Transition and Coherence in Intellectual Property Law: Essays in Honour of Annette Kur 9781108688529, 9781108484602, 9781108723367

Citation preview

transition and coherence in intellectual property law The nature and content of intellectual property (IP) law, which is heavily contingent on the state of technology and on social and market developments, has always been subject to ongoing transitions. How those transitions are effected and the shape they take is crucial to the ability of IP to achieve its stated goals and provide the necessary climate for investment in creativity, innovation and brand differentiation. Yet the need for change can run headlong into a desire for coherence. A search for coherence tests the limits of the concept of “intellectual property,” is imperiled by overlaps between different IP regimes, and calls for a unifying normative theme. This volume assembles contributors from across IP and the globe to explore these questions, including whether coherence is desirable. It should be read by anyone interested in understanding the conceptual underpinnings of one of the most important and dynamic areas of the law. Niklas Bruun is Professor emeritus, Hanken School of Economics Helsinki. He served as the Director of the IPR University Center in Helsinki from 2000-2018. He has been the leading scholar in IP-law in Finland for many years and is the author of numerous books and articles in the field, including Intellectual Property Law of Finland. Professor Bruun has also been the chair of several committees for law revisions of IP in Finland. Together with Professor Nari Lee, he led a research project on innovation and IP enforcement in China that brought together researchers from across Europe and China. Graeme B. Dinwoodie is the Global Professor of Intellectual Property Law at ChicagoKent College of Law. He remains a visiting Professor at the University of Oxford, where he held the IP Chair for nine years. His previous visiting appointments include serving as the Yong Shook Lin Visiting Professor of Intellectual Property Law at the National University of Singapore, and Global Professor of Law at New York University School of Law. Professor Dinwoodie was a John F. Kennedy Scholar at Harvard Law School and was elected to the American Law Institute in 2003. He is the co-author of A Neofederalist Vision of TRIPS: The Resilience of the International Intellectual Property Regime and five casebooks. Marianne Levin is Professor emerita, Department of Law, Stockholm University. Since 1995, she has served as the Chair of the Swedish Association for the Protection of Intellectual Property Rights. Professor Levin was the head of the Press and PR department of the EPO from 1986-1988, and was one of the four people involved in the 1990 draft for an EU Design protection. She served as the Director of the Institute of Intellectual Property and Market Law, and founded the Master’s Program in European Intellectual Property Law at Stockholm University. She holds an honorary doctorate from the Swedish School of Economics, Helsinki. Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Ansgar Ohly holds the Chair of Private Law, Intellectual Property and Competition Law at the Ludwig Maximilian University of Munich. He is also a Visiting Professor at the University of Oxford and a Visiting Senior Member of St Peter’s College Oxford. He has published widely on all areas of intellectual property law and the law of unfair competition law, with a special emphasis on European developments and the comparison of civil law and common law systems. He is the co-editor of GRUR, the leading German intellectual property journal, and of the commentary on German copyright law founded by G. Schricker.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

CAMBRIDGE INTELLECTUAL PROPERTY AND INFORMATION LAW

As its economic potential has rapidly expanded, intellectual property has become a subject of front-rank legal importance. Cambridge Intellectual Property and Information Law is a series of monograph studies of major current issues in intellectual property. Each volume contains a mix of international, European, comparative and national law, making this a highly significant series for practitioners, judges and academic researchers in many countries. Series Editors Lionel Bently Herchel Smith Professor of Intellectual Property Law, University of Cambridge Graeme Dinwoodie Global Professor of Intellectual Property Law, Chicago-Kent College of Law, Illinois Institute of Technology Advisory Editors William R. Cornish, Emeritus Herchel Smith Professor of Intellectual Property Law, University of Cambridge François Dessemontet, Professor of Law, University of Lausanne Jane C. Ginsburg, Morton L. Janklow Professor of Literary and Artistic Property Law, Columbia Law School Paul Goldstein, Professor of Law, Stanford University The Rt Hon. Sir Robin Jacob, Hugh Laddie Professor of Intellectual Property, University College London Ansgar Ohly, Professor of Intellectual Property Law, Ludwig-Maximilian University of Munich A list of books in the series can be found at the end of this volume.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Transition and Coherence in Intellectual Property Law essays in honour of annette kur Edited by

NIKLAS BRUUN Hanken School of Economics (Finland)

GRAEME B. DINWOODIE Chicago-Kent College of Law

MARIANNE LEVIN Stockholm University Department of Law

ANSGAR OHLY Ludwig Maximilian University of Munich Faculty of Law

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

University Printing House, Cambridge cb2 8bs, United Kingdom One Liberty Plaza, 20th Floor, New York, ny 10006, USA 477 Williamstown Road, Port Melbourne, vic 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781108484602 doi: 10.1017/9781108688529 © Cambridge University Press 2021 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2021 A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data names: Kur, Annette, honouree. | Bruun, Niklas, editor. | Dinwoodie, Graeme B., editor. | Levin, Marianne, 1942– editor. | Ohly, Ansgar, editor. title: Transition and coherence in intellectual property law : essays in honour of Annette Kur / edited by Niklas Bruun, Hanken School of Economics (Finland); Graeme B. Dinwoodie, Chicago-Kent College of Law; Marianne Levin, Stockholm University Department of Law; Ansgar Ohly, Ludwig Maximilian University of Munich Faculty of Law. description: Cambridge, United Kingdom ; New York, NY : Cambridge University Press, 2021. | series: Cambridge intellectual property and information law | Includes bibliographical references and index. identifiers: lccn 2020017900 (print) | lccn 2020017901 (ebook) | isbn 9781108484602 (hardback) | isbn 9781108723367 (paperback) | isbn 9781108688529 (ebook) subjects: lcsh: Intellectual property–European Union countries. | Intellectual property (International law) | Agreement on Trade-Related Aspects of Intellectual Property Rights (1994 April 15) | Kur, Annette. classification: lcc kje2636 .t73 2021 (print) | lcc kje2636 (ebook) | ddc 346.2404/8–dc23 LC record available at https://lccn.loc.gov/2020017900 LC ebook record available at https://lccn.loc.gov/2020017901 isbn 978-1-108-48460-2 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Contents

Preface Greetings to Annette Kur from the Second Floor Dieter Stauder Annette Kur: Toward Understanding Stacey Dogan

page xiii 1 3

part i transition a forms and institutions

13

1

Transitional Provisions in Intellectual Property Legislation Richard Arnold

15

2

Judicial Creativity and Transitions in EU Intellectual Property Law Lionel Bently

25

3

Before and after Designers Guild: Another Look at Appellate Deference in New Zealand’s Copyright Law Graeme W. Austin

46

EU Design Law: Transitioning Towards Coherence? Fifteen Years of National Case Law Estelle Derclaye

56

Copyright and the CJEU: Some Structural Deficits as Seen from a German Perspective Thomas Dreier

68

4

5

Downloaded from https://www.cambridge.org/core. vii , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

viii

6

Contents

b international commitments and constraints

83

Global Intellectual Property: Transition and Coherence through Rules of Interpretation Susy Frankel

85

7

Article 20 of the TRIPS Agreement: Up in Smoke? Daniel J. Gervais

8

Implementing Treaty Obligations at the National Level Sam Ricketson

107

9

Multiple and Overlapping Transitions in IP Marco Ricolfi

121

10

Transition and Continuity in the Private International Law of Intellectual Property Christian Heinze

133

From Nintendo Wii to Perfumes, Driving a BMW Car: A Tale of Transition to the Wrong Kind of Coherence Paul Torremans

145

c new agents and the challenge of new technologies

155

12

Transition through Automation Henning Grosse Ruse-Khan

157

13

Eye, Robot: Artificial Intelligence and Trade Mark Registers Dev S. Gangjee

174

14

Patent Protection of Inventions Involving Artificial Intelligence Stefan Luginbuehl

191

15

Automated Profiling in New Media and Entertainment Markets: What to Protect, and How? Megan Richardson

11

96

200

part ii coherence

16

a intellectual “property” and its limits

211

The (Lack of ) Coherence of Data Ownership with the Intellectual Property System Josef Drexl

213

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Contents

17

The Threefold Fictitiousness of Intellectual Property Alexander Peukert

18

An Intellectual Property Structural Engineer Extraordinaire and Her Lifelong Quest for Coherence Peter K. Yu

19

Open Yet Secret – Trading of Tangible Goods and Trade Secrets Nari Lee

20

From Smörgåsbord to New Nordic Cuisine: EU Harmonization of Trade Secrets Protection in the Nordic Countries Jens Schovsbo

21

Trade Mark Rights and Parallel Imports vis-à-vis the Never-Ending Evolution of the Behavior of Firms: Transition and Coherence Put to a Test Vincenzo Di Cataldo

22

Legal Concept of “Exhaustion”: Exhausted? Reto M. Hilty

23

Building Coherence in Technological Transitions: Putting Exploitation at the Core of IP Séverine Dussollier

24

25

26

27

28

ix

224

232

242

254

263

272

284

“Accessory Exhaustion” – and Use of a Work as a Work Ole-Andreas Rognstad

294

b ip overlaps

305

Intellectual Property in Transition: The Several Sides of Overlapping Copyright and Trademark Protection Jane C. Ginsburg and Irene Calboli

307

Cultural Heritage and the Public Domain: What the US’s Myriad and Mayo can Teach Oslo’s Angry Boy Rochelle C. Dreyfuss

322

Public Order in the Light of Aesthetic Theory: The Copyright/ Trademark Interface after Vigeland Martin Senftleben

332

Separability as Channeling: A Cautionary Tale Mark P. McKenna

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

344

x

29

Contents

Novelty, Idea or New Meaning as Criteria for Copyright Protection? Transitions in Swedish Design Law Jan Rosén

352

30

Examining Functionality Mark D. Janis

364

31

Substantial Value and the Concept of Shapes Antoon A. Quaedvlieg

373

32

Copyright and Patents on Software: The UPC’s Answer to an Old Problem of Intellectual Property Overlaps Axel Metzger

382

Chopping off Hydra’s Heads: Spare Parts in EU Design and Trade Mark Law Anna Tischner

392

c (un-)fairness

405

Geographical Indications as Intellectual Property Rights – Beyond Transition and Coherence Roland Knaak

407

Presence or Absence of Coherence in Trade Identity Protection in the European Union Alexander von Mühlendahl

416

33

34

35

36

Virtue Ethics and Private Law – A Sketch Marcus Norrgård

37

Closing the Gap: How EU Law Constrains National Rules Against Imitation Martin Husovec

436

European Union Law and Slavish Imitation: An Update in Honour of Annette Kur Matthias Leistner

447

38

39

The German Misappropriation Origins of Trademark Antidilution Doctrine: A Translation of the 1924 Odol Opinion of the Elberfeld Landgericht Barton Beebe

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

428

460

Contents

40

41

42

The Relationship between the Unfair Competition Regime and IP Law Gustavo Ghidini and Giovanni Cavani Comparative Advertising: Does Trade Mark Law Over- or Under-Protect the Average Consumer? A Couple of Recent Examples of Asian Jurisdictions Going their Own Way David Llewelyn

xi

467

478

conclusion

487

Transition and Coherence in Intellectual Property Law Graeme B. Dinwoodie and Ansgar Ohly

489

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Preface

As will be seen from the identity and location of the contributors to this volume, there is no doubt among intellectual property scholars anywhere that Annette Kur deserves a “Festschrift.” The four undersigned editors wholeheartedly agree. And the theme of “transition and coherence” is exceptionally well suited as a vehicle by which to honour Annette. Annette’s focus during her long, unique and successful academic career has been marked by transitions – starting from German law and gradually expanding from there. After having obtained her law degree (Zweites Staatsexamen) in autumn 1975, she was accepted as a Doctoral Candidate of the Max Planck Institute for Foreign and International Patent, Copyright and Competition Law in Munich (now the Max Planck Institute for Innovation and Competition). There she successfully completed her dissertation in 1980 with the title “Beweisfürung und Beweislast in Verfahren wegen unlauteren Wettbewerbs” (“Burden of Proof and Giving of Evidence in Proceedings Against Unfair Competition”) under the supervision of Professor Gerhard Schricker, one of the two Institute directors. Immediately after her dissertation, Annette was engaged as a member of the Institute’s research staff. At that time, the Institute was organized in a number of regional departments and from the start Annette was involved in the Nordic department. By the late 1970s, she had already spent time in the Nordic countries, being introduced to Niklas when she was hosted by Professor Berndt Godenhielm as a guest researcher in Helsinki. Over the following months, Annette broadened the German focus of her dissertation across the Baltic as well as the Atlantic to a comparison with the US and Swedish regimes of burden of proof: Beweislast und Beweisführung im Wettbewerbsprozeß: rechtsvergleichende Untersuchung zum deutschen, amerikanischen und schwedischen Recht (1981). In making this first transition from a national German perspective to take account of Nordic and US law, Annette was already intrigued by parts of the legal world

Downloaded from https://www.cambridge.org/core. xiii , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.001

xiv

Preface

other than Germany. The comparative approach nourished her curious mind and offered challenges in the identification and analysis of similarities and differences. At the time, the European Union had not developed an approach to unfair competition law or to procedural issues. But Annette’s forward-looking research agenda fitted well with the growing body of EU regulation of intellectual property rights, which took place more or less at the same time as the European Union enlarged towards the Nordic countries. European integration and harmonization of intellectual property rights proceeded most intensely in the fields where Annette had focused her ongoing research: trade mark and design law. At that time, Marianne had been a scholar at the Max Planck Institute and defended her thesis on design law in Sweden. Thus, Friedrich-Karl Beier, the other director of the Max Planck Institute, formed a task force which included these two formidable women. This task force was able to exercise considerable influence on the European Commission as it planned to shape European design law. Annette soon became a leading scholar on EU intellectual property law, not only in her “own” fields of trade mark and design law, but more generally. In her capacity as Nordic contact person in the Max Planck Institute, she was often welcomed and engaged in especially Finland and Sweden as a visiting fellow or guest professor, lecturing and supervising masters and doctoral students. In addition, she hosted many young Nordic researchers at the Max Planck Institute for shorter visits. Her close contacts and intense cooperation with Stockholm and Helsinki were later recognized by well-deserved Honorary Doctorate Degrees in both countries (Stockholm 2012 and Helsinki 2014). Annette’s Nordic activities did not hinder her from remaining active at the Max Planck Institute where her close collaboration with both Ansgar (then at Max Planck) and Graeme started in the 1990s. Annette also was the crucial contact link between the IPR University Center in Helsinki and the University of Bayreuth in Germany when these universities established a common graduate school “Intellectual Property and the Public Domain” in 2006–2015, which led to a close collaboration between Nordic and German researchers. Although one might think that it is more than enough work for one person to focus on German, Nordic and EU intellectual property law, Annette made a further transition to the global level in the 2000s. She pursued important research in public international intellectual property law and private international intellectual property law. Her focus on TRIPS resulted in the 2011 book Intellectual Property Rights in the Fair World Trade System (Kur and Levin, eds.). On matters of private international law, Annette not only participated as an Adviser to the American Law Institute’s project on principles governing conflicts issues in transnational disputes, but also headed up the counterpart CLIP Principles Project in Europe. The overarching feature of Annette Kur as an intellectual property researcher is her search for consistency and coherence within intellectual property. She has a German instinct to take systemic aspects into account even when dealing with Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.001

Preface

xv

concrete legal issues. But her approach is anything but dogmatic. Indeed, Annette could even be described as a Nordic legal realist in her open-minded approach to new solutions taking into account the law in action in real-world contexts. As a result, her views are often more reflective and analytical, and less absolute, than those of other intellectual property scholars. For example, her approach to some of the CJEU’s most controversial decisions, such as Baby-Dry in trade mark law or Svensson in copyright law, has been more relaxed than that of many colleagues. On the other hand, Annette does not withhold criticism when she comes across inconsistent or superfluous reasoning. Annette’s many interesting transitions and impressive activities (not least in ATRIP, ALAI and ECTA) in combination with her positive and unpretentious behavior have garnered her numerous friends among colleagues all over the globe. In this volume, we could only include a small number of potential authors. We thank all those who willingly contributed with excellent essays within a strict timetable, and we apologize to all those who would have wanted to participate but whom we could not include within the allotted pages. Niklas Bruun Graeme B. Dinwoodie Marianne Levin Ansgar Ohly

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.001

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.001

Greetings to Annette Kur from the Second Floor Dieter Stauder1

Finding the right words to salute Annette Kur is not easy, even though we worked in close proximity in the same Institute for a long time. We learned a lot about and from each other, and our professional lives had much in common. Trust developed. This is why we Max Planck people maintain our companionship and friendly familiarity. But a word of greeting in a festschrift should also contain an appropriate degree of scholarly distance. In the former buildings of the Max Planck Institute, Annette’s office was on the second floor of an old villa in Siebertstraße 3, a room which in earlier times was inhabited by the servants of the baronial family who had built the house. The second floor – not to forget the other floors, or the friends in the cellar – was one of the creative hotspots of the Institute, and a nucleus of friendships which were often to last for a long time. Over the years it accommodated an illustrious group of researchers and scholars, of which Rudolf Kraßer was the old master, and Kurt Haertel, one of the founding fathers of the European patent system, the celebrity. Annette became an important member of this team. All of us in Siebertstraße and, later, in the buildings elsewhere, became known as the ‘Max Planck Mafia’, now a fond expression for us old ones. Annette’s arrival in Siebertstraße marked the beginning of some gradual changes in the Institute, that could soon be discerned and that have had an enduring effect. Annette was the first female academic to head a department. She took over the Department for the Nordic Countries from Hans Peter Kunz-Hallstein, who originates from the same city as she does. Both are familiar with the languages, the culture and the law of the Nordic countries. We found this succession quite remarkable and decided to ‘wait and see’. The North was not so far away. And it made perfect sense that a woman should be in charge of countries which have always been progressive with respect to the position of women in society. 1

Dr. iur., ret. Professor, Université Robert Schuman, ret. Director, Section Internationale du CEIPI, Strasbourg. Translation: Ansgar Ohly.

Downloaded from https://www.cambridge.org/core. 1 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.002

2

Dieter Stauder

Something was changing in the Institute; things were slightly different from what they used to be. The unmistakable male predominance among the Institute members had become less sacrosanct. Gradually, the spirit of the second floor also percolated to the management floor. At first slightly smiled at, then taken seriously, the energetic lady came to our centre of attention. The change was brought about by the ombudsman, or, rather, the ombudswoman Annette. With the ombudsman, a new figure entered the Institute. He, or rather she, protected consumers from misleading practices and deception, engaged in empirical research, looked into new types of legal action and stressed the public good to which our area of law is committed. The protection of consumers and of the economically and socially weaker was a heartfelt concern for Annette, both in professional and private life. We felt that. Annette personified the apparent severity of the North with its commitment, fairness and humanity, as we noticed before long. But it also soon became clear to us that she was open to the South, to conservative Bavaria and its inhabitants – among whom some notable specimens adorned our Institute – and that she shared their zest for life. Annette celebrated heartily with us, including at Ringberg. The scholars, speakers and guests whom Annette invited to our old villa and who became members of our ‘family’ are as unforgettable to us as the Institute is to them. At Ringberg and elsewhere, Annette has made significant contributions to the future of European trade mark, design and unfair competition law. The design law proposal, written with Marianne Levin, Friedrich-Karl Beier and Kurt Haertel, became the blueprint of EU design law. Annette’s suggestions and arguments have often been the crucial ones. Her determination and her clear opinion matter and have effect. Annette has temperament. When she said ‘no’ she hit the table with her fist – at least that is how it seemed to me. Her ‘trade mark’ is her way of forming her own opinion, defending it and contradicting inconclusive arguments, although never in a doctrinal or patronising way. We learned this from her, and her attitude spread. In her hands, the ‘capitalist’ instruments of intellectual property lose their rigidity. With elegance and empathy, Annette has prepared the ground for her convictions of social justice. Individuals and enterprises find protection under the umbrella of the social purposes of the law. When we look at the impressive series of titles and honours that Annette has been awarded, at her almost endless list of publications and presentations, at her manifold contributions to and impacts on the world of intellectual property law, her companions of old times are full of admiration and respect. We honour our spirited and courageous Annette and her youthful verve from the bottom of our hearts.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.002

Annette Kur: Toward Understanding Stacey Dogan1

Any fool can know. The point is to understand. Albert Einstein

I remember the first time that I met Annette Kur, about a decade ago, at the annual Trademark Scholars’ Roundtable that Graeme Dinwoodie and Mark Janis had launched a few years earlier. As we settled into the conference room, I noticed an elegant newcomer among the familiar faces from the United States. Introductions revealed the newcomer as Annette Kur, a name that I recognized as a leading European intellectual property scholar. Beyond the name, I knew little about Annette or her work, because – like too many American intellectual property specialists – I had only the slightest familiarity with either the law or the legal literature from Europe. Over the next two days, I observed Annette engage in the conversation with a style I have come to recognize as classic Kur. Mostly, she listened, with intensity, to the intellectual sparring over how to define and address boundary problems in trademark law. Her spoken contributions were spare but incisive, and introduced me to a complicating and contrasting European lens for understanding trademark and unfair competition law. I remember struggling with the indeterminacy of a law that turned on undefined notions of ‘unfairness’, especially given the variability of that concept from Germany to France to Benelux. But I was struck by the depth and breadth of Annette’s knowledge of the various permutations, as well as of the historical, political, and philosophical forces that had brought them about. That knowledge, of course, comes in no small part from her position at the epicenter of European intellectual property debates over the past few decades. Perhaps the most memorable qualities that I recall from that first encounter with Annette were her 1

Associate Dean for Academic Affairs, Professor and Law Alumni Scholar, Boston University School of Law.

Downloaded from https://www.cambridge.org/core. 3 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.003

4

Stacey Dogan

open-mindedness and apparent lack of an ideological agenda. She appeared to be grappling honestly with the kaleidoscope of normative frames that we discussed; while she hailed from the European tradition, she took seriously the competitionrelated concerns that animated my more economically oriented approach. In the years since that first meeting, I have engaged with Annette at countless conferences and have read a smattering of her many articles. Each of these encounters reinforces my impression of a keen thinker with a deep command of law and theory, combined with the curiosity and philosophical flexibility that allow her to change her mind. Annette has approached her scholarly journey in the best possible way: as a quest for understanding. Indeed, one might argue that her willingness to reconsider is part of what has equipped Annette to play a central role in the quest for greater coherence in a time of transition in Europe. The remainder of this introduction describes two faces of Annette Kur: as a tutor on European IP laws and sensibilities, and as a role model for a scholarly life of relevance and meaning.

