Introduction
Barred from travel and the opportunity to experience cultural heritage sites in person during the pandemic of 2020, the world moved online. While archaeological sites stood empty, advances in three-dimensional (3D) technology allowed people to enjoy meaningful interactions with cultural heritage through virtual replicas. The technology to record heritage sites digitally and create 3D models from the data recorded through methods such as laser scanning and photogrammetry was developed over the past couple decades primarily for informational, archival, and preservation purposes.Footnote 1 Today, digital renderings also serve to reconstruct virtually sites that have been destroyed and reimagine sites in creative and interactive ways. The pandemic has shown us that, with such virtual experiences available, the audience for digital cultural heritage sites – so-called technoheritageFootnote 2 – is likely to grow, engaging lay persons and specialist scholars alike.Footnote 3 Out of necessity or preference, erstwhile travelling tourists may soon opt to explore cultural sites through technoheritage online rather than in person.
Although virtual availability affords democratized access to cultural heritage sites in theory, the process of digitizing heritage implicates questions of intellectual property rights and how they should be allocated, both from a legal and cultural heritage policy perspective. Much scholarship has been devoted to debating and analyzing intellectual property issues in the context of digitizing cultural property housed in museums and other collecting institutions – mostly raising concerns about the control that museums exert over digital copies of their collection, regardless of whether such replicas are subject to copyright protection by law.Footnote 4 But whether digital replicas of entire cultural heritage sites are subject to copyright protections by law requires a different analysis, as does the debate about who should control access to such technoheritage.
This article seeks to untangle the knotty intellectual property issues in the context of digital cultural heritage sites. The first part of the article provides background on the scope of the 3D digital models assessed in this article as well as how and why they are made. The second part identifies what intellectual property rights are at stake in technoheritage, analyzing to what extent copyright might exist in virtual models of heritage sites under US law, which, like many other copyright regimes, requires originality for copyright protection as well as the quasi-copyrights that emerge from owning and controlling access to the underlying data. The third part explains how these intellectual property rights might be allocated among the various stakeholders in cultural heritage data-recording projects as well as other factors that might affect broad public access to technoheritage. The fourth part analyzes the significance of intellectual property rights allocation in determining who has control over the historical and political narrative of a site.
With an increased public interest in digital cultural heritage site experiences, there has also been increased demand from the cultural heritage field to develop principles that govern how such technoheritage should be managed and by whom.Footnote 5 This article posits that understanding the intellectual property in and to such digital sites and addressing this issue in the complex manner that the law requires are crucial to finding solutions that are legally sound and flexible enough to balance concerns regarding appropriation of cultural heritage with open access to information.
3D virtual models of cultural heritage sites: An overview of the landscape
The field of technoheritage has grown significantly during the past decade. Scholars are rapidly deploying increasingly diverse technologies to recreate and understand the past, including augmented and virtual reality.Footnote 6 This article focuses on 3D virtual renderings of cultural heritage sites. Organizations that specialize in the digitization of cultural heritage, such as CyArk,Footnote 7 Rekrei,Footnote 8 Iconem,Footnote 9 and Factum Foundation,Footnote 10 have created renderings of various cultural heritage sites with a range of interactive features and run the gamut from 3D digital models to immersive virtual reality experiences. The following serve as just a few illustrative examples:
-
• Factum Foundation’s tomb of Seti I in the Valley of the Kings allows viewers to examine a digital rendering of the tomb from every angle.Footnote 11 Viewers can move to specific locations through a map overview, or they can virtually walk through the tomb in a direction of the viewer’s choosing.
-
• Rekrei offers access to 3D digital models of cultural property, and one of its seminal projects was a virtual reconstruction of the Mosul Museum that was destroyed by the Islamic State of Iraq and the Levant (ISIS), which users can enjoy through an immersive virtual reality experience.Footnote 12
-
• Iconem partnered with Microsoft’s “Artificial Intelligence for Cultural Heritage” initiative to digitally record and reconstruct ancient Olympia, which has been destroyed by centuries of “weather, war, earthquakes, and modernization.”Footnote 13 The project allows viewers to “explore Ancient Olympia as it stood in its glory more than 2,000 years ago.”Footnote 14 The digital site invites users to explore the ancient Olympia in several ways, including by clicking on the various reconstructed monuments on the site (each accompanied by a brief explanation), by selecting a “curated tour,” and by exploring how the ancients used the site on each day of the Olympic games.Footnote 15
-
• In collaboration with a nonprofit division of Google called Google Arts & Culture, CyArk has created a virtual rendering of the site of Bagan.Footnote 16 Audiences can explore this ancient city in much the same manner as the tomb of Seti. As one media outlet described, “[t]he digital renditions allow viewers to virtually wander the halls of the temple, look up close at paintings, and turn the building over to look up at its chambers.”Footnote 17 Audiences of this Bagan model have the option to listen to music, take an audio tour narrated by a historian, and click on various aspects of the site for didactic text.Footnote 18
-
• CyArk created an interactive virtual reality game called “Resonant,” which is based on CyArk’s documentation of the Balcony House cliff dwelling at Mesa Verde in Colorado.Footnote 19 In the game, “players are transported to the cliff dwelling site of Balcony House, guided via walkie-talkie and a notebook with the mission to photograph portions of the site in preparation for a documentary film.”Footnote 20
Understanding the intellectual property rights associated with such technoheritage requires some background in the technology used to digitize cultural heritage sites as well as the range of purposes such models serve.
