21 for 2021: Reflections on user-generated content

The 21 for 2021 project, a CREATe project within the AHRC Creative Industries Policy and Evidence Centre (PEC), offers syntheses of empirical evidence catalogued on the Copyright Evidence Portal, answering 21 copyright questions for the 21^st century. At the intermediate stage of the project, we have covered a range of topical issues: 

• User creativity in online platforms – copyright or contract? (Amy Thomas) 
• Journalists and press publishers – are their interests aligned? (Ula Furgał) 
• Copyright, re-use and digital business models (Christian Peukert and Margaritha Windisch)
• Term of copyright: optimality and reality (Paul Heald) 
• Notice-and-takedown in copyright intermediary liability (Kristofer Erickson and Martin Kretschmer) 
• Does copyright work differently in different industries? Music and television (Kenny Barr) 
• Performers’ rights (Ruth Towse) 
• Computational uses (Thomas Margoni) 
• Copyright reversion (Joshua Yuvaraj and Rebecca Giblin) 

In this post, Luis Porangaba (Lecturer in Intellectual Property Law, University of Glasgow, and Member of CREATe) reflects on the role of empirical evidence from the perspective of a legal scholar, and how to facilitate an interdisciplinary dialogue between law and social sciences.

Having been invited to take part in the ’21 for 2021’ project, I began musing on the possible ways a typical legal scholar such as myself could use and learn from the Copyright Evidence Wiki. Being someone not terribly familiar with empirical methods – though I have always valued (and admired) the insights that social science research may provide – I found the Wiki’s interface fairly intuitive and easy to operate. In a few clicks I had already been exposed to numerous studies, all helpfully presented in own wiki pages together with abstracts, main results and policy implications. So, I would say, there was no dearth of literature to be consulted.

Nonetheless, instead of offering a wide-ranging literature review of empirical research available on the platform – and, believe me, there is a plethora of studies covering the most varied copyright-related topics – I have decided for another approach, to engage differently. In this short freewheeling piece, I have resorted to some of the catalogued studies to illustrate how the Wiki may facilitate a healthy, desirable dialogue between law and social sciences. What empirical studies tell us about the assumptions we make in copyright law? How may we use such evidence? Against this backdrop, I set out to write on a heavily debated topic in contemporary copyright: user-generated content.

Anyone using online platforms, playing video games or reading fan fiction will have come across user-generated content of some kind. As part of a ‘remix culture,’[1] we have all experienced content being created, used, transformed and disseminated in myriad ways, from the open collaboration project Wikipedia to the anonymous meme. Amongst legal scholars, the term ‘user-generated content’ is now in common parlance, with a growing literature drawing attention to the effects that copyright may or may not have upon it. We seem particularly concerned with dissemination of this kind of content online, the liability that could or should be imposed on intermediaries (e.g., Facebook, YouTube), and ensuring that freedom of expression is preserved.[2] By looking into notice-and-takedown requests relating to parody videos on YouTube over a four-year period, Erickson and Kretschmer (2018) indicate some of these concerns may well be justified. It may seem that right holders’ stated policy concerns relating to commercial substitutability and risk of reputational harm are hardly considered in their sending takedown requests, which raises further questions about the increasing adoption of automated filtering systems.

Yet, the reality of legal practice may unfold another, arguably less coherent story. Given that user-generated content is not treated as a discrete doctrinal category,[3] several different aspects of copyright may encroach on one’s ability to build upon pre-existing works. For a copyright lawyer, frames from movies turned into memes and gifs would typically pose questions of subsistence, infringement and defences. In the UK, for example, whether a single frame or a sequence of frames are taken from a film could make a difference.[4]

Depending upon the relevant legal system (which may not always be apparent when we think of content sprawling over the Internet), the way such questions are framed and, of course, their answers will vary significantly.[5] And courts not always weigh – or even foresee – a potential spill over of their decisions across all such areas of copyright to user-generated content. European copyright is one such example. By ruling that any part of a work, no matter its size or significance, may enjoy protection so long as it is original,[6] the Court of Justice of the European Union may have extended the scope of copyright to otherwise unprotected or free uses.[7] Thus it is no wonder that empirical studies such as Fiesler, Feuston and Bruckman (2015) have identified that users are often concerned about – and understandably sceptical of their ability of – determining the content of the law, which may be chilling on creative activities.

So what else can we learn of user-generated content? Is there evidence of how user-generated content, as a social phenomenon, is understood and dealt with across the different industries and cultures?

