “Recent Second Circuit Opinions Indicate that Google’s Library Project is Not Transformative” (2007) 25 Cardozo Arts & Entertainment Law Journal, 303-331.

40 http://www.googlebooksettlement.com The terms of the Settlement Agreement (hereafter Settlement) are available at www.googlebooksettlement.com/agreement.html (accessed 31 July 2009).

41 Ibid, art 1.16. The definition of “book” applies only to paper copy books published before 5 January 2009, and excludes periodicals, sheet music and unpublished papers.

42 Ibid, art 2.1(b).

43 Ibid, art 2.1(a).

44 Ibid, art 6.

45 Ibid, art 7.

46 Ibid, art 2.4.

47 Ibid, art 3.5.

48 Ibid, art 17.33.

49 Ibid, art 6.3.

50 P Samuelson, “Legally Speaking – The Dead Souls of the Google Book Search Settlement” (2009) 52 Communications of the ACM, 28-30, at 29.

51 For example, European Union regulators have announced an inquiry into the Google Books Settlement to consider its effect on European books: “EU asks Publishers for Feedback on Google Books” (20 July 2009) Associated Press, available at http://finance.yahoo.com/news/EU-asks-publishers-for-apf-3599301835.html/print;_ylt=A0S00tDFmWRKF38AIBbeba9_?x=0 (accessed 31 July 2009).

52 Settlement, above note 40, art 6.3.

53 Z Elinson, “Fate of Orphan Works Stirs Opposition to proposed Google Books Settlement” New York Law Journal (2009) available on legaltrac.

54 B Kahle, “A Book Grab by Google” (2009) The Washington Post; Samuelson, above note 50, Grimmelman, above note 29.

55 See eg Grimmelman, ibid; R Picker, “The Google Book Search Settlement: A New Orphan-Works Monopoly?” (2009) 462 Journal of Competition Law & Economics, forthcoming; U of Chicago Law & Economics, Olin Working Paper, available at http://ssrn.com/abstract=1387582 (accessed 31 July 2009)

56 M Helft, “US Presses Anti-trust Inquiry into Google Books Settlement”, (9 June 2009) New York Times, at 35.

57 Kahle, above note 54, Samuelson, above note 50. The Internet Archive, of which Kahle is a principal, has filed a motion to intervene with the Court: Elinson, above note 53.

58 Lemley, above note 8.

59 Roy Blount, President of the Authors Guild, quoted in A Pham, “Authors Guild defends Google Books Settlement” (25 June 2009) LA Times available at http://latimesblogs.latimes.com/technology/2009/06./googlebookssettlementauthorsguild.html (accessed 31 July 2009).

60 Settlement, above note 40, art 2.4, Lemley, above note 8. However, others question whether the Book Registry will have sufficient incentive to licence to third parties: Fraser, above note 14, at 23.

61 Google Orphan Works Submission, above note 32, at 5.

62 Robert Thomson, editor, Wall Street Journal, quoted in J Schulze, “Google Dubbed Internet Parasite by WSJ Editor”, (6 April 2009) The Australian available at www.australianit.news.com.au/story/0,24897,25295383-15318,00.html (accessed 31 July 2009).

63 Darnton, above note 36, at 5.

64 For explanation of forms of collective licensing see T Kiskinen-Olsson & P Greenwood, IFFRO General Papers, ch II Reprographic Reproduction, (1997) International Federation of Reproduction Rights Organisations, available at www.ifrro.org/upload/documents/Reprographic-Reproduction-1997.pdf (accessed 31 July 2009).

65 For an explanation of these schemes see D Gervais, Application of an Extended Collective Licensing Regime in Canada: Principles and Issues Relating to Implementation (study prepared for the Department of Canadian Heritage, June 2003) available at aix1.uottawa.ca/~dgervais/publications/extended_licensing.pdf (accessed 31 July 2009).

66 B Lang, “Orphan Works and the Google Books Settlement – an International Perspective” (23 July 2009), available at http://www.datcha.net/ecrits/liste/orphan-gbs.pdf.

67 Berne, above note 3, art 9(2).

68 For a discussion of whether extended collective licensing schemes are required to satisfy the Berne three step test, see D Gervais (ed) Collective Management of Copyright and Related Rights (Kluwer Law International, 2006), ch II M Ficsor, “Collective Management of Copyright and Related Rights in the Digital, Networked Environment”.

69 Darnton, above note 36, at 5.

70 Ibid.

71 Grimmelman, above note 29, at 15. See also Fraser, above note 14, at 23, citing Picker, above note 55.

72 Macmillan CEO John Sargent has acknowledged that “at times the parties were working ‘at the edges of class action law’” during their negotiations to reach the Settlement: quoted in A Albanese & C Reid, “Google Supporters ready to state case”, (8 June 2009) Publishers Weekly available at http://publishersweekly.com (accessed 31 July 2009).

73 Fraser, above note 14, at 23-24.

74 Google Orphan Works Submission, above, note 32, at 3-4.

75 Ibid, at 3.

76 Ibid, at 3-4.

77 Above note 3, art 5(2).

78 See submission of the MPAA, above note 32. Google’s submission also acknowledges that the point would need to be considered if its submission were to be implemented.

79 Gervais, above note 65, at 19.

80 D Slater, “Google Book Search settlement and access to out of print books”, (2 June 2009) Google Public Policy Blog, available at http://www.crunchbase.com/bloggerboard/tech/publication/google-public-policy-blog (accessed 31 July 2009).

81 Charles Brown, advisor to the President of the US National Federation for the Blind, quoted in G Gross, “Civil Rights Activists Champion Google Book Deal”, (29 July 2009), IDG News Service, available at http://www.pcworld.com/businesscenter/article/169275/civil_rights_activists_champion_google_book_deal.html (accessed 31 July 2009).