Introduction

Museum ethics have now entered the digital age, as have images and the copyrights pertaining thereto. In general terms, the law is regarded as a kind of minimum ethical standard. As a result, one could be inclined to believe that copyright can provide guidance as to how to handle reproduction rights in an ethically satisfactory manner. However, two cases – one very recent but already famous, and another, slightly less recent but arguably even more famous and highly disputed – show that copyright and the logic of rights and reproductions have their limitations. It is suggested that current museum practice in view of copyright is to some extent unethical. To assess the ethics of using reproductions of works of art in the public domain, the public domain first has to be reinstalled.

This paper will begin by examining two case studies. The first case involved the National Portrait Gallery in London and the Wikimedia Foundation. The second case, Bridgeman v. Corel, was decided by a New York court in 1999. Next, the underlying copyright issues involved in the two cases will be assessed. The paper will conclude with some ethical considerations.

1. Case studies

1.1 National Portrait Gallery (UK) v. Wikimedia Commons

The National Portrait Gallery in London (NPG) was founded in 1856. It is not an art museum as such, but focuses on collecting portraits of historically important people. Since 1896, it has been based at St Martin’s Place in London, not far from the National Gallery. The gallery’s collections comprise more than 4,000 paintings, and some 6,800 works on paper, such as photographs and prints. Digitisation did not bypass the NPG, and it started making digital images of works from the collection available early in the twenty-first century. The user can view images of specific works on the gallery’s website. Before the dispute of 2009, it was even possible to zoom in on details, by a high-resolution image being divided into tiles. Zooming in displayed only the relevant tiles of the image, and never the complete image in its high-resolution glory. The main reason for the use of such technology is to save bandwidth. Admittedly, the National Portrait Gallery also had in mind a rudimentary form of protection against the images being copied in their maximum resolution.

In 2009, Derrick Coetzee, a graduate student at the University of California, Berkeley, wrote a script that downloaded the image tiles of more than 3,000 images from the National Portrait Gallery’s website and re-assembled them into complete, high-resolution images, which he then uploaded to the Wikimedia Commons repository on the Internet. In other words, the images that the NPG offered in tiles were now available in their entirety. Pursuant to UK law, these were all images in the public domain: the original copyright holders had been dead for more than 70 years (cf. Copyright, Designs and Patents Act 1988, section 12, subsection 2). In a demand letter to Coetzee, however, the NPG claimed that its new photographic reproductions of the works in its collections were copyrighted works in their own right, and that the uploading of the images infringed both the NPG’s database rights and its copyrights (see the article ‘National Portrait Gallery and Wikimedia Foundation copyright dispute’ on Wikipedia, and a discussion of the case in Stokes 2012: 156–164).

The matter never reached the courts, and the dispute was never properly settled. Instead, both parties eventually entered a dialogue in which they agreed to disagree. At a conference on galleries, libraries, museums and Wikipedia in the autumn of 2010, the head of the NPG’s rights and reproductions department, Tom Morgan, presented a paper analysing the unsatisfactory situation and the shared goals of both organisations, arguing for a new kind of relationship beyond the law (Morgan 2010). While the National Portrait Gallery allegedly wanted to avoid the Streisand effect, it’s reluctance to sue Coetzee might also have been the result of a case decided - twice - ten years earlier, Bridgeman v. Corel.

1.2 Bridgeman v. Corel

In 1998 and 1999, a District Court in the US state of New York decided the case of Bridgeman Art Library, LTD. v. Corel Corp., a decision that has divided opinion among experts ever since. Founded in 1972 by Harriet Bridgeman, the Bridgeman Art Library is an important archive for reproductions of works of art. Corel is a Canadian software company. In the UK, the USA and Canada, it sold a CD-ROM containing digital images of paintings by European masters, all of which were in the public domain. Bridgeman claimed that it held the copyright to these images, arguing that its own digital images were the only possible source of the images distributed by Corel, and that these copies were infringements of Bridgeman’s copyright. Corel denied that there was any copyright pertaining to the Bridgeman photographs.

While the details of the case and its rulings are intriguing in themselves, only parts are of interest here. In a first ruling on 13 November 1998 [The Bridgeman Art Library, Ltd. v. Corel Corp., 25 F. Supp. 2d 421 (SDNY 1998)], Judge Kaplan granted the defendant’s motion for a summary dismissal of the suit. He argued that the issue of the copyrightability of photographs was governed by the law of the United Kingdom – where the photographs had originally been published – while the question of whether an infringement had occurred in the United States was a matter of United States law. Judge Kaplan applied the originality test set forth in section 1, subsection 1 (a) CDPA (UK Copyright, Designs and Patents Act 1988), which protects ‘original literary, dramatic, musical or artistic works’ to the Bridgeman photographs. For a definition of originality, Kaplan cited Interlego AG v Tyco Industries Inc & Ors (Hong Kong) [1988] UKPC 3 (05 May 1988), RPC, 343, which had ruled that, in order to be original, works ‘need not be original or novel in form, but it must originate with the author and not be copied from another work’. (This was in itself a quote from another case, a summary by J Megarry in British Northrop Limited v. Texteam Blackburn Limited [1974] RPC 57, at 68.) Judge Kaplan concluded that the Bridgeman photographs ‘lacked sufficient originality to be copyrightable under United Kingdom law’. Indeed, they were meant to be faithful reproductions of the works of art that they represented. One could argue that such images are a priori not originals, as the Privy Council had pointed out in the Interlego case:

‘Take the simplest case of artistic copyright, a painting or a photograph. It takes great skill, judgment and labour to produce a good copy by painting or to produce an enlarged photograph from a positive print, but no one would reasonably contend that the copy painting or enlargement was an “original” artistic work in which the copier is entitled to claim copyright. Skill, labour or judgment merely in the process of copying cannot confer originality. […] [C]opying, per se, however much skill or labour may be devoted to the process, cannot make an original work.’ (loc. cit, at 371)

The case was re-argued and reconsidered after the court was – in the words of Judge Kaplan – ‘bombarded with additional submissions’ by Bridgeman. In a second decision [36 F. Supp. 2d 191 (S.D.N.Y. 1999)], the court exclusively applied American law, but ultimately affirmed the earlier decision, dismissing the case for the same reasons.

