A Culture of Copyright/Situating the study

by Andrea Wallace

1. Situating the study

“Rights need to be fronted. The idea of having joined up collections is great. But, in practice, the rights must be sorted out at the start.”

Staff at a large UK GLAM

1.1. Introducing the study

This report contributes a UK perspective of open GLAM while analysing it against the global open GLAM movement and recent legal developments that will dramatically alter the digital landscape of open access to cultural heritage collections in the public domain. It is directed at a range of audiences central to this landscape, including TaNC and the AHRC, UK government, lawmakers and policymakers, GLAM directors, staff and supporters, the general public and users, and even audiences outside the UK.

The research proceeded against a backdrop of similar studies spanning two decades of data on the benefits and drawbacks of copyright claims, commercial licensing and open access strategies adopted by cultural institutions and organisations:

• A 2002 report by Simon Tanner and Marilyn Deegan explored reproduction charging models for digital cultural heritage in the UK and Europe among 51 institutions, including in-depth interviews and observations of a representative sample of 15. Tanner and Deegan found “the most powerful deciding factor for price was the perceived market value of the item (as defined by what similar organizations are charging) rather than the actual cost of creation and provision.” None of the institutions fully recovered the costs of the services through licensing alone.^[1]
• Simon Tanner’s follow-on 2004 report explored reproduction charging models for digital cultural heritage in the US among 100 museums, including in-depth interviews and observations of a representative sample of 20. Tanner found most licensing departments operated at a loss. A few larger museums produced profits based on income generated around a relatively tiny group of popular works.^[2]
• Studies in 2006 and 2009 by Nancy Allen, Hillary Ballon and Mariet Westermann focused on the field of art history and art history publishing. The research found the dependence on high quality images and the copyright costs associated with them negatively impact the field of art history and are serious impediments to the productive development of digital publications for art history.^[3]
• A 2009 report by Prodromos Tsiavos mapped the flows of content, value and rights across the UK public sector. Tsiavos found rights clearance, the constant education of audiences and staff training on intellectual property are crucial to the success of digitisation initiatives and open access objectives, which correlate to the higher maintenance costs involved in implementing and sustaining such programmes.^[4]
• A 2010 report by Simon Tanner and Marilyn Deegan studied the opportunities, benefits and impacts of digitised resources for the UK, laying important groundwork for (re)evaluating how value can be transformed and measured through digital initiatives that provide access to the UK’s heritage collections.^[5]
• Work by Kenneth Crews and Melanie Brown between 2010-2012 examined the broad website terms of use and licensing policies of 50 art museums in the United States. Crews and Brown uncovered a range of “copyright overreach” strategies limiting reuse in ways that inhibit new creativity, scholarly exploration and pose a threat to the public domain.^[6]
• Kristin Kelly’s 2013 report surveyed 11 art museums in the United Kingdom and United States at various stages of opening up public domain collections. Kelly found that fears around loss of control and revenue loss either faded, or did not materialise, while the evidence of both internal and external benefits of open access steadily grew.^[7]
• The 2015 Striking the Balance Report studied the ways in which a lack of clarity, standardisation and funding around open licensing shaped the business models and access policies of various United Kingdom cultural institutions. The report highlighted the significant investment gap between goals for digital remits and the resources available to realise them.^[8]
• A 2016 paper by Effie Kapsalis revisited the museums featured in Kelly’s 2013 report and documented the impact of funding obligations and public expectations on increasing open access to collections and the demonstrated benefits for cultural institutions. Kapsalis’s work laid the groundwork for the Smithsonian’s own shift to open access in 2020.^[9]
• A 2018 report by Martine Denoyelle, Katie Durand, Johanna Daniel and Eli DoukaridouRamantani detailed the complex systems regulating the circulation of art images in French collections and the impact of copyright on downstream reuse, finding commercial objectives were at the root of the problem.^[10]
• In Australia, a 2021 report explored audiences’ expectations around, and engagement with, art through new technologies, including how COVID-19 has accelerated the shift to digital cultural engagement and the development of a growing ‘dual economy’.^[11]

These studies document almost two decades of research undertaken on how technology has changed the ways cultural institutions document and manage their collections, and the new barriers and opportunities that can arise through licensing and open access. All overlap with findings from this study.

1.2. Overview of the research scope

For reasons explained throughout this report, this research scope is limited to the open access policies and practices around collections in the public domain. It does not address the access parameters around in-copyright works and collections, as these are defined by national copyright laws and controlled by the works’ rightsholders, which is rarely the institution. This report studies access to materials for which the institution claims to be, and sometimes is, the rightsholder.