Technological methods to create 3D virtual models of cultural heritage sites
Digital cultural heritage site models are created using data that is recorded from two primary methods: laser scanning and photogrammetry. Laser scanning records data by sending laser light to the surface of an object.Footnote 21 There are several different kinds of lasers.Footnote 22 Those most commonly used to record cultural heritage sites are long-medium range scanners that use light detection and ranging (LiDAR) to record measurements.Footnote 23 Using systems based on the “time of light” or “laser pulse,” in which the laser calculates the distance to the target by bouncing a laser light off the target’s surface and recording the time it takes for the light to return, these scanners are most accurate for larger spaces such as monuments and archaeological sites.Footnote 24
Photogrammetry is an alternative method to record cultural heritage data by making measurements from photographs.Footnote 25 Historic England, the English governmental body that protects and manages England’s cultural heritage, defines photogrammetry as “the art, science and technology of determining the size, shape and identification of objects by analysing terrestrial or aerial imagery.”Footnote 26 In other words, photogrammetry is a method of extracting data about an object from photographs. Although photogrammetry technology for large spaces is not yet as well developed as LiDAR, photogrammetry can capture some information that LiDAR scanning cannot capture equally, such as color, surface textures and edges, and translucent materials such as alabaster and marble.Footnote 27 However, as photogrammetry relies on photographs instead of a physical scan, the primary advantage of photogrammetry is that it can be used to record cultural heritage that cannot be accessed because it is in danger or is lost.Footnote 28
The data garnered via laser scanning or photogrammetry must be processed in order to create a virtual model. In addition to the myriad decisions that are made in order to capture the best raw data from the site (for example, what technology to use and where to place the scanner or camera), data processing requires another level of artisanry. Processing the data can be time-consuming and painstaking and sometimes requires manipulating data points to produce the most accurate representation of the scanned subject.Footnote 29 This kind of manual manipulation includes removing unwanted features like adjacent buildings, people at the site, vegetation, obstructions, and reflections, correcting errors and holes in the surface, and adding a more realistic texture.Footnote 30 Therefore, processing can be “extensive and requires both skill and artistry.”Footnote 31
The intended purpose of 3D digital models of cultural heritage sites
Site stewards, non-governmental organizations, research institutions, nonprofits, and other stakeholders seek to record cultural heritage digitally for a range of purposes that fall into three broad categories: (1) to disseminate information about the site for archival and preservation purposes; (2) to reconstruct cultural heritage that has been destroyed; and (3) to offer creative interpretations of a cultural heritage site.Footnote 32
Archiving and disseminating information
Many cultural heritage digitization projects, such as the Factum Foundation’s 3D model tomb of Seti I, are meant to capture a cultural heritage site primarily for informational purposes. This “objective”Footnote 33 record aims to enrich the world’s cultural archive as well as “advance the understanding of cultural heritage, its values and its evolution.”Footnote 34 These initiatives also allow scholars and cultural heritage site practitioners to analyze a site closely, examine how a site changes over time, and identify conservation and preservation needs.Footnote 35 The organization Iconem, for instance, states that its 3D scans document and help preserve sites at risk and “further conservation of these endangered places by digitising them for exploration and study, today and tomorrow.”Footnote 36
Digitally reconstructing destroyed heritage
Cultural heritage data recording has been key to the restoration and digital conservation of sites that have been destroyed, whether by natural or human disaster. Rekrei, for instance, began as an initiative to recreate digitally cultural heritage that had been destroyed by ISIS in Iraq.Footnote 37 A pair of archaeologists launched a campaign to create 3D representations of heritage that had been lost through photogrammetry using crowd-sourced photographs.Footnote 38 Originally called Project Mosul, Rekrei has expanded its mission to lost heritage around the globe.Footnote 39 Iconem and Microsoft’s project to preserve and restore Ancient Olympia includes reconstructions of destroyed heritage as well. The 3D models of the buildings on the site are meant to be historically accurate, “recreated in lifelike detail, painstakingly researched by the Hellenic Ministry’s expert archaeologists to be as true as possible to their original forms” over 2,000 years ago.Footnote 40
Creative interpretations of the cultural heritage site
Some 3D renderings of cultural heritage sites are intended as creative interpretations of the site. CyArk’s game, Resonant, illustrates ways in which 3D virtual models can serve as a mere foundation for more imaginative projects. As explained on the CyArk website, “[t]he VR experience of Balcony House combines 3D data, lighting effects and audio cues with actual voices from descendant communities to convey a powerful sense of place at the site across time and cultures … [and it] provides an opportunity to interact with stories of the site in ways that would otherwise be impossible.”Footnote 41 3D models that serve as virtual tours tend to include creative interpretations of the site too. For example, the digital model of Bagan is accompanied by music, video content, and explanatory text regarding the site for viewers’ virtual tours.Footnote 42 Similarly, the digitized model of ancient Olympia offers virtual visitors the opportunity to tour the reconstructed site in different creative ways, also accompanied by music, graphics, and informational text.Footnote 43
As discussed further below, the purposes of the digital model (information, reconstruction, innovation), and the concordant originality or creative aspects of the technoheritage, affect the intellectual property analysis.