The two cases point at identical questions of copyright. Are photographic reproductions of works of art in the public domain protected by copyright? This is what Bridgeman and the National Portrait Gallery claim. But are these reproductions really ‘original’ in the sense of copyright law? What is the legal situation in countries other than the UK? The following section will analyse the legal framework concerned.

2. The copyright status of photographic reproductions of works of art in the public domain

The question is: are photographic reproductions of works of art in the public domain copyrightable? My analysis will centre around UK and EU law, with a few digressions into US law, taking up the facts of the Bridgeman and National Portrait Gallery cases.

2.1 The public domain

This paper will not touch on the question of works of art that are still under copyright. Generally, copyright protection ceases after a certain period, which is defined by the applicable law. Historically, the duration of copyright has been extended from 14 years in the first copyright act, the ‘Act of Anne’ (1709, 8 Anne c. 19), to fifty or seventy years after the author’s death in most countries. On expiry of this term, works are no longer copyright-protected but are in the public domain. Consequently, they are free to use, and anybody can do that which, before, was the author’s exclusive right: this includes the right to make copies and to reproduce the work, and to sell these reproductions. The Bridgeman Art Library offered reproductions of such works. If a work is in the public domain, these reproductions cannot be exclusive.

Indeed, the Bridgeman Art Library did not contest the public domain as such, but maintained that Corel had used Bridgeman’s images. However, it can be almost impossible to decide from which source a reproduction has been made, in particular if there are several images in circulation. The Bridgeman Art Library argued that some of the works of art were in private collections, to which Bridgeman had gained exclusive access, but Judge Kaplan dismissed the argument. It was impossible to establish similarities that could prove that Corel had copied Bridgeman’s images and not anybody else’s: ‘Because, by definition, public domain works are not subjects of copyright, this cannot serve as the requisite similarity for copyright infringement.’ [The Bridgeman Art Library, Ltd. v. Corel Corp., 25 F. Supp. 2d 421 (S.D.N.Y. 1998)]. In other words, ‘how would one discern whose copies are whose?’ (Bielstein 2006: 45). Even if it had been clear that Corel had used the Bridgeman images, there would not have been an infringement of the copyright in the original work, since this had expired, but only of the copyright in the reproduction.

The National Portrait Gallery case points to the heart of this issue: here it was uncontested that Derrick Coetzee had used the images on the NPG’s website. Furthermore, the works of art in question were part of the museum’s collections. Modern copyright has established that it is not the owner of a work of art who is the copyright holder, but the creator: possession does not automatically convey copyright. Historically, however, museums have been able to control reproductions of public-domain images from their collections (Butler 1998: 72). In the early nineteenth century, when copyright was still in its infancy, the president of a French commission debating the role and foundations of copyright observed it to be a rule that the Royal Museums acquired the copyright together with any object entering their collections (Commission de la propriété littéraire 1826: 278). In this constellation the museum acts as the defender of the author’s rights. Although this no longer reflects the copyright laws of the 21st century, part of the argument seems to inform museums’ actions to the present day: even if they do not own the copyright to the works in their collection, they might at least own the copyright of the reproductions of these works – at least, that is their stance (cf. Crews 2012: 806).

2.2 The museums’ view: copyright in reproductions and factual monopolies

A Guide to Copyright for Museums and Galleries, published in 2000 in association with the UK Museums Copyright Group, and supported and endorsed by the Museums and Galleries Commission and the Scottish Museums Council, states that ‘where there is sufficient skill and labour, copyright protection can be given to duplicates’, pointing out technical skills such as the prevention of glare, careful light meter readings and faithful colours (Wienand, Booy and Fry 2000: 16). While it is admitted that this ‘is a highly contentious area’ (ibid.), the common practice of museums and galleries evidently remains unchanged (cf. Crews 2012: 806; Reese 2009: 1040). In Britain, the Museums Copyright Group concluded that the Bridgeman cases had ‘not influenced the ways museums negotiate or license rights’ (Museums Copyright Group 2004). As a rule, museums claim on their websites that they own copyright to the images of the works in their collection, as The Bridgeman Art Library still does: ‘The Bridgeman Art Library is the copyright holder, or authorised representative of the copyright holder, for all photographic transparencies in the Bridgeman Art Library.’ For example, the website of the Museum of Fine Arts in Boston claims: ‘The Images are not simple reproductions of the works depicted and are protected by copyright.’ (MFA Boston 2013; see also Crews 2012: 822; Mazzone 2011: 17). Such statements, it has been remarked, are in complete opposition to the court’s decision in the Bridgeman cases (Cameron 2006: 48; Crews 2012: 808–09; Dobson 2009: 335; Matz 2000: 4; Mazzone 2011: 15; Wojcik 2008: 259, 271).

2.3 Originality, photography, Graves’ Case and recent UK decisions

As Bielstein asks, ‘What does it mean […] to claim that one can copyright a copy?’ (2006: 41) Although the standard required for a work to qualify as original is indeed very low (Garnett 2001: 229–30; Stokes 2012: 50), a frequently cited case, University of London Press Ltd v. University Tutorial Press Ltd ([1916] 2 Ch. 601), also sets a minimum standard: ‘that the work must not be copied from another work – that it should originate from the author.’ (ibid., per Peterson J at 609). This test of originality is also part of other countries’ copyright legislation. Originality has been called the ‘bedrock principle of copyright’ by the US Supreme Court (Feist Publications, Inc. v. Rural Tel. Serv. Co, 499 U.S. 340, 347 [1991]). Indeed, despite their diverging traditions and concepts of copyright, British and US standards may not be so different (cf. Cameron 2006: 55–6), and the same is true of continental copyright laws in the light of European legislation. Specifically regulating photographs, Art. 6 of the European Council Directive 93/98/EEC provides that ‘[p]hotographs which are original in the sense that they are the author’s own intellectual creation shall be protected […]. No other criteria shall be applied to determine their eligibility for protection.’