1 the teacher Before describing my experience of Annette the teacher, it may help to picture the student body I have in mind: scholars such as myself, steeped in the US tradition of intellectual property law and theory. At least nominally, US intellectual property law – including trademark – hews to a utilitarian model, with a heavy emphasis on economics and debate over the social welfare justifications for various legal rules. Of course, the literature is replete with challenges to that model, from both normative and descriptive perspectives.2 But even skeptics concede that, as of the early twentyfirst century, the language of economics, consumer welfare, and markets has played a central role in judicial and scholarly writings on intellectual property.3 In copyright and patent law, this translates into an emphasis on incentives. In trademark law, it leads to a heavy focus on marketplace competition and concepts like

2

3

For examples in trademark law, see, e.g., Deven Desai, Bounded by Brands: An Information Network Approach to Trademarks, 47 U.C. Davis L. Rev. 821 (2014); Chad J. Doellinger, A New Theory of Trademarks, 111 Penn St. L. Rev. 823 (2007) (contending that ‘the economic approach’ to trademarks ‘has gradually eroded the true normative – and moral – foundation of trademark law’); Mark P. McKenna, The Normative Foundations of Trademark Law, 82 Notre Dame L. Rev. 1839 (2007); Mark P. McKenna, A Consumer Decision-Making Theory of Trademark Law, 98 Va. L. Rev. 67 (2012); Jeremy Sheff, Veblen Brands, 96 Minn. L. Rev. 769 (2012); cf. Ariel Katz, Beyond Search Costs: The Linguistic and Trust Function of Trademarks (2010) B.Y.U.L. Rev. 1555 (proposing a modified version of the economic model that distinguishes between the ‘linguistic function’ and the ‘trust function’ of trademarks). See Barton Beebe, The Semiotic Analysis of Trademark Law, 51 UCLA L. Rev. 621, 623–624 (2004) (concluding that the search-costs theory ‘has long offered a totalizing and, for many, quite definitive theory of American trademark law . . . The influence of this analysis is now nearly total’).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.003

Annette Kur: Toward Understanding

5

consumer search costs.4 My own scholarship has embraced this economic lens, emphasizing competition as the principal goal of trademark law, and the legal protection of marks as a tool for achieving that goal.5 Because the utilitarian approach focuses on outcomes rather than deontological notions of right or wrong, scholars and jurists in the utilitarian tradition tend to frame their analysis by reference to the overall societal impact of legal rules.6 We ask, for example, whether extending trademark protection to new subject matter would leave consumers better off (by protecting against confusion and possible mistaken purchasing decisions) or worse off (by limiting price and quality competition in product markets).7 The law fiercely protects the right of comparative advertisers – even when their advertising may ‘free ride’ on trademark holders’ reputations – because it enables more informed and competitive markets.8 More generally, debates over proposed changes to intellectual property laws center on whether they will improve societal outcomes, rather than favoring a particular interest group or sector.9 Or at least that’s the theory. In reality, developments in US trademark law can lead to much head-scratching for those who measure the law against its stated utilitarian goals. Critics have pointed out that doctrines such as post-sale confusion,10 initial interest confusion,11 merchandising rights,12 and dilution13 have little justification in a competition-centered trademark regime. Yet courts (and sometimes legislatures) embrace these doctrines, often dressing them up in the language of confusion.14 Countless scholars have pointed out the incoherence of some of these 4

5

6

7 8

9

10 11 12 13 14

See Qualitex Co. v. Jacobson Products Co., 514 U.S. 159, 163–164 (noting that trademark law ‘reduce[s] the customer’s cost of shopping and making purchasing decisions’ and ‘helps assure a producer that it (and not an imitating competitor) will reap the financial, reputation-related awards associated with a desirable product’) (internal citations omitted); Ty Inc. v. Perryman, 306 F.3d 509, 510 (7th Cir. 2002) (‘The fundamental purpose of a trademark is to reduce consumer search costs by providing a concise and unequivocal identifier of the particular source of particular goods.’). E.g., Stacey L. Dogan and Mark A. Lemley, The Merchandising Right: Fragile Theory or Fait Accompli?, 54 Emory L.J. 461 (2005); Stacey L. Dogan and Mark A. Lemley, Trademarks and Consumer Search Costs on the Internet, 41 Hous. L. Rev. 777 (2004). See generally Graeme B. Dinwoodie, The Death of Ontology: A Teleological Approach to Trademark Law, 84 Iowa L. Rev. 611 (1999). Id. See, e.g., Stacey L. Dogan and Mark A. Lemley, A Search-Costs Theory of Limiting Doctrines in Trademark Law, 97 Trademark Rep. 1223 (2007). See generally Stacey Dogan, Greeted with a Shrug: the Impact of the Community Design System on United States Law, in Annette Kur, Marianne Levin and Jens Schovsbo (eds.), The EU Design Approach: A Global Appraisal (Edward Elgar, 2018), 207, 209. See Sheff, Veblen Brands, note 2 above. See Dogan and Lemley, Trademarks and Consumer Search Costs, note 5 above. See Dogan and Lemley, The Merchandising Right, note 5 above. 15 U.S.C. § 1125(c). E.g., Boston Prof’l Hockey Ass’n, Inc. v. Dallas Cap & Emblem Mfg., Inc., 510 F.2d 1004 (5th Cir. 1975) (concluding that, despite the district court’s opinion that the defendant’s use of

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.003

6

Stacey Dogan

explanations. As Greg Lastowka observed, ‘Courts sometimes perform their multifactored analysis in such bizarre ways that some sort of doctrinal mischief is clearly afoot’.15 Courts, in other words, are rationalizing their decisions by reference to traditional legal standards, but their outcomes seem in tension with the normative goals that those standards were designed to promote. At least in some of these cases, the outcome cannot be justified based on utilitarian principles. There’s something else going on. It is against this backdrop that I encountered Annette Kur for the first time. I had spent the last several years (often with Mark Lemley) grappling with the ‘doctrinal mischief’ committed by courts in internet-related trademark cases.16 Like many other US scholars, we framed our critique within the utilitarian tradition, pointing out that the courts’ decisions (which effectively expanded trademark holders’ rights) would reduce competition, impoverish the information marketplace, and leave consumers worse off. Defendants made these same claims in litigation, but often to no avail.17 In our view, these courts had seized on expansive notions of confusion to allow claims of infringement designed to prevent reputational free riding rather than to promote functioning competitive markets.18 This anti-free-riding impulse struck us (and others) as inconsistent with the competition-based foundation of US trademark law. Before meeting Annette, I was generally aware that Europe approached trademark law through a different conceptual lens. Over the course of that first Roundtable with Annette, however, I began to understand the history and political economy of European trademark law, and how those forces shaped trademark doctrine in the region. I came to appreciate, for example, the role of the French perfume industry in inserting a curious provision into the Comparative Advertising Directive that prohibits certain forms of truthful comparative advertising that would clearly benefit consumers.19 I grew familiar with the explicit anti-free-riding protection contemplated by the European Trademark Directives.20 And I came to realize

15 16

17

18 19

20

trademarks did not confuse consumers ‘[w]here the consuming public had the certain knowledge that the source and origin of the trademark symbol was’ the mark-holder, the use constituted infringement). Cf. Greg Lastowka, Trademark’s Daemons, 48 Hous. L. Rev. 779, 790 (2011). Dogan and Lemley, The Merchandising Right, note 5 above; Dogan and Lemley, A SearchCosts Theory, note 7 above. E.g., Brief for Defendant-Appellee, Rescuecom Corp. v. Google Inc. (2d Cir., filed Feb. 27, 2007, 2007 WL 6475452); cf. Rescuecom Corp. v. Google Inc., 562 F.3d 123 (2d Cir. 2009). Dogan and Lemley, A Search-Costs Theory, note 8 above. Directive 2006/114/EC of the European Parliament and of the Council of 12 December 2006 concerning misleading and comparative advertising, Art. 4(g); see generally Annette Kur, Lionel Bently, and Ansgar Ohly, Sweet Smells and a Sour Taste – the ECJ’s L’Oreal Decision, Max Planck Institute for Intellectual Property, Competition & Tax Law Research Paper Series No. 09-12, available at http://ssrn.com/abstract=1492032. See Directive 2008/95/EC of the European Parliament and of the Council of 22 October 2008 to approximate the laws of the Member States relating to trademarks, Art. 5(2) (the Directive in

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.003

Annette Kur: Toward Understanding

7

that the feature of US trademark law that I had long criticized as incompatible with the economic view of trademarks – in particular, the persistent intuition that free riding on trademark holders’ reputation was somehow wrong – was baked into the trademark law of Europe. While I initially reacted with dismay at what I viewed as Europe’s failure to appreciate the importance of competition in unfair competition law, I have realized, under Annette’s tutelage, that European notions of ‘unfairness’ resemble the antiexploitation impulse that lurks beneath the surface of so many US trademark opinions. Free riding, in other words, troubles courts in both the USA and Europe, but Europe offers a vehicle for addressing it; one might argue that European law is simply more honest, and avoids the doctrinal distortions so common in US law. Upon closer examination, however, the impression of convergence toward a global anti-free-riding norm appears premature. For one thing, even though courts seem to overlook it at times, US trademark law’s emphasis on competition gives advocates a tool to challenge courts’ expansionist impulses, and often results in corrections over time. In the internet context, for example, courts have written more nuanced opinions as they have appreciated the benefits to consumers from various unauthorized uses of marks.21 And the Supreme Court has used competition as a reason to limit trademark rights in other ways, including to add rigor to the secondary meaning and functionality standards in the product design context.22 In other words, the absence of a right against ‘taking unfair advantage’ matters in US law; while courts may try to circumvent it at times, the normative framework at least partially constrains them. Even so, the fact that courts in both systems have visceral reactions to free riding says something about basic human intuitions regarding notions of reputation, identity, and exploitation. I have learned from Annette to think critically about my assumption that markets and economics offer the only rational model for trademark

21

22

place when I met Annette); see also Commission Regulation 2017/1001 of 16 June 2017 on the European Union Trade Mark, Art. 9(2)(c); Directive 2015/2436 of the European Parliament and of the Council of 16 December 2015 to approximate the laws of the Member States relating to trademarks, OJ L 336, Art. 10(2)(c). E.g., Network Automation, Inc. v. Advanced Systems Concepts, Inc., 638 F.3d 1137 (9th Cir. 2011). European case law reflects a similar trend. Compare, e.g., Google France SARL v. Louis Vuitton Malletier SA, 2010 ECJ (C-237-08) (holding that keyword-based ads linking to competing products may interfere with the source-identifying function of a trademark and therefore constitute infringement), with Interflora Inc. v. Marks & Spencer plc, 2011 ECJ (C-323-09) } 45 (‘[O]ne of the blessings of the internet is precisely that it greatly enhances consumers’ possibilities to make enlightened choices between goods and services.’). Indeed, rather than convergence toward a global anti-free-riding norm, perhaps we are witnessing a convergence toward an approach that balances trademark-holder concerns with those of consumers and competitors. E.g., Wal-Mart Stores, Inc. v. Samara Bros., Inc., 529 U.S. 205 (2000); TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 (2001).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.003

8

Stacey Dogan

protection. The fact that culturally rich Europe judges fairness by a different metric has at least opened me to the possibility of a different world view.

2 the role model While I found Annette’s substantive instruction in European trademark law eyeopening and educational, her more lasting impact came through the model that she provided of a scholarly life of relevance, incisiveness, and integrity. It must have been heady for Annette, as a young scholar, to play such a central role in the crafting of the European Design System. She approached the task with vigor, proposing with her co-authors a framework for protecting the ‘design approach’.23 At the time, she and her colleagues were convinced that ‘the needs of the design industry’ mandated a sui generis design right. They forged ahead with a proposal that the European Parliament accepted almost whole cloth.24 For scholars (as opposed to lobbyists) to have such a dramatic impact on the law is almost unprecedented. Having achieved such spectacular success, one might have expected Annette either to rest on her laurels or to spend the ensuing years defending and justifying the Design System’s virtues. Annette, however, viewed her part in the European Design System as anything but complete. After the system went into effect, she shifted into an evaluative role, observing the law’s rollout with a dispassionate eye. She hosted workshops at the Max Planck Institute to assess the law’s economic impact and to gauge its interpretation by the courts. She wrote articles exploring some of the unanticipated effects of the design law, such as its use to inhibit competition in the market for spare automobile parts.25 And she grappled with competition concerns associated with allowing designers to tack on multiple forms of legal rights over the same design feature.26 Most recently, in a book reflecting on the design experience from the Max Planck study through the present, Annette frankly acknowledged that the original proposal came at a time of ‘strongly affirmative’ views on intellectual property protection, and gave inadequate attention to limitations and exceptions.27 Annette’s openness, non-defensiveness, and intellectual flexibility have allowed 23

24 25

26

27

See Max Planck Institute for Foreign and International Patent, Copyright and Competition Law, Towards a European Design Law (1991). For a history of the project and its impact on European law, see Annette Kur and Marianne Levin, The Design Approach Revisited: Background and Meaning, in Annette Kur, Marianne Levin and Jens Schovsbo (eds.), The EU Design Approach: A Global Appraisal (Edward Elgar, 2018), 1–27. See Kur and Levin, id. See, e.g., Josef Drexl, Reto Hilty and Annette Kur, Design Protection for Spare Parts and the Commission’s Proposal for a Repairs Clause, 36 IIC 448–457 (2005). E.g., Annette Kur, What to Protect, and How? Unfair Competition, Intellectual Property, or Protection Sui Generis, in Na Ri Lee et al. (eds.), Intellectual Property, Unfair Competition and Publicity (Edward Elgar European Intellectual Property Institutes Network Series, 2014), 11–32. Id. at 25.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.003

Annette Kur: Toward Understanding

9

her to play an ongoing, constructive role in shaping the design system – and intellectual property laws more generally – at a time of transition in Europe. Her combination of vision, vigor, humility, and insight make Annette a model of a scholar who seeks not only attention, but relevance over time. In an industry that often rewards the most provocative and polarizing ideas, Annette opts for the pragmatic and incremental. In a community of people who like to hear themselves speak, Annette listens. As she wrote in 2014, in discussing the tension between unfair competition and other forms of intellectual property: For the academic community, it follows that long-standing efforts by many to shed more light on this complicated and complex area should be continued and, where possible, deepened. Moreover, we should definitely become more ‘European’ in the way in which we try to arrive at a deeper understanding, and a more transparent and consistent treatment, of the many issues posed [by the tensions that her work explored.]

May we all become more ‘European’ in the sense that Annette describes it – in a way that seeks deeper understanding, transparency, and consistency across different areas of law. May we all, in other words, be a little more like Annette Kur.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.003

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.003

par t i

Transition

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

a

Forms and Institutions

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

b

International Commitments and Constraints

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

c

New Agents and the Challenge of New Technologies

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

1 Transitional Provisions in Intellectual Property Legislation Richard Arnold1

1.1 introduction Intellectual property rights endure for long periods. Under current European laws, patents last 20 years, registered designs up to 25 years, a copyright can easily last over 100 years and a trade mark may remain registered indefinitely. The law cannot remain static for such long periods, however. Changes in technology, business models and social attitudes, among other things, mean that intellectual property legislation must be periodically revised if it is to remain fit for purpose.2 Although the speed with which legislatures react to the need for change varies greatly, most countries do introduce new or amended laws from time to time. In addition, the desire for increased harmonisation, both internationally and regionally, is another factor that leads to legislative change. What happens when new legislation comes into force? If the legislation is creating an entirely new right, such as the Community trade mark (now EU trade mark) or Community registered design, then it is possible for the legislature to start with a clean sheet.3 More usually, however, there will be existing rights to which the legislation may apply. In these circumstances the legislature is faced with a series of choices. Should the new legislation only apply fully prospectively, that is to say, to rights that come into existence after the legislation comes into force? Or partially prospectively, that is to say, both to existing rights and to new rights, but in the case of existing rights only in respect of future acts of infringement? Or retrospectively (which may extinguish existing causes of action if the right is narrowed or turn lawful acts retrospectively into unlawful ones if the 1

2

3

Judge of the Court of Appeal of England and Wales. I am grateful to Lionel Bently, Estelle Derclaye, Dev Gangjee and the editors for comments on earlier drafts. See Richard Arnold, The Need for a New Copyright Act: A Case Study in Law Reform (Herchel Smith Intellectual Property Lecture 2014) [2015] QMJIP 110 (calling for a new UK Copyright Act to replace the Copyright, Designs and Patents Act 1988). Even then, care should be taken over commencement provisions, since new rights may adversely affect existing arrangements or plans.

Downloaded from https://www.cambridge.org/core. 15 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

16

Richard Arnold

right is broadened)? Should there be some kind of phasing-in period? Should there be protection for people who have started an activity and may now be turned into infringers, and if so what form should such protection take? How should existing transactions such as assignments and licences of existing rights be treated? It is the function of transitional provisions to address these questions.4 In practice, transitional provisions in intellectual property legislation are of considerable importance, but legislatures frequently give them insufficient attention and the subject has been little studied by scholars. This chapter represents a small attempt to redress that neglect, although space does not permit a full treatment of this important subject.

1.2 what happens when there are no transitional provisions? It is regrettably all too common for legislatures not to include transitional provisions in new legislation, or to leave gaps in the transitional provisions that are included. The effect of this is to create uncertainty as to how the new legislation is to be applied, which ultimately the courts must resolve as best they can.5 This is unsatisfactory for at least three reasons. First, it imposes a cost on users of the relevant system, who must take legal advice and may have to spend money on litigation. Second, it means that the courts are left to infer the legislative intention, which may not be straightforward. Third, it is more difficult for courts than legislatures to apply graduated solutions such as phasing-in periods.6 To illustrate the problems, I will give two examples of failure to include transitional provisions, the first from the UK and the second from the EU. Until 1 October 2014, section 213(1) and (2) of Part III of the UK Copyright, Designs and Patents Act 1988 provided as follows: (1) Design right is a property right which subsists in accordance with this Part in an original design. (2) In this Part “design” means the design of any aspect of the shape or configuration (whether internal or external) of the whole or part of an article. 4

5 6

See R v. Secretary of State for Social Security ex p. Britnell [1991] 1 WLR 198 at 202 (Lord Keith of Kinkel) citing G. C. Thornton, Legislative Drafting (Butterworths, 3rd ed., 1987) at 319: “The function of a transitional provision is to make special provision for the application of legislation to the circumstances which exist at the time when that legislation comes into force.” Another situation where transitional provisions are required in the intellectual property context is where a country becomes a qualifying one for the purposes of copyright or a related right, and rights spring up in works which were formerly in the public domain. See Bennion on Statutory Interpretation (LexisNexis, 7th ed., 2017) at 176–178. The House of Lords held in Re Spectrum Plus Ltd [2005] UKHL 41, [2005] 2 AC 680 that the courts of England and Wales had power prospectively to change the law by judicial decision, at least where overruling a previous decision; but the power has not yet been exercised.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

Transitional Provisions in IP Legislation

17

On 1 October 2014 the UK Intellectual Property Act 2014 came into force. Section 1(1) amended section 213(2) of the 1988 Act by deleting the words “any aspect of”. Section 1(3) amended section 213(4) of the 1988 Act, but section 1(4) provided that section 1(3) “applies only to designs created after the commencement of that subsection”. The same qualification was applied to other amendments by sections 2(3) and 3(6). It is clear from the absence of any such provision relating to section 1(1) that section 1(1) applies to designs created before 1 October 2014. Thus it is not fully prospective, and could be partially prospective. But is it retrospective, that is to say, does it apply to allegedly infringing acts committed prior to 1 October 2014? The 2014 Act is silent on this question, even by implication. In DKH Retail Ltd v. H. Young Operations Ltd7 it was common ground between the parties that section 1(1) had retrospective effect. Doubt was expressed about this in Whitby Specialist Vehicles Ltd v. Yorkshire Specialist Vehicles Ltd,8 and in Neptune (Europe) Ltd v. DeVOL Kitchens Ltd9 it was held that section 1(1) of the 2014 Act did not apply to allegedly infringing acts committed before 1 October 2014. This conclusion was based partly on the general principle of statutory interpretation that, unless the contrary intention appears, an enactment is presumed not to be intended to have retrospective effect,10 and partly on the legislative intention indicated by the Explanatory Notes accompanying the 2014 Act, which was to “limit” protection. Article 7(1)(e) of Council Regulation 40/94/EC of 20 December 1993 on the Community trade mark prohibited the registration of: signs which consist exclusively of: (i) the shape which results from the nature of the goods themselves; (ii) the shape of goods which is necessary to obtain a technical result; (iii) the shape which gives substantial value to the goods.

The same provision was contained in Article 7(1)(e) of Council Regulation 207/ 2009/EC of 26 February 2009. It was amended by Article 1(9) of European Parliament and Council Regulation 2015/2424/EU of 16 December 2015 so as to insert the words “or another characteristic” after the words “the shape”. Article 4 of Regulation 2015/2424 provided that it entered into force on 23 March 2016. It also provided that certain provisions in Article 1 applied from 1 October 2017, but these did not include Article 1(9). Accordingly, the amendment to Article 7(1)(e) applied from 23 March 2016. Clearly, the amended provision applies to EU trade marks registered on or after that date. But does it apply to EU trade marks registered before that date? And does it make a difference whether a declaration of invalidity which is sought in respect of such a trade mark was claimed before or after that date? (The amended Regulation has subsequently been replaced by European Parliament and Council 7 8 9 10

[2014] EWHC 4034 (IPEC), [2015] FSR 21 at [10]–[18]. [2014] EWHC 4242 (Pat), [2016] FSR 5 at [41]. [2017] EWHC 2172 (Pat), [2018] FSR 3 at [29]–[42]. See Bennion (note 5 above) at 181–191.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

18

Richard Arnold

Regulation 2017/1001/EU of 14 June 2017, but that does not affect these questions.) Since Regulation 2015/2424 was silent on these questions, the Svea Hovrätt (Swedish Patent and Market Court of Appeal) was forced to refer them to the Court of Justice of the European Union (CJEU) in Case C-21/18 Textilis Ltd v. Svenskt Tenn AB. The CJEU applied the principle that, in order to ensure observance of the principles of legal certainty and the protection of legitimate expectations, substantive rules of EU law must be interpreted as applying to situations existing before their entry into force only in so far as it clearly follows from their terms, objectives or general scheme that such effect must be given to them. In the case of Article 7(1)(e)(iii) as amended, there was nothing in the terms, purpose or scheme of the legislation to suggest that it was intended to apply to trade marks registered before 23 March 2016.11 The effect of this ruling is to apply different requirements of registrability to trade marks depending on whether they were registered before or after that date, despite the legislature having deliberately changed the law. The absence of transitional provisions in Regulation 2015/2424 to address points such as this may be contrasted with the detailed transitional provisions included in Schedule 3 to the Trade Marks Act 1994 to address the way in which the new law applied to existing registered trade marks. These provided, in particular, that existing marks were to be treated as if registered under the 1994 Act (and hence vulnerable to invalidity or revocation on grounds provided under that Act);12 but that it was not an infringement of an existing mark to continue after commencement of the 1994 Act any use which did not amount to infringement under the old law.13

1.3 continuity of the law In devising appropriate transitional provisions, a fundamental guiding principle that has often been applied by the UK Parliament is that of continuity of the law.14 The concept of continuity of the law is not easy to define with precision, but in broad terms it refers to the idea that the law should continue to apply in as seamless a manner as possible after a change in the law.15 A simple example is where legal proceedings are extant, but have not yet been determined, as at the date when the law changes. The principle of continuity of the law means that such proceedings should not fall into some legal abyss, but should continue in being. As this example shows, however, the principle of continuity of the law can be given effect to in more 11

12 13 14

15

[EU:C:2019:199] at [29]–[33]. See also Case T447/16 Pirelli Tyre SpA v. European Union Intellectual Property Office [EU:T:2018:709] at [14]–[28], where the General Court held that the amended Regulation did not apply to a case in which the EU trade mark was registered in 2002 and the application for a declaration of invalidity was filed in 2012. Trade Marks Act 1994, Sch. 3 para. 2(1). Trade Marks Act 1994, Sch. 3 para. 4(2)(a). This is specifically referred to in the heading to Sch. 1 paras. 3 and 4 to the Copyright, Designs and Patents Act 1988. See Jowett’s Dictionary of English Law (Sweet & Maxwell, 4th ed., 2015).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

Transitional Provisions in IP Legislation

19

than one way. The first way, and the way that is generally accepted as representing continuity of the law in the truest sense in this context, is for the old law to continue to apply to the subject matter of the existing proceedings even if it has otherwise been repealed or replaced. A second way is for the new law to apply to the subject matter of the existing proceedings. A third way is for the old law to apply to some aspects (e.g. damages for past infringements) and the new law to others (e.g. the availability of an injunction to restrain future infringements). As this example shows, there is a need for policy choices to be made as to when the new law applies.

1.4 legitimate expectations Another fundamental guiding principle that has often been applied by the UK Parliament in making such policy choices is that of legitimate expectations. The legislature recognises that both right holders and users of intellectual property rights have legitimate expectations about the way in which changes to the law affecting their rights should be implemented.16 These expectations often revolve around the question of retrospectivity. Right owners expect that they will not be deprived of existing rights, while users expect that existing uses will not be turned into infringements. Such expectations are not necessarily all-or-nothing ones, however. Thus right owners may be prepared to accept limitations on existing rights, at least in the future, particularly if such limitations are imposed, say, as part of the price for an extension of the term of their rights. Similarly, users may be prepared to accept certain activities being turned into infringements if they are given time to sell off existing stocks (in which they have acquired property rights) and to transfer their businesses to new areas of activity (where they no longer have the freedom to conduct a business involving public domain works).

1.5 fundamental rights It is well established that intellectual property rights are “possessions” within the meaning of Article 1 of the First Protocol to the European Convention on Human Rights.17 If a new law reduces the scope and/or duration of an intellectual property right, then right holders may argue that they have been deprived of their possessions.18

16

17 18

The EU legislature explicitly recognised the legitimate expectations of users in recital (25) of European Parliament and Council Directive 2006/116/EC of 12 December 2006 on the term of protection of copyright and certain related rights (codified version) (“the Term Directive”). See e.g. Anheuser-Busch Inc v. Portugal (2007) 45 EHRR 36 at [42]. Similarly, they may argue that their investments have been expropriated for the purposes of an investor-state dispute settlement claim under a bilateral investment treaty or free trade agreement: see Rochelle Dreyfuss and Susy Frankel, Reconceptualizing ISDS: When is IP an Investment and How Much Can States Regulate It? (2018) 21 Vanderbilt J of Ent & Tech Law 376.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

20

Richard Arnold

In Neptune v. DeVOL,19 the court considered that partially prospective application of section 1(1) of the 2014 Act was not contrary to Article 1 because the amendment merely restricted the future scope of design right and did not extinguish it. This seems a little simplistic. In general, however, it would appear that passing legislation which amends the law applicable to an intellectual property right will fall within the State’s competence to control the use of property in accordance with the general interest, provided that the change has an objective justification and is proportionate to that justification.20 The inclusion of appropriate transitional provisions which take account of the legitimate expectations of affected right holders may assist in demonstrating that the legislation is proportionate. If a new law extends the scope and/or duration of an intellectual property right, then there can be no question of any complaint by right holders under Article 1 of the First Protocol; but other parties affected by the change may argue that it amounts to an interference with another fundamental right, such as freedom of expression (protected by Article 10 of the ECHR and Article 11 of the Charter of Fundamental Rights of the European Union) or the freedom to conduct a business (protected by Article 16 of the Charter). In Sky plc v. SkyKick UK Ltd,21 the defendants argued that the amendments to Article 12(a) of Regulation 207/2009 by Article 1(13) of Regulation 2015/2424 to restrict the own name defence to trade mark infringement to natural persons was invalid since it was an unjustified or disproportionate interference with these fundamental rights. The argument was rejected in any event, but the case in favour of the amendments might have been strengthened by the inclusion of a suitable transitional provision.