Around this, the research cast a wide net to understand the scope of digital collections and policies that UK GLAMs publish on their own websites and external platforms (e.g., Art UK, Europeana, Flickr Commons, Wikimedia Commons).

1.3. Defining GLAM, open and other terms

GLAM refers to Galleries, Libraries, Archives and Museums and is used as shorthand to refer to any national, regional or local cultural heritage institution or similar organisation.

User refers to many individuals or groups including members of the general public, educators, researchers, GLAM staff, and even GLAMs themselves, located in the UK or elsewhere.

Work refers to an item, information or output. Sometimes a work requires creative input and attracts copyright protection; when it does not, the work is in the public domain from the moment of creation.

Licence refers to the legal conditions under which the work is provided.

Open, open access or open licence carries the meaning of “open” as defined by the Open Knowledge Foundation: “Open means anyone can freely access, use, modify, and share for any purpose (subject, at most, to requirements that preserve provenance and openness).”^[12] Under international open access statements, materials must be made available for commercial reuse to qualify as open.^[13]

Public domain conveys an absence of copyright or similar restrictions on use. For the purposes of this report, ‘public domain’ should not be equivocated or conflated with terms like ‘published’ or ‘publicly available’ in reference to digital media. Also note that references to ‘open’ can include knowledge and materials in the public domain since anyone can freely access, use, modify, and share them for any purpose.

Public domain tools are used to mark public domain media. These are not licences, because no rights exist to support the application of a licence (or rights have been waived).

Closed licence denotes some rights have been released, but the rights to prohibit modification or commercial use remain reserved.

Digital media and/or digital collections refers to the range of content produced during the digitisation and management of physical collections, and may include data, metadata, paradata, text and images (i.e., digital surrogates).

Open media and/or open collections refers to open digital media and open digital collections produced during the digitisation and management of physical collections, as described immediately above. Not all digital media and digital collections published online are open.

Digital surrogate refers to a digital reproduction of an object. In its public task, the National Archives defines digital surrogate as “a representation of a record, usually an image, stored in digital form.”^[14]

All eligible data describes when GLAMs release all digital surrogates of public domain works under open licences and public domain tools as a matter of policy.

Some eligible data describes when GLAMs release some digital surrogates of public domain works under open licences and tools on an individual project or output basis as a matter of practice.

Open GLAM is an independent movement associated with the Open Knowledge Foundation, Wikimedia Foundation, and Creative Commons. Open GLAM relates to and overlaps with other open initiatives, like open access, open culture, open science, open data, open source, and open innovation.

Instance refers to a GLAM-level policy or practice on open access. For example, instance can refer to the Smithsonian Institution, which releases almost four million high resolution images to the public domain (CC0) on its own website, or the Bath Postal Museum, which releases one jpeg at 550×685 pixels at 72dpi to the public domain (Public Domain Mark) on Art UK. Both are treated as one instance for the purpose of tabulating an open access policy or practice.

Volume refers to the number of digital surrogates published by or across GLAMs. Using the examples above, the volume of open assets published by the Smithsonian is nearly 4,000,000 assets (CC0); the volume of open assets published by the Bath Postal Museum is 1 asset (Public Domain Mark). Volume does not imply unique assets. There can be overlap where GLAMs contribute open assets to more than one platform.

Reuse refers to both use (first use) and reuse (downstream use) of digital media.

Data aggregator refers to an organisation that collects data from one or more sources, provides some value-added processing, and repackages the result in a reusable form.

Technical protection measures are actions taken to block or limit access to a work, such as watermarking, disabling download or uploading the lowest quality of images.

Moral rights refer to the noneconomic rights that protect the personal and reputational, rather than monetary, value of a work to its creator.

1.4. Methodology

This report draws on several types of information:

• Existing empirical data on open GLAM globally, specifically the ‘Survey of GLAM open access policy and practice’, (‘Open GLAM Survey’);^[15]
• New empirical data on UK GLAMs, including data from:
  • A survey of access to the digital collections of 195 UK GLAMs across internal and external platforms (‘UK GLAM Sample’);^[16]
  • A survey of the copyright and open access policies of 63 GLAMs from the UK GLAM Sample (‘UK GLAM Policies’);^[17]
  • 30 one-hour interviews with TaNC project investigators, UK GLAM staff, external platform staff and open GLAM advocates;
• A review of case law and policy developments in the United Kingdom, the United States and the European Union; and
• A literature review of scholarly writing on copyright and open access to digital collections.

1.4.1. Data sources, collection and scope

All surveys adopted a user perspective in reviewing and extracting publicly available information from GLAM websites and external platforms. For the purposes of data collection, this involved taking on the role of a user and the ways in which they encountered rights information on a given website.

