The problem that orphan works pose for cultural heritage institutions who want to enable the widest possible digital engagement with our shared cultural heritage is substantial. There has been a twin-track response to this problem in the UK, in the form of the exception implemented under the Orphan Works Directive operating in tandem with the Orphan Works Licensing Scheme (OWLS). This approach offers greater opportunities than many other European countries for developing digitisation initiatives that can be tailored to institutional collections, needs, budgets and ambition. Crucially, however, both schemes are tethered to the requirement of diligent search.
Within Europe, Member States are required to determine what sources should be consulted as part of the diligent search requirement; at the same time, their guidance must include 'the relevant sources listed in the Annex' to the Directive. Different jurisdictions have taken different approaches to the implementation of this obligation. Some countries have simply transposed the list of sources set out in the Directive into their national copyright regime, without providing any further details or guidance. Other jurisdictions have chosen to articulate a more complete list of sources to be consulted within their enabling legislation. In the UK, the sources listed in the Annex are replicated verbatim in the UK's Orphan Works Regulations. In addition, the Intellectual Property Office (IPO) has produced three sets of guidelines that are 'primarily intended' for those wanting to make an application through OWLS, although each does state that the guidelines may be helpful to those conducting a diligent search in relation to the Directive. Each of the guidelines is accompanied by a Diligent Search checklist. Taken together, the guidelines and the checklists provide very useful information and signposting when undertaking a diligent search.
These different approaches to implementing the Directive underscore the lack of clarity that envelopes diligent search, both conceptually and a matter of practical implementation across the EU. Academics disagree on the nature of the obligation diligent search imposes. It has been suggested that if a Member State has provided guidance on diligent search incorporating a list of potentially relevant sources then all those sources should be consulted; otherwise, the search could not be considered diligent. In other words, this concept of diligence requires an exhaustive search of identified sources. We disagree. We argue that a more purposive interpretation of the diligent search requirement should be adopted across the EU. Diligence should not be characterised by an unthinking adherence to a check-list of sources, however useful and well-crafted. Much will depend on the content of the work and the context in which it is found, as well as the expertise and the knowledge-base of the person conducting the search. Within the UK, the IPO have clearly signalled in its guidance and practice that diligence must be context-specific: that is, there is no minimum requirement to be followed in every case. We encourage other jurisdictions to adopt a similarly practical, pragmatic approach.
However, the benefit of this context-specific approach is complicated by the fact that the enacting legislation within the UK has introduced (or seems to have introduced) two different diligent search standards: good faith diligent search and reasonable diligent search, for the Directive and OWLS respectively. At present, it is unclear whether the imposition of a different obligation was intended by the legislature, and the IPO do not address the issue in their guidance. That said, the IPO have structured their guidance in a way that suggests, albeit tacitly, a different standard of care is implied, and that the standard of reasonableness is set at a higher threshold than that of good faith. From our perspective, we do not think it desirable, or necessary, to complicate the UK orphan works regime with different standards regarding diligent search. We prefer an approach that treats good faith and reasonableness as synonymous in this context, rather than introducing a hierarchy of practice. This would provide a simplicity of process, while adequately safeguarding the interests of unknown and unlocatable rightholders. What is needed, however, is clear direction from the IPO on this issue.
The IPO might also influence the public understanding and practice of diligent search in another way. At present, across most Member States, details of diligent search remain with the institution availing of the exception under the Directive. Only the details of the work and relevant rightholders are formally reported to the EUIPO, although information on the search must, of course, be retained by the institution itself. As such, the Directive does not encourage or enable a shared practice of diligent search. In the UK, however, there is an opportunity to do just that: to make available documentation evidencing the practical reality of diligent search activity in relation to a variety of works, contexts and institutions. As helpful as the IPO's guidelines are in providing an overview of sources that may be useful when conducting a search, the sheer volume of potential sources is likely to overwhelm many archivists or institutions contemplating making use of OWLS or the Directive. But, supplementing those guidelines with examples of real-life search activity - search activity that has been considered and approved by the IPO as part of the licensing process - would prove invaluable in helping the heritage sector better understand the nature of the diligent search requirement in both theory and practice. It would allow a shared practice of diligent search to emerge, one that could prove instrumental in positively shaping the heritage sector's engagement with the orphan works regime in the UK and across Europe.