1.6 what kinds of transitional provisions may be adopted? As can be seen from the discussion above, the essential question involved in selecting transitional provisions is the desired degree of prospective or retrospective application of the new law, but the choice is not restricted to two alternatives. On the contrary, it is often the case that a range of potential solutions is available. Illustrations of the choices available abound in the most detailed set of transitional provisions contained in any UK intellectual property statute, namely those contained in Schedule 1 to the Copyright, Designs and Patents Act 1988, which runs to

19 20

21

Note 9 above at [41]. See Lester, Pannick & Herberg, Human Rights Law and Practice (LexisNexis, 3rd ed., 2009) at 4.9.16–4.9.30. For discussion of these principles in the context of a proposal to reduce further the term of protection of certain unpublished copyright works, see Ronan Deazley and Frankie McCarthy, Perpetual Copyright and Constitutional Property [2018] IPQ 253 at 271–275. [2018] EWHC 155 (Ch), [2018] RPC 5.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

Transitional Provisions in IP Legislation

21

forty-six paragraphs and many sub-paragraphs.22 Space does not permit a full account of these provisions,23 but the following are worth highlighting. General. Paragraph 3 of Schedule 1 lays down a general rule that the provisions contained in the 1988 Act “apply in relation to things24 existing at commencement as they apply in relation to things coming into existence after commencement”, but this is subject to any express provision to the contrary. Thus the default position is that the new law applies to existing rights.25 Paragraph 4 contains provisions that substitute the 1988 Act for the Copyright Act 1956 in various contexts, for example references in enactments or documents which would otherwise be interpreted as referring to copyright under the 1956 Act are to be interpreted as referring to copyright under the 1988 Act. Again, this is subject to any express provision to the contrary. The effect of this is to reinforce the default rule under paragraph 3. Subsistence. Paragraph 5 of Schedule 1 provides that copyright subsists in an existing work after commencement only if copyright subsisted in it before commencement, except where an existing work qualifies for copyright protection by virtue of first publication or by virtue of the application of the 1988 Act to new countries under international arrangements. The effect of this is that, notwithstanding the general rule laid down by paragraph 3, the 1988 Act does not in itself create new copyrights in existing works. This is followed by a series of specific provisions excluding copyright in certain types of works created before 1 June 1957 (the commencement date of the 1956 Act). Authorship. Paragraph 10 of Schedule 1 provides that the question of who was the author of an existing work is to be determined in accordance with the law in force at the time the work was made. This is except for moral rights, where the 1988 Act is to be applied. The explanation for this exception is that the moral rights provisions in the 1988 Act are largely new. When read together with paragraph 3, the effect of this is to confer moral rights on existing works (but paragraph 22(1) provides that no act done before commencement is actionable on moral rights grounds). First ownership. Paragraph 11(1) of Schedule 1 provides that the question of who was the first owner of copyright of an existing work is to be determined in accordance with the law in force at the time the work was made. This is subject to a limited exception for certain works commissioned before, but made after, commencement. Duration. Paragraph 12 of Schedule 1 contains an elaborate set of provisions dealing with the duration of copyright in existing works. Sub-paragraph (2) provides 22

23

24 25

Sch. 3 to the Trade Marks Act 1994, which is probably the runner-up, contains a mere twentytwo paragraphs. Note that some of these provisions have been amended since they were originally enacted, but I am referring to the original versions. The result of the amendments is to make the law even more complex. Thus, even with transitional provisions, the cost to users of ascertaining the law can be considerable. In context, this expression must include works and rights. Novello & Co. Ltd v. Keith Prowse Music Publishing Co Ltd [2004] EWHC 766 (Ch), [2004] EMLR 16 at [8] (affd. [2004] EWCA Civ 1776, [2005] EMLR 21).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

22

Richard Arnold

that, in many cases, copyright continues to subsist until the date on which it would have expired under the 1956 Act (but no longer). There is then a series of subparagraphs that make provision for the duration of copyright in unpublished works (preventing such works from continuing to enjoy indefinite copyright protection, and thus introducing a term limit on such rights26). Sub-paragraph (6) provides that copyright in any other existing work endures for the period it would endure under the 1988 Act. This applies to typographical arrangements of published editions;27 it is not clear that it applies to anything else. Infringement. Paragraph 14(1) of Schedule 1 provides that the infringement provisions in the 1956 Act continue to apply to acts done before commencement, while the infringement provisions in the 1988 Act apply to acts done after commencement. Thus the infringement provisions of the 1988 Act have partially prospective effect in the sense discussed above. This is subject to a series of provisions which clarify the position in particular circumstances. Similarly, paragraph 31(1) provides that the remedies for infringement in section 17 of the 1956 Act (but not those in section 1828) apply to acts done before commencement, while the remedies in the 1988 Act apply to acts done after commencement. Again, there are additional clarificatory provisions.

1.7 interpretation of transitional provisions Although transitional provisions are better included than omitted for the reasons explained above, it will be appreciated that transitional provisions may give rise to their own problems of interpretation. These must be resolved using the same techniques as for any other kind of legislation.29

1.8 judicial review of transitional provisions It appears that, at least in some circumstances, transitional provisions are subject to judicial review under English law. Section 52 of the Copyright, Designs and Patents Act 1988 provided that, where an artistic work had been exploited by or with the consent of the copyright owner by making articles, using an industrial process, which were copies of the work and marketing them, the work could be copied by making articles without infringing copyright in the work after the end of the period of twenty-five years. Section 74 of the Enterprise and Regulatory Reform Act 2013 repealed section 52 of the 1988 Act with effect from a date to be appointed 26 27 28

29

See Deazley and McCarthy (note 20 above). Copinger and Skone James on Copyright (Sweet & Maxwell, 17th ed., 2016) at 6–89. Conversion damages. Thus conversion damages ceased to be recoverable when the 1988 Act came into force on 1 August 1989. See e.g. Novello v. Keith Prowse (note 25 above), a case on the interpretation of a transitional provision contained in Sch. 7 para. 28(3) to the Copyright Act 1956.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

Transitional Provisions in IP Legislation

23

under section 103(3), thus extending the effective period of protection for works of applied art from twenty-five years to life of the author plus seventy years.30 The repeal was undertaken because the Government believed31 (rightly or wrongly) that it was obliged to bring UK law into line with the EU Term Directive.32 The Government also believed that it was obliged to do so even for existing works, thus reviving copyright protection for works that had ceased to be protected many years before. The Government therefore decided to provide a transitional period before the repeal took effect so as to give those affected time to sell off stocks and make other adjustments to their businesses. Following a call for evidence in October 2013 and a public consultation from 15 September 2014 to 27 October 2014 in which a three-year transitional provision was proposed,33 the Government decided in February 2015 to bring section 74 of the 2013 Act into force on 6 April 2020.34 This was provided for by the Enterprise and Regulatory Reform Act 2013 (Commencement No. 8 and Saving Provisions) Order 2015 made on 10 March 2015.35 That gave persons who were adversely affected by the repeal five years in which to rearrange their affairs. The making of the 2015 Order was challenged, however, by three furniture manufacturers by means of an application for judicial review lodged on 18 May 2015. The applicants contended that the five-year transitional provision was contrary to EU law.36 Rather than contesting the application for judicial review, the Government repealed the 2015 Order, by the Enterprise and Regulatory Reform Act 2013 (Commencement No. 8 and Saving Provisions) (Revocation) Order 2015 made on 20 July 2015.37 The Government then ran a second public consultation on transitional arrangements from 28 October 2015 to 23 December 2015 in which a six-month transitional period was proposed with an additional six months for selling off existing stock.38 On 21 April 2016 the Government announced that it had decided, in the light of the responses to the consultation, to extend the transitional period to nine months. As a result, section 74 was brought into force with effect from 28 July 2016 by Article 3 of the Enterprise and Regulatory Reform Act 2013 (Commencement No. 10 and Saving Provisions) Order

30

31 32 33

34

35 36 37 38

For criticism of this change in the law see Lionel Bently, The Return of Industrial Copyright? [2012] EIPR 654. As a result of Case C-168/09 Flos SpA v. Semeraro Casa e Famiglia SpA [2011] ECR I-181. Note 16 above. Consultation on Transitional Provisions for the Repeal of Section 52 of the Copyright, Designs and Patents Act 1988 (Intellectual Property Office, 2014). Transitional Provisions for the Repeal of Section 52 of the Copyright, Designs and Patents Act 1988: Government Response to the Consultation and Announcement of Transitional Provisions (Intellectual Property Office, 2015). SI 2015/641. Relying on what the CJEU had said in Flos (note 31 above) at [55]–[63]. SI 2015/1558. Consultation on Transitional Arrangements for the Repeal of Section 52 of the Copyright, Designs and Patents Act 1988 (Intellectual Property Office, 2015).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

24

Richard Arnold

2016 made on 18 May 2016.39 In addition, Article 4 provided that sales of existing stocks prior to 28 January 2017 would not infringe, nor would the making or importation of copies pursuant to a contract entered into before 28 October 2015 (i.e. the date the second consultation opened). The effect of this change was to give persons who were adversely affected by the repeal of section 52 of the 1988 Act more than three years less in which to sell off stocks and make adjustments to their businesses than was originally proposed. There was no attempt to challenge the new provisions by those affected, however. This episode demonstrates two things. First, it demonstrates the need for careful thought as to the appropriate design of transitional provisions when making changes to the law of this kind. This is particularly important where it is proposed to extend or enlarge existing rights. Proposals need to be compatible with EU law (where applicable), evidence based and subject to public consultation. Second, it demonstrates that those affected by transitional provisions may be able to challenge them by way of judicial review. Although the ground of challenge in this case was compatibility with EU law, a challenge to transitional provisions introduced by secondary legislation would also appear to be open on rationality grounds if the evidence base for a proposal is lacking.40 If the change involves a restriction on existing rights, then a challenge based on Article 1 of the First Protocol and the Human Rights Act 1988 may be possible.

1.9 conclusion Transitional provisions are an important feature of intellectual property legislation which no responsible legislature should neglect. The first and most basic requirement is that it should be made clear whether the new law applies only to future rights or whether it also applies to existing rights. Where the new law is to be applied to existing rights, careful thought needs to be given by the legislature to the design of appropriate provisions, taking into account the principle of continuity of the law and the legitimate expectations of both right holders and users. The key question is the extent to which the new law is to have prospective or retrospective application. Any interference with existing rights must be objectively justified and proportionate. Proposals for transitional provisions should be evidence based and subject to public consultation. Those adversely affected by them should have the possibility of challenging them by way of judicial review if proper grounds exist.

39 40

SI 2016/593. Cf. R (on the application of British Academy of Songwriters, Composers and Authors) v. Secretary of State for Business, Innovation and Skills [2015] EWHC 1723 (Admin) and [2015] EWHC 2041 (Admin), [2015] Bus LR 1435.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.006

2 Judicial Creativity and Transitions in EU Intellectual Property Law Lionel Bently1

2.1 introduction As Sir Richard Arnold capably explains elsewhere in this volume,2 when legislative changes are made to intellectual property laws, provision needs to be made for mediating discontinuities between the old and new regime. In the United Kingdom, this is typically done by so-called “transitional provisions,” which specify the extent to which the old law is maintained in relation to existing rights, interests and relationships, or how far (and from when) the new law becomes applicable. In intellectual property statutes, these typically provide that rights in existence before the change continue to subsist, and that rightholders benefit from expansion of rights but may be subject to broadened or altered exceptions and limitations. In EU law, general principles operate to produce similar effects (without the specific intervention of the legislator),3 though these are sometimes made explicit in EU IP legislation.4

1

2 3

4

Herchel Smith Professor of Intellectual Property Law, University of Cambridge. My thanks go to Graeme Dinwoodie, Martin Husovec, Jonathan Griffiths and Matthias Leistner for their extremely insightful observations on the issues raised herein and to Annette Kur for her consistently inspiring scholarship. Chapter 1, “Transitional Provisions in Intellectual Property Legislation,” herein. Absent an indication otherwise, the Court of Justice of the European Union (CJEU) assumes that legislation does not apply to past acts, but does apply to continuing situations that arose under past regimes: Case 68/69, Bundesknappschaft v. Élisabeth Brock [1970] ECR 171, ECLI: EU:C:1970:74. Council Directive 91/250/EEC of 14 May 1991 on the legal protection of computer programs (now codified as Directive 2009/24/EC of the European Parliament and of the Council of 23 April 2009 on the legal protection of computer programs, but without the transitional provisions), Art. 9(2); Directive 96/9/EC of the European Parliament and of the Council of 11 March 1996 on the legal protection of databases, Art. 14 (hereafter, Database Directive, 96/9/ EC); Council Directive 93/98/EEC of 29 October 1993 harmonizing the term of protection of copyright and certain related rights, Art. 10 (replaced by Directive 2006/116/EC of the European

Downloaded from https://www.cambridge.org/core. 25 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

26

Lionel Bently

Things are different when it comes to judicial transitions. Under English orthodoxy, judges declare the law – whether it be the common law or the interpretation of statute.5 Similarly, as the CJEU has explained on numerous occasions, its judgments interpret, clarify and define the meaning and scope of existing rules, and thus specify how those rules ought already to have been understood.6 The necessary corollary is that such rulings apply from the date of entry into force of the measure that the court is interpreting, rather than (for example) the date of the ruling.7 As a consequence, and subject to national rules on limitation and res judicata,8 such rulings apply to legal relationships arising and established before the judgment.9 Their effect is immediate and retrospective. The problems raised by declaratory theory are significantly magnified when a tribunal actively seeks to develop the law, as the CJEU has frequently done with European law: the more unexpected the interpretation, the more disruptive will be this retrospective effect.10 Two examples, drawn from some of the Court’s most surprising rulings, will serve to illustrate the potential in relation to intellectual property and raise questions as to how the resulting implications for innocent parties can best be mitigated. Both examples come from the context of copyright law, partially harmonized by a series of directives, and relate to the question of whether there are general EU rules on subject matter or originality. In both examples, the matter had not been explicitly harmonized by any particular directive, and it might have been thought that there had been a decision not to harmonize the matter, precisely because three individual directives dealt with issues of subject matter and originality in relation to their specific fields. Indeed, as late as 2004, the Commission itself stated in a Staff Working Paper that the matters remained unharmonized.11

5 6

7

8

9

10

11

Parliament and of the Council of 12 December 2006 on the term of protection of copyright and certain related rights). In re Spectrum Plus Ltd (in liquidation) [2005] UKHL 41, [2005] 2 AC 680 (HL). Gerhard Bebr, Preliminary Rulings of the Court of Justice: Their Authority and Temporal Effect (1981) 18 CMLR 475, 489 ff; Joined Cases 66, 127 and 128/79, Salumi, ECLI:EU:C:1980:101, [9]. In the context of intellectual property law, this was explained in Case C-501/15P, EUIPO v. Cactus SA, ECLI:EU:C:2017:383, [AG54] (Advocate-General Wahl). In the Latin to which the CJEU frequently resorts, its rulings are said to be effective “ex tunc” rather than “ex nunc.” Case C-453/00, Kühne & Heitz NV v. Produktschap voor Pluimvee en Eieren, ECLI:EU: C:2004:17, [2004] ECR I-837, [24]. Case C-61/79, Amministrazione delle finanze dello Stato v. Dekavit Italiana, ECLI:EU: C:1980:100, [1980] ECR 1205, [16]; Case C-453/00, Kühne & Heitz NV v. Produktschap voor Pluimvee en Eieren, ECLI:EU:C:2004:17, [2004] ECR I-837, [25]–[27]. In a classic, if not universally approved, analysis Trevor Hartley categorized leading Court of Justice decisions into those “within the text,” “outside the text,” and “contrary to the text”: T. Hartley, The European Court, Judicial Objectivity and the Constitution of the European Union (1996) 112 LQR 95. For a rapid riposte, see A. Arnull, The European Court and Judicial Objectivity: A Reply to Professor Hartley (1996) 112 LQR 411. Commission Staff Working Paper on the Review of the EC Legal Framework in the Field of Copyright, SEC(2004) 995, 14 (“These special cases aside, Member states remain free to

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

27

Nevertheless, following the CJEU decision in Softwarová,12 it became likely that those countries that operate “closed list” systems of subject matter for copyright are acting incompatibly with EU law in so far as the effect is to prevent copyright arising in works which are original expressions (their author’s own intellectual creations) of a type that falls outside the closed list. The effect is, it seems, to require the recognition of rights in pre-existing intellectual creations. For example, even if it is not a sculpture under UK law,13 the effect could be that the denial of copyright protection to the helmet of “stormtroopers” from George Lucas’s Star Wars film is no longer permitted under EU law. If so, is there any protection for a person who in good faith and on the basis of the House of Lords’ decision in Lucasfilm had invested in a factory making replica helmets? Conversely, following the CJEU decision in Infopaq,14 it seems that UK copyright law should not protect any works that are a mere product of labour, skill and judgement but which do not reach the threshold of “intellectual creations” involving free creative choices.15 On its face, the ruling has the potential to require that existing rights be extinguished. For example, the creation of a simple logo might have generated protection under UK copyright law, and such rights may have been the foundation of various transactions. Would the creator of the logo be entitled to continue to enforce such a copyright? Does it matter when the logo was created? If it had been licensed, would the creator still be able to claim royalties? Or must we proceed as if the copyright had never existed? English courts have acknowledged the potentially disruptive effects of declaratory theory in situations where courts reach conclusions that would have surprised many operators in the field. In In re Spectrum Plus Ltd (in liquidation), when considering the possibility of prospective overruling, Lord Nicholls observed that from time to time court decisions on points of law represent a change in what until then the law in question was generally thought to be . . . A court ruling which changes the law from what it was previously thought to be operates retrospectively as well as prospectively . . . People generally conduct their affairs on the basis of what they understand the law to be. This "retrospective" effect of a change in the law of this nature can therefore have disruptive and seemingly unfair consequences.16

12

13 14

15 16

determine what level of originality a work must possess for granting it copyright protection . . .”). Case C-393/09, Bezpeˇcnostní softwarová asociace – Svaz softwarové ochrany v. Ministerstvo kultury, ECLI:EU:C:2010:816, [45]–[46]; Case C-310/17, Levola Hengelo v. Smilde Foods, ECLI:EU:C:2018:899, [45]. Lucasfilm v. Ainsworth [2011] UKSC 39. Case C-5/08, Infopaq International A/S v. Danske Dagblades Forening, ECLI:EU: C:2009:465, [45]. Ibid., [45]. [2005] UKHL 41, [6]–[8] (Lord Nicholls) (emphasis supplied). See further Ronald Sackville, Continuity and Judicial Creativity – Some Observations, 20 UNSWLJ 145, 159 (1997).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

28

Lionel Bently

Indeed, as the examples illustrate, new rights may spring up, and existing activities, previously legal, may become “illegal overnight.”17 Alternatively, old rights (hitherto believed to exist) may be extinguished, and the basis for existing investments and transactions thrown into jeopardy.18 What freedom does national authority have to mediate that effect? Are such mitigating measures a matter for the legislature, administrative agencies or the judiciary?19 In what follows, I want to explore three possible responses which could be used to mitigate the disruption that might otherwise flow from adhering strictly to declaratory theory: first, that the court treat some of its rulings as prospective only; second, where judicial changes are retrospective, that the legislature (or administration) implement transitional measures; and third, that when formulating remedies, courts take account of the “reliance interests” of those that acted on the basis of what turned out to be incorrect assumption about EU law.

2.2 prospective effect only In relation to general EU law, since the time of the Defrenne v. Sabena decision concerning equal pay (and direct effect of the Treaty),20 the CJEU has recognized that in exceptional circumstances it has power to limit the effects of its judgments, in particular so that the rules only apply to the litigants and prospectively.21 Such an exceptional course of action will be taken only if four conditions are met. First, there must have been “objective, significant uncertainty regarding the implications of Community provisions.” Second, that uncertainty must have given rise to the generation of rules which were considered to be valid. Third, a “large number” of legal relationships must have been entered into in good faith on the basis of rules. Fourth, as a result of the preceding circumstances, it must be shown that retrospectively deviating from those rules presents a “risk of serious

17

18

19

20 21

AG Bot in Case C-168/09, Flos SpA v. Semararo Casa e Famiglia SpA, ECLI:EU:C:2010:371, [AG78] (“those undertakings whose economic activities become illegal almost overnight”). See e.g. R. Mallinson, “Colour Combinations: Getting Back to WYSIWYG,” https://unitedkingdom.taylorwessing.com/documents/get/796/colour-combinations-getting-back-to-wysiwyg .pdf/show_on_screen, at 3 (“many, if not virtually all, of the colour combination EUTMs registered to date face potential cancellation. It poses a real threat to some very significant brand value within the EU”). For some examples of ways that the common law accommodated reliance interests, see M. L. Friedland, Prospective and Retrospective Judicial Lawmaking, 24 UTLJ 170, 183–186 (1974). Case C-43/75, Defrenne v. Sabena, EU:C:1976:56, [1976] ICR 547. Case C-24/88, Vincent Blaizot v. University of Liège and others, ECLI:EU:C:1988:43, [34]; Case C‑437/97, Evangelischer Krankenhausverein Wien v. Abgabenberufungskommission Wien and Wein & Co. HandelsgesmbH v. Oberösterreichische Landesregierung, EU:C:2000:110; Cases C‑191/14, C‑192/14, C‑295/14, C‑389/14 and C‑391/14 to C‑393/14, Borealis Polyolefine GmbH v. Bundesminister für Land- und Forstwirtschaft, Umwelt und Wasserwirtschaft, EU:C:2016:311, [101]–[111].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

29

economic repercussions.”22 It is for the Court itself to decide whether these conditions apply and it is usually said to be necessary that, if the Court considers limiting a judgment so that it has prospective effect, it states this expressly in its ruling.23 In intellectual property law, the most prominent prospective ruling was in Chartered Institute of Patent Attorneys/IP Translator.24 Here, in the context of an application to register “IP Translator” as a UK trade mark for services defined by reference to the class headings of Class 41,25 the United Kingdom’s trade mark appeal tribunal (“the Appointed Person”) referred questions to the CJEU on whether the Trade Mark Directive must be interpreted as harmonizing rules relating to possible specifications of goods and services and their interpretation.26 In particular, the referring court wanted guidance on whether the application necessarily covered – and thus needed to be rejected as descriptive of – “translation services,” which are goods listed in Class 41, though this is not apparent from a mere reading of the class headings. In relation to EUTMs,27 the EUIPO had issued a “communication” indicating that an application referencing all the headings in a particular class of the Nice classification regime would be treated as covering all goods or services listed in the class.28 The Court indicated that while there was nothing specific in the Directive, rules on the formulation of specifications were necessary to ensure that the law achieved 22

23

24 25

26

27

28

Cases C-367/93 to C-377/93, F. G. Roders BV and others v. Inspecteur der Invoerrechten en Accijnzen, EU:C:1995:261, [43]; Case C-209/03, R (Bidar) v. Ealing London Borough Council, ECLI:EU:C:2005:169, [69]. Case C-61/79, Dekavit Italiana, ECLI:EU:C:1980:100, [1980] ECR 1205, [18]; Case C-24/88, Vincent Blaizot v. University of Liège and others, ECLI:EU:C:1988:43, [28]; C‑437/97, Evangelischer Krankenhausverein Wien v. Abgabenberufungskommission Wien and Wein & Co. HandelsgesmbH v. Oberösterreichische Landesregierung, EU:C:2000:110, [57]. Case C-307/10, ECLI:EU:C:2012:361. Nice Agreement Concerning the International Classification of Goods and Services for the Purposes of the Registration of Marks, concluded at the Nice Diplomatic Conference on 15 June 1957, last revised in Geneva on 13 May 1977 and amended on 28 September 1979 (United Nations Treaty Series, Vol. 1154, No. I 18200, 89), hereafter “the Nice Agreement.” Directive 2008/95/EC of the European Parliament and of the Council of 22 October 2008 to approximate the laws of Member States relating to trade marks. This had replaced First Council Directive 89/104/EEC of 21 December 1988 to approximate the laws of the Member States relating to trade marks (OJ 1989 L 40, p. 1), and has subsequently itself been replaced by Directive (EU) 2015/2436 of the European Parliament and of the Council of 16 December 2015 to approximate the laws of the Member States relating to trade marks. At the time of the case these were still referred to as “Community Trade Marks” and governed by Council Regulation 207/2009 of 26 February 2009 on the Community Trade Mark (replacing Council Regulation (EC) No. 40/94 of 20 December 1993). The Community Trade Mark Implementing Regulation, Commission Regulation (EC) No. 2868/95, rule 2(2) stated that the list of goods and services must be worded in such a way as “to indicate clearly the nature of the goods and services.” Communication 4/03 of the President of the Office of 16 June 2003 concerning the use of class headings in lists of goods and services for Community trade mark applications and registrations, OJ OHIM 2003, p. 1647.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

30

Lionel Bently

its harmonizing intent.29 The Court therefore held that an applicant must define the goods or services with “clarity and precision,”30 and that the use of Nice class headings may or may not be a suitable reference point to achieve this effect: some class headings are clear and precise, others are not. The Court went on to consider principles of interpretation, noting the suggestion by the referring court that it should adopt a literal approach, “which seeks to give the terms used in those indications their natural and usual meaning.” Rather cryptically,31 the Court skirted the question of interpretation, focussing instead on the need for the applicant to be clear about what they intended to cover: an applicant employing class headings must specify whether they intend that to cover all the listed goods/services, and if only some, to identify which. An application that leaves the matter unclear “cannot be considered clear and precise.”32 Although the Court’s ruling purported to be an interpretation of the Trade Marks Directive,33 the judgment left unclear the question of how to deal with existing registrations in relation to goods or services that lacked the necessary clarity and precision.34 Were these necessarily invalid? If so, on which of the absolute grounds? If not, was the requirement one which only affected applicants – rather like the statutory rules in patent law that claims be “clear and concise” which are preconditions to a valid application but not grounds of revocation? Moreover, for EU trade marks, what was to be done in respect of applications which had referenced the socalled “class headings” at a time when the EUIPO treated such specifications as encompassing “all the goods/services in the class”?35 Although the Court in IP Translator nowhere addressed the issue of prospective effect,36 let alone the conditions that permitted a prospective-only ruling, the circumstances fell easily within the four Defrenne conditions. First, in the years preceding the reference that there was “objective, significant uncertainty” as to whether the rules applicable to the formulation and interpretation of specifications had been harmonized, and, if so, how.37 The legislative starting 29 30 31

32 33 34

35

36 37

Case C-307/10, ECLI:EU:C:2012:361, [41]. Ibid., [49]. Cf. Case C-577/14P, Brandconcern BV v. European Union Intellectual Property Office (EUIPO) and Scooters India, ECLI:EU:C:2016:571, [AG58] (“the Court politely declined that invitation”). Case C-307/10, ECLI:EU:C:2012:361, [60]–[61]. Case C-307/10, [1]. In Case C-371/18, Sky plc v. Skykick UK Ltd, EU:C:2020:45, the CJEU held that no such ground of invalidity exists in national or EU trade mark law. In effect, this removes the issue of retrospective effect at least as it bears on invalidity. The Max Planck Institute study conducted, inter alia, by Annette Kur, had anticipated the issue and suggested transitional provisions: [4.63]. This was noted by AG Wahl in Case C-501/15P, Cactus, [AG56]. The judgments sometimes add the phrase “to which the conduct of other Member States or the Commission may even have contributed,” and many emphasize the action of the Commission in encouraging Member State reliance on a particular view of EU law. See Case

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

31

point was that procedural matters had been left to national authorities.38 However, uncertainty as to how far this remained the position arose in part from the CJEU’s rulings under the Directive, for example on the criteria by which graphic representations were to be judged,39 as well as some aspects of the definition of goods and services.40 Nevertheless, in IP Translator the Court acknowledged that “there is no provision of Directive 2008/95 which directly governs the question of the identification of the goods and services concerned.”41 Second, that uncertainty had clearly given rise to the generation of rules by the EUIPO and national offices which were considered to be valid. It was well known that these diverged and, as already noted, the President of the EUIPO had sought to clarify the position at the EUIPO. The Belgian, Finnish, Greek, Italian and Maltese offices treated the use of all the class headings as a specification encompassing “the entire class”; Hungary, Lithuania and Romania treated such as a specification covering the goods/services mentioned in the headings and all other goods/services listed in the Nice classification as falling within the class; whereas the remaining Member States interpreted class headings literally.42 The OHIM president’s communication of 2003 acknowledged this diversity but insisted that it was “proper” for it to allow the use of headings, and that where an applicant did so, “[u]se of all the general indications listed in the class heading of a particular class constitutes a claim to all the goods or services falling within this particular class.” Moreover, the General Court had given rulings that referred to, and assumed the validity of, the diverse national practices.43