Open GLAM Survey
Managed by Douglas McCarthy and Dr Andrea Wallace, the Open GLAM Survey is an ongoing informal survey of open access policies and practices in the global GLAM sector. It collects all known instances of open collections published online by GLAMs and other organisations.^[18] This involves personally reviewing GLAM websites, data aggregators and other platforms on a periodic basis and manually updating the Google spreadsheet.^[19] As of 7 October 2021, the survey featured 1208 instances of open GLAM, 80 of which are in the UK.

UK GLAM Sample
Created for this report, this sample replicates and expands the Open GLAM Survey data extraction and methodology to include a range of GLAMs across the UK and new data points. The initial sample of 350 organisations included Independent Research Organisations (IROs) and Research Centre Institutes (RCIs), GLAMs associated with TaNC Foundation and Discovery projects, UK GLAMs in the Open GLAM Survey, and other UK GLAMs and related organisations. An initial review was performed to identify and remove organisations outside the scope of inquiry (e.g., no permanent collections). The final sample included 195 organisations.

From the final sample, 24 are IROs (all RCIs were removed). Another 32 are Universities (including GLAMs within universities). This brings the total number of organisations eligible for AHRC funding to 56 (or 28.6%). The remaining 140 include public and private GLAMs at national, regional and local levels (e.g., councils, historic buildings) and research initiatives or data aggregators (e.g., Portable Antiquities Scheme, Culture Grid, Archaeology Data Service).^[20] Organisations are distributed across the UK as follows: Channel Islands (1 total); England (154 total); Isle of Man (1 total); Northern Ireland (5 total); Scotland (28 total); Wales (6 total).

Data collection for the UK GLAM Sample extended to:

Data collected	Scope
Entry fee	Any entry fee charged for only Galleries and Museums, including those within Universities
Location	City, country, Wikidata link and Q code (for mapping purposes)
Funding eligibility	General eligibility for AHRC funding and participation in a Foundation or Discovery project
Own website and digital collections	Extracted from the GLAM's own website: Digital collections presence: (1) whether the GLAM has a collection online, and if so whether (2) images are purely illustrative or (3) searchable as objects via a database or (4) searchable only as records via a database; Digital collections volume (if stated or searchable): (1) volume of all digital collections and (2) volume of open data; Rights claimed and technical protection measures: whether (1) copyright is clearly claimed, (2) inconsistent, (3) unclear and/or (4) no copyright policy exists on the website; and (5) whether any technical protection measures are used to prevent download and reuse (e.g., watermarks, download disabled, etc); Rights statements: for (1) the digital asset, (2) underlying work and (3) metadata; along with links to policies for (4) reuse and (5) cultural sensitivity; Commercial licensing: whether the GLAM (1) manages their own commercial licensing operations or (2) outsources licensing; (3) the quantity of digital collections within Bridgeman Images and (4) the quantity of digital collections within Google Art & Culture; Public task: how 'documents' and access to information is outlined by GLAMS with the Re-Use of the Public Sector Information Regulation 2015 obligations[21];
Art UK	Engagement with Art UK and (1) the scope of digital surrogates available (2D and/or 3D works), (2) the total volume, (3) open data volume, (4) closed licence volume, (5) status of the tool(s) used (i.e., public domain or CC0, open, closed, both, or All rights reserved) and (6) the rights statement primarily used
Platforms used	Platforms, if any, where open data is published: (1) Own website, (2) Art UK, (3) Europeana, (4) Flickr, (5) Flickr Commons, (6) Sketchfab; and (7) Wikimedia Commons; in addition to (8) the volume of data released and (9) rights statements used across each platform
Open access status	Comparing statements used across all platforms to code the GLAM via (1) the majority approach to digital collections and (2) 'most open' level at which digital collections have been published across platforms: All eligible data, no new rights (public domain or CCO0); All eligible data, open-compliant (CC BY, CC BY-SA); Some eligible data, no new rights (public domain or CC0); Some eligible data, open-compliant (CC BY, CC BY-SA); Closed licences (CC BY-ND, CC BY-NC, CC BY-NC-SA, CC BY-NC-ND and equivalents); Closed licences by exception (All rights reserved, except for photographs of sculptures produced for an Art UK project funded by The National Lottery Heritage Fund); and All rights reserved
Engagement	Date of first engagement with open access, if known
Commercial licensing	Commercial licensing of collections via the GLAM and/or a third party, if stated
Total open data volume	Total volume of (1) open assets online across all known platforms and, of those, (2) the volume that is legally data volume compliant with the public domain (i.e., CC0, Public Domain Mark or No Known Copyright Restrictions)

UK GLAM Policies

Following data collection for the UK GLAM Sample, 63 GLAMs were selected for a more in-depth dive into copyright and open access policies and practice.^[22] This survey created a dataset extending to GLAM policies published on their own website related to copyright, open access, sensitive materials, and the public task, as well as policies on external platforms where digital collections are published.