Intellectual property rights in digitized cultural heritage sites
The process of transforming cultural heritage into technoheritage converts a little piece of the world into digital content, in which intellectual property rights may reside. There are two potential categories of intellectual property that might emerge from the digitization of cultural heritage sites: (1) the copyright in original aspects of the visualization of the site and (2) the site data reflected in the digital replication of the site, which, though not protected as intellectual property by law, functions as a kind of intellectual property when the right to use such information is restricted and controlled.
A copyright analysis of digitized cultural heritage sites
Copyright protection under US law exists only for works that are original, not for works that simply represent facts about the world.Footnote 44 Because much of digitized cultural heritage amounts to mere visualized data, careful analysis is necessary to determine to what extent such technoheritage might constitute copyrightable expression, if at all.
The originality requirement
The Copyright Act grants creators protection for “original works of authorship fixed in any tangible medium of expression.”Footnote 45 American courts have interpreted “original” to require some “modicum of creativity.”Footnote 46 What distinguishes a copyrightable work from a compilation of facts is some minimal level of creativity.Footnote 47 Even if a particular work requires immense time, skill, and resources to create, without some creativity, copyright protections will not apply, regardless of the author’s “sweat of the brow.”Footnote 48
Photographic copies are not original
Soon after the medium of photography was invented, courts grappled with how to apply the originality requirement, as the camera seemed to capture reality mechanically. Courts found that the level of creativity required to show originality is low and, in photography, could include elements such as “posing the subjects, lighting, angle, selection of film and camera, evoking the desired expression, and almost any other variant involved.”Footnote 49 Photographs that simply replicated its subject, however, devoid of such minor creative choices, merely captured a fact of the world and were not protected by copyright.
This issue was first addressed in Bridgeman Art Library v. Corel Corp. Footnote 50 The Southern District of New York held that photographic copies of public domain paintings were not subject to copyright protection.Footnote 51 Although the court recognized that most photographs express the minimal creativity to achieve copyrightability, “one need not deny the creativity inherent in the art of photography to recognize that a photograph which is no more than a copy of the work of another as exact as science and technology permit lacks originality.”Footnote 52 The court rejected the argument that transforming the paintings into a new medium (photography) or the endeavor to color correct the photographs evidenced creativity.Footnote 53 Upon a motion for re-argument and reconsideration, the court reaffirmed its decision that, despite the effort it took to create the photographic reproductions, they were not copyrightable. The court pointed to the intent of the plaintiff claiming copyright protection: “[P]laintiff by its own admission has labored to create ‘slavish copies’ of public domain works of art. While it may be assumed that this required both skill and effort, there was no spark of originality indeed, the point of the exercise was to reproduce the underlying works with absolute fidelity.”Footnote 54
3D copies are more complicated
While Bridgeman remains the leading case on the copyrightability of two-dimensional replicas of public domain cultural property, a few other courts have faced this question in the context of 3D copies. Applying Bridgeman’s approach to determine originality based on intent of the creator, these courts also examined the 3D models objectively to decide whether they were indeed mere “slavish copies” or something more. In Meshwerks, Inc. v. Toyota Motor Sales U.S.A., the Tenth Circuit found that the plaintiff, Meshwerks, could not assert copyright in its 3D digital models of Toyota’s cars.Footnote 55 Meshwerks had been hired by Toyota’s advertising agency to digitize and model Toyota’s cars. The court described the labor-intensive process of creating the 3D digital models, involving meticulous data collecting and processing: “Meshwerks took copious measurements of Toyota’s vehicles … [b]ased on [which], modeling software then generated a digital image resembling a wire-frame model.”Footnote 56 And, because “the on-screen image remained far from perfect,” Meshwerks personnel processed the data to achieve an accurate representation of the vehicle.Footnote 57 Indeed, the court specified that “[a]pproximately 90% of the data points contained in each final model … [were the result of] the skill and effort its digital sculptors manually expended…. For example, Meshwerks added details like the wheels, headlights, and door handles.”Footnote 58
But, despite the time and skill involved, the digital models were not independent copyrightable creations based on both (1) Meshwerks’s purpose in creating them and (2) an objective assessment of the models. Citing the Bridgeman decision, the Tenth Circuit found that Meshwerks’s intent to reproduce the Toyota cars with absolute fidelity reflected a lack of original copyrightable expression.Footnote 59 Then, too, the court noted that the models did not include any original, individualizing features such as color, background, or contextual elements but, rather, simply portrayed the “raw facts in the world,” which are not copyrightable.Footnote 60
The court rejected Meshwerks’s argument that its models were analogous to commercial photographs of a utilitarian, non-copyrightable vodka bottle, deemed copyrightable in a Ninth Circuit case.Footnote 61 In that case, the Ninth Circuit found that the commercial photographs were copyrightable to the extent that they reflected the photographer’s creative decisions around issues such as lighting, angle, and background.Footnote 62 Such artistic judgment resulted in a very “thin” layer of copyright protection such that a different photograph of the same vodka bottle (which is itself not protected by copyright) would not necessarily infringe the plaintiff’s photograph.Footnote 63 Unlike the photograph of the vodka bottle, however, Meshwerks’s models, according to the court’s judgment, did not evidence any similar authorial decisions about “lighting, shading, the background in front of which a vehicle would be posted, the angle at which to pose it, or the like.”Footnote 64 The works therefore failed to manifest the minimal originality to warrant even a thin layer of protection.Footnote 65