38

39 40

41

42

43

C-24/88, Vincent Blaizot, ECLI:EU:C:1988:43, [32]–[33] (letters between European Commission and Belgian state would have led the latter to believe its rules were compatible with Treaty, Art. 7). The recitals were explicit: “Whereas Member States also remain free to fix the provisions of procedure concerning the registration, the revocation and the invalidity of trade marks acquired by registration.” Case C-273/00, Ralf Sieckmann v. Deutsches Patent- und Markenamt, ECLI:EU:C:2002:748. Case C-363/99, Koninklijke KPN Nederland NV v. Benelux-Merkenbureau (Postkantoor), ECLI:EU:C:2004:36, [112] and [114]; Case C-418/02, Praktiker Bau- und Heimwerkermärkte AG, ECLI:EU:C:2005:425, [50]. The marginalization of procedural autonomy has continued with the recent decision in Case C-705/17, Patent- och registreringsverket v. Mats Hansson, ECLI:EU:C:2019:481 (ruling national practices of using disclaimers cannot affect substantive assessment of similarity of marks). Ibid., [38]. See also Brandconcern, ECLI:C:EU:2016:571, [AG60] (the obligation of clarity and precision was “deduced” from “all the articles and recitals of that legislative text”). See European Trade Mark and Design Network, Common Communication on the Implementation of ‘IP Translator’ v1.1 (20 November 2013) at https://www.tmdn.org/network/documents/ 89965/f7f4a715-6170-4be4-9d89-efcda0ad50b4. Case T-186/02, BMI Bertollo srl v. OHIM (Opposition by Diesel SpA), ECLI:EU:T:197, [42] (CFI, Ch) (treating an Italian registration which referred to class headings as covering all goods/ services in the class); Case T 575/08, 4care AG v. OHIM (Laboratorios Diafarm SA intervening), ECLI:EU:T:2010:359, [26] (GC) (opposition based on earlier Spanish registration that referenced the class headings of Class 5).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

32

Lionel Bently

Third, a “large number” of legal relationships must have been entered into in good faith on the basis of rules. Although it is impossible to know or even guess how many, decisions about registrability, validity, opposition, infringement and revocation would already have been given based on the assumed validity of different national practices. Article 13 of the Directive (and Article 38 of the Regulation) make clear decisions about matters such as distinctiveness and descriptiveness must be given in relation to each of the goods or services claimed.44 Fourth, retrospectively deviating from those rules presented a “risk of serious economic repercussions.”45 Applying the test of clarity and precision to existing registrations would have generated very considerable uncertainty, especially in those jurisdictions where it had been common to utilize class headings. Declaratory theory might have resulted in the invalidation of many thousands of registrations, and called into question decisions, including those of the General Court, that had assumed that a registration in relation to class headings covered all goods or services in the class.46 Subsequent decisions have confirmed the prospective operation of IP Translator. In Brandconcern, the CJEU interpreted IP Translator as only applying in relation to new applications,47 so that a registration dating from before the decision was to be interpreted on the basis of the understanding prevalent at the EUIPO at the time of the application. In Cactus SA,48 the CJEU took that approach one step further in a case concerning opposition based on a 2002 registration for the class headings in Class 35. The Court found that the registration covered all goods in that class (not just those listed), and that the earlier CJEU decision in Praktiker (to the effect that an applicant for a trade mark in respect of “retail services” must specify the goods or types of goods to which those services relate) also only had prospective effect, so as not to affect trade marks that had been granted before the judgment in the latter case, namely July 2005.49

44

45

46

47

48 49

Case C-363/99, Koninklijke KPN Nederland NV (Postkantoor) 2004:EU:C:2004:36, [73]; Case C-239/05, BVBA Management v. Benelux-Merkenbureau, ECLI:EUC:2007:99, [37]. Cases C-367/93 to C-377/93, F. G. Roders BV and others v. Inspecteur der Invoerrechten en Accijnzen, EU:C:1995:261, [43]; Case C-209/03, R (Bidar) v. Ealing London Borough Council, ECLI:EU:C:2005:169, [69]. The EUIPO subsequently received 23,000 declarations relating to registration that had used class headings: see Verena von Bomhard (Chapter 11) in Annette Kur and Martin Senftleben (eds.), European Trade Mark Law: A Commentary (Oxford University Press, 2015), 582, [11.41]. Case C-577/14P, ECLI:EU:C:2017:122. Advocate-General Sanchez-Bordona acknowledged that the IP Translator case did not express itself as a “prospective” ruling (as hitherto had been required under Defrenne): Case C-577/14P, ECLI:EU:C:2016:571 (AG). Case C-501/15P, EUIPO v. Cactus SA, ECLI:EU:C:2017:750. Case C-412/02, Praktiker Bau- und Heimwerkermärkte AG, ECLI:EU:C:2005:425, [50] (allowing registration for retail services, but requiring applicant to specify the goods or types of goods to which those services relate).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

33

The CJEU treated both Praktiker and IP Translator as concerned with applications and therefore prospective in operation.50 By interpreting the CJEU judgment in IP Translator as prospective,51 the CJEU left space for Member States and the EUIPO to introduce remedial measures (which would have been difficult had the Court held its judgment retrospective so as to render earlier applications void or ineffective). Indeed, the EU legislature inserted a new Article 28(8) into Regulation No. 207/2009, allowing the proprietors of trade marks registered before 22 June 2012 to clarify, by 24 September 2016, their intention on the date of filing in relation to the goods or services mentioned in the heading of the class concerned.52 Affording prospective effect only to a legal change (whether legislative or judicial) comes at significant costs – first, that the desired change does not apply to a cohort of cases; and second, in terms of the operational costs of a system where the legal position is contingent on some historic fact (such as whether an application or registration was made, or an intellectual artefact was created or published before a certain time). Given that many of the changes in EU law (both by the legislature and the courts) are driven by an urgent desire to transform multiple distinct national regimes into a single unit (and in turn to create an internal market), it is understandable that this option is rarely attractive. Leaving intact a cohort of cases that is subject to residual (national) rules obviously reduces the benefits of harmonization. Not only do the Defrenne conditions establish a high threshold before the Court can limit a ruling to be prospective only, but even if they were established, the CJEU might reasonably prefer to insist on the application of its ruling to all existing cases. IP Translator is a rare example of a ruling with only prospective effect, and it might be explained, in part, by the fact that the operational costs are less profound: it is easy for concerned parties to identify the date of application for any trade mark registration and it should be possible to obtain access to the relevant rules applicable at the time.

2.3 retrospective effect – acquired rights and legitimate expectations In circumstances where prospective application is not available or not regarded as desirable, it may be that transitional provisions can, or indeed must, be introduced.53 50

51

52

53

See, further, Cases C-155/8P to C-158/18P, Tulliallan Burlington Ltd v. EUIPO, ECLI:EU: C:2019:538 (AG Hogan) (registrations for retail services made before the date of the Praktiker decision were unaffected by it). Phil Johnson, So Precisely What Will You Use Your Trade Mark for? Bad Faith and Clarity in Trade Mark Specifications (2018) 49 IIC 940, 951. This process is not intended to alter the substance of the coverage, so decisions made by the Office in relation to pre-IP Translator specifications are applicable even once the specification is amended: Case C-653/17P, VM Vermögens-Management GmbH v. EUIPO, ECLI:EU: C:2019:406. Case 84/78, Ditta Angelo Tomadini Snc v. Amministrazione delle finanze dello Stato, ECLI: EU:C:1979:129, [20].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

34

Lionel Bently

When EU legislation fails to specify transitional arrangements, the Court has indicated that general principles of EU law are to be applied and these include the protection of acquired rights and legitimate expectations. This protection of acquired rights and legitimate expectations has been repeatedly described as a “fundamental principle” of Community or EU law.54 As such, these principles apply both to the Union institutions and to national authorities implementing EU law.55 Despite being a “fundamental principle,” considerable uncertainty surrounds the circumstances in which action must be taken to secure the protection of acquired rights and legitimate expectations, and even what these seemingly straightforward terms mean. In the few cases that consider transitional provisions,56 the Court has drawn on jurisprudence developed primarily in the EU administrative law context.57 There, issues of acquired rights have arisen frequently in cases concerning agriculture and food products (where regulations are subject to “constant adjustment”). Here the concepts are invoked typically as part of a claim that a particular act or decision is invalid and thus should be annulled, or that the tribunal should award compensation in favour of a person whose “acquired right” or “legitimate expectation” has been frustrated. As a result of the frequency of such changes, a high threshold is adopted. In contrast, a more generous characterization of legitimate expectations when considering judicial harmonization of property rights; in this context, the Advocate-Generals have referred to a wide “margin of discretion.”58 Under the EU administrative law case law, a right is considered to be acquired “when the event giving rise to it occurred before” the relevant change.59 It should be observed that this is distinct from the requirements for the existence of property or a possession protected under Article 1 of the First Protocol of the European Convention on Human Rights or Article 17 of the Charter. Interestingly, the notion of “acquired rights” closely parallels the distinction between acts which are “complete” 54

55 56

57

58

59

Term Directive, recital 27; Case C-104/97P, Atlanta v. European Community, ECLI:EU: C:1999:498, [1999] ECR I-6983, [52]; Case C-168/09, Flos SpA v. Semararo Casa e Famiglia SpA, ECLI:EU:C:2011:29, [50]. Case C-381/97, Belgocodex SA v. Belgian State, ECLI:EU:C:1998:589, [1998] ECR I-8153, [26]. In both Butterfly and Flos, which concerned transitional measures relating to implementation of the EU Copyright Term and Designs Directives, 93/98 and 98/71, respectively, the CJEU reference a number of decisions concerning legitimate expectations in annulment proceedings: Case C-60/98, Butterfly Music Srl v. Carosello Edizioni Musicali e Discografiche Srl (CEMED), ECLI:EU:C:333, [25], ECLI:EU:C:1999:166, [AG36] (AG Cosmas); Case C-168/ 09, Flos, ECLI:EU:C:2011:29, [53]. However, the concepts of legitimate expectation under A1P1 and under administrative law, were elided, if not treated as equivalent, by AG Kokott in Case C-477/14, Pillbox 38 (UK) Ltd v. Secretary of State for Health, ECLI:EU:C:2015:854, [AG196]–[AG197]. Case C-60/98, Butterfly Music, ECLI:EU:C:1999:166, [AG26] (AG Cosmas); Case C-168/09, Flos, ECLI:EU:C:2010:371, [AG68] (AG Bot). Case 443/07P, Isabel Clara Centeno Mediavilla and others v. Commission of the European Communities, ECLI:EU:C:2008:767, [62]–[63]; Case C‑496/08P, Pilar Angé Serrano v. European Parliament, ECLI:EU:C:2010:116, [84].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

35

or “definitive” and those which are “continuing,” used to determine ex nunc application of legislation.60 In a very different context, the Commission once gave as an example of an act which was concluded “the registration of a trade mark.”61 Although one might have thought this debateable, given that registration only confers a presumption of validity, a more recent decision has confirmed the Commission’s intuition: changes to Article 7(1)(e)(iii) of Regulation 207/2009 by Regulation 2015/2424 have been held “not applicable” to marks registered before 23 March 2016, the date of entry into force of the latter Regulation.62 There is no obvious reason why other registered rights – patents, designs or plant breeders’ rights – would be any different. It is not clear whether the same logic would apply to copyright or unregistered Community design rights, rights which arise on origination/first publication or when a design is “first made available to the public” in the EU.63 Even greater uncertainty surrounds the concept of “legitimate expectations.” Within the EU administrative law context,64 a leading treatise explains, “it is difficult to provide general guidance as to the conditions that have to be met to invoke the principle.”65 Nevertheless, EU administrative lawyers seem confident that a legitimate expectation might arise where a relevant authority has made a promise or representation.66 In such cases, a legitimate expectation depends on the representation having four characteristics: first, it must be by a competent authority; second, it must be specific; third, it must not be in conflict with an unambiguous provision of EU law; and fourth, it must generate an expectation on the part of a “prudent and alert” operator.67 What would be an “authorised and reliable source”68 which could give rise to relevant expectations? Perhaps most authoritative might be a legislative provision,

60

61

62 63 64

65

66 67

68

C‑321/97, Andersson and Wåkerås-Andersson, EU:C:1999:9, [AG58] (AG Cosmas) (when a legal state became “permanent and fixed”); Case C-60/98, Butterfly Music, ECLI:EU: C:1999:166, [AG25] n 15 (AG Cosmas) (Term Directive would only be retroactive if it applied to situations that were “permanently fixed,” that is, where a person had already “pressed, released and sold” a copy of a sound recording). Case 303/13P, European Commission v. Jørgen Andersen, ECLI:EU:C:2015:340, [AG19] (AG Wathelet). Case C-21/18, Textilis Ozgur Keskin v. Svenskt Tenn AB, ECLI:EU:C:2019:199. Berne Convention, Art. 5; Council Regulation 6/2002, Art. 11. The concept is, of course, a well-known (if not uncontroversial) one within many national administrative laws, but it also finds a place in areas such as competition law and international investment law: see e.g. Case C‑170/13, Huawei Technologies Co. v. ZTE Corp. and ZTE Deutschland GmbH, EU:C:2015:477, [53]–[54]. Herwig Hofmann, Gerard Rowe and Alexander Tu¨rk, Administrative Law and Policy of the European Union (Oxford University Press, 2011), 178–190. Ibid., 187–190. Case C‑545/11, Agrargenossenschaft Neuzelle eG v. Landrat des Landkreises Oder-Spree, EU: C:2013:169, [23]–[26] (setting out principles). Case T-101/15 and 102/15, Red Bull v. EUIPO, EU:T:2017:852, [125] (GC).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

36

Lionel Bently

such as a statement that a particular field is left to Member States.69 Slightly less authoritative might be so-called “joint statements” or “joint declarations” of the Council and Commission made during the legislative process.70 Just below this in the hierarchy would be an official statement by the European Commission published in the Official Journal.71 Least authoritative might be statements made by the Commission in evaluation reports and staff working papers, as well as press releases. An equivalent source, in relation to registered rights, might be a representation from the EUIPO or Community Plant Variety Office. In Brandconcern, the General Court held that administrative communications with regard to registrability can ground a “legitimate expectation.” Referring to the president’s 2003 Communication, the General Court found that “[a] legitimate expectation had been created by OHIM itself, since it had given assurances as regards the treatment of trade mark applications using the complete heading of a particular class.”72 The General Court went on to remark that such a legitimate expectation existed even though the application preceded the president’s Communication of 2003. The reasons for this included the fact that the communication was intended to clarify existing practice and that this practice had been approved by the General Court.73 Of course, as with Commission communications, so with those of the Office: whether a statement has the capacity to generate a “legitimate” expectation may depend on other factors.74 For example, the EUIPO’s “examination guidelines” explicitly state that they are “intended to help both our users and our staff in charge of the various procedures,” 69 70

71

72

73 74

Case C-168/09, Flos, ECLI:EU:C:2011:29 (on Article 17 of Directive 98/71). A number of such statements were entered in the minutes of the meeting of the Council of the European Union at which Regulation No. 40/94 was adopted, OJ OHIM, 1996(5). For example, Statement No. 10, relating to Article 15, recorded that “[t]he Council and the Commission consider that use which is genuine within the meaning of Article 15 in one country constitutes genuine use in the Community.” When the CJEU ruled this not to be the case, in Case C-149/11, Leno Merken BV v. Hagelkruis Beheer BV, EU:C:2012:816, however, no transitional measures were thought necessary. Annette Kur offered some thoughts as to why in “What Kind of Use Is This? Open Questions After ONEL/OMEL,” Max Planck Institute for Intellectual Property & Competition Law Research Paper No. 13–15, available at SSRN: https:// ssrn.com/abstract=2337463. Statement by the Commission concerning Article 2 of Directive 2004/48/EC of the European Parliament and of the Council on the enforcement of intellectual property rights, 2005/295/ EC, OJ L 93/37 (“The Commission considers that at least the following intellectual property rights are covered by the scope of the Directive . . .”). Case T-51/12, Scooters India Ltd v. Office for Harmonisation in the Internal Market, ECLI:EU: T:2014:844, [27]. Ibid., [30]–[31]. Although in Cactus SA, Advocate-General Wahl played down the significance of such communications, observing that the comments “intended to explain and clarify the EUIPO practice in examining Community trade mark applications” and “are not legally binding” should not be understood as casting doubt on the capacity of such statements to generate “legitimate expectations.” Rather, the Advocate-General and Court, at [40], were keen to explain why the 2012 Communication, issued in response to IP Translator, could not affect the existing rights of trade mark proprietors.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

37

are “the main point of reference . . . on our examination practices,” but “are not legislative texts and, therefore, . . . are not binding.”75 Even though we are concerned with aspects of EU law, a relevant representation can originate with a national authority. Legitimate expectations thus might be developed, for example, by national legislation purporting to implement EU law, so that even if it turns out to be incorrect, it may be appropriate for that national legislature to adopt transitional measures.76 A relevant government office, such as the United Kingdom’s IPO, might offer guidance on its understanding of the law. As we will see, the other conditions mean that a representation at national level is less likely to give rise to a “legitimate” expectation: for example, if there is an obvious conflict with EU law, or it is not reasonable for an operator to rely on a statement from a national authority. Indeed, within the jurisprudence on EU administrative law, one senses a marked reluctance to treat national representations as generating legitimate expectations, perhaps for fear that doing so would undermine central decision-making. The second requirement, that of a specific representation,77 was in issue in Red Bull GmbH v. EUIPO, where the CJEU rejected an argument that Red Bull’s twocolour trade mark application, submitted to the EUIPO on 15 January 2002, should be treated in accordance with an understanding of the law that existed prior to the Court’s decision in Heidelberger Bauchemie.78 Red Bull argued that the EUIPO’s guidelines suggested that a mark in the form submitted would be registrable and, indeed, the EUIPO had registered it on 25 July 2005. It was only following the Heidelberger case that an application for cancellation had been made. The General Court had rejected Red Bull’s argument,79 in part noting that “in so far as such assurances did not comply with the applicable provisions, they could not sustain a legitimate expectation.”80 However, the CJEU found that it was right to have done so, not because the EUIPO guidelines were illegal, but rather because Red Bull had

75

76

77

78

79 80

Case C-149/11, Leno Merken BV, EU:C:2012:816, [48] (“the OHIM Guidelines are not binding legal acts”). Case C-168/09, Flos, ECLI:EU:C:2011:29, [47] (“national legislation transposing Directive 98/ 71 . . . may give rise to a legitimate expectation, on the part of third party producers of those products, that they will be able to continue using those designs”). The specificity condition is frequently coupled with other associated characteristics, in particular that an assurance is “precise, unconditional and consistent.” For this reason, “Joint Statements” might not be the basis of a legitimate expectation, at least when they express themselves as “without prejudice to the interpretation of that text by the Court,” as did those accompanying the adoption of the EU Trade Marks Directive and Regulation. See further Case C-149/11, Leno Merken BV, EU:C:2012:816, [47]. Case C‑49/02, Heidelberger Bauchemie GmbH, EU:C:2004:384. On the various sets of guidelines, see e.g. R. Mallinson, “Colour Combinations: Getting Back to WYSIWYG,” https:// united-kingdom.taylorwessing.com/documents/get/796/colour-combinations-getting-back-towysiwyg.pdf/show_on_screen. Cases T-101/15 and T-102/15, Red Bull GmbH v. EUIPO, ECLI:EU:T:2017:852. Ibid., [134].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

38

Lionel Bently

failed to identify “any specific part of EUIPO guidelines to the effect that EUIPO informed the public that it considered it unnecessary for a mark consisting of a combination of colours to be systematically arranged by associating those colours in a predetermined and uniform way.” 81 The third requirement, that the representation is not in conflict with an unambiguous provision of EU law, is often coupled with a further qualification: a legitimation expectation cannot be generated when a “national authority responsible for applying EU law . . . acts in breach of that law.”82 It might seem from this additional element that no legitimate expectation can be generated by any national rule, or ruling, that subsequently transpires (under “declaratory theory”) to have been in breach of EU law. However, that does not appear to be correct. Rather, a national rule should be regarded as capable of generating legitimate expectations in circumstances where it is not obviously in conflict with “a precise provision of EU law.”83 Support for this conclusion can be found in the notorious case of Flos v. Semararo,84 in which the CJEU considered various legislative provisions in Italian law regulating the application of copyright law to designs.85 Article 17 of the Designs Directive had sought to regulate the relationship between the law of registered designs (which the Directive harmonized) and the law of copyright, which had by then only been dealt with in a few directives.86 Article 17 required Member States to offer copyright protection even where a corresponding design was registered, but left to Member States the conditions under which such copyright was to arise and the extent of protection afforded.87 In these circumstances, Italian implementation had initially introduced copyright protection for industrial designs which possess in themselves creative character and artistic value, but created a ten-year transition during which the newly recognized copyright was not enforceable against those who engaged before that date in the manufacture, supply or marketing of products based on designs that had entered into the

81 82

83

84 85 86

87

Case C-124/18P, Red Bull v. EUIPO, EU:C:2019:641, [82]. Case C-5/82, Hauptzottamt Krefeld v. Maizena [1982] ECR 4601, [22]; Joined Cases C-31/91 to C-44/91, Lageder SpA v. Amministrazione delle Finanze dello Stato, ECLI:EU:C:1993:132, [34] (“a practice of a Member State which does not conform to Community rules may never give rise to a legal situation protected by Community law”). Case C-630/11P, HGA srl v. Commission, EU:C:2013:387, [132]. In contrast, where a rule was unclear and an authority interpreted it in a particular manner, that created a legitimate expectation that prevented it from reversing its initial interpretation: Case C-159/82, Angélique Verli-Wallace v. Commission, ECLI:EU:C:1983:212, [1983] ECR I-2711, 2724 (AG Slynn). Case C-168/09, Flos, ECLI:EU:C:2011:29, [44]. Legislative Decree No. 95/2001 entered into force on 19 April 2001. Directive 98/71/EC of the European Parliament and of the Council of 13 October 1998 on the legal protection of designs. For background, see A. Kur, M. Levin and J. Schovsbo, The EU Design Approach: A Global Appraisal (Edward Elgar, 2018). Ibid., Art. 17.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

39

public domain.88 In 2007 these provisions were broadened further so as to exclude from their scope products made in accordance with designs that entered into the public domain before the date of entry into force of Legislative Decree No. 95/ 2001.89 In proceedings relating to the design of a lamp which before 2001 would not have been protected under Italian copyright law, the Tribunale di Milano asked the CJEU whether such an exclusion from protection of works that had fallen into the public domain could be justified under Article 17. The Court, interpreting the term “extent” in the light of the Term Directive, concluded it could not.90 Nevertheless, the Court did recognise that transitional provisions would be appropriate. “Legitimate expectations” existed because the Italian law implementing the Directive had represented that those works in the public domain before 2001 were to remain unprotected.91 Although the Court did not adopt precisely that same reasoning as Advocate-General Bot, the latter had argued that it was not merely the national legislation that gave rise to the legitimate expectations, but also the fact that the EU legislation lacked clarity: [AG75] In the light of these matters and of the fact that Directive 98/71 is silent on the application of copyright to designs that are in the public domain, I believe that the entry into force of the directive in the national legal order is likely to have created in Semeraro a legitimate expectation as to the maintenance in force of the national legislation existing at the time, namely that all rights in the Arco lamp were extinguished and that it was therefore entitled to continue to produce and market Fluida lamps. [AG76] Economic operators could reasonably have doubts, in my view, as to an interpretation of Article 17 of Directive 98/71 such as the one that I am proposing to the Court, namely that designs that entered the public domain before the entry into force of the national provisions transposing that directive are eligible for copyright protection.92

Of course, one difficult question will be precisely how a national authority (or a reliance party) is to gauge when such incompatibility is established. The easiest case is, of course, an enforcement proceeding or preliminary reference which calls for consideration of a national rule – as Flos itself was in relation to the Italian rules. In such cases, the operative provisions of the CJEU judgment will likely establish conclusively the invalidity in the approach taken in national law. However, other examples demonstrate that this question is not always so easily resolved. 88

89 90 91 92

Legislative Decree No. 164 of 12 April 2001 on the implementation of Directive 98/71 adding Article 25a to Legislative Decree No. 95/2001. Decree-Law No. 10 of 15 February 2007, Art. 4(4). Case C-168/09, Flos, ECLI:EU:C:2011:29, [44]. Ibid., [47]. ECLI:EU:C:2010:371, [AG75]–[AG76].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

40

Lionel Bently

The fourth condition concerns whether a prudent operator would have expected the subject of the representation to remain valid.93 In much of the traditional administrative case law, the key question is whether the claimant could reasonably expect the precise rules of quotas or levies to be maintained – and the answer is frequently that the regular adjustment of such rules renders such an expectation unreasonable.94 In relation to the CJEU’s interpretation of EU intellectual property law, those who follow the Court’s case law closely know that it is not realistic to assume that the Court will defer to any particular national interpretation,95 nor the joint statements of the Council and Commission adopted during the legislative process.96 However, it is less obvious that traders might not rely on such representations. Mere questions of textual interpretation are unlikely to generate legitimate expectations, so that transitional measures would not be inappropriate to protect someone who had relied on a particular textual interpretation that the CJEU ultimately rejected. Indeed, the European Court of Human Rights, interpreting Article 1 of the First Protocol, has stated that “no legitimate expectation can be said to arise where there is a dispute as to the correct interpretation and application of domestic law and the applicant’s submissions are subsequently rejected by the national courts.”97 This is primarily because the range of possible interpretations is foreseeable. As Lord Devlin observed long ago, “judge-made change in the law rarely comes out of a blue sky.”98 While the CJEU is occasionally more active than the UK courts, in most instances the outcome is one on a range that might have been contemplated. Here, it seems, the question of what a trader can reasonably expect takes one into the difficult territory of distinguishing between a ruling that is reasonably predictable and one that is surprising. In addition to satisfying these four requirements, a relevant operator must have some interest in the protection of the “legitimate expectation.” This normally requires some sort of reliance on the representation.99 Moreover, any such acts 93