Indeed recent decisions by British courts broach the issue of harmonisation between EU and UK laws. Temple Island v. New English Teas and Nicholas John Houghton, a case in the England and Wales Patents County Court heard in November 2011, included an insightful appreciation of the correlations between European and UK regarding the standard of photographic originality, remarking that there was ‘no difference in substance’ between the law as applied by the Austrian Supreme Court, the Court of Justice of the European Union, and UK law, recapitulating that a photograph is original ‘in terms of copyright law if the task of taking the photograph leaves ample room for an individual arrangement’ ([2012] EWPCC 1, at para. 20). This takes up the definition of originality from the Infopaq decision of the European Court of Justice, stating that a work is original when it is an ‘author’s own intellectual creation’ (Infopaq International A/S v. Danske Dagblades Forening, Case C-5/08, [19 July 2009] ECR I-6569, at para. 37), in turn reiterating Art. 6 of the European Council Directive 93/98/EEC. The definition has since been confirmed in Painer v. Standard Verlags-GmbH and others (Case C-145/10, [1 December 2011] ECR I-12594, at para. 87).

In view of these recent decisions, the resurrection of Graves’ Case [(1869) LR 4 QB 715] by authors arguing in favour of copyright protection for reproductive photographs (Garnett 2000: 231) becomes even more doubtful. In this case, Henry Graves, a well-known publisher of art reproductions in Victorian Britain, had registered photographs of engravings to obtain protection under the Fine Arts Copyright Act of 1862 (25 & 26 Vict. c.68). J. B. Walker, who had sold pirated copies of these reproductions, was charged with the infringement of Graves’ copyright and applied to the court to delete the entries in the register, arguing that there was no copyright in the photographs of the engravings. Blackburn rejected the argument, observing that ‘all photographs are copies of some object, such as a painting or a statue. And it seems to me that a photograph taken from a picture is an original photograph, in so far that to copy it is an infringement of this statute.’ [(1869) LR 4 QB 715, at 723]. The persuasive power of this reasoning is limited; apparently Blackburn merely asserted the conclusion, which he supported only with his personal view: ‘it seems to me that […]’. Judge Kaplan discussed the authority of Graves’ Case in the second Bridgeman decision, concluding that it did not reflect the subsequent development of the concept of originality [36 F. Supp. 2d 191 (S.D.N.Y. 1999), at 198].

The limits of the authority of Graves’ Case become even clearer if one takes into account the fact that Henry Graves was also the copyright owner of the reproduced paintings, as well as the corresponding engravings (cf. Lupton 1988: 257). To deny Graves copyright protection for his photographs would have effectively undermined his copyright in the paintings and the engravings, which he also owned. In other words, the works in question were not, as in the Bridgeman and National Portrait Gallery cases, in the public domain; what was at stake was the integrity of Graves’ bundle of exclusive rights. Since section 4 of the Fine Arts Copyright Act of 1862 provided protection only for works listed in the ‘Register of Proprietors of Copyright in Paintings, Drawings and Photographs’, it was a prudent move on Graves’ part to register all media in order to obtain complete protection for all types of use. For these reasons, Graves’ Case is not good authority on the originality of photographs, but rather on the enforcement of an effective copyright protection covering all types of use (cf. the cautious reflection by Stokes that the specific facts in Graves’ Case and the case’s antiquity could make it a less than persuasive authority, and it ‘may no longer be good law’ (2012: 137)). Both Bridgeman v. Corel and the facts in the dispute of the National Portrait Gallery with Wikimedia Commons should be clearly distinguished from Graves’ Case.

The historical distance to Graves’ Case highlights one particular aspect of those voices resting their argument on that case - an argument, by the way, that attempts to take advantage of the historical campaign for the copyright protection of photographs as such – which is that this issue was the subject of considerable and highly emotional debate at the end of the nineteenth century. The prevailing view on photography was far more mechanistic than it is today. Eastman Kodak’s slogan, ‘You press the button, we do the rest’, was a popular advertisement, which summed up the common belief that a photograph was the result of an ingenious piece of machinery, the pre-determined product of an optical-chemical and mechanical process. Artist-photographers had to fight to obtain copyright protection for their works (see for example Bigeon 1894). However, emphasising, on the one hand, that photographs deserve the same protection as other works of art while, on the other, asserting that photography has suffered from its supposedly industrial nature, and arguing that every photograph is ‘more or less a copy of something else’ (Garnett 2000: 229) belies a sophisticated understanding of what photography is. Speaking of a portrait painter ‘copying the physical attributes and appearance’ of a sitter (Ong 2010: 170) hardly does justice to the process (cf. the refined approach by Malkan 2005: 427–8). In photography, the relationship between originals and copies is complicated by the dual origin of the photograph itself, which comprises both creative and mechanical elements (Malkan 2005: 439). The historical argument that photography is a potentially artistic medium should not be confused with a claim that all photographs are original. While it is true, as Allan (2007: 977) observes, that just because a process is laborious and requires technical skill does not make it unoriginal and uncreative, the opposite conclusion would be equally as flawed: labour and skill alone do not constitute an original process. A photograph can be as artistic as a painting, but it is not always, just as it can be of an industrial nature, but is not always. To state that every photograph is also a copy, would mean reiterating the mechanistic standpoint that was criticised in the first place. Taken together, the statements are self-contradictory; interpreted separately, they project elements of an individual photograph on the concept of photography as a medium. But it is not only the concept of originality that has made progress since 1869, it is also our understanding of what photography is. This convoluted approach to photography by some legal authors is irritating, because Garnett (2000: 234) himself observes that the medium itself is not the entire message: ‘it depends on how it is used.’

2.4 Originality v. skill and reproductive translations

In an obiter dictum of Interlego, Lord Oliver had questioned whether Walter v. Lane’s skill and labour standard should be applied to copies, averring that ‘[s]kill, labour or judgment merely in the process of copying could not confer originality.’ ([1988] RPC, 343, at 371; cf. Judge and Gervais 2009: 395). In Walter v. Lane ([1900] AC 539), the defendant had re-published the plaintiff’s – The Times’ – transcripts of five parliamentary speeches. The House of Lords emphasised the irrelevance of originality in the decision on authorship, but, as Deazley (2001a: 180) remarks, the relevant Copyright Act of 1842 did not incorporate the requirement of originality, which renders the precedent unconvincing. Nevertheless, in their most recent edition, the authors of Copinger & Skone James on Copyright (2010: 3–144) claim that Lord Oliver’s oft-quoted obiter dictum is inconsistent with Walter v. Lane and therefore ‘must now be considered, as a generality, to be wrong’. The latter decision being more than one hundred years old, the use of the word ‘now’ in Copinger is surprising, all the more so as Interlego had made reference to Walter v. Lane.