The court in Meshwerks clarified that there is nothing inherent in the medium of 3D digitization to preclude copyright protection, leaving open the possibility that a 3D model could constitute copyrightable expression.Footnote 66 And, indeed, two years later, in Osment Models v. Mike’s Train House,Footnote 67 the Western District of Missouri found that a certain model train station, based on a real historic train station but with several visible alterations, was indeed sufficiently original to survive a motion for summary judgment as to the question of copyrightability.Footnote 68 The plaintiff seeking copyright protection in that case, Osment, had not merely copied the real train station into another medium.Footnote 69 Osment’s models were meant to reflect historic accuracy but not to “replicate actual buildings.”Footnote 70 Then, too, Osment made visual changes to the real train station, such as altering certain coloring, the number of chimneys, the window frames, and the awnings.Footnote 71 The court clarified, however, that Osment’s copyright protection extended only so far as it had added these original elements, creating just a very thin layer of copyright protection.Footnote 72 Unlike the models in Meshwerks, therefore, Osment’s models in this case were analogous to the photographs of the vodka bottles.Footnote 73
Applying the case law to 3D virtual models of cultural heritage sites
With little relevant case law, it is not clear whether 3D virtual models of cultural heritage sites would be copyrightable under current precedent. It seems that the closer a model “slavishly” adheres to portraying the site as an exact replica, the less likely the model would qualify for copyright protection, regardless of the creativity exercised in the data collecting or processing. However, to the extent that the model veers away from reproducing facts or replicating reality, such deviations would be protected by copyright.Footnote 74
The courts’ emphasis on intent suggests that the level of copyright protection (if any exists) might depend on the purpose of the model as discussed above. First, courts might treat models created to disseminate purely objective information for scholarly purposes, like the Factum Foundation’s tomb of Seti, similarly to the photographs in Bridgeman or the scan in Meshwerks, and therefore may not be subject to any copyright protection. According to the attorney who argued the Meshwerks case, for example, “digital 3D models of cultural heritage are ‘awesome, impressive, incredibly labor intensive, highly detailed, and skillful. One thing they are not, however, is copyrightable.”Footnote 75 Second, even reconstructed sites, like Rekrei’s models or the reconstructed buildings in ancient Olympia, might fail to qualify as copyrightable works because the goal is to replicate something that used to exist, which is subject to public domain. According to copyright specialist Charles Cronin, “[r]estoring a public domain work to its original reification, whether an ancient bronze or a score by Mozart, requires time and intellectual application, but these investments yield no copyrightable interest to those making them, at least under United States law.”Footnote 76 Finally, creative interpretations of sites, by contrast, like CyArk’s video game or creative virtual tours of the sites of Bagan and ancient Olympia, are more likely to enjoy copyright protection, at least insofar as such models incorporate original elements and do not simply portray a faithful reproduction or reconstruction of the site.Footnote 77
Intent, however, is only one part of the originality inquiry under Meshwerks and Osment. And the purpose of a cultural heritage site model may not be dispositive to the question of whether it constitutes copyrightable expression. Meshwerks indicates that all of the technical labor, or “artistry” as the Factum Foundation phrases it,Footnote 78 required to produce a 3D model will not render it copyrightable so long as the result simply reflects facts of the world.Footnote 79 According to this case law, none of the judgment calls made throughout the laser scanning, photogrammetry, or data processing to make a faithful reproduction of a site can give rise to copyright protection. But while 3D models of discrete objects like Toyota’s cars might simply portray fact, the copyright analysis for a 3D scan grows more nuanced the larger and more complex the subject recorded.Footnote 80 Digitizing an entire cultural heritage site inherently requires some diversion from reality that might imbue a layer of copyright protection.
As an initial matter, unlike 3D replicas of cultural property objects that are viewed and understood ex situ in museum collections, models of cultural heritage sites must always detach the site from the real environment in which it exists. When embarking on a laser scanning or photogrammetry project, the creator of the model must decide where to draw the boundaries of the site for the model. The creator must also decide whether to include representations of the people who usually occupy the site (for example, guards and tourists) or to portray any given tree, shrub, or other environmental element that might obstruct the view of the site from certain angles, not to mention all the other objects that are integral to real cultural heritage sites: signage, refreshment stands, ticket kiosks, and even trashcans and litter. The creator therefore makes decisions that affect how all the elements of a lived, real-world physical heritage site are portrayed in virtual reality. All these slight alterations are possibly enough to warrant copyright protection.Footnote 81 But even if that were true, any model that portrays a relatively realistic rendering of a site would enjoy a very thin layer of copyright protection that would only go so far as those minor creative choices.