94

95

96

97 98 99

Indeed, the CJEU has sometimes embraced a “prudent and circumspect” trader test: Case C17/03, Verenigung voor Energie Milieu en water (VEMW) v. Directeur van de Dienst uitvoering en toezicht energie, ECLI:EU:C:2005:362, [74]. Case C-78/77, Johann Lührs v. Hauptzollamt Hamburg-Jonas, ECLI:EU:C:1978:20; Case C‑545/11, Agrargenossenschaft Neuzelle, EU:C:2013:169, [30]–[33]. Joined Cases C-108/97 and C-109/97, Windsurfing Chiemsee Produktions- und Vertriebs v. Boots- und Segelzubehör Walter Huber and Franz Attenberger, ECLI:EU:C:1999:230, [35] (rejecting German doctrine of Freihaltebedürfnis). Case C-104/01, Libertel Groep BV v. Benelux-Merkenbureau, ECLI:EU:C:2003:244, [25]. Indeed, Annette Kur predicted the outcome of Case C-109/11, Leno Merken v. Hagelkruis Beheer, EU:C:2012:816: A. Kur, Unitary Rights in Fragmented Markets? Some Thoughts on the CTM System and its Interaction with National Law, in M.-Ch. Janssens and G. van Overwalle, Harmonisation of European IP Law: From European Rules to Belgian Law and Practice – Contributions in Honour of Frank Gotzen (Bruylant/Larcier, 2012), 119–136. Kopecký v. Slovakia [GC], Appn. No. 44912/98, ECHR 2004-IX [50]. P. Devlin, Judges and Lawmakers (1976) 39 MLR 1, 10. Case C-168/09, Flos, EU:C:2011:29, [58] (“the measure may be regarded as appropriate only if it is directed at a category of third parties entitled to rely on the principle of protection of

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

41

must have been performed “in good faith.”100 Only then would an EU body or national authority be required to protect a legitimate expectation. Where such an interest exists, or there are acquired rights, proportionate measures should be taken to minimize that prejudice.101 How should such acquired rights or legitimate expectations be given effect? To date, the main guidance comes from two cases in which intellectual property rights have been revived, either as a result of explicit legislation (the copyright Term Directive,102 discussed in the Butterfly case) or judicial ruling (the Flos decision). In those cases, the CJEU indicated that the measures must be tailored so as only to give protection to those with acquired rights or legitimate expectations. Moreover, the measure should only extend as far in duration as was needed to protect those interests. In Flos, this was conceptualized in terms of the time necessary to sell off stock,103 though in principle it might be necessary to accommodate other commitments (such as ongoing contracts). Most importantly, it must be limited, so as not to prevent the application of EU law as interpreted. What about where the effect of a judgment of the CJEU is to take rights away? What is the effect of the Football Dataco decision,104 for example, regarding protection granted by Dutch law to so-called “other writings” where such protection arose in writings created after the Database Directive came into operation but before the CJEU ruling, or even between the CJEU ruling and the repeal of that law?105 What is the effect of the Sieckmann ruling on national trade mark rights based on “graphic representations” that did not meet the requirements established in that case (assuming those to be interpretations of substantive law)?106 If such changes were

100 101 102 103 104 105

106

legitimate expectations – that is to say, persons who have already performed acts of exploitation”). Term Directive, recital 27. Flos, [56]. Directive 93/98, Art. 10(3). Flos, [59]. Case C-604/10, Football Dataco v. Yahoo! UK, EU:C:2012:115. On the removal, see P. B. Hugenholtz, “Goodbye, Geschriftenbescherming!,” at http:// copyrightblog.kluweriplaw.com/2013/03/06/goodbye-geschriftenbescherming/. Case C-273/00, Ralf Sieckmann v. Deutsches Patent- und Markenamt, ECLI:EU:C:2002:748. The Court in this case knew full well of the presence of a small number of existing registrations at the OHIM (as it then was, now EUIPO) and the UKIPO: see ECLI:EU:C:2001:594, [AG32]–[AG33]. The existing OHIM registration for the “smell of fresh cut grass” for tennis balls remained on the register until 2006, when the proprietor did not renew it. Of the two UK registrations then in place, No. 2001416 for “floral fragrance/smell reminiscent of roses” for vehicle tyres was assigned by Sumitomo Rubber Industries to Dunlop in 2002, renewed in 2004 and assigned to Goodyear in 2009, before being allowed to expire in 2014, whereas No. 2000234 for the “strong smell of bitter beer” applied to the flights of darts has been renewed twice, in 2004 and 2014, and is now not due for renewal until 2024. This activity may suggest that the operators have not considered the Sieckmann decision to have applied to existing registrations. In contrast, in Case T-305/04, Eden Sarl v. OHIM, ECLI:EU:T:380, the General Court applied the Sieckmann ruling when refusing an application dating from 26 March 1999, an application which had been transferred to the appellant in July 2004.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

42

Lionel Bently

legislative, they would likely be (or be seen to be) prospective. With declaratory theory, however, it is difficult to escape the conclusion that such rights should be deemed not to have ever existed. Indeed, in response to Sieckmann, Annette Kur has pronounced that earlier decisions “have been overruled.”107 When a national authority comes to implement such a conclusion, as noted, rights which were registered under national law before a CJEU decision might well be regarded as “acquired rights.” More significantly, perhaps, the European Union’s Charter specifies the limited circumstances in which a person may be deprived of the possessions and the use of property regulated.108 It famously adds that “intellectual property shall be protected” and thus, at the very least, shielded from illegitimate expropriation.109 As is well known, the CJEU has (albeit in very different contexts) invoked that provision of the Charter in case law concerning registered trade marks,110 copyright,111 and patents.112 In the parallel context of Article 1 of the First Protocol to the European Convention on Human Rights, the European Court of Human Rights treated a trade mark application as a “possession,”113 though – not unimportantly – held there had been no expropriation by Portugal because the Supreme Court had carefully examined the factual and legal basis of its claim when it held that the application was invalid. Consequently, if a legal change would establish a novel rule that would eliminate an existing national or EU intellectual property right (or even an application for such), then it seems that provision of some sort should be made for a transition. The measure, importantly, should not have the effect of leaving the contested right in place indefinitely: to do so would not bring about the required alteration. Moreover, it would leave a Member State vulnerable to an action for compensation from anyone who suffered damage from the continued maintenance of the right. However, perhaps surprisingly, there are a number of possible “mediating” positions between ownership of a valid right and its invalidity. To begin, in relation to registered rights, there may be the possibility of allowing amendment. For example, if, following the CJEU decision on the requirements for graphic representation in Libertel,114 it was decided that trade mark rights already 107 108 109 110

111

112 113 114

Kur and Senftleben, note 46 above, 98, [4.23]. Charter of Fundamental Rights of the European Union, Art. 17. Ibid., Art. 17(2). Case C‑580/13, Coty Germany, EU:C:2015:485; Case C-477/14, Pillbox 38 (UK) Ltd v. Secretary of State for Health, ECLI:EU:C:2015:854 (AG Kokott); Case T-106/12, Cytochroma Development International, ECLI:EU:T:2013:340, [45]–[46]. Case C-516/17, Spiegel Online Gmbh v. Volker Beck, ECLI:EU:C:2019:625, [56]; Case C‑314/12, UPC Telekabel Wien, EU:C:2014:192, [61]; Case C‑70/10, Scarlet Extended, EU:C:2011:771, [43]; Case C‑360/10, SABAM, EU:C:2012:85, [41]. Case C‑170/13, Huawei Technologies, EU:C:2015:477, [42], [58] (patents). Case T-106/12, Cytochroma Development International, ECLI:EU:T:2013:340, [45]–[46]. Case C-104/01, Libertel Groep BV v. Benelux-Merkenbureau, ECLI:EU:C:2003:244, discussed by Annette Kur in Kur and Senftleben (eds.), note 46 above, 100, [4.30].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

43

registered for colours but without utilizing an international colour code were no longer valid, the Member State might have introduced a process for amendment of such representations.115 Oddly, and in contrast with patent laws, mechanisms to amend trade mark representations (as opposed to specifications) are rarely present in modern trade mark laws: such a response, however, would potentially have been a sensible transitional measure to comply with the new law, but to minimize the harm to existing rightholders. In some cases, of course, no amendment will save a registered right. For example, it is recognized that there is no form of representation that will enable a smell to be registered in a clear and precise manner. Thus, following the CJEU ruling in Sieckmann, there was no way to give effect to the ruling apart from treating the registrations as invalid. In such a case, the act of expropriation would have been justified on the basis that it was in the public interest and in accordance with law. The Member State (and indeed the OHIM) would then be left with the task of compensating the former “right-owner.” In practice, such registrations could have had little real economic value, and, if the marks were in fact being used, some protection would still have remained possible under many national unfair competition or passing off regimes.116 One obvious measure of compensation would be to refund fees involved in acquiring the “rights,” or, where they have been assigned, the price the assignee has paid. There will be some situations where the compensation option will not be attractive, and transition might best be given effect to by national legislation phasing out the right. This might be seen as the converse of a “selling off stocks” transitional provision for those relying on the absence of a right which suddenly “springs up” as a response to a CJEU ruling. Inevitably, rules phasing out rights will have an arbitrary component to their formulation, but they are not unknown in intellectual property law. When, in 1989, the UK brought an end to its absurd “industrial copyright” protection (which gave the full, life-plus-fifty years, term to design drawings for functional articles that were incapable of registration under the registered design regime), it replaced the existing right with a five-year period of exclusivity, followed by five years in which the rights were subject to compulsory licences.117 115

116

117

Indeed, the Brussels Court of Appeal proposed such a course of action in Brussels Court of Appeal, 9th Chamber, 24 June 2004 (Veuve Cliquot), but the Benelux Trademarks Office would not permit amendment: INTA, Non-Traditional Trademarks in Europe – Shape and Colour Trademarks – Common Issues with Obtaining, Exploiting and Enforcing Rights (March 2005), at https://www.inta.org/Advocacy/Documents/INTAShapeColorEurope2005.pdf. Cf. L’Oreal v. Bellure [2006] EWHC 2355 (Ch), [164], Lewison J (smells could not be protected under the law of passing off ). Copyright, Designs and Patents Act 1988, Sch. 1, para. 19. Copyright was not removed in its entirety, but only in respect of the making of articles to a design document. For a discussion of some of the attractions of so-called “liability rules,” see Annette Kur and Jens Schovsbo, Expropriation or Fair Game for All? The Gradual Dismantling of the IP Exclusivity Paradigm” (1 November 2009), Max Planck Institute for Intellectual Property, Competition & Tax Law

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

44

Lionel Bently

2.4 remedial solutions Where no legislative action is taken to introduce transitional provisions in response to judicial innovations, courts may be able to find ways to protect reliance parties through other means, in particular by tailoring remedies. In cases where the issue falls within the remit of the Enforcement Directive (2004/48/EC), some assistance – and guidance – may be derived from Article 3, which establishes overarching principles applicable to measures, procedures and remedies for the enforcement of intellectual property rights and lays down five sets of criteria. According to Article 3, the measures etc. must: (i) be effective, proportionate and dissuasive; (ii) be fair and equitable; (iii) not be unnecessarily complicated or costly; (iv) not entail unreasonable time limits or unwarranted delays; and (v) be applied in such a manner as to avoid the creation of barriers to legitimate trade. Recital 17 further elaborates that the measures, procedures and remedies “should be determined in each case in such a manner as to take due account of the specific characteristics of that case, including the specific features of each intellectual property right and, where appropriate, the intentional or unintentional character of the infringement.” The “specific characteristic” of a case and the “specific features of each intellectual property right” could certainly include any previous uncertainty surrounding the recognition of the right itself and any action taken by one party in the belief that such a right did not exist. As a result, one can quite imagine a court staying injunctive relief, for example to allow the defendant to sell off stocks. Indeed, such delays have been deployed in cases unaffected by legislative or judicial change, but merely applying established law to particular facts.118 In addition, where a person relied on national law that turns out to be inconsistent with a subsequent CJEU interpretation, it should be possible (within EU law) for national courts to exempt them from an award for compensatory damages. This is because under Article 13 of the Enforcement Directive, damages need only be awardable where a defendant knew or had reason to believe they were infringing. It cannot be said that a person who acted on a reasonable interpretation of the law at the time knew or had reasonable grounds to have known they were infringing. Of course, that would not be so once the CJEU has established the conclusive interpretation of the legal rule: a Member State which permitted such a person to sell off stock (made at a time when they believed they were entitled to manufacture products that later turn out, under CJEU-developed interpretation, to be infringing) would nevertheless seem to be obliged to award compensatory damages (by way of royalties, for example).

118

Research Paper No. 09-14. Available at SSRN: https://ssrn.com/abstract=1508330 or http://dx .doi.org/10.2139/ssrn.1508330. E.g. Edwards Life Sciences LLC v. Boston Scientific Scimed Inc (No 3) [2018] EWHC 1256 (Pat), [2018] FSR 31.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

Judicial Creativity and Transitions in EU IP Law

45

2.5 conclusion In many respects, the question of how to handle the effects of judicial innovations in intellectual property is a new one, very much associated with the failure of the EU legislative process to harmonize in full intellectual property law, the slowness of Member States to adjust understandings of harmonized law, and the drive of the CJEU to eradicate difference. This chapter has sought to suggest some ways in which the associated problem of reconciling acquired rights and legitimate expectations with the declaratory nature of such judicial innovations might be achieved. The real solution to these problems of transition is, however, full legislative transition to an EU IP law. It looks as if the UK will not be party to such a solution and, no doubt, Brexit will create its own transitional issues.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.007

3 Before and after Designers Guild: Another Look at Appellate Deference in New Zealand’s Copyright Law Graeme W. Austin1

3.1 introduction In Designers Guild Ltd v. Russell Williams (Textiles) Ltd,2 the House of Lords held that the Court of Appeal had erred by assessing anew the decision of the trial judge as to the substantiality of the defendant’s copying of the plaintiff’s fabric and wallpaper designs. As Lord Hoffmann explained, because that decision involved “the application of a not altogether precise legal standard to a combination of features of varying importance,” analysis of substantial copying “falls within the class of case in which an appellate court should not reverse a judge’s decision unless [the judge] has erred in principle.”3 Their Lordships agreed that there had been no such error. The New Zealand Supreme Court endorsed this approach in 2006. Citing Designers Guild, it said: “[I]t is appropriate for appellate Courts to give the trial judge’s assessment the degree of latitude that conventionally applies to appellate review of a discretion.”4 This chapter considers the influence of Designers Guild in New Zealand’s copyright law, using as its springboard two decisions of the New Zealand Court of Appeal: one before and one shortly after the Supreme Court’s endorsement. Steelbro NZ Ltd v. Tidd Ross Todd Ltd5 was a 2007 decision involving the copyright in the design drawings for machines used to transport shipping containers. Mayceys Confectionery Ltd v. Ivor Beckmann6 was a 1994 decision involving an allegation 1

2

3 4 5 6

Chair of Private Law, Victoria University of Wellington; Professor, Melbourne University Law School. Thanks to Paul Sumpter for insightful comments on an earlier draft, and to Hugo Lawrence for excellent research assistance. Designers Guild Ltd v. Russell Williams (Textiles) Ltd (t/a Washington DC) [2000] 1 WLR 2416 (HL). Id., at 2423. Henkel KGaA v. Holdfast New Zealand Ltd [2007] 1 NZLR 577, 593 and fn 37. [2007] NZCA 486 (Steelbro). (1995) 33 NZIPR 543 (Mayceys).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms 46 of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

Appellate Deference in New Zealand’s Copyright Law

47

of infringement of copyright in moulds used for manufacturing crocodile-shaped chewy candy (or “jubes”) known as “Killer Crocs.” In the latter, the Court overturned the trial judge’s findings on copying. In the former, the Court declined to overturn the trial judge’s findings even where it acknowledged the potential for the enforcement of the plaintiff’s copyright to create a monopoly in a non-patented functional product. In New Zealand, the deference to be accorded to lower courts’ decisions is of particular importance, given that copyright is frequently used to protect the designs of functional products. Copying of the three-dimensional object is often held to infringe the underlying design drawings.7 Unlike some other mature copyright systems, New Zealand copyright law does not withhold copyright protection from functional objects,8 or from designs that have, or could have, been registered as such.9 Many New Zealand copyright cases involve relatively prosaic, but highly functional, objects, such as toilet pan connectors,10 blister packs for kiwifruit,11 asparagus sorter cups,12 and, perhaps most famously, the popular flying discs known as “frisbees.”13 Unlike the fabric designs that were the focus of Designers Guild, for products involved in the typical New Zealand case substitutes were not readily available. Accordingly, in New Zealand, decisions about the scope of a copyright can have more direct impact on competition for ordinary products. Without seeking to decry the importance of aesthetics, New Zealand’s copyright protection for functional objects heightens the stakes for consumers. Requiring one fabric designer not to imitate another’s particular juxtaposition of stripes and flowers is one thing; it is quite another if copyright in design drawings requires more extensive and expensive “design around,” thereby increasing the costs of market entry imposed on competitors. In the first kind of case, the available alternatives are limited only by designers’ imagination.14 In the second kind of case, design elements are more

7 8

9

10 11 12 13 14

Frank M Winstone (Merchants) Ltd v. Plix Products [1985] 1 NZLR 376, at 379. Copyright Act (1976) 17 USCS § 101: “the design of a useful article, as defined in this section, shall be considered a pictorial, graphic, or sculptural work only if, and only to the extent that, such design incorporates pictorial, graphic, or sculptural features that can be identified separately from, and are capable of existing independently of, the utilitarian aspects of the article.” See, e.g., Copyright Act 1968 (Cth), s. 75 (Australia). On the differences between New Zealand and Australian law, see Jeanswest Corporation (New Zealand) Ltd v. G-Star Raw [2015] NZCA 14. In the EU context, compare Cofemel-Sociedade de Vesturário SA v. G-Star Raw CV (Case C-683/17) EU:C:2019:721. P S Johnson & Associates Ltd v. Bucko Enterprises Ltd [1975] 1 NZLR 311. Frank M Winstone (Merchants) Ltd v. Plix Products [1985] 1 NZLR 376. Okara Technologies Ltd v. Geostel Vision Ltd [2016] NZHC 1188. Wham-O MFG Co v. Lincoln Industries Ltd [1981] 2 NZLR 628. In Designers Guild, it was noted that the defendant’s fabric design was closer to the claimant’s design than any alternative available in the market.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

48

Graeme W. Austin

likely to be limited by functional imperatives,15 including, in some instances, manufacturing constraints.16 In commentary on the scope of protection for non-patented functional aspects of products, an idea that is persistently reinforced is coherence as between different modes of intellectual property protection. For example, a 2002 report of the WIPO Standing Committee on the Law of Trademarks, Industrial Designs and Geographical Indications reported that “[i]t appears to be a generally accepted principle that, in respect of inventions, patent protection preempts any other form of protection under intellectual property.”17 It explained that copyright does not cover product inventions “because an invention will normally not be regarded as an artistic expression, a reflection of the ‘author’s personality.’”18 The report also explained that protection under designs law may also be precluded for configurations “entirely or solely dictated by technical functions.”19 Aesthetic choices are constrained by a designer’s imagination. Solutions to functional problems offer fewer options. Putatively “factual” decisions on issues such as “copying” of a “substantial part” do not just affect the positions of the parties inter se: their economic impact can be much wider. Typically, these concerns find expression as limitations on the protection of functional aspects of products. They are quite strongly reinforced, for example, in the European Union designs system, with the prohibition in the EU Designs Regulation against the subsistence of design rights in features of appearance of a product which are solely dictated by the product’s “technical function.”20 The Court of Justice of the European Union recently reinforced this prohibition with its holding in the Doceram reference that the prohibition against protection applies even when the function could be achieved through a different design.21 The appropriateness of New Zealand’s copyright protections for functional objects can be left for another day. For now, and while the current substantive rules 15 16 17

18 19

20 21

Burge v. Swarbrick [2007] HCA 17, at [51]. See, e.g., Beckmann v. Mayceys Confectionery Ltd (1995) 33 NZIPR 543 (CA) (Mayceys). World Intellectual Property Organization, Standing Committee on the Law of Trademarks, Industrial Designs and Geographical Indications, Industrial Designs and their Relation with Works of Applied Art and Three-Dimensional Marks (2002) SCT/9/6, para. 39. Id., at [41]. Id., at [42]. Accord Case C- 833/18 SI and Brompton Bicycle Ltd v Chedech/Get2Get, (Fifth Chamber) ECLI:EU:C:2020:461. Regulation 6/2002/EC, Art. 8. Doceram GmbH v. CeramTec GmbH, Case C‑395/16 (2018). In contrast, the Australian designs legislation provides that a protectable visual feature “may” serve a “functional purpose.” Designs Act 2003 (Cth), s. 5. The language followed a recommendation of the Australian Law Reform Commission: ALRC Report 74, 31 August 1995 aimed at avoiding litigation over whether a design feature was dictated by function. At the same time, the Australian Copyright Act 1968 (Cth) withholds copyright protection from many works that have been, or could have been, registered designs. The exception for works of artistic craftsmanship has been narrowly construed by Australia’s highest court, with the result that only those artistic works whose design was “unconstrained by functional considerations” are likely to receive copyright protection. Burge v. Swarbrick, above note 15, at [83].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

Appellate Deference in New Zealand’s Copyright Law

49

are in place, highlighting this distinction between New Zealand’s law and that in other jurisdictions might provoke some reconsideration of the appropriate level of deference to be accorded in New Zealand copyright appeals.

3.2 designers guild and deference Steelbro v. Tidd Ross Todd is a striking illustration of the implications of adopting the Designers Guild approach to deference.22 There, the Court of Appeal upheld the trial judge’s holding that the defendant had infringed copyright in the drawings of a side-loading trailer used to transport shipping containers. By copying the plaintiff’s trailer, the defendant was held to have indirectly copied the underlying design drawings. The plaintiff had designed a trailer that allowed two 20-foot shipping containers to be independently loaded and transported. Prior to the plaintiff’s innovation, trailers were able to transport 40-foot containers or two 20-foot containers that were locked together. Independent transportation of two 20-foot containers had been impossible. The Court of Appeal was certainly aware of the policy implications of its decision, including those flowing from allowing copyright to protect a monopoly in a functional product. It rehearsed in some detail the arguments for the defendant that had highlighted these concerns. The defendant had argued that the lower court’s decision that copyright had been infringed “strikes at fair competition and stifles technological progress”:23 Particularly in the area of industrial design, a certain degree of copying and comparison is inevitable and desirable in a competitive market. Innovation based on prior design should not be unduly constrained.

Counsel for the defendant had also cited Sir Hugh Laddie’s extrajudicial comment that: [t]he whole of human development is derivative. We stand on the shoulders of the scientists, artists and craftsmen who precede us. We borrow and develop what they have done; not necessarily as parasites, but simply as the next generation. It is at the heart of what we know as progress . . . [B]orrowing and developing has always been acceptable.24

These ideas bolstered other arguments trained on the level of originality in designs whose dimensions were largely dictated by function. The plaintiff, in turn, advanced competing arguments, noting, for example, that New Zealand’s approach to industrial design protection is different from the approaches of its major trading partners, and that this must have been deliberate.25 22 23 24 25

Steelbro, above note 5. Id., at [98]. Id., at [98]. Id., at [99].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

50

Graeme W. Austin

While the Court of Appeal acknowledged the policy conundrum, it concluded that resolution of the case “depends . . . on the factual findings.”26 Citing Designers Guild, it deferred to the trial court’s conclusion on the substantiality of the defendant’s copying, characterizing this issue as a “question of fact,”27 that is, whether the judge below was “correct in finding that [the defendant] had illegitimately crossed the line into copying not only the concept of [the crane] but also the expression of that concept.”28 The Court concluded unanimously that the trial judge had not erred in finding that a substantial part of the design drawings had been copied. It is instructive to contrast Steelbro with a decision of the New Zealand Court of Appeal that predated the endorsement of Designers Guild by the New Zealand Supreme Court. While Mayceys v. Beckmann did not involve the kinds of functional products involved in the cases mentioned above, it illustrates the prior willingness of the New Zealand Court of Appeal to revisit lower courts’ findings on copying and substantial similarity. The parties were competitors in the confectionery market. They had settled earlier proceedings involving the defendant’s marketing of confectionery which had directly copied the plaintiff’s crocodile-shaped product. Undeterred, the defendant again sought to compete in the crocodile jube market, this time alleging that it had commissioned a new designer who had not seen the plaintiff’s product. Because the factual evidence on derivation was equivocal, the Court of Appeal focused principally on whether the defendant’s new crocodile jube was substantially similar to the plaintiff’s artistic work: “the plaster model or sculpture of a crocodile which is the master from which moulds were made for use in producing its ‘Killer Crocs.’”29 Overturning the findings of the first instance judge, the Court held not only that the two products were not substantially similar, but also that “there was no real similarity as would suggest other than each has been devised independently.”30 Mayceys is suggestive of the kinds of opportunities that might be missed as a result of the more deferential attitude required by Designers Guild.31 As to the assessment of copying by the first instance judge, the Court of Appeal observed that it had “given careful consideration” to the reasons expressed by the trial judge for his finding that there was significant objective similarity between the plaintiff’s and defendant’s products. At the same time, it suggested that the trial judge’s comparison “perhaps was made without giving sufficient consideration to the fact that both articles are representations of common animals.” In passing, the Court observed 26 27 28 29 30 31

Id., at [101]. Id., at [106]. Id., at [103]. Mayceys, above note 6, at [544]. Id., at [548]. The distinction between the two cases cannot be explained on the basis that in Steelbro there had been a finding of derivation: if appellate courts are encouraged to be reticent about revisiting holdings on substantial similarity, a mixed question of law and fact, it would logically seem to follow that they should be even more so with derivation, a pure question of fact.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

Appellate Deference in New Zealand’s Copyright Law

51

that, while the features of crocodiles might be distinctive, this was “not because of any creativity on the part of the copyright owner.”32 This discussion was in the context of an analysis that emphasized the “purpose for which resemblance is assessed” in a copyright infringement action, and a view that whether similarities are sufficient to be regarded as an unlawful reproduction should be assessed in the light of “the originality in distinctiveness of the copyright work.”33 That was immediately followed by a passage hinting at a concern with the risks of monopoly effects that would accompany a finding of infringement: There can be little doubt that the original designer of the crocodile had a flair for the distinctive. But that was not [the plaintiff’s designer]. He simply set out to produce a model representing as realistically as possible the well known reptile. His skill and effort were in producing the particular representation in the form of a plaster model. That gives him no monopoly in models of crocodiles but only a right not to have his particular model copied. Anyone else is free to make a model of a crocodile and to use it to make moulds for jube production so long as they do not make directly or indirectly a copy of [the plaintiff’s] model or a substantial part of it.34

In the Court’s view, it was therefore necessary to examine whether the similarities were such that would go beyond what would be expected of two people working to the same end. Of course, real crocodiles do not miraculously transmogrify into plaster models depicting crocodiles, suitable for use as moulds for gelatinous confectionery. The plaintiff’s designer was therefore an author, notwithstanding his considerable debt to the quirks of reptilian evolution. The difficulty in a case such as this is that the features in which copyright is claimed will often be needed by subsequent entrants to the market for the same kind of product. Many such features will inevitably be dictated by the design problem of depicting a crocodile, as well as more specific design problems associated with designing a shape that enables release of the product from the mould. Authorial flair in the depiction of reptilian appendages – tails and claws, etc. – can be constrained by design exigencies, such as the ease with which the products can be removed from the moulds during the manufacturing process. As a result, and as the Court appeared to recognize, albeit without engaging the point directly, decisions as to the substantiality of copying set the height of barriers to competition. Decisions made by a primary judge have significant potential to increase costs of design around, which, in turn, may be passed on to consumers. At the same time, a 32 33

34

Mayceys, above note 6, at [548]. Id., at [547]. Here, the Court of Appeal anticipated by well over a decade the similar analysis by the High Court of Australia in IceTV Pty Limited v. Nine Network Australia Pty Limited (2009) 239 CLR 458. Mayceys, above note 6, at [546].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

52

Graeme W. Austin

finding that a substantial part had not been taken might have significant implications for prices in consumer markets. All of this is more acute in the New Zealand context because of the conspicuous absence of legislative exclusions from copyright protections of functional designs or functional aspects of products.35 The subsistence and enforcement of copyrights is very likely to raise the costs associated with getting competing productions to market. The stakes are sufficiently high that a full appellate review could be helpful. The deference required by Designers Guild does not necessarily always preclude a New Zealand appellate court from departing from a lower court’s assessment. As Lord Hoffmann’s statement (quoted at the beginning of this chapter) indicates, deference is not required where a lower court “has erred in principle,” a point that is further explored below. As the approach of the Court of Appeal in Steelbro reflects, however, questions of copying and the substantiality of the part taken are not regarded as points of principle, even if, as is adverted to by the Court in Mayceys, how these statutory tests are applied can affect the level of barriers to market entry. While the Court was quite aware of the possible implications of its decision for competition in the container crane market, drawing on Designers Guild allowed it to sidestep these issues by characterizing them as matters of fact: “on the facts, Steelbro’s conduct falls on the impermissible side of that line.”36 It did so notwithstanding Designers Guild’s remote factual context. Thus, given the current state of New Zealand copyright jurisprudence, any scope offered by the “erred in principle” exception to the no deference rule could be quite limited.