Furthermore, the suggested analogy between the photographer reproducing two-dimensional works of art and the ‘reporter who faithfully reproduces the spoken word of a speaker, and whose work was protected in Walter v. Lane’ (Garnett 2000: 236; Garnett is one of the authors of the recent edition of Copinger & Skone James) does not hold: the reporter translates the spoken word into a text, whereas the photographer reproduces an image. The reporter creates a work in a new medium directed at a different sense: speech is for hearing, the report is for seeing / reading. In contrast, the photographer of a painting only reiterates its visual message. The case of the photographer of a sculpture further clarifies the difference: he translates a three-dimensional object, which can also be appreciated by the sense of touch, into an image that can be appreciated only visually. Conversely, in Martin v. Polyplas Manufacturers Ltd. ([1969] NZLR 1046) the plaintiff had produced three-dimensional plastic engravings from two-dimensional drawings, which incidentally is why that case cannot be compared to Bridgeman v. Corel as Stokes (2012: 138) proposes to do. Following Alexander Baumgarten, who defined aesthetics as the science of the experiences of the senses (cf. Baumgarten 1750: 6), these sensory translations and potentials become relevant to the distinction between original and other works.

The case of photographs seems even more unclear given that mezzotint reproductions of paintings have been held to be original derivative works under US law (Alfred Bell & Co. v. Catalda Fine Arts, 191 F.2d 99, at 104). However, non-photographic means of reproduction, such as engraving and mezzotint, do include an element of translation, namely from colour to black and white, and from shades of colour to the graphic language of lines, stipples, or cross-hatching (cf. Butler 1998: 84). Engravings are created using lines and stipples in a variety of densities, while mezzotint is a technique that uses shades of grey from a pre-treated copper-plate. Both involve personal intellectual choices in terms of a design following the visual grammar of the medium employed. Modern photography does not involve the same kind of choices; its aim is a faithful reproduction based on the test of visual consistency, a comparison of original and reproduction. Photographs serve as surrogate images (Butler 1998: 116). Indeed, as Ginsburg (2003: 1083) observes, the easier the technology makes it to produce quality reproductions, ‘the less the copyist’s skill should be equated with authorship even in a Commonwealth jurisdiction.’ This does not mean that the originality standard has shifted, as Cameron (2006: 37) suggests, rather it signifies a distinction based on different reproductive techniques, which is consistent with the originality test as defined by Lord Atkinson in MacMillan & Co. Ltd v. Cooper ([1924] 130 L.T.R. 675, at 678–79): ‘labour, skill and capital expended should be sufficient to impart to the product some quality or character, which the raw material did not possess, and which differentiates the product from the raw material.’ Photography is still the benchmark technology when it comes to visual reproduction, because it does not add quality or character to the work of art reproduced, and it does not call for an original translation into another artistic medium such as engraving or mezzotint, where the categories of Walter v. Lane would be appropriate. Instead of translating the visual impact of a work of art, a photographic reproduction attempts to conserve that impression with the help of its technical qualities and mechanisms.

2.5 Why the subject matters: characteristics of photographic copies of works of art

While the authors mentioned are correct in their understanding of photography as a medium that is a mixture of creative and mechanical processes, and in one sense always a copy of its object of representation, they fail to acknowledge that the subject of a photograph does matter when deciding on its originality. Indeed, it might even matter in other media, as the case of Appropriation Art suggests. Appropriation Art is a direction in postmodern art, which, among other intentions, aims to undermine copyright concepts and the corresponding aesthetics of originality by seemingly faithfully copying other works of art. In the second Bridgeman v. Corel decision, Judge Kaplan referred to the non-copyrightability of works of Appropriation Art – since they are not original, they cannot be protected. The reality of copies in the visual arts is more complex than the brief paragraph in Copinger & Skone James indicates. Their statement that ‘the photograph ought to be entitled to copyright protection, irrespective of its subject matter’, is a bold one, because it generalises and oversimplifies. The short taxonomy offered by Judge and Gervais (2009: 377) is helpful in this context:

‘Literary and artistic works are the result of one or more of three types of choices: technical choices, those that are essentially dictated by the technique used (e.g., in painting or photography, or certain forms of poetry); functional choices, those dictated by the function that a utilitarian work will serve (e.g., a chair must not collapse when someone is sitting on it); and finally creative choices, those that truly stem from the author and where, if someone else has produced the work, there would most likely have been a different result.’

Reproductive photographs are not the result of the third type of choices but only of the first and second: they are the outcome of technical preconditions such as light, distance, etc., and the functional purpose of rendering the original work of art as faithfully as possible (see Butler 1998: 119–22, quoting the handbook ‘How to Photograph Works of Art’ by Sheldan Collins). Antoine Latreille (2002: 300) suggests a similar model; he distinguishes (1) an active composition of the photographer, who creatively decides on the subject and its constellation, (2) a passive composition, where the photographer decides on technical issues such as angle, focus, light, etc., corresponding to the functional choices mentioned above, and (3) a fixed composition, limiting the photographer’s choices to technical issues. In conclusion, in certain situations, such as a faithful reproduction, a photograph is not ‘un œuvre d’esprit’ (Latreille 2002: 299). In their 1969 hand-book The Graphic Reproduction and Photography of Works of Art, John Lewis and Edwin Smith clearly state for the reproductive photography of two-dimensional works of art: ‘Most forms of photography involve choice, at least of a view-point, of lighting and of exposure. Here there is none.’ (Lewis and Smith 1969: 49).