Data ownership as an intellectual property right
Regardless of whether the digitized models of cultural heritage are protected by copyright, control over access to, and use of, the underlying data itself can be very powerful. Although data alone is not subject to copyright protection or any other intellectual property doctrine by law, gatekeepers to data can control by contract how that information is used. This is crucial in determining the allocation of intellectual property rights over digital cultural heritage sites, particularly as any copyright in the models that might exist would typically be very thin. Even though the cultural heritage sites themselves are not subject to copyright, the ability to control the exploitation of the site data used to create digital renderings allows the data owner to exercise a quasi-copyright over such depictions.
Drawing from museums’ approach
A great deal of scholarly discussion has explored the phenomenon of “copyright overreach” that museums exert over digitized public domain works of art in their collections.Footnote 82 Even though Bridgeman, Meshwerks, and Osment held that such photographs and 3D scans are not likely to be protected by copyright, museums exercise ownership rights all the same.Footnote 83 A frequently cited example of this position involves the bust of Nefertiti in the Neues Museum in Berlin.Footnote 84 Although this ancient work of art is in the public domain, the Neues Museum did not permit access to its digital information about the piece.Footnote 85 Undeterred, two trench coat-wearing artists sneaked a 3D scanner hidden under a cashmere scarf into the museum and surreptitiously scanned the world famous sculpture in 2015.Footnote 86 The artists, Nora Al-Badri and Jan Nikolai Nelles, released the digital replica online under a creative commons license,Footnote 87 allowing the public to reproduce or remix the work for their own purposes.Footnote 88 Al-Badri and Nelles’s digital heist was meant to serve as a statement against Western nations’ historical spoliation of Egyptian patrimony still displayed in colonialist narratives and to protest restrictive policies on sharing information relating to museum collections.Footnote 89
The artists’ digital replica was so detailed and precise, however, that some experts recognized that it was not based on the scan from the heist but, rather, on the Neues Museum’s own professional digital scan, leaked or stolen from the museum’s servers.Footnote 90 One specialist sought confirmation, but the museum refused to disclose its scan, and it was only after a German-equivalent Freedom of Information Act request and years of legal battles that the museum finally released its digital model under a creative commons license.Footnote 91 The museum’s digital bust of Nefertiti appears to be exactly the same as Al-Badri and Nelles’s version except that it has a copyright notice digitally carved into the flat underside of the base.Footnote 92
This story of the Nefertiti bust illustrates the power that museums wield by controlling access to information even when the object to which the information pertains is in the public domain. By controlling access, museums also control the right to use such digital photos and scans. Museums exercise such control through license terms, which are enforced through various mechanisms such as formal contracts, website terms of use, or an application of a creative commons license (as the Neues Museum ultimately did).Footnote 93 In some license terms, museums hold themselves out as the copyright owners to such digitized cultural property; in others, the museums’ claims are more vague.Footnote 94 But, ultimately, whether the museums’ rights are legally based in copyright is largely irrelevant. Copyright specialists Kenneth Crews and Melissa Brown conducted a study of museums’ approach to their digitized collections and concluded: “Whether or not the copyright in the reproduction is ultimately found to be legally valid, license terms and conditions of use are also often implicated through contract law.”Footnote 95 Thus, what copyright does not provide, museums can achieve through contracts.Footnote 96 The effect is the same; because museums serve as the gatekeepers to this information, they are able to control its use.Footnote 97
Although most analyses of museums’ attitudes toward their digitized collections critique museums’ restrictive approach, there are many reasons why museums are compelled to exert such control.Footnote 98 For example, as trustees of their collections, museums must ensure the works’ integrity and take measures to avoid disseminating information that, on the one hand, could be inaccurate as to aspects like attribution, size, and medium or, on the other hand, might be too copious and precise so as to allow forgeries. Museums generate income through these licenses, relating to both research permissions and licenses for resale products. Museums also desire credit for their collections, which augments their prominence and reputation. And, finally, museums may need to restrict use as required by donors.Footnote 99
Applying this discussion of “copyright overreach” by museums to the context of digital cultural heritage sites, a key lesson is that controlling access to public domain technoheritage can be exercised through contract law and, indeed, restrictions may be necessary for a host of reasons. Whoever owns the data to the 3D virtual models of cultural heritage sites sits in a similar position to the museums. So long as the relevant provisions are included in the contract governing digitization of a cultural heritage site, the data owner can control how the information is portrayed and whether and how it can be shared with others.