3.3 deference in context The UK Supreme Court reaffirmed Designers Guild in Lucasfilm v. Ainsworth,37 a 2012 decision that held that helmets worn by the “stormtroopers” in the Star Wars movies were not “sculptures” for the purposes of the Copyright, Designs and Patents Act 1988 (UK).38 Noting that the Court of Appeal had been correct to invoke Designers Guild, the Court said:39 During the 17 days of the trial [the trial judge] heard evidence about the helmet and the other artefacts from numerous different witnesses. Long and thorough as his judgment is, he may not have recorded every nuance that contributed to his conclusion. He did not err in law or reach an obviously untenable conclusion, and the Court of Appeal was right to uphold his decision on this point. 35

36 37 38 39

The Copyright Act 1994 (NZ) limits the term of copyright protection for industrially applied articles, but does not withhold protection from them. See, id., s. 175. Steelbro, above note 5, at [100]. [2012] 1 AC 208, at [48]. Id., at [48]. Id., at [47].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

Appellate Deference in New Zealand’s Copyright Law

53

In their Lordships’ view, the point had particular force in the case of a second appeal.40 While there is much wisdom in their Lordships’ observation that “[t]he type of judgmental conclusion that often has to be reached in intellectual property cases – on issues such as obviousness, inventiveness, and copying – are matters on which appellate courts should be slow to interfere with the judgment of the trial judge,”41 the approach should be understood in context. If Designers Guild involved purely aesthetic choices, in Lucasfilm the Supreme Court set out some fairly clear policy barriers to the encroachment of copyright protections into the domain of functionality. Referring to the legislative context, it opined: “[i]t is possible to recognise an emerging legislative purpose (though the process has been slow and laborious) of protecting three-dimensional objects in a graduated way, quite unlike the protection afforded by the indiscriminate protection of literary copyright.” The Court also noted that the different periods of protection afforded to three-dimensional objects reflected a carefully delineated hierarchy, with works of art receiving the strongest protection, and functional objects being toward the other end of the spectrum. Within the UK statutory framework, their Lordships were able to observe that “[t]here are good policy reasons for the differences in the periods of protection, and the Court should not, in our view, encourage the boundaries of full copyright protection to creep outwards.”42 The strong endorsement of the deference principle thus arose from a particular statutory context, further buttressed by judicial policy statements, that is already directed at constraining some of the anti-competitive effects of copyright protections for functional objects. As is noted above, New Zealand’s designs and copyright laws hew to a different approach. While the definition of protectable design in the Designs Act 1953 excludes features of shape or configuration which are “dictated solely by the function” of the article to which the design is applied, that prohibition’s significance is diminished by the scope of New Zealand copyright protections. As Steelbro shows, functional aspects of designs can be fully protected by copyright law.43 While the current New Zealand legal framework remains in place,44 copyright protections will continue to be extended to functional products or aspects of products through the protection of the underlying design drawings. Cases involving purely aesthetic choices arise relatively seldom in New Zealand. Without radical legislative reforms,

40 41 42 43

44

Id., at [46], citing Biogen Inc v. Medeva plc [1997] RPC 1. Id., at [46]. Id., at [49]. Albeit often for a shorter term: typically sixteen years in the case of industrially applied sculptures with a purely utilitarian function, with twenty-five years of protection for works of artistic craftsmanship. See Copyright Act 1994, s. 75. The New Zealand government is currently in the process of reviewing New Zealand’s Copyright Act 1994. It released an issues paper in November 2018, Review of the Copyright Act 1994 (Ministry of Business, Innovation and Employment, November 2018).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

54

Graeme W. Austin

that will not change. The kind of appellate review available in this jurisdiction is thus especially important. Doctrinal developments after the New Zealand Supreme Court’s 2006 endorsement of Designers Guild will inevitably require New Zealand’s appellate courts to revisit the issue of deference. One year later, and just over a month after Steelbro was decided, in a case that, incidentally, involved a finding on deceptive similarity in the context of a trade mark opposition, the Supreme Court clarified the role of appellate courts. It held that, in general, “those exercising general rights of appeal are entitled to judgment in accordance with the opinion of the appellate Court, even where that opinion is an assessment of fact and degree and entails a value judgment.”45 It would seem to follow that the continued relevance of the Designers Guild deference point in New Zealand copyright law need not be determined in the light of the Court of Appeal’s endorsement of the principle in Steelbro, but, instead, in the light of the Supreme Court’s restatement of general principles of appellate deference a year later. Given these doctrinal developments, it is probably time for a reassessment of the role of appellate courts in copyright cases. Earlier in 2019, the UK Supreme Court returned to the question of appellate deference in a high-profile case involving a dosing patent, Actavis Group PTC EHF & Ors v. ICOS Corporation.46 In the context of the “obviousness” test for patentability, the Court discussed what an error of principle might involve:47 An error of principle is not confined to an error as to the law but extends to certain types of error in the application of a legal standard to the facts in an evaluation of those facts. What is the nature of such an evaluative error? In this case we are not concerned with any challenge to the trial judge’s conclusions of primary fact but with the correctness of the judge’s evaluation of the facts which he has found, in which he weighs a number of different factors against each other. This evaluative process is often a matter of degree upon which different judges can legitimately differ and an appellate court ought not to interfere unless it is satisfied that the judge’s conclusion is outside the bounds within which reasonable disagreement is possible [. . .] Thus, in the absence of a legal error by the trial judge, which might be asking the wrong question, failing to take account of relevant matters, or taking into account irrelevant matters, the Court of Appeal would be justified in differing from a trial judge’s assessment of obviousness if the appellate court were to reach the view that the judge’s conclusion was outside the bounds within which reasonable disagreement is possible.

45

46 47

Austin, Nichols & Co v. Stichting Lodestar [2008] 2 NZLR 141, at [16]. For discussion in the intellectual property context, see Duncan Laing and Ian Finch, James & Wells: Intellectual Property Law in New Zealand (Thomson Reuters, 3rd ed., 2017), at [4.13]. [2019] UKSC 15 (27 March 2019). Id., at [80]–[81].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

Appellate Deference in New Zealand’s Copyright Law

55

The malleable concept of “bounds within which reasonable disagreement is possible” combined with the approach of New Zealand courts to general appeals perhaps offers support for adoption of a more testing approach to the decisions of lower courts in copyright cases. That proposition would, however, need to be set against the New Zealand Supreme Court’s view that greater deference is required in cases involving discretion.48 So far, the New Zealand Supreme Court has not had the opportunity to affirm that, by characterizing an appeal as to the substantiality of copying as “appellate review of a discretion,”49 it meant copyright infringement decisions to be in the category of cases where greater deference is required, or whether these cases should be regarded as “general” appeals.50 As the New Zealand Supreme Court has acknowledged, the distinction between a general appeal and the exercise of a discretion “is not altogether easy to describe in the abstract.”51 The New Zealand Supreme Court has also noted that a requirement to make factual evaluations and value judgments does not in itself make a decision discretionary. Even so, as this chapter has attempted to show, “factual” decisions can be acutely important to competitors’ freedom to operate.52 The stakes are high enough to warrant another look.

48

49 50

51 52

K v. B [2011] 2 NZLR 1. In some statutory frameworks in other intellectual property contexts, the appellate role is described as discretionary. See, e.g., Ministry of Health v. Healthtrio Inc [2008] NZHC 27 (patent); Amadeus Global Travel Distribution SA v. Sabre Inc [2003] NZHC 188 (trade mark). See text above at note 3. Leading New Zealand commentators have suggested that copyright cases might be characterized as discretionary: Laing and Finch, James & Wells: Intellectual Property Law in New Zealand, above note 45, at [4.13]. K v. B, above note 48, at [32]. In the US context, two leading academics have argued that the implications of enforcement of copyright to freedom of expression are sufficiently significant to encourage the exercise of appellate review. See Eugene Volokh and Brett McDonnell, Essay: Freedom of Speech and Independent Judgment Review in Copyright Cases, 107 Yale L.J. 2341 (1998).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.008

4 EU Design Law: Transitioning Towards Coherence? Fifteen Years of National Case Law Estelle Derclaye1

4.1 introduction The Design Directive and Design Regulation2 (“EU design law”) are now almost twenty years old. Annette Kur was part of the drafting committee of the Max Planck proposal for an EU design law.3 The European Commission’s proposal for a legal protection of designs was close to the Max Planck draft for the most part,4 and the draft is also quite close to the actual texts of the directive and regulation.5 1

2

3

4

5

Professor of Intellectual Property Law, School of Law, University of Nottingham. This study was made possible thanks to funding from the Research Priority Area ‘Data Driven Discovery’, University of Nottingham awarded to the author and Gilles Stupfler. My thanks go to Gilles Stupfler and Oliver Church for their precious guidance on statistical issues. I am also grateful to a team of research assistants who helped summarise the decisions in English and provided answers to queries. Some wish to remain anonymous, the others are: Mikko Antikainen, Denisa Assefova, Carlos Sebastian Barreto Cifuentes, Tristan Branellec, Lara Burghgraeve, Antoine Camilleri, Jack Darroch, Stephanie de Potter, Linnéa Dejemo, Thomas Espeel, Paweł Halwa, Helena Haag, Vibeke Huus Rosenquist, Edita Ivanauskiene, Florian Jacques, Dimitar Karaikov, Ursa Kranjc, Oscar Lisshagen, Mireia Moreso, Dorottya Paku, Maria Papavasileou, Michaela Papouskova, Maxence Poivre, Razvan Popa, Andrea Radonjanin, Lars Segato, Romy Siebelink, Eva Skufca, Gabor Soos, Sona Surmova, Kätlin Taimsaar, Eugenia Tonello, Alexandra van Lier, Catarina Videira Louro and Laura Valtere. My sincere apologies if I missed anyone in this very long list. Last but not least, I am grateful for Graeme Dinwoodie’s comments on a previous draft of this article. All errors remain mine. Directive 98/71/EC of 13 October 1998 on the Legal Protection of Designs, OJ L289/28, 28/10/ 1998; Regulation 6/2002 of 12 December 2001 on Community Designs, OJ L3/1, 5/1/2002. Max Planck Institute for Foreign and International Patent, Copyright and Competition Law, Towards a European Design Law (1991). The draft law was also drafted by M. Levin, F.-K. Beier and K. Haertel. See Green Paper on the Legal Protection of Industrial Design, Brussels, June 1991, III/F/5131/91EN, 11 and 144. The system is described in Towards a European Design Law, note 3 above, 5 and 6, and adopts a design approach because the patent approach (in the main the requirement of absolute novelty and a strict check of all validity conditions by the relevant IP office) is suitable for big business but not SMEs and this is the category which most needs efficient protection of

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms 56 of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

Fifteen Years of Design Case Law

57

In assessing Annette’s many contributions it is therefore fitting to examine the impact of the Design Directive and Design Regulation. This chapter examines the case law of the twenty-eight EU Member States since the entry into force of the Design Directive and Design Regulation until August 2017, to see how countries have been transitioning from no design law (e.g. Greece) or from their old design law to the new EU design law and whether the law has achieved coherence over the first years it has been applied. Because space is limited, the chapter concentrates on one aspect, though arguably the most important (and definitely the most interpreted at national level), namely the notion of individual character and its corresponding use in infringement analysis. The chapter examines whether national courts have applied this concept and associated tests (mainly the concept of informed user) adequately or not and concludes that, overall, they have.6 Very few courts have misapplied the law, and with time the misapplications have decreased. Nonetheless, there are some courts that still misapply these notions, so the chapter proposes remedies to achieve full coherence.

4.2 methodology To perform this analysis, all the decisions adopted by judicial and administrative courts at all levels in the twenty-eight Member States that were published between 28 October 2001 (the date of entry into force of the Design Directive) and 31 August 2017, inclusive, were collected. No decisions were found in Luxembourg. This encompasses only decisions handed down on substantive aspects of design law (namely, Articles 1–13 and 15 of the Design Directive and corresponding articles in the Design Regulation). Therefore, this analysis includes decisions on the merits and preliminary proceedings but not decisions which are purely procedural or criminal proceedings. Court decisions doing no more than transmitting references to the Court of Justice of the European Union (CJEU) were also excluded. In some countries, first instance and sometimes second instance decisions are not

6

design. The system adopted for designs is a two-tier system, namely a short (three years’) unregistered right along with a registered right. The right had a grace period, relative novelty, a duration of an initial five years renewable up to twenty-five years. The Legal Review on Industrial Design Protection in Europe, final report, MARKT2014/083/D (DG Growth, 2016), http://ec.europa.eu/growth/content/legal-reviewindustrial-design-protection-europe-0_en, at 68 (hereafter ‘Legal Review’) came to the same conclusion on the two notions assessed here though it surveyed only fifteen Member States. It must be noted that because of space limitations, this chapter does not examine how national courts have interpreted the notions of product, design sector and design corpus in their assessment of individual character and infringement, but the Legal Review suggested (at 68) that some national courts were confused on these notions and recommended (at 156 and 159) codifying the case law of the General Court in relation to the product’s nature and purpose, how the product is used and the industrial sector to which it belongs, to improve the determination of a design’s individual character.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

58

Estelle Derclaye

published, so information was gathered from second and third instance decisions when possible. This resulted in a dataset consisting of 2,214 decisions. This number takes into account all design rights (national registered design right (RDR), Community registered design right (CRDR) and Community unregistered design right (CUDR)) on which the claimant relied in the same case, and so is higher than the number of judgments handed down by courts. The dataset contains 2,084 judgments. Decisions involving designs that were not based on CUDR, CRDR or RDR have not been included in the sample; in other words, decisions based on protection and/or infringement of a design only by copyright, only by a patent, only by a trade mark or only by unfair competition law are not included. But decisions involving one or more of those four design rights and also involving copyright, patent, trade mark, unfair competition or other rights are included in the sample. Decisions based on UK unregistered design right (UKUDR) are not included. Decisions after 28 October 2001 applying the old law are not included in the sample unless they applied both the old law (on validity) and the new law (on infringement).7 The gathering, reading and summarising of the decisions was done by research assistants who were given common detailed instructions including how to search for decisions in databases and what information to summarise in a judgment. The assistants were all law students with training in intellectual property law or practising lawyers specialised in intellectual property law. They had excellent command of English and were all native speakers of the language of the decisions on which they were assigned to work. The assistants summarised the main facts and rulings of the decisions in English and noted whether the court or judge was specialised in intellectual property law or not. By specialised, we mean that the court is an intellectual property court or a court that has intellectual property in its competences. In this latter case, the judges are not only dealing with intellectual property cases but also with other areas of the law. However, they routinely and often hear intellectual property cases as their court/chamber is competent exclusively to deal with them even if it hears non-intellectual property cases too. Below is a table of the decisions per country. For obvious reasons of size, the case law is more abundant in countries with a large population. An exception is the UK, 7

French decisions applying the old law wrongly, i.e. where the new law should apply, were not included because they were not helpful to determine the application of the law. Because of the large amount of decisions in France and budget limitations for this research, first instance court decisions from France that were not appealed were not collected. Since the number of decisions in France is highest among all Member States even with those decisions excluded, the French data was kept for the analysis. However, this means that the wrong application of the individual character and infringement test in France may be higher. However, if those French decisions had been included, because of the sheer number of decisions in the Member States the overall percentage would not increase by much and would be skewed because France is by far the country where courts have most often misapplied the law on these aspects.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

Fifteen Years of Design Case Law

59

which comparatively has far fewer litigations than other big states such as Italy, France and Germany. This can be explained mainly by designers’ perception that the cost of litigation in the UK is high8 and the relatively low success rate on infringement.9 But it is also affected by the existence of the UKUDR, which does not exist in any other EU country, and which is heavily relied on in litigation (41 decisions on UKUDR in addition to the 54 on the other design rights). Austria Belgium Bulgaria Croatia Cyprus Czech Republic Denmark

42 50 70 24 1 54 35

Estonia Finland France Germany Greece Hungary Ireland

10 24 718 237 38 20 2

Italy Latvia Lithuania Malta Netherlands Poland Portugal

211 1 16 2 238 46 24

Romania Slovakia Slovenia Spain Sweden UK

95 35 21 123 23 54

4.2.1 Notion of Individual Character and Test of Infringement Before analysing how the national courts have interpreted the notion of individual character and infringement, a brief reminder of the notions as interpreted by the EU courts is in order. The test of what is sufficient individual character and what amounts to infringement each involve an assessment of the overall impression made by the claimant’s design (or at the infringement level, the defendant’s design) on the informed user. For a design to have individual character, the overall impression must be different; for a design to be infringed, it must not be different.10 But the basic question asked is the same in both contexts. The infringement test is different from tests used in parallel contexts in trade mark and patent law. First, as the CJEU said in PepsiCo,11 the informed user is not a designer or technical expert (as in patent law),12 nor an average consumer (as in trade mark law). He lies somewhere in between and is a particularly observant person “either because of his personal experience or his extensive knowledge of the 8

9

10 11 12

J. Moultrie and F. Livesey, Design Right Case Studies, in UK IPO (ed.), The Economics of Design Rights. An Intellectual Property Office Report (UK Intellectual Property Office, 2011), www.gov.uk/government/publications/the-economics-of-design-rights, 1; BOP Consulting, Design Rights, An International Comparison, in UK IPO (ed.), The Economics of Design Rights. An Intellectual Property Office Report (UK Intellectual Property Office, 2011), https://www.gov.uk/government/publications/the-economics-of-design-rights, 22; A. Carter-Silk and M. Lewiston, The Development of Design Law: Past and Future (UK Intellectual Property Office, 2012), https://www.gov.uk/government/publications/the-devel opment-of-design-law-past-and-future, 71. Neither the Court of Appeal nor the Supreme Court has ever found a CRDR or RDR infringed since their introduction in the UK. Only one CUDR has been found infringed at the appellate level. See Landor & Hawa International v. Azure Designs [2006] EWCA Civ 1285. Articles 5 and 9 of the Design Directive and Articles 6 and 10 of the Design Regulation. Judgment of 20/10/2011, Case C-281/10P, PepsiCo v. Grupo Promer [2011] ECR I-10153. See also Shenzhen Taiden Industrial v. OHIM, Case T-153/08 [2010] ECR II-02517, para. 47.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

60

Estelle Derclaye

sector in question”; he “knows the various designs which exist in the sector concerned, possesses a certain degree of knowledge with regard to the features which those designs normally include, and, as a result of his interest in the products concerned, shows a relatively high degree of attention when he uses them”.13 According to the General Court (GC) in the same case, the informed user is also not a manufacturer or seller of products in which the design is incorporated.14 As implied in PepsiCo, the informed user is a legal fiction15 and thus there is only one such user. However, the GC’s later case law is unclear on this. Indeed, after the CJEU PepsiCo decision, the GC carried on giving a human personality to the informed user and also sometimes identified two informed users (often the final consumer and the distributor).16 Second, unlike trade mark law, there is no place for (risk of ) confusion in design law.17 If there were, depending on the case, there may or may not be more instances of infringement. Third, unlike in trade mark law, which relies on the imperfect recollection of the average consumer, the comparison has to be made directly (side by side) unless it is impossible.18 Finally, the degree of freedom of the designer is taken into consideration to determine the overall impression. The literature agrees that since the text of Articles 6 and 10 of the Design Regulation and corresponding Articles 5 and 9 in the Design Directive suggest a quasi-identical assessment, their interpretation must be consistent.19 There are a few small differences between the assessment performed to determine individual character and infringement, but for the purposes of this chapter they are not important.20 13 14 15

16

17

18 19

20

PepsiCo, note 11 above, paras. 53 and 59 (CJEU). Case T-9/07; see also El Hogar Perfecto Del Siglo XXI v. OHIM, case T-337/12 (CJEU 6th ch.). D. Stone, European Union Design Law: A Practitioner’s Guide (Oxford University Press, 2nd ed., 2016), 221–222, paras. 12.56–12.61; L. Bently, B. Sherman, D. Gangjee and P. Johnson, Intellectual Property Law (Oxford University Press, 5th ed., 2018), 774. G. Hasselblatt (ed.), Community Design Regulation: A Commentary (Beck, 2015), 81–82, citing Shenzhen, note 12 above; Kwang Yang v. OHIM, case T-10/08; and El Hogar, note 14 above. Stone, note 15 above, 254. As early as 2006, the Austrian Supreme Court, in only its second case on the new EU design law (14 February 2006, 4 Ob 177/05s, Construction Site Advertising II), held that design law’s task is not to protect the public against product confusion. The court relied significantly on the literature. It repeated this finding in the Febreze case, 22 May 2007, 4 Ob 43/07p. Also in 2006, a German court held that risk of confusion was not the test (CA Hamburg, Az 5 U 135/05, 20 December 2006 (Cell Phone Design), and later again, CA Hamburg, Az 5 U 152/11, 29 August 2012 (Crystal Skull Bottle)). PepsiCo, note 11 above, paras. 55, 57 (CJEU). Hasselblatt, note 16 above, 80, 151; D. Musker in C. Gielen and V. von Bomhard (ds.), Concise European Trade Mark and Design Law (Kluwer, 2011), on Article 10 CDR and D. Stone, European Union Design Law: A Practitioner’s Guide (Oxford University Press, 2012), paras. 18.32 and 11.04; Bently, et al., note 15 above, 795. The main one is that Article 10 does not specify the date of reference concerning the informed user’s knowledge, though the literature and at least one court agree that it must be the same as that for individual character, namely the date of filing or priority. See Dyson v. Vax (2010)

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

Fifteen Years of Design Case Law

61

4.2.2 National Case Law on Individual Character and Infringement How have national courts interpreted the notions of individual character and infringement? Over the period examined, 62 out of 2,214 decisions or 2.8 per cent21 misapplied the individual character requirement and/or the test of infringement. This is 4.1 per cent of the 1,496 decisions which assessed either validity or infringement. By misapplication of the law, we mean that those courts used a confusion-based criterion to judge individual character and/or overall impression in infringement cases, or used the average consumer instead of the informed user. The percentage is very low, showing that, overall, national courts have applied the law well. Countries can be grouped broadly into three categories. There are twelve countries where no court has used the trade mark test of likelihood of confusion nor the concept of average consumer (Austria, Cyprus, Finland, Germany, Ireland, Latvia, Malta, Lithuania, Poland, Slovenia, Sweden and the UK).22 In the second category, there are five countries where some courts did make the mistake at the start but stopped relatively quickly (Croatia, Denmark, Greece, the Netherlands and Romania). In the last category, there are ten countries where, even if the number of courts making the mistake is low, courts repeat the mistake over time and/or have done so recently. Many of those making the mistake are specialised courts (in Belgium, Bulgaria, Czech Republic, Estonia, France, Hungary, Italy, Portugal, Slovakia and Spain). The breakdown is as follows:23 Croatia (1 NS),24 Denmark (1 S),25 Greece (1 NS),26 the Netherlands (1 S, 3 NS),27 Romania (1 S),28 Belgium (2 S, 1 NS),29 Bulgaria