It is here that the ‘visual significance’ put forward by Lord Oliver in Interlego ([1988] RPC, 343, at 374) becomes legally significant as a response to the view harking back to Graves’ Case, ‘that every photograph is more or less a copy of something else’ (Garnett 2000: 229; cf. Graves v. Walker (1869) LR 4 QB 715, at 723: ‘All photographs are copies of some object […].’), because this view does not take into account that some photographs are nothing more than a copy of something else, or, in the words of Latreille (2002: 304), the photographer of a reproduction does not bring forward his own vision but the artist’s whose work he is reproducing (cf. Butler 1998, 102–03; Malkan 2005: 458; contra Allan 2007: 972; Bently and Sherman 2009: 110; Lehment 2008: 36–7). Even an opponent of the Bridgeman decisions admits that ‘the reproductions really do look like the underlying works’. (Allan 2007: 977; cf. Ong 2010: 180: ‘the idea expressed is the copyrightable expression found in the antecedent work’ – in other words, not a personal expression).

The European Court of Justice has emphasised very clearly in Painer v. Standard Verlags-GmbH and others that ‘copyright is liable to apply only in relation to a subject-matter, such as a photograph, which is original in the sense that it is its author’s own intellectual creation. As stated in recital 17 in the preamble to Directive 93/98, an intellectual creation is an author’s own if it reflects the author’s personality. That is the case if the author was able to express his creative abilities in the production of the work by making free and creative choices […].’ (Case C-145/10, [1 December 2011] ECR I-12594, at paras. 87–89). These criteria are also reflected in Temple Island v. New English Teas and Nicholas John Houghton ([2012] EWPCC 1, at s.20), where Judge Colin Birss QC stated: ‘A photograph […] can have the character of an artistic work in terms of copyright law if the task of taking the photograph leaves ample room for an individual arrangement. What is decisive are the arrangements (motif, visual angle, illumination, etc.) selected by the photographer himself or herself.’

Faithful reproductions, which can be measured against a technical standard, do not allow for these creative choices and individual arrangements. Comparing different reproductions with each other and with the original, it will be possible to agree on the best, which will be the most faithful. For a variety of original photographs, this category is likely to be fruitless. In the terminology of semiotics, copyright does not protect pure icons, only those signs that are also symbols.

The question has been posed why an ‘opportunistic photograph’ of somebody happening to be at the right place at the right time, such as a tourist’s photograph of Buckingham palace, is original enough to be copyright protected, but the reproduction of a two-dimensional artwork, which requires considerably more skill and labour, should not (Garnett 2000: 235). The answer is that the tourist’s photograph is unique and personal; the reproduction is not. It is a copy. It does not originate with the photographer, it originates with the author of the original – a fact that Allan (2007: 972) seems to disregard. The minimum personal intellectual achievement is not given – the achievement is intellectual, but not personal (Nordemann 1987: 18; contra Garbers-v. Boehm 2011: 155). The comparison of the snapshot and the reproduction tries to reintroduce an aspect of artistic quality to the analysis of originality, which is alien to copyright (on this aspect cf. Rahmatian 2005: 373). The ‘opportunistic photograph’ might not be original in its technique or aesthetics, but it will always be original in that it conveys the personality of its maker (cf. Temple Island v. New English Teas and Nicholas John Houghton [2012] EWPCC 1, at s.51). If a reproduction conveys the personality of its maker, it is not a good reproduction (Malkan 2005: 434). While a reproductive ‘photograph is different in quality and character from the painting [and] has different attributes and different uses’ (Garnett 2000: 236; cf. Allan 2007: 978), these differences stem exclusively from the relation between an original and its copy. For example, colour differences are not a personal contribution from the photographer but an inadequacy (Butler 1998: 111). These changes do not transform the copy into an original, and to repeat the contrary by introducing concepts such as the authorship of a copy (Ong 2010: 179) does nothing to make the argument any more convincing (see Meshwerks, Inc. v. Toyota Motor Sales USA., Inc., 528 F.3d 1258 (10th Cir. 2008), at 1267).

In other words, while acknowledging that the artistic level required for copyright protection is low, and that copyright has been conferred on simple works (Garnett 2000: 230), reproductive photographs are not of a low artistic quality: they are not artistic at all. Nor are they meant to be; as Malkan (2005: 425) puts it: ‘Copies can be authorized, but only works can be authored.’ The very faithfulness to the original of the Bridgeman Library’s copies was their greatest asset – ‘artistic’ or ‘original’ reproductions would not be marketable, because they would not serve their purpose. The commercial value of photographic reproductions is based on their unoriginality. In consequence, copyright is not an appropriate means of protecting their market.

2.6 Two- and three-dimensional works

However, an important distinction must be made between three-dimensional and two-dimensional objects (cf. Bielstein 2006: 40). True, a philosopher might argue that no such thing as a two-dimensional object exists; it would be more appropriate to speak of objects, of which only one side needs to be reproduced to represent it in its entirety. Plainly, this is the difference between sculpture (three-dimensional) and works such as paintings, drawings, and prints (commonly referred to as two-dimensional). Reproductions of three-dimensional objects necessitate a creative choice that goes beyond purely technical decisions, for example from which angle the object is photographed, or the photograph(er)’s focus (cf. Antiquesportfolio.com plc v. Rodney Fitch & Co, [2001] E.C.D.R., 51, at 59; Cameron 2006: 42–3; Garbers-v. Boehm 2011: 140–50; Stang 2009: 177; but cf. also the doubts raised by Wojcik 2008: 266–7). Each reproductive photograph of a sculpture is likely to be different. Each reproductive photograph of a painting or print can be different, but for purely technical reasons – reproductive photographs of two-dimensional works of art are meant to be exact and identical, or what is commonly referred to as a ‘slavish copy’. While there are also works that are more difficult to classify as two- or three-dimensional, such as paintings including three-dimensional objects, like Kurt Schwitters’ collages (cf. Allan 2007: 978), Stokes’ comment that it ‘is hard to see how a court could reasonably make a distinction between photographs of two- and three-dimensional items’ (2012: 139) is solely relying on the argument of the common labour and skill needed for both sorts of photograph and thereby missing the point of originality.

2.7 Conclusion

There is no copyright in photographic reproductions of two-dimensional works of art in the public domain. Consequently, under US law, the Bridgeman cases ‘still stand for, at the very least, the important and distinct principle that exact replications of public domain works warrant no copyright protection’ (Dobson 2009: 347). Taking into account recent European and UK case decisions, it is reasonable to state that the same principles are also valid for UK law.