Conflation of copyright and data ownership in cultural heritage
Like museums, practitioners in the cultural heritage sector tend to treat data as copyrightable content too. Although copyright in original elements of a 3D virtual model is separate and distinct from rights to the data underlying the replica, cultural heritage practitioners often lump the two together in a vague all-encompassing reference to intellectual property that seems to be subject to copyright protection. For example, CyArk’s website content, including all its virtual models, is tagged with a copyright notice for “CyArk & Partners, all rights reserved,”Footnote 100 which ambiguously covers both copyrightable content and public domain information on the website. Then too, under its Data Use Policy, CyArk licenses the point cloud data and photogrammetric imagery in its “Open Heritage Program” models under a creative commons copyright license. Adopting a creative commons license, however, implies that the content covered by the license is copyrightable. CyArk thus treats non-copyrightable information as copyrightable material.Footnote 101
This conflation of data and copyright could result from various factors at play. For example, because copyright protection in 3D virtual models of cultural heritage sites is usually flimsy at best, the most significant intellectual property is the data, which can function essentially as copyrightable content through contractual terms, like in many museum licenses. Treating data as part of the copyrightable content would be preferable for organizations recording the heritage sites because it allows such organizations to sweep up data ownership into their otherwise very thin copyright ownership. Scholars critiquing museum “overreach” have ascribed this purposeful protectionism to museums. In the museum context, however, it is clear that should any intellectual property exist in the digitized versions of a museum’s collection, the museum would be the owner. In the cultural heritage field by contrast (as discussed further below), there are many more potential rights holders involved in a digitization project, including the organization recording the data and the site steward. The additional stakeholders render the intellectual property analysis of digital cultural heritage sites all the more complex and complicates any assumption that, like museums, organizations creating the 3D models are intentionally adopting a proprietary approach.
Cultural heritage practitioners might simply find it easier to treat everything as the same category of intellectual property, all subject to copyright protection, rather than parse out the copyrightable content from pure data. The position that data is copyrightable may also emerge from a lack of understanding in the cultural heritage field about this technical area of the law and may be the by-product of the field’s effort to graft cultural heritage into the digital landscape. For instance, Rekrei provides its models under a creative commons license merely due to its use of the 3D content platform Sketchfab. Rekrei users add their own content to Rekrei’s database of technoheritage by creating models on Sketchfab and then tagging them as part of the Rekrei community.Footnote 102 Under the Sketchfab terms of use, content creators own their content; in order to allow others to download and use it, creators must license their content under a creative commons license.Footnote 103 Using Sketchfab for models of public domain cultural property thus imbues such models with a copyright patina, regardless of whether they are really copyrightable by law. Ironically, this results in organizations that seek to increase access to this information restricting access by subjecting such information to copyright protections that it does not necessarily enjoy by law. Finally, it is also possible that the conflation of data and intellectual property stems from a genuine confusion about intellectual property law and a misconception that it protects content that constitutes mere data. Whatever the rationale or motivation for conflating data and copyright, the outcome is a proprietary approach to 3D virtual models of cultural heritage sites, which are treated as copyrightable intellectual property, regardless of how accurate or unoriginal.
Allocating intellectual property rights in the digitization of cultural heritage sites
Despite a growing call to develop consistent ethical and sustainable protocols for digital cultural heritage recording, there is no uniform policy or default practice governing how rights to the data or copyrights are allocated among the various collaborators involved in a cultural heritage digitization project. The negotiations over intellectual property rights to digital models of cultural heritage sites involve several stakeholders with often countervailing viewpoints. The primary stakeholders are (1) the site stewards who own or are otherwise authorized to care for the site and may therefore take a proprietary approach to content regarding their site and (2) the organizations recording the data and creating the models, which often take the stance that their work preserves and disseminates the world’s shared heritage.Footnote 104 In addition, these projects are frequently funded at least in part by private entities such as charitable foundations that affect the negotiations over intellectual property too, by requiring specific intellectual property structures in their grants.Footnote 105 Sometimes, the site steward does not reflect the “source community” – the people who originally created or used the sites or who consider it a vital part of their cultural identity – which could further complicate the analysis over intellectual property rights.Footnote 106 For example, the US federal government serves as the custodian for various Native American cultural sites.Footnote 107 Finally, these negotiations over intellectual property rights are in the context of creating a digital model that will be used by a public audience, with a desire for open access and broad latitude on how they can interact with and consume the digital cultural heritage. All these diverse perspectives result in a complex analysis of how intellectual property should be allocated, and the result differs for each project depending upon the specific facts and dynamics.Footnote 108
The approach of many organizations in the cultural heritage sector is that the site data are generally owned by the site steward.Footnote 109 This position might stem from the fact that site stewards hold the key (sometimes literally) to accessing the physical site itself in order to scan it properly. However, the extent to which site stewards leverage their role as gatekeeper varies drastically depending on the circumstances of a project, including the site stewards’ experience and resources. Site stewards that have greater resources to negotiate digitization of their heritage sites, more experience with such projects, and a history of using cultural patrimony as a political tool tend to take a restrictive, retentionist position regarding their site data. For example, under Egypt’s Antiquities Protection Law, all Egyptian cultural heritage sites are guarded, and only those specialized institutions granted a license by the Supreme Council of Antiquities may “study, draw, [or] photograph” a cultural heritage site.Footnote 110 Cultural heritage practitioners cannot legally access or record a site without a license from the Egyptian government. Similarly, the Italian Code of Cultural Heritage and Landscape dictates that any reproduction of cultural property requires authorization of the relevant site custodian.Footnote 111 Under these legal regimes, organizations seeking to create virtual models of Egyptian or Italian sites must comply with whatever restrictions the site steward (a government agency) might require, including ownership of the data extracted.