21 22

23

24 25 26 27

28 29

EWHC 1923, paras. 42–45 and Bently et al., note 15 above, 795–796; Hasselblatt, note 16 above, 152–153; Musker, note 19 above, 377. 2 per cent if France, the country where the law has been most misapplied, is discounted. Note, however, that in Latvia and Cyprus there is only one decision in each country. In Ireland and Malta, there are two decisions but the second is an appeal so there is only one litigation. S means that the court or judge is specialised and NS means it is not specialised. CFI = Court of First Instance and CA = Court of Appeal. CA Zagreb, 21/10/2008, case no. Pž-6155/08-3. Maritime and Commercial High Court (CFI), 22/5/2008, Reisenthel Accesoires v. Zebra A/S. CFI Piraeus, case no. 2445/2009, date unknown. CA ‘s-Hertogenbosch, 13/5/2003, KG C0201329-BR (NS); CFI Haarlem, 10/6/2004, 101429 – KG ZA 04-192, Hasbro International Inc. et al. v. Tulip Toys B.V., Incoparts Services (NS); CFI Almelo, 16/6/2004, 64498 – KG ZA 04-124, Tribù v. Trip Trap Denmark (NS); CFI ‘s-Hertogenbosch, 2/5/2006, 140897 – KG ZA 06-235, Inter Shop v. Free Time Products et al. (NS); CFI The Hague, BMW & Rolls Royce v. ABC, 7/7/2010, IEPT20100707 (S). CFI Bucharest, 10/05/2011, case no. 866. CFI Kortrijk, 15/5/2003, Modular Lighting Instruments v. Wever and Ducre (NS); Commercial Court Brussels, 29/10/2014, Mars et al. v. Gedon Pet (S); Chairman Commercial Court Brussels, 15/4/2015, Smart Trike Pte v. Fun Belgium (S). All CFI.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

62

Estelle Derclaye

(1 S, 3 NS, 3 unknown if S or NS),30 Czech Republic (1 S, 1 NS),31 Estonia (2 S),32 France (14 S, 4 NS),33 Hungary (7 S),34 Italy (5 S),35 Portugal (1 S, 1 NS),36 Slovakia (4 S),37 and Spain (1 S, 5 NS).38 In some of these countries, it is the same courts which make the mistake (Sofia City Court and Sofia Administrative Court, Budapest Court of First Instance, Budapest Court of Appeals, Hungarian Supreme Court, two Paris Court of Appeal judges), but in most it is different courts.39 30

31

32

33

34

35

36

37

38

39

Sofia City Court, 23/5/2012 (not known if specialised or not); Sofia Administrative Court, 21/12/ 2012 (NS); Sofia City Court, 4/1/2013 (not known if specialised or not); Sofia City Court, 5/4/ 2013 (not known if specialised or not); Sofia City Court, 4/10/2013 (S); Sofia City Court, 20/7/ 2015 (NS); Sofia Administrative Court, 9/5/2016 (NS). All CFI. Supreme Court, 26/3/2013, case file no. 23 Cdo 1960/2014 (NS); Prague High Court (CA), 8/4/ 2013, case file no. 3 Cmo 322/2012-129 (S). Tartu Circuit Court, 16/11/2007, case no. 1-06-8492/19; Tallinn Circuit Court, 18/6/2010, case no. 2-08-13577 (both CFI). Courts applying risk of confusion test: CA Paris, 18/10/2006, Modascoop v. Vivarte et Creations Rivers (S); CA Paris, 8/11/2006, Poltronesofa v. Cinna (S); CA Bordeaux, 2nd ch., 29/11/2006, no. 05/05101 (NS); CA Douai, 2nd ch., 1st sec., 10/5/2007, no. 05/02309 (NS); CA Paris, 21/2/ 2014, Vetstar v. Territoire Redskins (S); CA Paris, 5th pôle, 2nd ch., 5/6/2015, Tintamar v. Nocibe France Distribution (S); CA Paris, 5/6/2015, Air Refreshing Control et Areco Finances et Technologie v. Bernard Ribière Organisation (S). Courts applying average consumer test: CA Paris, 4th ch., sec. A, 14/2/2007, no. 06/09813 (S); CA Paris, 30/3/2007, Stolar v. Guy Ella et LGE (S); CA Paris, 7/2/2014, Sprintex v. Bahadirlar Tekstil Paris (S); CA Paris, 5th pôle, 2nd ch., 16/10/2015, Evergreen Land et Evergreen Land Europe v. City Green (S); CA Paris, 5th pôle, 2nd ch., 29/9/2017, S.A.R.L. Hipanema v. S.A.R.L. Créa’ Mode distribution and S.A.R.L. Dress Code (S). Courts applying both risk of confusion and average consumer test: CA Paris, 8/11/ 2006, Sté Manufacture Générale Horlogerie v. Sté Guy Ellia, Auchan et al. (S); CA Paris, 2nd ch., 9/12/2011, Sté Anna Maria Blome v. Stéphane Corler et al. (S). CA Budapest, 1/4/2010, case no. 8.Pf.22.149/2010/6 (S); Court of Justice of Budapest, 11/12/2009, case no. 3.P.20.082/2009/9 (S), aff’d by CA Budapest, 30/9/2010, case no. 8.Pf.20.430/2010/4 (S); Court of Justice of Budapest, 8/11/2010, case no. 3.P.26.788/2009 and CA Budapest, 12/5/2011, case no. 8.Pf.20.413/2011/3; Court of Justice of Budapest, 13/12/2010, case no. 3.Pk.23.715/2010/2 (S), aff’d by Supreme Court, 8/2/2012, case no. Pfv.IV.21.763/2011/5 (S). CA Milan, 20/12/2007, Giochi Preziosi v. Luca Dinelli (S); CFI Turin, 4/6/2010, Alessi v. Barazzoni (S); CFI Venice, 22/6/2010, Fratelli Bovo v. Punto Oro IV (S); CFI Milan, 21/11/ 2014, Monica Bianco v. Loxi Life (S); CFI Naples, 30/1/2017, Victor Talavera and Talavera Hair Products v. Diffitalia Group (S). CA Guimarães, 15/12/2009, case no. 1798/07.0TBFLG.G1 (“All Star”) (NS); Intellectual Property Court, 26/5/2015, case no. 1241/11.0TYLSB, Leitão & Irmão v. Hermès Sellier (S). Regional Court of Banska Bystrica (CFI), 15/2/2013, case no. 24S/129/2012, Ryba Košice, spol. v. Industrial Property Office of the Slovak Republic (S); District Court of Kosice I (CFI), 15/1/ 2014, case no. 27CbPv/4/2012, Crocs v. Trico Line SK (S); District Court of Banska Bystrica (CFI), 20/2/2017, F. and Z. Batal Company, Aleppo, and Alaska Foods v. RiD (S), aff’d by Regional Court of Bratislava (CA), 8/6/2017, case no. 43CoPv/7/2017 (S). CFI Alicante, 28/2/2007, Imatex v. Sanpere (S); CA Granada, 30/9/2013, Milade v. Juan Pedro and Eurotex Import-Export SL and Direccion000, C.B. (NS); Administrative Court of Valencia (CFI), 5/2/2014, Abanicos Gil Y Blay v. Dirección General de la Oficina de Patentes y Marcas and Grupo Osborne (NS); CA Barcelona, 23/9/2014, Barcino R. Mas de Forme y Colors y Abel. v. Creaciones Jumi, Clemencia y Juan Ramón (NS); CA Burgos, 26/1/2015, Montes Design v. Industria Rodrigo (NS); CA La Coruña, 3/11/2016, Rosana v. Objetivo Producciones (NS). In all but one Member State, the legislation is not the issue as it complies with the directive and regulation. But in Estonia, the problem is legislative, because the Estonian Industrial Design

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

Fifteen Years of Design Case Law

63

What else can be gleaned from analysing the 62 decisions misapplying the law? First, rulings are sometimes very short, so even if it appears that the courts were using trade mark law notions, it is not always known whether courts also implicitly used the concept of average consumer, which involves an imperfect recollection, or a side-by-side comparison as design law mandates. Second, 39 of the 62 decisions (62.9 per cent) were handed down by courts specialised in intellectual property law compared to 20 by non-specialised courts (with 3 where it is not known whether the court is specialised or not). Third, it is mostly the first instance courts that misapply the law (20 specialised as opposed to 9 non-specialised (i.e. 68.9 per cent)); there are fewer at the court of appeal level (6 specialised as opposed to 7 non-specialised (i.e. 46 per cent)). That said, there are more first instance decisions and some are not appealed, so the proportions may simply reflect this fact.40 Fourth, it does not seem that court specialisation makes any difference, i.e. specialised courts do not apply the law better.41 Indeed, in most countries, most courts deciding design cases are specialised (72.2 per cent of courts are specialised overall in all Member States42); only in Bulgaria is the number of non-specialised courts deciding design cases higher than that of specialised ones. Overall, a greater number of specialised courts (62.9 per cent) misapply the law. The worst ‘offenders’ both in terms of the number of wrong decisions and of court specialisation are, in decreasing order: France, Hungary, Bulgaria, Spain and Italy. That said, the overall rate of misapplication has gone down with time. In addition, in some countries (Croatia, Denmark, Estonia, Greece, the Netherlands and Romania), all the ‘wrong’

40

41

42

Act states: “§ 6(1) An industrial design is deemed to be new if no identical or confusingly similar industrial design is disclosed in the Republic of Estonia or any foreign state prior to the filing date of the registration application for the industrial design (hereinafter registration application) or, if priority is claimed, prior to the date of priority. In determining the novelty of an industrial design, identical or confusingly similar registered industrial designs are also taken into consideration after they are disclosed if such industrial designs have an earlier filing date of the registration application or earlier date of priority” (our emphasis). The confusion test is reiterated in the Estonian provision on infringement. Pursuant to § 16(1) and (2) of the Industrial Design Act, “the owner of an industrial design has the exclusive right to manufacture products according to the industrial design, and to distribute, sell, offer for sale, or import, export or store for the aforesaid purposes products which are manufactured according to the registered industrial design and has the right to prohibit other persons from manufacturing without authorization products according to an identical or confusingly similar industrial design and from distributing, selling, offering for sale or importing, exporting or storing for the aforementioned purposes products which are manufactured according to the registered industrial design” (our emphasis). See WIPO Lex, consolidated text in force as of 1/1/2015. As French first instance courts were not included, it is worth noting the numbers for France separately: 14 specialised against 2 non-specialised, all courts of appeal. The proportion of specialised against non-specialised stays relatively high if we only take the countries of the third group: 24 S (38 with France) against 15 NS, 3 unknown. This percentage does not include those (69 decisions) where this characteristic is unknown.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

64

Estelle Derclaye

decisions predate the CJEU landmark decision in PepsiCo,43 which may indicate that the courts in these countries are now applying the CJEU precedent well. Looking at these decisions in more detail, a few more mistakes in the relevant assessments can be highlighted. Some courts also make mistakes regarding related aspects, such as using a memory test rather than a comparison side by side,44 comparing the claimant’s and defendant’s products without checking that the claimant’s product is the same as that in the registration documents,45 and, less problematically, as the EU courts’ case law is unclear on this point, using several informed users.46 Two courts have also erroneously used an expert as the informed user.47 More questionably, some French courts have used unclear tests.48 In these French decisions, it is not clear whether several informed users are used, and if so, whether some are experts or not. That said, the application of the tests in France has greatly improved over the years. Even if some courts initially used risk of confusion and average consumer, in 43 44

45

46

47

48

See note 11 above. In the Febreze case, the Austrian Supreme Court held that the differences between the two objects to be compared are so considerable that they are apparent for an informed user who resorts to his or her memory without a detailed comparison of the two designs. Hence this assessment clashed with that of the Court of Appeal of England and Wales in the same litigation. CFI Milan, 4/7/2017, Stefanplast S.p.a. v. Plana S Proizvodnja in Prdodaja D.o.o. (S); CA Paris, 7/2/2014, Sprintex, note 33 above (S) (average consumer who does not see the claimant’s design and the defendant’s product simultaneously). Tallinn Circuit Court, 18/6/2010, note 32 above; Banska Bystrica District Court (CFI), 20/2/ 2017, F. and Z. Batal Company, note 37 above (S), aff’d by Bratislava Regional Court (CA), 8/6/ 2017, note 37 above (S); CA Barcelona, 23/9/2014, Barcino, note 38 above (NS). In decision 48, 11/1/2017, the Sofia City Court (unknown if specialised) uses three different informed users: the person who produces, sells and transports the food products; and in decision 1624, 13/3/2017, the Sofia Administrative Court (NS) also uses three informed users: the person who uses the forklift trucks, the professional trader who sells these products and the prospective buyer. Commercial Court of Brussels, 21/3/2013, Del Ponti NV v. M. Gluszek (the informed user is a designer colleague) (S); CA Barcelona, 23/9/2014, Barcino, note 38 above (NS) (the buyersellers and not the “final consumers” – so the traders, who are expert in the area). CA Versailles, 1st ch., sec. 1, 30/1/2014, no. 11/08203 (NS) (it is necessary to make a comparison of the two models in question, assessing them, not in relation to the isolated elements but in a global manner based on the overall impression produced in the eyes of the customer, in this case urban furniture professionals); CA Bordeaux, 1st ch., sec. A, 12/2/2016, Villeroy & Boch v. Diodon (NS) (the informed user designates the user who possesses particular attention as of his/her professional experience or his/her knowledge of the field; it can also be the average consumer or the professional who buys and sells the product which incorporates the design, provided they use it properly); CA Paris, 5th pôle, 1st ch., 27/1/2015, Keter Holdings v. Shaf (S) (the informed user can be both the average consumer and the professional who buys and resells the product incorporating the design, as long as he/she uses the product according to its purpose); CA Paris, 5th pôle, 1st ch., 19/5/2015, Kontiki, Depesche Vertrieb v. Editions Play Bac (S) (young girls from 7 to 10 years old or 5 to 12 years old and people who have an extensive knowledge of creative notebooks because of their quality or function); CA Paris, 5th pôle, 2nd ch., 27/11/2015, Piganiol v. Publicis Conseil et L’Oréal (S) (the public interested and able to detect detailed differences in an area where the freedom of the creator is limited, and also the professionals in the sector concerned).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

Fifteen Years of Design Case Law

65

2008 the Supreme Court held that “it is not necessary to investigate whether there is a risk of confusion between the disputed products”49 and that the moderately attentive consumer is the wrong test.50 Later, in 2015, the Court was even clearer: “the risk of confusion is irrelevant to find infringement”.51 After these Supreme Court decisions, lower courts’ incorrect application of the notion of individual character and infringement test decreased.52 However, one specialised judge carried on making the same mistake,53 and another specialised judge sometimes used the wrong test and sometimes the right test.54 Finally, there are a few isolated decisions which apply the individual character and infringement test wrongly in other ways. A Polish non-specialised court confused the requirement of novelty and individual character,55 whereas the EUIPO Boards of Appeal and the GC have repeated that the two tests are different.56 However, these Polish decisions are old and the mistake does not appear later, so these decisions may now be historical. In interim proceedings, the Court of Milan held that the test of infringement is more stringent for CUDR, in other words requires perfect identity,57 whereas the scope is exactly the same as for CRDR, the only difference being the proof of copying.58 It is unclear whether this was an isolated case, and it seems that some judges at the Court of Milan have followed this line of thinking in later cases, while others

49

50

51

52

53

54 55

56

57

58

Cass.com., 26/3/2008, Frendo v. Établissements Tarrerias Bonjean – PIBD 2008, no. 875, III, 349 (SJ). Cass.com., 23/9/2008, Manufacture Générale Horlogerie v. Auchan France, Guy Ellia, Société Générale de Marques, Société Lge – no. 07/15053 (S); Cass.com., 17/3/2009, Tuileries Briqueteries du Lauragais Guiraud Frères v. Tejas Borja (S) (the observer of average attention is the wrong test). And also Cass.com., 15/6/2010, no. 08-20.999, PIBD 2010, no. 923, III, 553 (S) and Cass., 3/4/2013, Coline Diffusion v. Ida 2000 – no. 12-13356 (S). Cass.civ., 1st ch., 15/5/2015, Longchamp et al. v. Vivadia et al., followed later by CA Paris, 5th pôle, 1st ch., 26/1/2016, Smoby Toys v. Splash-Toys (SJ). For correct applications, see e.g. CA Rennes, 2nd comm. ch., 1/9/2009, Isnard Jezequel v. Kenzo (NS) (unlike trade mark imitation, which is based on the risk of confusion in the mind of a consumer of average attention, design imitation requires a comparative examination of the designs in the presence of an informed observer); CA Paris, 21/9/2012, SCPE v. Jalons Éditions et Cogenor (S); CA Paris, 11/1/2013, Normandie Structure v. Catherine Dalo et Dalo Frères (S); CA Paris, 13/6/ 2014, ST Dupont v. Aker (S); CA Paris, 5th pôle, 2nd ch., 4/3/2016, Terdis, Villeroy et Boch et Diodon v. Cora et Confiserie du Tech. (S). Decision of 19/5/2015, note 48 above, though it came only four days after the Supreme Court decision so maybe the judge was not yet aware of it when writing his judgment. Compare decision of 27/11/2015 with that of 13/6/2014, notes 48 and 52 above. CA Warsaw, 11/12/2008, case no. VI SA/Wa 498/09; and CA Warsaw, 15/12/2009, case no. VIII SA/Wa 332/09 (an industrial design may be considered as not satisfying the requirements of novelty only if it differs in immaterial details from an earlier design, i.e. if it is not individual). See mainly cases R-1456/2006-3, Imperial International v. Handl Cookware and T-68/11, Erich Kastenholz v. OHIM. CFI Milan, 6/5/2011, Dolce e Gabbana et al. v. Progetti Marchi Moda e Gaudì Trade (S – Zana). See e.g. Stone, 2nd ed., note 15 above, 452, 456.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

66

Estelle Derclaye

have not.59 In other precautionary proceedings, the same court, but differently composed, also used a stricter test of individual character, stating the protected shape must have a level of individuality not only to attract the attention of the consumer, but also to be a reason for preferring to buy the claimant’s product.60

4.3 conclusion Overall, in relation to the important concepts of individual character and infringement, national courts transitioned to the new EU design law well and coherently. With the CJEU PepsiCo decision, over time, those which did not apply the law well finally did so. That said, coherence in this area of EU design law is not fully achieved. There are a few courts in the third group that are still misapplying the law and, somewhat worryingly, most of these are specialised courts (38 out of 54, i.e. 70.3 per cent61). However, it may be that there are more specialised courts misapplying the law because, as we saw, most national courts deciding design cases are specialised anyway. The number of courts making a mistake per country is also generally very small, so conclusions are hard to draw. Thus, specialisation is probably not the problem here and we are certainly not suggesting getting rid of specialised intellectual property courts. One can think of several solutions. To remedy the continuing, though sporadic, misapplication of those tests in the third group of countries, both lawyers’ and judges’ training should be ramped up in addition to the judges’ training already organised by EUIPO.62 It is mainly the Hungarian and Bulgarian courts, and a few French courts, that need more guidance. Estonia should modify its Design Act to remove any confusion-based assessment. The GC should clarify that only one informed user must be used to clear up the confusion still remaining in some national courts. Ideally, the European Commission should revise the directive and regulation, codifying the EU courts case law or, falling short of this, issue guidelines to clarify that risk of confusion is not the test and that the informed user is not the average consumer but is a unique, fictitious, person. An opportunity to do so has arisen, as the Commission launched, in December 2018, a public consultation on 59

60

61 62

3/5/2012, Arsmoda v. Renè Caovilla (judges Tavassi, Perrotti and Zana find infringement as only one marginal detail is different); 21/11/2014, Monica Bianco, note 35 above (judge Giani seems to use the identity test); 27/11/2015, Diesel et al. v. Zara Italia et al. (judge Tavassi holds that the claimant’s design and the defendant’s product are essentially identical); 21/11/2016, Fan Europe Lighting v. Leroy Merlin Italia e Novecento (judge Perrotti also seems to apply the identity test). CFI Milan, 20/9/2012, Crocs v. De Fonseca (precautionary proceedings) (S – Tavassi/Gandolfi/ Marangoni). This number includes three courts for which it is unknown if they are specialised or not. The EUIPO organises seminars and symposia for national judges to meet, learn and exchange knowledge, information, judgments and best practice, and also to achieve greater judge specialisation. See https://euipo.europa.eu/ohimportal/en/web/observatory/knowledge-building and https://euipo.europa.eu/tunnel-web/secure/webdav/guest/document_library/observatory/ documents/Judges_events/Judges_Seminar_09–2017_sum_en.pdf.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

Fifteen Years of Design Case Law

67

design protection in the EU to inform it on the potential need for improvement of EU design law.63 In October 2019, DG GROW launched a tender for a study on certain aspects of design law, including spare parts. Therefore, it probably means that it considers reform necessary, but any changes in the law may not occur for some time after the study is completed.

63

See https://ec.europa.eu/info/law/better-regulation/initiatives/ares-2018-3527248/public-consult ation_en.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.009

5 Copyright and the CJEU: Some Structural Deficits as Seen from a German Perspective Thomas Dreier1

5.1 introduction Some years before the EU Commission undertook a renewed effort to further harmonize existing copyright law by proposing, in 2016, its so-called copyright package,2 the German Association for the Protection of Intellectual Property (GRUR) had organized a panel including a representative of the Commission and a member of the Court of Justice of the European Union (CJEU). The question to be considered was the role of the CJEU in the process of harmonizing EU copyright law. The representative of the CJEU answered that, of course, the CJEU limits itself to merely interpreting existing EU law and certainly does not fill in the gaps left by the EU legislature. To this answer the representative of the Commission replied, with a smile on her face, that in view of the difficulties of further harmonizing the copyright laws of the EU Member States the Commission was quite happy with the way in which the CJEU had been “interpreting” the existing copyright Directives. Although both statements correctly reflect the division of competencies between the law maker on the one hand, and the CJEU on the other, it was, of course, clear to both speakers that undisputedly the CJEU does not limit itself to interpreting existing EU law, but that acting in this way the CJEU likewise further develops existing EU copyright law beyond the text enacted by the EU legislature. As is well known, the most obvious and prominent examples to this effect are: the extension of 1

2

Professor Dr.; MCJ (New York University); Director, Institute of Information and Economic Law, Center for Applied Legal Studies, Karlsruhe Institute of Technology (KIT), Karlsruhe, and Honorary Professor, Faculty of Law, University of Freiburg, Germany. https://ec.europa.eu/digital-single-market/en/modernisation-eu-copyright-rules. The package comprised proposals for a Regulation and a Directive regarding the implementation of the WIPO Marrakesh VIP Treaty, as well as for a Regulation on certain online transmissions of broadcasting organisations and retransmissions of television and radio programmes, and, most importantly, for a Directive on copyright in the Digital Single Market, Documents COM(2016) 593 to 596 final of 14.9.2016.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms 68 of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

Copyright and the CJEU: Some Structural Deficits

69

the definition of the “author’s own intellectual creation,” which was contained only in three Directives, to all copyrighted works;3 the detailed definition of “fair compensation” in Article 5(2) and (3) of the InfoSoc Directive 2001/29/EC;4 the statement that in order to achieve further harmonization, Member States which have chosen to provide for an optional exception from the list of Article 5(2) and (3) in their national law must do so to the full extent as defined in the InfoSoc Directive;5 and, last but not least, the development by the CJEU of a new definition of the so far well-established concept of “communication to the public.”6 The first part of this chapter discusses the scope and some methodological aspects of the decisions with which the CJEU has so far contributed to further harmonizing EU copyright law beyond the exact wording of the statutory language adopted by the EU legislature. In the second part, it is argued that the referral procedure as currently defined in Article 267 TFEU is not well suited to enable the CJEU to develop a fully coherent copyright system as expected by the Member States. Hence, one might consider modifying the current procedure in order to reduce the uncertainty of the further development of EU copyright law by the CJEU.

5.2 the cjeu and the harmonizing effort of its copyright decisions 5.2.1 Interpretation Versus Lawmaking Admittedly, as the anecdote reported in the opening paragraph demonstrates, the line between lawmaking and interpretation is a fine one that cannot easily be drawn. 3

4

5

6

Art. 1(3) of Directive 2009/24/EC on the legal protection of computer programs, Art. 3(1) of Directive 96/9/EC on the legal protection of databases, and Art. 6 of Directive 2006/116/EC on the term of protection of copyright and certain related rights; for the extension to all works protected by copyright see CJEU, cases C-5/08, Slg. 2009, I-6569 – Infopaq and C-145/10, Slg. 2011, I-12533 – Painer. CJEU, cases C-245/00, Slg. 2003, I-1251 – SENA; C-271/10, Slg. 2011, I-5815 – VEWA; C-467/08, Slg. 2010, I-10055 – Padawan; C-462/09, Slg. 2011, I-5331 – Stichting de Thuiskopie; C-457-460/ 11, ECLI:EU:C:2013:426 – VG Wort et al.; C-521/11, ECLI:EU:C:2013:515 – Amazon.com International Sales et al.; C-435/12, ECLI:EU:C:2014:254 – ACI Adam; C-463/12, ECLI:EU: C:2015:144 – Copydan Båndkopi; C-572/13, ECLI:EU:C:2015:750 – Hewlett-Packard Belgium; C-470/14, ECLI:EU:C:2016:418 – EGEDA et al.; C-110/15, ECLI: EU:C:2016:717 – Nokia Italia et al. CJEU, cases C-510/10, ECLI:EU:C:2012:244 – DR und TV2 Danmark; C-201/13, ECLI:EU: C:2014:2132 – Deckmyn and Vrijheidsfonds; but see also the less strict C-462/09, Slg. 2011, I5331 – Stichting de Thuiskopie and C-145/10, Slg. 2011, I-12533 – Painer. In particular, CJEU, cases C-135/10, ECLI:EU:C:2012:140 – SCF Consorzio Fonografici; C-607/ 11, ECLI:EU:C:2013:147 – ITV Broadcasting et al.; C-466/12, ECLI:EU:C:2014:76 – Svensson; C-348/13, ECLI:EU:C:2014:2315 – BestWater International; C-117/15, ECLI:EU:C:2016:379 – Reha Training; C-151/15, ECLI:EU:C:2015:468 – Sociedade Portuguesa de Autores; C-160/15, ECLI:EU:C:2016:644 – GS Media; C-527/15, ECLI:EU:C:2017:300 – Stichting Brein (Filmspeler); C-641/15, ECLI:EU:C:2017:131 – Verwertungsgesellschaft Rundfunk; C-161/17, ECLI:EU:C:2018:634 – Renckhoff.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

70

Thomas Dreier

The reason simply is that whenever an abstract legal norm has to be applied to a given set of specific and concrete facts, the legal norm has to be interpreted by spelling out details of the norm which do not, as such, show up in the statutory text itself. It follows that to the extent to which the courts are entrusted with the power to interpret statutory texts, any further development of the law by way of such interpretation is legitimate. In other words, interpretation and lawmaking are not opposites, but they overlap in part. The question to ask is therefore not whether a given court confines itself to the activity of interpretation and abstains from lawmaking. Rather, one would have to ask whether and at which point the interpretation of the statutory text detaches itself from the legal text in a way which transcends the limits of mere interpretation and ventures into the territory of norm-setting and hence lawmaking. The extent to which the CJEU rightly claims for itself the power to develop the statutory text of existing copyright exceptions by way of interpretation has been subject to extensive academic analysis.7 True, the concept of “autonomous” interpretation of rules and notions of the Directives as evoked by the CJEU in many, albeit not in all, of its decisions8 results directly from the supremacy of EU law over the national laws of the EU’s Member States. Hence, the CJEU may well resort to, and is even under an obligation to apply, EU law. However, for reasons further explained in Section 5.3, the CJEU would be well advised to exercise its supreme power of interpreting secondary EU legislation only for purposes of interpretation in the proper sense, rather than for filling in gaps left by the EU legislature. After all, the “autonomy” is one of the existing concepts as laid down by the legislature, not one of the CJEU to create new concepts, and certainly not one to develop a completely new set of legal norms derived from interpretation of more or less unspecified definitions, words and notions used in the language of the Directives.