Despite these opportunities for potentially improving and enhancing engagement with the Directive and OWLS, the simple fact remains that for anything other than small-scale digitisation initiatives, the long-term efficacy and relevance of both schemes appears fatally compromised by the demands of diligent search. The costs and challenges of rights clearance activity are a significant barrier to the digitisation of cultural heritage collections. Existing literature evidences this reality, and our research - the first major UK study concerning the concept of diligent search since the introduction of the Directive and OWLS - confirms that diligent search and mass digitisation are fundamentally incompatible, however light-touch the nature of the diligent search obligation. With respect to the 16 scrapbooks created by Edwin Morgan, we estimate that it would take one researcher over 8 years to undertake the diligent search activity alone, at a cost of more than £185,000. And of course, this would not guarantee that the scrapbooks could be made available in their entirety. Known and locatable rightholders might make permission contingent on the payment of unaffordable licence fees, or might simply withhold permission on other grounds. No cultural heritage institution, however well-resourced, would ever take on such a speculative and costly venture.
As for the scrapbooks, from 1953 Morgan made sustained efforts to get them published without success. Similarly, the Edwin Morgan estate and the University of Glasgow are keen to share them online with the rest of the world. However, they will remain accessible only within the physical confines of the University Library building for the foreseeable future, apart from a selection of the pages that we digitised for this project. In this respect, they evidence the very real phenomenon of the so-called 20 th century black hole. Because copyright status shapes selection processes regarding the digitisation of heritage materials, collections containing work from the mid- to late 20th century are less likely to be digitised for public consumption. In short, our digital historic record is skewed towards material created in the nineteenth and early twentieth century, material that resides in the public domain.
Moreover, the orphan works regime may unintentionally introduce another variable likely to distort digitisation selection processes. We have argued that the diligent search requirement must be interpreted to require less than a non-exhaustive search of all possible relevant sources: that is, the nature of the obligation will be determined by the content and context of the work. But in addition, the experience of this project is that when presented with a work lacking any meaningful contextual information, diligence will often require little more than a google search (for text), or the use of reverse image search technology (for image). For one black and white photograph, the subject of an exploratory application to OWLS, reference to six sources only was deemed necessary to satisfy the diligent search requirement, with the time spent on search activity and completion of the application totalling less than an hour. In other words, the less you know about the work in question the easier and less costly it can be to conduct an effective diligent search (whether good faith or reasonable). Just as institutions tend to privilege public domain material when making selections for digitisation, so too institutions might be tempted to select the low-hanging fruit in the copyright garden: material that is in copyright but that lacks any contextual metadata and so minimising the search burden. Put simply, items may be selected for digitisation on the basis that little to nothing is known about them.
Whenever choosing to make copyright-protected material available online, risk also plays an inevitable role in shaping digitisation strategies. Indeed, the solution to the problem of orphan works adopted by the European Directive requires taking risks, in that reliance on the exception provides no guarantee of immunity from future litigation. A reappearing rightholder might challenge the robustness of the diligent search carried out in relation to their work; if successful, what was perceived to be a lawful use would be deemed to fall outside the scope of the Directive. OWLS on the other hand directly mitigates this risk: it provides users with a reassurance about regulatory compliance which the exception cannot, as well as a shield against future liability should any rightholders reappear. It would be entirely understandable if risk-averse institutions opted for the safety-net of the OWLS regime instead of the Directive, even if it comes with an additional financial cost attached, and even if use turns on having to renew the relevant licence every seven years.