Interviews

Interviews sought information on any aspects of copyright or open access related to a TaNC project, the individual’s role and/or GLAM policies and practice. Interviews also provided an opportunity to verify information in online policies and understand the challenges that shape them. Interview data has been generalised or presented in the aggregate to maintain participants’ anonymity. Any direct quotes have been approved for use.

1.4.2. Data accuracy

Due to the nature of web-based research using publicly available information, there may be some inaccuracies and limitations in the data due to their sources.

Every effort was made to ensure no inaccurate or misleading data appears in this report, but the author cannot guarantee absolute accuracy. Some may be inherent to data, legal interpretations, information or other statements produced by the data sources themselves. These inaccuracies reflect the reality of rights management and interpretation of law within and across GLAMs.

For example, GLAM policy statements are often conflicting, subject to change and may not expressly claim or disclaim copyright in digital collections. Many GLAMs lack formal policies, and/or online collections or searchable collections, which can impact high-level categorisation and quantitative analysis. Additional factors related to resources, capacity, technology, platforms, legacy data, funding obligations, senior management, sustainability and staff turnover can result in inconsistent approaches taken across internal and external platforms, even with a given GLAM.

For these reasons, qualitative discussions of policies and data reflect holistic assessments accounting for these contradictions. Rather than presenting a definitive value, the quantitative discussions below represent an ‘at least’ approach to measuring digital collections and open collections published online.

1.5. GLAM distribution across the datasets

The figures below show the distribution of galleries, libraries, archives and museums across two datasets:

• the Open GLAM Survey (‘Global instances in Open GLAM Survey’, 1,208 total) with the UK’s representation for comparison (‘UK instances in the Open GLAM Survey’, 80 total); and
• the UK GLAM Sample (195 total, including the 80 UK instances in the Open GLAM Survey).

Figure 1. Distribution in proportions organised by dataset^[23]

Figure 2. Distribution organised by sector^[24]

 

---

1. https://kclpure.kcl.ac.uk/portal/en/publications/exploring-charging-models-for-digital-cultural-heritage-in-europe(5f0e70c0-8753-4d71-bbf9-881dfa5352a9)/export.html
2. https://kclpure.kcl.ac.uk/portal/en/publications/reproduction-charging-models--rights-policy-for-digital-images-in-american-art-museums(95d04077-f8ec-4094-b8c1-d585c6b16d9b).html
3. http://cnx.org/content/col10728/1.1/; http://cnx.org/content/col10376/1.1/
4. https://sca.jiscinvolve.org/wp/files/2009/04/sca_2009symp_ipr_casestudies-final.pdf
5. https://kdl.kcl.ac.uk/what-we-do/consultancy/strategic-thinking-and-practice/inspiring-research-inspiring-scholarship/
6. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2026476
7. https://www.clir.org/pubs/reports/pub157/
8. https://www.nationalmuseums.org.uk/media/documents/publications/striking_the_balance.pdf
9. http://siarchives.si.edu/sites/default/files/pdfs/2016_03_10_OpenCollections_Public.pdf
10. https://www.inha.fr/fr/recherche/le-departement-des-etudes-et-de-la-recherche/domaines-de-recherche/programmes-en-cours/images-usages.html
11. https://australiacouncil.gov.au/advocacy-and-research/in-real-life/
12. https://opendefinition.org/
13. Examples include the 2002 Budapest Open Access Initiative, the 2003 Bethesda Statement on Open Access Publishing and the 2003 Berlin Declaration on Open Access to Knowledge in the Science and Humanities.
14. https://www.nationalarchives.gov.uk/information-management/re-using-public-sector-information/about-psi/public-task/
15. Douglas McCarthy and Andrea Wallace, “Survey of GLAM open access policy and practice,” http://bit.ly/OpenGLAMsurvey
16. Appendix 1. UK GLAM Sample, https://doi.org/10.5281/zenodo.6242179
17. Appendix 2. UK GLAM Policies on copyright and open access, https://doi.org/10.5281/zenodo.6242559
18. For more information: https://docs.google.com/document/d/15U__Z50WCUM_OWQ9HKLvLMlkcMoCN68FLVl9OKJQ8yY/edit
19. http://bit.ly/OpenGLAMsurvey
20. The full list is available in Appendix 1.
21. Discussed in Section 2.1.
22. The full list is available in Appendix 2. UK GLAM Policies on copyright and open access.
23. Figures: https://doi.org/10.5281/zenodo.6242179
24. Figures: https://doi.org/10.5281/zenodo.6242179