It is possible that an analysis by the National Portrait Gallery’s lawyers came to the similar conclusions, which might have been one of the reasons behind the decision not to sue Derrick Coetzee. While the fact that the high-resolution images provided by the National Portrait Gallery were accessible only using the Zoomify software and had to be stitched together might constitute a circumvention of technical protection measures (cf. sect. 296 CDPA), works in the public domain do not actually fall within the scope of this regulation (Garbers-v. Boehm 2011: 217–18; see also Art. 6, ect. 3 of the Directive 2001/29/EC, linking technological measures to protected works). Furthermore, the Zoomify technology is not a very effective means to control access to the higher-resolution image (cf. Stokes 2012: 160: ‘clearly a matter of fact’).

3. The business of copyright, copyfraud, and museum ethics

In view of these findings, what is the significance of museums’ and agencies’ disregard for the law (see above, 2.2)?

3.1 Re-appropriation of the public domain as a fundraising strategy

Museums re-appropriate works in the public domain through their practice of claiming copyright protection for their reproductions, thereby arbitrarily extending the copyright term. Susan Bielstein (2006: 38) and others speak of the ‘privatisation of the public domain’, Kathleen Butler (1998: 74) of a circumvention of the public domain, and Ronan Deazley (2001a: 183) of a ‘de facto perpetual monopoly over the commercial reproduction of publicly owned works of art’. The opposite view, namely that copyright protection for reproductions does not change the public-domain status of the original work is admittedly based on the assumption ‘that the antecedent work was accessible to the public’ (Ong 2010: 185), a condition that is, however, rarely met in real life, thereby devaluing this consideration (cf. Cameron 2006: 57–8; Deazley 2001b: 601; Reese 2009: 1047; Stang 2009: 208).

As Tuchmann (2001: 313) observes, the core issue for museums has always been access and aesthetic control, not copyright, except as an instrumentality of that access and control. Similarly, in her analysis of French museums and their position vis-à-vis copyright law, Sonia Tapissier-Gicquel concludes that it would be more transparent to charge service fees for reproductions instead of copyright fees. From the perspective of museums, however, this would imply a loss of control of the image as is currently provided by copyright regulations. Furthermore, she observes, the declaration as copyright fees allows the RMN to charge higher fees (Tapissier-Gicquel 2004: 130–131). This hints at the real problem. The reasons behind the museums’ claim on copyright are factual and prosaic. As the Guide to Copyright for Museums and Galleries bluntly states: ‘In order for museums to raise income from merchandising, licensing or the commercial “development” of any part of their collections, the museum must quite simply build, somehow, exclusive rights to its collections.’ (Wienand, Booy and Fry, 2000: 52). While copyright is not something that can be ‘built’, and the factual construction of copyright in works of art in the public domain remains a pretension, one ‘can understand the museums’ motivation to take this position.’ (Reese 2009: 1040; see also Latreille 2002: 302). Licence fees may become an increasingly important source of income in times of economic crisis, when museums and galleries are forced to raise money.

However, while it has been stated that licensing fees constitute ‘an essential source of revenue’ (Cameron 2006: 32), the facts remain unclear in most cases, and some revenue figures seem to indicate a rather limited market and significance (cf. Deazley 2010: 109–10; Hamma 2005). In our digital age, servicing image orders has become less labour-intensive and time-consuming, thus reducing service costs (cf. Allen 2009: 29; Ballon and Westermann 2006: 31). It has become doubtful whether a truly public domain encompassing the reproductions of its two-dimensional works of art would, as Deazley (2001a: 184) asserts, have serious financial implications for the cultural institutions concerned (but cf. Kelly 2013: 29). The current WIPO Guide on Managing Intellectual Property for Museums analyses recent developments in museums’ business models and concludes that ‘the making available of museum images is merely a means to a commercial end, and not the end itself’, pointing out that ‘providing unfettered access to museum images is actually good business’ (Pantalony 2007: 98–99). One of Kelly’s conclusions is very clear in this regard as well: ‘Real and perceived gains far outweigh the real and perceived losses for every museum in the study that has made a transition to an open access approach’ (Kelly 2013: 24).

The present situation is an unsatisfactory double-bind, as described by Cameron (2006: 61): because the revenue from licensing images ‘will enable the museum to continue protecting the public domain works already in its collection […], museums are both serving the public by preserving this art and harming the public by falsely claiming copyright in public domain images.’ The opposition is formulated differently in the ICOM Code of Ethics for Museums, 2006, in section 1.9 – ‘The governing body should ensure that there are sufficient funds to carry out and develop the activities of the museum.’ – and section 3.2: ‘Museums have a particular responsibility for making collections and all relevant information available as freely as possible, […].’ To return once more to the National Portrait Gallery, their website states: ‘The Gallery’s image licensing department raises money by licensing reproductions, thus supporting both the free entry policy and the Gallery’s main functions caring for its Collection and engaging people with its works.’

3.2 Copyright overreaching, museum ethics and the public mission of museums

Nonetheless, to quote a recent essay on copyright, ‘Law is not the solution to business problems’ (Patry 2011: 141). Furthermore, the codes of ethics for museums also include an acknowledgement of the legal position, for example stating that ‘All those who work for or govern museums should ensure that they […] make all policy and practice at the museum comply with the law.’ (section 2.8). The ICOM Code of Ethics includes similar provisions, stating (in section 7.1): ‘Museums should conform to all national and local laws and respect the legislation of other states as they affect their operation.’ At present, however, the suggestion is that many museums’ practice in rights and reproductions does not comply with UK copyright law. Instead, it is a widespread practice to contract around copyright law with terms and conditions regulating access to works in the public domain and their reproductions (Mazzone 2011: 95–100; see also Bohannan 2008 on contract clauses and copyright policy), notwithstanding that such contracts cannot legitimately be used to bypass copyright law but must respect it (Wojcik 2008: 269, 279–84). As Crews (2012: 821) has clearly stated: ‘any restrictions beyond the reach of copyright are in defiance of the law and the social and intellectual objectives that copyright aims to serve.’