By contrast, site stewards with fewer resources and where cultural property may be less politicized sometimes do not negotiate control or ownership of the intellectual property at all. The site of Bagan in Myanmar, for example, was recorded by CyArk for a project in collaboration with Google Arts & Culture.Footnote 112 Myanmar does not have a robust history of protecting its cultural patrimony or the administrative and regulatory infrastructure that such a commitment to national cultural heritage entails. Although the negotiation details of the CyArk and Google Arts & Culture project are confidential, the result as reported in the media is that CyArk owns the scanned data itself, not just the copyrightable content that it might have created.Footnote 113 While the specifics of this negotiation are not available to be analyzed, the outcome suggests that perhaps a lack of experience and financial resources may have played a role. The Myanmar Department of Archaeology may not have fully appreciated the power it could wield or may have decided that the value gained in publicizing and increasing engagement in its site outweighed ownership rights.
Finally, while controlling access to a cultural heritage site often renders site stewards the gatekeepers to site data, this is not always the case. Photogrammetry technology allows for the construction of virtual 3D models without needing access to the physical site. As scholar Erin Thompson summarizes, “[a]n owner could seek to prevent digitalization by restricting access to an artifact, for example, by prohibiting photography by visitors, as many museums do. But this is a moot point once sufficient photographs have been taken to permit 3D modeling.”Footnote 114 For example, initiatives such as Rekrei digitally reconstruct destroyed heritage based on crowd-sourced photos.Footnote 115 Determining who owns, or should own, the scans in these contexts is all the more complex.
Further complicating this analysis is the fact that, while site stewards are usually the natural candidates to own the data, the organizations creating the 3D model data, under US law, would by default own the copyrights to the extent that they exist in the model itself. However, a powerful site steward could require ownership by assignment. And, as discussed above, copyright ownership in 3D replicas is at most relatively thin and does not afford the owner to use or sublicense the underlying data. Therefore, owning the copyrights in a 3D model does not necessarily provide much leverage in negotiations over intellectual property allocation. Still, the more creative and innovative the technoheritage is, and the more sophisticated the creator, these copyrights could shift the dynamics in deciding which party should own what.
Allocating ownership among the stakeholders, however, does not necessarily dictate the availability or accessibility of the intellectual property and the data. An owner can decide to open access or restrict access to its intellectual property or data. Even when ownership rights have been decided, how that owner permits others, including public audiences, to use the technoheritage and its data also depends on the context of the specific project.
Governing laws or regulations
The site steward’s locality may have laws or regulations governing the exploitation of intellectual property involving a cultural heritage site. For example, under Italy’s cultural patrimony law, not only must any reproduction of Italian cultural heritage be authorized by the relevant governmental bodies, but any commercial exploitation of such reproductions is also subject to concession fees to the government.Footnote 116
Concerns regarding site looting
Security concerns over a particular site may also prompt restrictions in the use of information. Specifics about an archaeological site can provide a roadmap for potential looters, for instance.Footnote 117 This is particularly sensitive for sites in conflict zones.Footnote 118 Site stewards may therefore wish to prohibit certain uses of the data or to permit only redacted versions of the information. For example, the Digital Archaeological Record (tDAR), a database that compiles myriad digitized records on cultural heritage from third parties, states in its agreement with these contributors that tDAR “will make reasonable efforts to obfuscate from public view any highly precise, mapped site locations contained in the contributor’s metadata,” in addition to information the contributor designates as “confidential,” including any other information “that would be reasonably expected to endanger in situ archaeological sites if it were made publicly available.”Footnote 119
Cultural and political sensitivities
A plethora of cultural or political sensitivities might warrant limiting access to cultural heritage data information or controlling how it is portrayed to the public. Some Indigenous peoples or local communities who care for cultural heritage may request restrictions to access due to “concerns about cultural privacy, sensitivity, or misappropriation.”Footnote 120 For example, when the Smithsonian created a series of 3D replicas of objects repatriated to the Tlingit community of southeast Alaska, Tlingit clan leaders allowed such reproductions to be shown to the public so long as no copies would be accessible online without their consultation.Footnote 121 In addition, in anthropologist Kimberly Christen’s project with the Warumungu community in Australia, she ensured that certain images were accessible only by people with the proper ritual standing.Footnote 122
Cultural heritage does not need to be sacred in order to be sensitive. In the summer of 2020, CyArk responded to the Black Lives Matter (BLM) movement by documenting how this historic moment and the “messages of racial justice, outrage, solidarity and change” were reflected in the landscape of American cities. CyArk aims to “tell the stories of these ongoing events and amplify the voices of activists and marginalized communities … [and] to engage with communities to document the physical manifestations of the struggles against racial injustice.”Footnote 123 If accessible to the public, however, the data collected in this project could be contextualized or manipulated in a manner that would produce a very different message. The same data in the hands of white supremacists, for instance, could degrade the BLM movement’s important work. Due to this risk, CyArk left the decision to publish the data to the community.Footnote 124 Some site stewards might face other political challenges regarding their cultural heritage sites and may need to control the context in which the site is portrayed accordingly. For example, when sites that are in the running to be considered a United Nations Educational, Scientific and Cultural Organization World Heritage Site are documented, the site steward may wish to control the narrative about the site to ensure that it is portrayed in a manner that optimizes, or at least does not diminish, its chances of selection.Footnote 125