For anything beyond a small-scale digitisation initiative, risk-management will not just shape a digitisation strategy, it will lie at its very core. The Wellcome Library's Codebreakers project provides a compelling illustration. To enable online access to as much work and correspondence of the pioneers of genetic research as possible, a risk-managed approach to copyright compliance was adopted, and necessarily so in the opinion of the Wellcome Library. In nearly four years since the project was first launched, the Wellcome have only received two requests to remove Codebreakers material from their online resource. For both, no reasons were given for the takedown request, no compensation was sought, and no litigation ensued. The initial risk taken on by the Wellcome in deploying their strategy was significant; but the benefits have far outstripped any potential negative consequences whether financial or reputational.
And within this project, concerning just 432 works meticulously arranged across 30 pages of a scrapbook in the mid-20th century, we too have assumed a certain quantum of risk. At heart, this project is concerned with an academic exploration of the practicalities - and implications - of diligent search when digitising unique cultural artefacts. Many of the 432 works from our initial sample are orphans. But, for only five of those orphans did we comply with the requirements of the Orphan Works Directive; for a further five, we secured a licence for non-commercial use through OWLS. In short, our use of the overwhelming majority of orphan works that we have made available through this resource lies outside the scope of both OWLS and the Directive. In addition, many of the works were not orphans. We identified and contacted 32 rightholders, 15 of whom (47%) agreed to our proposed use without asking for payment; notably, those granting permission for free included the Herald and Times Group (in relation to three different publications within their stable), Trinity Mirror-Sunday Mail, Newsweek and DC Thomson. However, a further seven rightholders (22%) requested fees, from £15 to $2000. No licence fees were paid.
So, on what basis do we make the majority of orphans available? And, on what basis do we make available those works for which we have been refused permission by the rightholders? We reproduce that material in accordance with sections 29 and 30 of the Copyright Designs and Patents Act 1988 permitting use for non-commercial research and for the purposes of criticism, review and quotation. Wherever reasonably practicable, we have acknowledged and attributed the original author of the work and the original source of publication. But, as with reliance on the orphan works exception, our claim to lawful use carries risk: indeed, risk and uncertainty are structurally embedded within all copyright exceptions. A rightholder might object that our use is not critical, or does not constitute quotation or non-commercial research. They might also argue that, in any event, our use is not fair. These are known unknowns. In good faith, we consider that our use of this material falls within the scope of sections 29 and 30, but we do not know that is does; nor can we. And these will remain known unknowns unless and until threatened litigation proceeds to court. Only with a judicial pronouncement might the fairness or lawfulness of our activity within this project be determined definitively; only then might these known unknowns become known knowns.
And yet, there were certain risks that we were not willing or able to take; or, to put it more accurately, there were risks that the project team were not able to agree upon. Attentive readers will have noticed our concession that only a selection of pages from this project have been made available online. Why not all? Because of extracts from two poems that featured in our sample: The Vanishing Ballad by Charles Madge and The Clearing by Robert Graves. The publisher was approached for permission and requested only £15 per poem for use of the extracts online. The complicating factor was that the publisher in question - Carcanet Press - is also Edwin Morgan's publisher. With events planned to commemorate Morgan's centenary in 2020, the University Library had already established a positive relationship with Carcanet; as such, sensitivity was required. Different options were discussed. One option was to pay the licence fee to Carcanet, given the circumstances and strategic importance of the centenary plans. Alternatively, the extracts in question might be permanently redacted from our online version of the sample pages, either on their own or in combination with all material that had been designated high risk. From the library's perspective, it was essential not to lose Carcanet's cooperation and support for the events in 2020 and beyond.
From an academic perspective, neither payment nor redaction on a selective basis was considered appropriate. This is an academic research project specifically concerned with rights clearance, risk, and how the copyright regime addresses the challenges of enabling access to digital cultural heritage. To deal with one rightholder differently from any of the others that refused permission without payment - however modest the fee that had been requested - would compromise the intellectual integrity of this project. The extracts in question featured in the last 12 pages of our 30-page sample. The decision was taken to make only the first 18 pages available online, just as Morgan had originally assembled them, with every work visible in colocation to each other; in this way, the integrity of Morgan's work is also preserved. And of course, the 12 pages from our sample that are not online remain available to view in the Level 12 reading room of the University Library, along with the other 3,570 pages that make up Morgan's scrapbooks.