Some of us may feel content in knowing that we are not alone, that law (albeit not without reservation) may not be that far apart from reality. Indeed, social science appears to be facing very similar, if not the same issues. By examining some of the studies in the Wiki, one could even say that what may seem a haphazard, incoherent approach in copyright merely reflects the complexity of the social phenomenon that is user-generated content.

Erickson (2014), for example, indicates there is no clear or widely accepted definition for user-generated content amongst stakeholders, with rights holders, online platforms and users expressing equivocal, if not plainly conflicting views. Normatively, user-generated content is perceived as a threat to moral rights and licensing revenue; as something imbued with intrinsic economic value; or a means of expression which is (or perhaps, should be) largely non-commercial and non-infringing. Interestingly, these different conceptions of user-generated content can be mapped onto the theoretical justifications for copyright often employed by legal scholars, judges and practitioners as rhetorical devices shaping legal discourse. We expect copyright to provide incentives for the creation and dissemination of works, and, yet many strongly believe such works ought to be protected as an extension of the personality of the author. Hence, the utilitarian route defended by, say, online platforms would favour a less interventionist approach (user-generated content is, after all, content being created and disseminated online), whereas an author-centric conception of copyright is framed around the basic tenet that the author should exert control over all aspects of the exploitation of her work (even if, in many instances, the rights holder is no longer or never had been the author). More recently, we have seen this tension assume more tangible form within the legislative process leading to the European Parliament’s approval of the Copyright in the Digital Single Market Directive.[8] While user-generated content permeated much of the debate surrounding the Article 17 provision,[9] it was never incorporated into statutory language.

It is therefore not surprising that we have yet to arrive at a legal definition of user-generated content which, a cynical legal scholar would dare say, may be left for courts to adjudicate. This begs the question: if user-generated content is not a uniform concept, with no single answer to be derived empirically, then how can or should it be regulated?

Here studies catalogued in the Copyright Evidence Wiki would reveal that, in sectors such as fan fiction and video games, user-generated content is already regulated in other ways; the gap left in copyright being largely supplanted by self-regulation, social norms, and contract (e.g., website terms and conditions, EULAs).[10]

Arai and Kinukawa (2014), for example, have examined the practice of ‘dojinshi’ in Japan, which refers to self-published fan fiction created predominantly by amateurs. Although such creations would ordinarily infringe copyright in the originating works, rights holders seldom seek legal recourse. The authors further argue that, at least in Japan, unauthorised use appears to produce externalities beneficial to rights holders, while contributing to social welfare. By looking into fandom sites of the famous book series which inspired the TV show Game of Thrones, Sarikakis, Krug and Rodriguez-Amat (2017) suggest that the romantic, author-centric conception of copyright has more weight than many of us would like to believe; in fact, it may inform social norms which reinforce or might even go further than the existing legal framework. While attribution and non-commerciality are encouraged, some of those norms run the risk of stifling creativity by imposing additional constraints which are neither reflected in nor derived from copyright law. In some fora, the originating work is held as authoritative, and attempts at creating or disseminating alternative, fan fiction-like narratives are reproached.

In the video game industry, Lastowska (2011) investigates the phenomenon of Minecraft, which became a surprising success by building upon – and openly encouraging – amateur creativity. While the model is premised upon rights holders expressly allowing (by means of a licence) users to modify, copyright may still act as a deterrent. Likewise, Poor (2012) examines four different platforms where ‘copying’ is encouraged, which are populated with reproductions of and homages to cultural intellectual property. Interestingly, users seem to understand they may be engaging in some form of copying, but they also feel part of a remix culture, struggling to distinguish between infringing and permitted uses. Fiesler, Feuston and Bruckman (2015) similarly draw attention to the inherent complexity of copyright as a potential barrier, which should be ‘considered more carefully in the design of user-generated content platforms.’ In his subsequent work, Lastowka (2013) finds that a significant proportion of user-generated content in videogames is wholly original, with only 25% of works surveyed being categorised as derivative or ‘referential’ of pre-existing works.

It may be trite to say that more research is needed, that we do not know or understand nearly enough of user-generated content. If anything, the literature covered in this piece addresses specific questions of limited scope which may not be sufficiently representative, and we should be wary of the risk of (over)generalisation. It would be interesting to know whether behaviour identified in this literature would be pervasive, or even if it would or could extend to other areas or across different cultures. Is the more permissible stance of right holders on doujinshi and its positive externalities a phenomenon limited to Japan, and why? To what extent is user behaviour towards fan fiction of George R R Martin’s work identified in an online discussion board representative of an entire class or sector? What, if anything, makes fan fiction and video games different?