What is at stake in the allocation of intellectual property rights over technoheritage
The allocation of intellectual property rights in digital cultural heritage implicates who has control over the historical and political narrative of a site. No depiction of cultural heritage is devoid of a particular perspective. Although digital reconstructions can purport to present objective views of the past, technoheritage is embedded with assumptions and decisions. Even the decision of which site to digitally record can reflect certain (usually Western) ideas of which world heritage is most important.Footnote 126 In fact, many of the same creative decisions that data recorders make which might be subject to copyright also imbue their virtual models with some perspective or bias. For example, the decision to include or exclude depictions of the surrounding community or other human beings at the site affects the site narrative. The absence tends to strip the site of its cultural context, and the dynamics between the site and the community that lives there – depicting a sort of terra nullius (or terra omnis, depending on your perspective). As Erin Thompson has explained, “[t]he empty spaces of digital reconstructions can appear ready for conquest in the same way, for example, that early European settlers imagined the landscape of America to be empty. Paintings, drawings, and later photographs showed vast stretches of land without any evidence of human inhabitants.”Footnote 127 Removing site signage such as didactic panels also allows for other interpretations of the site, which could be enforced by the manner in which the end user interacts with the digital model. For example, virtual models that follow a specific path through the site would direct one particular vision of the site.Footnote 128
The more that any party has a proprietary right to the intellectual property in a digital cultural heritage site, the more it can dictate how that digitized site is portrayed to the public. Therefore, the question of who should have what intellectual property rights closely maps the question of who should have domain over the site’s narrative and meaning. This debate reflects the broader discourse about cultural property ownership. Technoheritage, and the propriety rights to such digitized heritage, functions as a “meta-cultural property,” triggering many of the same questions about ownership.Footnote 129 Intellectual property rights to digital cultural heritage is a new frontier of the cultural property ownership debate, and the distinction between cultural property and the intellectual property rights therein can grow blurry.
On the one hand, because mostly Western groups record data and create virtual models, some scholars voice concerns about the “digital colonialism” that results from those organizations exercising too much control over how the information is disseminated.Footnote 130 Moreover, digital colonialism perpetuated through intellectual property ownership by Western organizations is not necessarily mitigated by allowing open access to such cultural heritage data. The colonialism is enacted in the ownership and control over the narrative as well as how it is disseminated to the public.Footnote 131 The concern is that, mediated through Western eyes, digital cultural heritage reflects Western notions and priorities, while excluding the perspectives of the communities for whom such heritage is a daily part of their lives or a key element to their cultural identity.
On the other hand, however, site stewards, particularly governmental entities, may have their own agendas that do not necessarily align with ideals of furthering knowledge about a shared world heritage. For example, in 2000, a group of American university students and faculty created a digital 3D scan of Michelangelo’s David. However, based on the alleged negotiated requirements with the Italian government, in order to gain permission to access the model, end users must promise to “‘keep renderings and use of the data in good taste’ because the artifacts in question ‘are the proud artistic patrimony of Italy.’”Footnote 132 Retentionist policies that lead to these kinds of restrictions have a chilling effect on creative and potentially controversial interpretations of cultural heritage.Footnote 133 Regardless of who has control of the narrative, however, it is critically important to understand the power dynamics over the intellectual property in order to contextualize the version of history depicted in a digital model of a cultural heritage site.
Conclusion
While there are a few supranational bodies and initiatives that have issued ethical standards for the digital recording of cultural heritage, there is no recent universal set of principles. The International Council of Monuments and Sites’s Principles for the Recording of Monuments, Groups of Buildings and Sites was drafted in 1996 and is long overdue for an update.Footnote 134 And there is a major push in the field to create such uniform guidance that would properly address modern technology.Footnote 135 It is beyond the scope of this article to suggest which supranational body takes the helm to issue such guidance. But to achieve policies and practices that reflect all stakeholders’ concerns, any such guidance must address intellectual property rights as they are central to determining how such information about cultural heritage sites is used as well as determining the interpretation and the cultural significance of these sites more generally.
Although digital models may have some thin layer of copyright, the underlying information reflected in the precise replica is separate and distinct from such copyrightable content. However, currently, data and whatever copyrighted content exists are treated altogether as copyrightable material. Therefore, one party ultimately controls everything, exacerbating the tension between the site stewards and the Western organizations digitizing the sites. But this clunky approach is not legally sound, nor is it particularly productive. For example, in many circumstances, it may be most appropriate for the site steward to own the data and for the organization recording the site to own any copyright in creative elements added. Depending on other stakeholders involved, however, such as private funders or source communities, alternative arrangements might be preferable. In addition, ownership by one party or another is not the end of the discussion; various project factors affect to what extent an owner may decide to provide others, including public audiences, access to its copyrights or data. A more complex analysis of the intellectual property affords creative, flexible solutions in allocation of rights that can protect site stewards’ needs regarding their heritage, promote the interests of data recorders to document and preserve the world’s shared cultural heritage, and, at the same time, further the public’s desire to enjoy, study, and participate in technoheritage.
Acknowledgments
Attorney, New York, USA. Now at Google.