Within the European Union, this means that contracts bypassing essential principles of copyright law can be problematic in view of the Council Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer contracts (cf. s.6 (1) in the UK’s Unfair Terms in Consumer Contracts Regulations 1999). The museums’ ‘abusive practices’ (Latreille 2002: 302) with regard to reproductions of two-dimensional works of art in the public domain are in breach of copyright law, irrespective of these institutions’ dire economic situation, and irrespective of whether this practise is called ‘copyfraud’ (Mazzone 2006: 1041–2; Mazzone 2011, passim) or ‘copyright overreaching’ (Crews 2012: 795). ‘Wrong life cannot be lived rightly’ (Adorno: 1974, 39).

On the other hand, is it ethical to download digital reproductions and to make them available in an open repository such as Wikimedia Commons? From the perspective of copyright law, Derrick Coetzee copied reproductions of two-dimensional works of art in the public domain, and was entitled to do so. However, he did not act in accordance with the terms and conditions as set out on the National Portrait Gallery’s website. Then again, these terms were not actually in line with copyright law. Metaphorically, Coetzee did not set somebody else’s chicken free; he freed a chicken that did not belong to any specific farmer and should not even be living on a farm in any case. Thus, the time has come to move past the logic of copyright pretensions in regard to reproductions of works of art in the public domain.

To return to Adorno’s ‘right life’, museums need to set aside their rhetoric of copyright and wake up to their mission. ‘Citizens should be encouraged to reproduce and make use of public domain materials, not treated as though they are breaking the law.’ (Mazzone 2006: 1060). This is true whether the copyfraudster is a private company, such as the Bridgeman Art Library, or a public museum. Neither holds any copyright in reproductions of two-dimensional works of art in the public domain. However, public institutions such as museums are on a public mission, as Ronan Deazley (2001a: 183) reminds us. Their ‘pictures belong to the public, and are (for the most part) […] in the public domain.’ The ICOM Code of Ethics for Museums states in section 2: ‘Museums that maintain collections hold them in trust for the benefit of society and its development.’ Meanwhile, the Code of Ethics for Museums published by the UK Museums Association states: ‘Society can expect museums to hold collections in trust on behalf of society, […] encourage people to explore collections for inspiration, learning and enjoyment, […] research, share and interpret information related to collections.’ Public collections’ trusteeship includes the dissemination of reproductions for non-commercial purposes. Marie Cornu and Nathalie Mallet-Poujol (2006: 314) argue that the public display of an object includes the willingness to communicate that object’s appearance and a fortiori its image, with the reservations of a possible – real, not merely fictitious – copyright in the object. Indeed, in Allan’s words (2007: 983) museums are ‘the proper institutions to entrust with the public dissemination of high-quality reproductions of a broad range of works of art.’ Recently this view has gained prominence with an interview given to the New York Times by the Director of Collections at the Rijksmuseum, Taco Dibbits, who is quoted saying ‘If they want to have a Vermeer on their toilet paper, I’d rather have a very high-quality image of Vermeer on toilet paper than a very bad reproduction’ (Siegal 2013). ‘Keeping the World Safe from Naked-Chicks-in-Art Refrigerator Magnets’ (Butler 1998) is no longer a realistic option. The recent developments in open access policy in the UK, Europe and the USA will put further pressure on museums and public collections to make their images accessible, as the underlying rationale is essentially identical: in the last five or so years major research funders have implemented policies to promote open access to the published findings of the research they fund (see Finch 2012: 34). Publicly funded research should be made accessible to the public. In a similar way, publicly funded art collections should be made accessible. For works in the public domain this means permitting unrestricted access to their reproductions, unhampered by copyright overreaching.

4. Reproductions reconsidered

So how could the mission of cultural partnership be developed in post-copyright categories?

4.1 Accessibility of the works for private photography

The question of a right to access to the public collections is still in a ‘chaotic’ state (Cornu and Mallet-Poujol 2006: 308). Many museums allow photography in their rooms (but the use of flashes is in most cases prohibited due to conservatorial concerns). In France, an administrative decree dating from 1979 (Journal officiel, 6 April 1979: 3047) established that photography for private use was permitted as far as permanent collections of public museums were concerned, but not for works on show in temporary exhibitions (see Cornu and Mallet-Poujol 2006: 312). In contrast, Wienand, Booy and Fry (2000: 55), advise museums in the UK not to allow photography in their galleries. The National Portrait Gallery runs a strict policy: ‘We also exert strict controls on all photography in the Gallery, which is allowed only on the understanding that copyright rests with us and that any further reproduction deriving from resulting photographic materials is subject to our written permission.’ (NPG Website: Copyright). Photography is not permitted at the Tate either, whereas it is at the Victoria & Albert Museum and the British Museum. In view of public museums’ educational mission, Bullinger (2006: 395–6) and Garbers-v. Boehm (2011: 292) raise doubts about a general ban on photography in museums.

A short, non-representative survey on websites such as Flickr and Picasa provides a mixed impression. Very few images of works of art could serve publishing purposes. Most private photographs are too individual and far from representing a professional reproduction. These images combine the subject matter of a reproduction with the stylistic characteristics of ‘occasional photographs’ – a warped picture plane, uneven sharpness and lighting, unfaithful colours. These kinds of photographs are not serious competition in the market of professional art reproductions. Not being of a commercial concern in terms of the market, private photographs without flash and tripods should therefore be permitted in public museums, at least in rooms that display works in the public domain. More professional photography may interfere with a museum’s everyday exhibition and display routines. Paintings might have to be taken out of their frames, a suitable room would have to be found, where the photography would not disturb other visitors. These procedures are cost-intensive. While it should be possible for professional photographers to create such high-quality reproductions, there is no need for a public museum to subsidise these activities (cf. Bullinger 2006: 395).