The available evidence nonetheless raises a set of intriguing, though not necessarily novel questions. We may ask ourselves to what extent copyright law is an efficient, responsive regulatory tool, for example. If our main objective or purpose is to incentivise the creation and dissemination of works, some of the evidence may suggest this is being achieved by copyright or, perhaps, irrespective of copyright. Indeed, the emergence of contract and self (de)regulation responses to user-generated content, such as in the video game industry, could be indicative of an increasing appetite for or interest in new forms of interaction and collaboration between right holders and users.

Conversely, empirical studies also suggest that personhood rationales rooted in author-centric, arguably romantic conceptions of copyright may prevail or otherwise influence behaviour in areas such as fan fiction, and social norms may hinder or frustrate regulatory interventions aiming at promoting user innovation. An approach exclusively focused on law reform, or even on the (re)interpretation of existing law, could prove suboptimal in some of those instances. Rather, there may be room for collaboration between legal scholars and social scientists, possibly exploring other (or additional) means of intervention by way of education and access to information for example. Similarly, a social science approach to memes and gifs, which are seemingly developing into a language of its own, could give rise to interesting interfaces with legal scholarship grappling with fundamental rights such as freedom of expression.

[1] Laurence Lessig, Remix: Making Art and Commerce Thrive in the Hybrid Economy (Bloomsbury Academic 2008).

[2] See for example Martin Senftleben, ‘Bermuda Triangle: Licensing, Filtering and Privileging User-Generated Content under the New Directive on Copyright in the Digital Single Market’ (2019) 41 European Intellectual Property Review 480; Giancarlo Frosio, ‘It’s All Linked: How Communication to the Public Affects Internet Architecture’ (2020) 37 Computer Law & Security Review 105410.

[3] There have been proposals for specific user-generated content defences or, even, statutory licensing schemes which, however, have not come to fruition. See for example Christophe Geiger, ‘Statutory Licenses as Enabler of Creative Uses’ in Kung-Chung Liu and Reto M Hilty (eds), Remuneration of Copyright Owners: Regulatory Challenges of New Business Models (Springer 2017).

[4] Nicholas Caddick, Uma Suthersanen and Gwilym Harbottle, Copinger and Skone James on Copyright (17 edn, Sweet & Maxwell 2019) para 7-120 (‘… the definition of a photograph excludes any image which is part of a film, a film being defined as something from which a moving image may be produced. So, whereas a photograph of a single frame of a film will amount to a copy of the film, a copy consisting of a number of photographs of frames will not do so if a moving image, however short, can be produced from that copy.’).

[5] From a US law perspective, see for example Daniel Gervais, ‘The Tangled Web of UGC: Making Copyright Sense of User-Generated Content’ (2009) 11 Vanderbilt Journal of Entertainment and Technology Law 841.

[6] Case C-5/08 Infopaq International A/S v Danske Dagblades Forening EU:C:2009:465.

[7] See for example Case C-145/10 Painer v Standard Verlags GmbH EU:C:2013:138, paras 90-98, stating that originality in a portrait photograph may lie in the choice of background, the subject’s pose, lighting, framing of view, and the overall atmosphere created. See also Case C-476/17 Pelham v Hutter EU:C:2019:624, rejecting the German doctrine of ‘free use’ under European copyright in the context of music sampling.

[8] See for example Catherine Stihler, ‘Reflections on the Making of Copyright Law’ (CREATe Symposium 2019, 9 October 2019)  <https://www.create.ac.uk/blog/2019/11/19/create-symposium-2019-catherine-stihler-public-lecture/> accessed 28 July 2021.

[9] See for example Ula Furgal, Martin Kretschmer and Amy Thomas, ‘Memes and Parasites: A Discourse Analysis of the Copyright in the Digital Single Market Directive’ CREATe Working Paper 2020/10 (October 2020) <https://zenodo.org/record/4085050/files/2020_10_Memes%20and%20Parasites.pdf> .

[10] For a review of some of the evidence relating to those themes, see Amy Thomas, ‘21 for 2021: User Creativity in Online Platforms – Copyright or Contract?’ (CREATe Blog, 28 May 2021)  <https://www.create.ac.uk/blog/2021/05/28/21-for-2021-user-creativity-in-online-platforms-copyright-or-contract/> accessed 12 August 2021.