4.2 An offer that can’t be refused

Reese (2009: 1041) has pointed out certain common interests in reproductions of works of art in the public domain. All parties have a preference for high-quality reproductions, sharp, in a high resolution and with the correct colours. He proposes ‘a sui generis system of limited protection for art reproduction photos’ with a shorter term of protection of between five and twenty-five years, a limitation of the protected rights to reproductions, and the duty to deposit the image in question in a repository, where, on the expiry of the term of protection, the public could access the image without having to deal with the museum (Reese 2009: 1048–9). While the latter idea would strengthen the public domain, the proposal also denotes an endorsement of the factual prolongation of the copyright term as is current practice, which is why other solutions are preferable. An alternative regime to copyright in regard to reproductions of works in the public domain might prove beneficial to the museums. Taking the approach to ‘Commercialise Authoritative Content’ (Pantalony 2007: 80–82) seriously, the museums’ market position with regard to reproductions can be strengthened if they offer a surplus value such as the best quality, additional art-historical information, and the aura of the original (cf. Tuchmann 2001: 315; see also Pfennig 2009: 141; Stang 2009: 216). Some users such as art historians may be willing to pay a premium for authenticity and accuracy (Reese 2009: 1042); other users such as advertising agencies may be willing to pay for the highest possible quality (Matz 2000: 19). Commercial users still come to the National Portrait Gallery to obtain the best reproductions available (cf. Morgan 2010: 39’00”). Part of a service fee for the use of a collection’s reproductive image might be advertised as supporting the educational aims of the museum. Awareness of social values has grown in recent years, and a considerable proportion of the public is prepared to support charities with donations (cf. Reese 2009: 1042). Exploitation of the collection for merchandising products such as coffee mugs, tea towels, underwear or inflatable figures can be protected by trademark laws (Reese 2009: 1042–43) or unfair competition laws (Matz 2000: 22) rather than copyright law.

The securement of exclusivity, as proposed by Wienand, Booy and Fry 2000: 53, like the music industry’s battle against downloads, is bound to fail. Apple’s iTunes and internet platforms such as Amazon, which are easy to use, offer a transparent system of rights management, and fair pricing, thereby providing a proof of concept for the management of museum reproductions (cf. Tuchmann 2001: 315, mentioning the ease of access as a reason why services such as Bridgeman’s will endure; Morgan and Minet 2012: 55, advocating a flagship centralised image licensing agency for museums and galleries in the UK). The fees furthermore have to be reasonable in order to comply with the public museums’ mission (Garbers-v. Boehm 2011: 295–6). As Maria Economou has shown in her paper for this conference, museums can enter successful partnerships with web 2.0 communities such as Flickr; examples include the Brooklyn Museum and the Smithsonian Institution (see also Allen 2009: 25). The loss of control of images is a fact in the digital age, but not necessarily a commercial risk, as a curator of the Walter Art Museum in Baltimore admits: ‘We have lost almost all control, and this has been vital to our success’ (quoted in Kelly 2013: 28).

Museum ethics are not one-sided. The visitors and website users are obliged to contribute to a successful partnership. After all, the museum is working on behalf of the public. In his talk at the Wikipedia conference in November 2010, the National Portrait Gallery’s Tom Morgan justifiably appealed for the public’s respect for the cultural content offered by the gallery (Morgan 2010: 42’10”). A museum should be credited when its reproductions are used and re-published. Interestingly, the concept of moral rights, which has only recently entered the spheres of Anglo-American copyright legislation, seems appropriate, in that the producer of a reproduction of a work of art in the public domain should have the right to be credited as reproducer (cf. Crews 2012: 816–17; Smiers and van Schijndel 2009: 64). Moreover, there are historical precedents of photographic reproductions being famous for their quality, for example Franz Hanfstaengl’s photolithographs or Goupil’s photogravures in the nineteenth century.

If museums offer this kind of service, the question of a service fee arises. If museums provide the public with high-quality reproductions, they can expect to be paid for their services, not least as part of the desired respect (cf. Crews 2012: 806). However, commercial and non-commercial users should be charged different rates. If a global player needs a high-quality reproduction of the Mona Lisa to promote his product, he should pay a fee to support the cultural field he is making use of: the museums and the visual heritage they administer. Underpaid academics publishing books with an average readership of a few dozen experts should not have to pay as much, since they pursue non-commercial interests. In the USA, where the doctrine of ‘fair use’ is more prominent than its British equivalent, ‘fair dealing’, the Metropolitan Museum of Art in 2007 launched an ‘Images for Academic Publishing’ initiative to provide scholars with high-quality images from its collection free of charge (see Allen 2009: 15–21). The Library of Congress makes clear that it does not own the copyright on the materials in its collections also made available on its website (cf. Mazzone 2011: 18). The National Gallery of Art in Washington DC follows an open access policy for their collection, and the list of American institutions offering open access to their collections is growing and includes highlights such as the Los Angeles County Museum of Art and the Yale Center for British Art (see Kelly 2013). The recent reopening of the Rijksmuseum in Amsterdam was accompanied by a website relaunch comprising freely accessible digital archives (cf. Siegal 2013).

In the UK, important institutions have changed common museum practice and have begun to make their collections available digitally, with the option to download a high-quality image suitable for publishing. The Victoria & Albert Museum and the British Museum offer this service for non-commercial purposes, including academic publishing (on the V&A’s ‘Search the Collection’ initiative cf. Deazley 2010: 109–10; Allen 2009: 21–23). If the image of an object in the British Museum’s collections has not yet been photographed, a reasonable fee of £60 (approx. €70) is incurred. In this way, the first-time user becomes the sponsor of a reproduction, since it will be included in the digital collection and be freely available. In 2012 the National Portrait Gallery changed its policy and implemented a new licensing model (see Atkinson 2012), offering professional and academic licences for high-resolution images suitable for publication and a creative commons licence for private use suitable for digital presentations. This model reflects the considerations of this article. Commercial projects incur a fee, whereas, as a rule, the fees for illustrations for academic publications are waived. (For recent surveys of art museum image licensing policies in the UK and the USA see Allen 2009; Crews 2010–12; Kelly 2013; all with further references.)

These examples are encouraging. From personal experience I know that when it comes to non-commercial publishing, many museums’ departments of rights and reproductions are extremely helpful and provide concessionary fees wherever possible. However, even lots of good individual examples cannot substitute a consistent policy. A satisfactory balance between the interests of public museums and the interests of the public in reproductions cannot be achieved using the categories of copyright. A rethink of the approach to reproductions of works in the public domain has only just begun.