5.2.2 Areas of Far-Reaching and Possibly Norm-Setting Interpretation However, this general theoretical discussion shall not be taken up here, since it is not really of importance in view of the point to be made in this chapter. Suffice it to say that the limits of interpretation can be said to be transcended where the CJEU defines the content of legal notions and specifies details of statutory rules upon 7

8

See V. Roder, Die Methodik des EuGH im Urheberrecht – Die autonome Auslegung des Gerichtshofs der Europa¨ischen Union im Spannungsverha¨ltnis zum nationalen Recht (MohrSiebeck, 2016), claiming a broad power of the CJEU to interpret existing statutory rules as autonomous concepts of EU law. See, e.g., CJEU, case C-467/08, ECR 2010, I-10055, ECLI:EU:C:2010:620 – Padawan (“fair compensation” an autonomous concept of EU law); case C-201/13, ECLI:EU:C:2014:2132 – Deckmyn and Vrijheidsfonds (“parody” as an autonomous concept). It might be added, however, that as of now it is still not quite clear to what extent the rights granted in the Charter mandate certain exceptions or whether they only have to be taken into account when interpreting existing limitations.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

Copyright and the CJEU: Some Structural Deficits

71

which Member States had not been able to agree at the time of adoption of the respective Directive. Regarding the InfoSoc Directive,9 the four main areas mentioned in the introduction shall be briefly discussed. The first of these areas relates to the concept of the standard of “originality,” which is expressly mentioned as the “author’s own intellectual creation” in only three Directives, but which has been extended by the CJEU without further ado to all works covered by existing – and without doubt also future – Directives and, invariably, Regulations in the area of copyright.10 Although this was probably the most outright case of the CJEU’s venturing into the realm of legislative activity, it seems to have met with criticism only at a theoretical and academic level. In practice, however, the effect of this generalization of the notion of originality might only be of little, if any, importance. Since the criterion ultimately eludes further formalization, the national courts of the Member States still seem to have a wide margin of discretion in determining whether the object at issue is for copyright protection. The second of these areas relates to the definition of “fair compensation” and to the procedural rules to be followed whenever it has to be paid once a Member State has exempted acts of analog reproduction and private copying activities from the authors’ exclusive reproduction right. In this respect, it should not be forgotten that despite several attempts by the EU Commission, the EU legislature was not able to harmonize the system of copy levies. As a result, the introduction of the limitations laid down in Article 5(2) and (3) of the InfoSoc Directive is not mandatory, having been left to Member States’ discretion. In view of this legislative history, it may reasonably be argued that the definition of the modalities of such systems was likewise left to the Member States’ discretion. However, the CJEU followed a different path. Although, as a starting point, it seems logical to conclude that any compensation to be paid according to Article 5(2) and (3) of the InfoSoc Directive can only compensate acts of copying which are permitted by these norms, the CJEU went further in forbidding Member States to provide for compensation for acts of reproduction which are not covered by the said exceptions. Moreover, the CJEU required that Member States provide for a claim of repayment of compensation which had been paid unduly (as impractical as such a claim might be in practice), as well as for a number of legal presumptions in order to determine whether or not the uses made were covered by the InfoSoc Directive’s private copying exceptions. In addition, since the CJEU classified the levies to be paid under a private copy exception as “compensation” (as stated in the legislative text), rather than as “remuneration” (as stated in recitals 38 [not referred to in the decision] and 39 of the 9

10

Directive 2001/29/EC of the European Parliament and of the Council of 22 May 2001 on the harmonisation of certain aspects of copyright and related rights in the information society, OJ L 167 of 22 June 2001, 10 et seq. Art. 1(3) of Directive 2009/24/EC on the legal protection of computer programs; Art. 3(1) of Directive 96/9/EC on the legal protection of databases; and CJEU, cases C-5/08, Slg. 2009, I-6569 – Infopaq and C-145/10, Slg. 2011, I-12533 – Painer.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

72

Thomas Dreier

InfoSoc Directive), the Court established a link between “compensation” and economic “harm,” based on recital 35, in which the “harm” to the rightholders resulting from the act in question is considered “a valuable” and hence not the only criterion.11 This interpretative decision has rather far-reaching normative effects. It not only disassociates the amounts to be paid from the idea of a contractual relationship on which copyright is based, but it likewise links the claim to the basis of a legal right, rather than to considerations of who has made the investment in the production of the copyrighted work.12 In sum, this “interpretation” does in fact define the details of national systems of “fair remuneration,” on which Member States did not agree. It may therefore well be argued that this “interpretation” was a rather far-reaching venture into the area of norm-setting. The third area where the CJEU arguably has stretched the concept of interpretation so that it turned into an activity of norm-setting is the case law concerning the interpretation of the list of optional limitations and exceptions contained in Article 5(2) and (3) of the InfoSoc Directive. This is at least true with regard to those decisions in which the CJEU held that, while Member States might be free to adopt in their national laws any number of the said optional exceptions, if they did so they had to do this in a verbatim way. From a logical point of view, this holding seems rather odd. Rather, since the EU legislation does not oblige Member States to adopt any of these exceptions in their national laws, it would seem logical to conclude that the legislature did in fact allow a Member State to opt for a more limited exception. This holds true if one considers the exceptions and limitations as circumscribed in the InfoSoc Directive as the maximum to which Member States are allowed to cut back existing exclusive rights. Again, the CJEU was obviously guided by a desire to harmonize Member States’ copyright laws further than the EU legislature had been able to achieve.13 Finally, the fourth example of where the CJEU in all likelihood switched from an interpretation mode into a norm-setting mode can be made out in the Court’s attempt to develop its own definition of what constitutes a “communication to the public.” Although judges of the CJEU have claimed that their definition is an authentic interpretation of international copyright conventions,14 this view has been 11

12

13

14

CJEU, cases C-457–460/11, ECLI:EU:C:2013:426, paras. 31, 37, 39, 41, 54, 75 and 78 – VG Wort et al. CJEU, case C-572/13, ECLI:EU:C:2015:750 – Hewlett-Packard Belgium. Following at the national level, for print publishers, see the German Federal Supreme Court (Bundesgerichtshof, BGH) case I ZR 198/13 of 21.04.2016, BGHZ 210, 77 – Verlegeranteil, and for music publishers case I ZR 267/16 of 18.10.2017. CJEU, cases C-467/08, ECR 2010, I-10055, ECLI:EU:C:2010:620 – Padawan; C-510/10, ECLI: EU:C:2012:244 – DR und TV2 Danmark, and again C-201/13, ECLI:EU:C:2014:2132, para. 16 – Deckmyn and Vrijheidsfonds. See, in particular, judge Malenovský’s explanation of the CJEU’s concept of “public communication,” summarized at https://www.create.ac.uk/blog/2016/01/18/the-quest-for-balance-incjeu-copyright-jurisprudence-as-explained-by-judge-malenovsky.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

Copyright and the CJEU: Some Structural Deficits

73

vigorously contradicted by, among others, a former WIPO Assistant Director General.15 As a matter of fact, the interpretation given to the notion of “communication to the public” by the CJEU is not free of contradictions and mysteries. For example, although the CJEU defines a “new public” as a public “that was not already taken into account by the copyright holders when they authorised the initial communication to the public of their work,”16 thus linking the exclusive right to the payment of a contractual remuneration, in some cases the CJEU without further explanation considers hotel guests and visitors to pubs and bars who listen to music or watch TV as a “new public,” although copyright fees have already been paid for the service providing the very same music and/or TV channel at the customers’ homes. It seems that the concept of “new public” morphs between an objective limitation (as in Svensson) and another word for “consent” (as in Soulier and Doke and Renckhoff).17 In addition, the importance of the different factors to be taken into account (most notably that of the “profitmaking nature”18) seems to have changed over time. Moreover, in the cases dealing with the legal nature of different sorts of hyperlinks, the CJEU in GS Media took the subjective element of “pursuit of financial gain” into consideration in order to answer the question whether an objective act of public communication is to be found.19 Finally, the CJEU even went as far as considering the mere provision of physical devices as acts of public communication,20 which as such, was never part of the pre-existing understanding of what public communication meant in both international and national copyright law.

15

16 17

18

19 20

See Dr. Mihály J. Ficsor, “Svensson – Honest Attempt at Establishing Due Balance Concerning the Use of Hyperlinks – Spoiled by the Erroneous ‘New Public’ Theory,” available at http://www .copyrightseesaw.net/uploads/fajlok/Svensson%20(black)%20and%20the%20new%20public% 20theory.doc. See, e.g., CJEU case C-160/15, ECLI:EU:C:2016:644, para. 37 – GS Media. CJEU, cases C-466/12, ECLI:EU:C:2014:76 – Svensson, C-301/15, ECLI:EU:C:2016:878 – Soulier and Doke and C-161/17, ECLI:EU:C:2018:634 – Renckhoff. For critical discussion see, e.g., P. B. Hugenholtz and S. C. Velze, Communication to a New Public? – Three Reasons Why EU Copyright Law Can Do Without a “New Public”, (2016) 47 IIC 797. Beginning with CJEU case C‑135/10, ECLI:EU:C:2012:140, para. 88 – SCF; for additional references see case C-160/15, ECLI:EU:C:2016:644, para. 38 – GS Media. However, the criterion of “profit-making nature” wasn’t even discussed in, e.g., case C-161/17, ECLI:EU: C:2018:634 – Renckhoff. CJEU, case C-160/15, ECLI:EU:C:2016:644, para. 38 – GS Media. CJEU, cases C-162/10, ECLI:EU:C:2012:141 – Phonographic Performance (Ireland) (CD and DVD players); C-527/15, ECLI:EU:C:2017:300 – Stichting Brein (Filmspeler) (multimedia player with pre-installed add-ons containing hyperlinks which upon operation connect to streaming websites operated by third parties). It should be noted, however, that the German Federal Supreme Court, in BGH I ZR 21/14 of 17.12.2015, GRUR 2016, 679 - Königshof, held that free-standing TV sets (as opposed to TV monitors linked to a hotel network) do not amount to an act of public communication.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

74

Thomas Dreier

5.3 a proposal for restructuring the competencies of the cjeu 5.3.1 CJEU Decisions Completing the Lacuna left by the EU Legislature As stated above, the interest here is not to discuss in theory the line of demarcation between legitimate interpretation and illegitimate norm-setting by the CJEU. The reason for not focusing on this dividing line is less that it is difficult to draw anyway, but that in my opinion the core of the problem is not one of possibly stretching interpretation to the point where it results in an activity of norm-setting, nor one of the degree of “harmonization” of Member States’ copyright laws within the EU. Rather, the core of the problem seems to be that the CJEU answers the questions referred to it by interpreting the legislative texts by way of a unified interpretation of legal concepts that have merely been harmonized. As the CJEU itself states: “it follows from the need for uniform application of EU law and from the principle of equality that the terms of a provision of EU law which makes no express reference to the law of the Member States for the purpose of determining its meaning and scope must normally be given an autonomous and uniform interpretation throughout the European Union.”21 Obviously, this understanding of the legal nature of the harmonizing Directives contradicts an understanding of the harmonizing process as leaving the areas that have not been harmonized to the Member States’ legislative discretion. It might be added that the CJEU repeatedly refers to recitals as if these recitals were part of the Directives’ statutory provisions, which, of course, they are not. As a matter of fact, several recitals contain language on issues that were important to some, but not all Member States, and which found their place in the recitals rather than in the articles of the Directive for the very reason that Member States could not agree on them. This creation of insular unification of an otherwise merely harmonized system of copyright law entails a number of problematic consequences for the consistency of the Member States’ national legal systems, where it leads to several disruptions. The first of these disruptions has to do with the fact that quite often the legal questions to be answered in a particular case referred to the CJEU by the courts of an individual Member State are embedded in particular national legal rules which are not to be found in an identical form in the legislation of other Member States. To cite just one example: it is by no means clear whether, and if so to what extent, the CJEU decision in the Soulier and Doke case, which struck down the French system of granting rights with regard to out-of-commerce works via collecting societies,22 also affects the slightly differently organized system set up under German law. Another example might be Hewlett-Packard Belgium (the Reprobel case), which left 21 22

CJEU, case C-201/13, ECLI:EU:C:2014:2132, para. 14 – Deckmyn and Vrijheidsfonds. CJEU, case C-301/15, ECLI:EU:C:2016:878 – Soulier and Doke.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

Copyright and the CJEU: Some Structural Deficits

75

unanswered the question whether the ruling prohibiting a statutory allocation of a part of the fair compensation payable to rightholders to the publishers likewise applies to contractual arrangements of the parties concerned.23 The problem is even more acute whenever the case to be decided by the CJEU not only concerns legal rules already harmonized, but which also concerns or borders to legal rules which have not (yet) been harmonized by EU legislation. A major example is the issue of secondary liability, a concept which, apart from some isolated instances,24 and maybe immunity therefrom under the E-Commerce Directive, has not been harmonized by EU law and which, in addition, is deeply rooted in each Member State’s national civil law regime of secondary liability. Whenever the CJEU decides on such matters – and, in the area of copyright, it will have to do so in view of the referral of several core questions regarding the German system of Störerhaftung25 – any attempt to devise, even by way of interpretation, a new system, will invariably lead to frictions in the different laws of Member States which will have to be overcome in a lengthy and cumbersome discussion and lawmaking effort. The second of these disruptions has to do with the limited number of cases referred to, and hence decided by, the CJEU. On the one hand, as mentioned above, already this limited number of cases decided by the CJEU – which does not have a long-standing tradition of building upon, and citing, prior case law – is not free of internal contradictions. More important, however, is the fact that on the basis of the limited number of cases – on the selection of which the CJEU has no influence since, according to Article 267 TFEU, it can only decide questions that have been referred to it by the national courts of the EU Member States – the CJEU cannot and will never be able to develop a complete copyright system which would satisfy the need for legal security indispensable for both the individual parties concerned and the copyright system as such. The following figures demonstrate that a far greater number of cases than the ones decided by the CJEU on a yearly basis would be required in order to develop a complete copyright system. For example, in 2017, the German Federal Supreme Court handed down around 35 judgments in copyright matters. In addition, in the very same year roughly the same number of copyright decisions was authored

23

24

25

CJEU, case C-572/13, ECLI:EU:C:2015:750 – Hewlett-Packard Belgium. Subsequently, the German Federal Supreme Court (Bundesgerichtshof, BGH) did in fact apply the holding also to an agreement between authors and publishers within a collecting society: case I ZR 198/13 of 21.04.2016, BGHZ 210, 77 – Verlegeranteil. In particular, the protection against acts committed by third parties enabling the unauthorized circumvention of technical protection measures: Articles 6(2) of the InfoSoc Directive 2001/29/ EC and 7(1)(c) of the Computer Program Directive 2009/24/EC. German Federal Supreme Court, BGH I ZR 140/15 of 13.09.2018 – YouTube and I ZR 53/17 of 20.09.2018 – uploaded, registered with the CJEU as cases C-682/18 – Youtube and C-683/18 – Elsevier.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

76

Thomas Dreier

by the Courts of Appeal of the German Länder.26 In other words, to keep the system of legal copyright rules running, some 70 or so major decisions are handed down per year in Germany. For France, the annual figures seem to be even higher, with around 40 cases decided by the Cour de cassation plus an additional estimate of some 200 cases by the Cour d’appel de Paris together with another 100 cases decided by other courts of appeal.27 Of course, such high figures may represent the peak for EU Member States. But it should be noted that likewise for Austria28 and Belgium29 figures are to be reported which are well above those for the copyright decisions handed down in the year 2017 by the CJEU. It may be added that the lack of legal certainty resulting from the limited number of cases decided is even exacerbated by the increasing tendency of the CJEU of resorting to a balance of conflicting fundamental rights in order to arrive at a decision in the case at bar. In order not to be misunderstood, there is nothing wrong with this approach, in case the answer to the questions referred to the CJEU truly hinges upon such a balance. However, an all too far-reaching balancing by the CJEU on the basis of the facts of each individual case has the major disadvantage that the outcome of the balancing activity in the next case handed down by the CJEU is hardly foreseeable.30 In order to remedy this situation, it is suggested that the CJEU – similarly to the German Constitutional Court – should resort to a balancing of its own only in cases when the balance struck by the legislature appears to be arbitrary or at random.

26

27

28

29

30

Results of the Beck-online database (search commands: Rechtsprechung, BGH, UrhG OHNE Anm, 01.01.2017–31.12.2017). For the cases decided between 2008 and 2012, as well as the methodological difficulties of ascertaining the exact numbers, see T. Dreier, Erbsenzählerei oder Erkenntnisgewinn – Anmerkungen zu einer statistischen Rechtswissenschaft am Beispiel des Urheberrechts, in K.-I. Wöhrn, W. Bullinger, E. W. Grunert and C. Ohst (eds.), Festschrift fu¨r Arthur-Axel Wandtke (de Gruyter, 2013), 21 et seq. I am thankful to Professor André Lucas for having communicated these estimates (pointing out that a great number of these cases concerns copyright protection of designs, posing questions in particular of originality and of similarity). According to statistics provided by Professor Christophe Caron, the Cour d’appel de Paris receives around 150 new cases each year. In Austria, between 2012 and 2017, respectively 15, 9, 9, 13, 8, 14 cases have been decided by the Austrian Supreme Court (Oberster Gerichtshof, OGH), i.e., an average of slightly more than 11 cases per year (figures supplied by Professor Michael Walter). In Belgium, in 2017, 122 cases were pending at the level of Court of Appeals, of which 28 had been newly introduced and 26 decided by way of judgment, to which a single-digit number of decisions handed down by the Cour de cassation has to be added (figures provided by Gunther Albrecht, Federal Public Service [Ministry of Economy], Belgium). It has been pointed out that the tendency to answer copyright questions by way of a balancing of fundamental rights might have to do with the fact that judge Malenovský has a background as a constitutional law judge, namely in an area of law which, far more than civil law, is based in its decisions on a balancing of conflicting fundamental rights.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

Copyright and the CJEU: Some Structural Deficits

77

5.3.2 Solving the Problem? What follows from all this? Theoretically, two solutions to the problem of piecemeal decisions by the CJEU with disruptive effects on Member States’ copyright legislation might be envisaged. On the one hand, the number of cases decided by the CJEU might be increased in order to be able to devise a complete copyright system. On the other, the number of cases decided by the CJEU might be decreased in order to reduce the disruptive and negative effects on legal certainty just outlined. As regards an increase in the number of cases to be decided by the CJEU, however, although the number of cases referred to the Court could theoretically be increased by a stricter application of the CILFIT-criteria,31 the CJEU is already hardly able to shoulder the ever-increasing number of cases referred. Unless one would dramatically increase the number of judges and eventually introduce a specialization of the different chambers of the CJEU, increasing the number of cases to be decided by the Court is in all likelihood not a viable option. Moreover, although according to Article 267 TFEU the CJEU does not decide the case at issue as such, but rather merely gives a preliminary ruling on the questions referred to it for clarification, in practice the ruling of the CJEU can be seen as a ruling of a “fourth instance” on top of the three-tier court system as it exists in many of the EU Member States (entry instance/première instance, appeal instance/Cour d’appel and the revision/cassation instance). In Member States such as Germany, where the decision by the highest national court is subject to constitutional review by that Member State’s Constitutional Court, a fifth court – and, in cases where the decision handed down by the CJEU requires facts to be ascertained which so far had not been considered to be relevant by the national courts, eventually a sixth court to which the case is remanded by the Member State’s highest court, plus another round at the Member State’s highest court32 – has to sit and have a say in the outcome of the matter. Hence, to be realistic, only a reduction in the number of cases to be decided by the CJEU appears to be a viable option in order to avoid the negative consequences 31

32

ECJ, case 283/81, ECR 1982 at 3415, ECLI:EU:C:1982:335 – CILFIT. In Germany, the Constitutional Court (BVerfG) has indeed considered a non-referral by the Federal Supreme Court as withholding the right to the lawful judge as guaranteed by the German constitution; see the German Constitutional Court (Bundesverfassungsgericht, BVerfG 1 BvR 1631/08 of 30.08.2010, GRUR 2010, 999 – Drucker und Plotter. Indeed, in a case concerning the legitimacy of sampling under the phonogram producers’ exclusive rights, which set German band Kraftwerk in opposition to producer Moses Pelham and singer Sabrina Setlur, the start of the litigation, which concerns a song produced by the defendants in 1997, dates back to 1999. After decisions by the entry level court (2004), the Court of Appeal (2007), the Federal Supreme Court (2008), and upon remand, again the Court of Appeal (2011), the Federal Supreme Court (2012) and the Constitutional Court (2016), the Federal Supreme Court referring its questions to the CJEU (2017), the CJEU handing down the answers (2019) and the Federal Supreme Court remanding the case to the Court of Appeal (2020), the final decision is still awaited – more than twenty (!) years on.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

78

Thomas Dreier

outlined above.33 If the hypothesis just presented – that the CJEU cannot, and will not be able to, build a complete and, more importantly, a coherent system of copyright law on the basis of the few cases decided – is true, this leads to a twofold proposal. First, the CJEU is probably well advised to limit itself to the activity of merely interpreting existing harmonizing legislation and refrain from trying to fill in the gaps left by the existing Directives in a unified way. In particular, in view of the limited number of cases – which in addition depend on the cases that are referred to the CJEU by the courts of the EU Member States – the CJEU should not attempt to develop a complete and unified legal system. The CJEU’s decisions will always remain piecemeal and the cases referred to some extent arbitrary. Moreover, in my opinion the CJEU should be more attentive to the fact that the answers to the questions referred to the Court often arise out of a particular legal situation in the referring state, especially in cases which are linked to other, unharmonized areas of law, and therefore have effects on the coherence of other Member States’ legal systems as well. Second, if it is true that the CJEU will not be able to meet the expectation of the Member States as regards the development of a fully fledged, complete and noncontradictory copyright system, one should abandon the idea of expecting the completion of such a task from the CJEU. The CJEU is not, and for the time being will never be, a court which is supreme regarding all EU copyright matters. Rather, it might be useful to compare the CJEU to the US Supreme Court. Certainly, the CJEU is positioned in addition to the national courts of the national Member States’ three-tier court system, whereas in the USA, the US Supreme Court is the proper third instance of the Federal Court System. However, it might be appropriate to functionally compare the CJEU to the US Supreme Court in view of their respective role with regard to a population which is roughly comparable in size (some 500 million inhabitants in the EU and some 325 million in the USA), as well as in view of the fact that both the EU and the USA are defined as a single internal market. In spite of these similarities, the average number of copyright cases handed down by the US Supreme Court is much lower – it rarely takes more than one copyright case a year, and there are years in which it takes no copyright cases at all34 – than the number of cases decided by the CJEU on an annual basis. It may well be surprising that even if one takes into consideration that litigation is a good deal more expensive in the USA than in the EU, and that the US Supreme Court already operates in a largely harmonized system, the USA can apparently live much better with legal discrepancies between the jurisprudence of their Federal Courts of Appeals, most notably between the Court of Appeal of the first (New York) and of 33

34

G. Dinwoodie, Trade Mark Harmonisation – National Courts and the European Court of Justice (2010) 41 IIC 1, reaches a similar conclusion with regard to trademark cases. Information provided by Professor Jane Ginsburg from Columbia University School of Law.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

Copyright and the CJEU: Some Structural Deficits

79

the tenth circuit (California), than the EU. This becomes particularly apparent with regard to the “fair use” defense, where a very limited number of cases handed down by the Supreme Court have to be seen against a backdrop of a great and not always consistent number of cases decided by the lower courts. These thoughts – and I am sure there are several more – demonstrate that total unification of the details of the copyright system in a federal system is not a necessary condition for the copyright system to properly function and satisfy the needs of those affected by it. Hence, because the CJEU should not aim at such unification, a reduction in the number of cases to be decided would not be against the idea of the Single Market. Moreover, another lesson that can be learned from the US example is that if it might well support a unitary EU copyright title, it would still point to less intervention by the CJEU.

5.4 conclusion How could a reduction in the number of cases to be decided by the CJEU be achieved? In theory, two possibilities are available. There could be a reduction in either the number of cases referred to the CJEU by national courts, and/or the number of cases decided by the CJEU. As regards the number of cases that are referred by the national courts to the CJEU, the framework is set by the current wording of Article 267(2) and (3) TFEU. Whereas national courts against whose decisions there still is a judicial remedy under national law are not forced to make a referral, a court of a Member State against whose decisions there is no judicial remedy under national law is under an obligation to bring the matter before the CJEU, “unless it has established that the question raised is irrelevant or that the community provision in question has already been interpreted by the court of justice or that the correct application of community law is so obvious as to leave no scope for any reasonable doubt,” as has been specified by the CJEU in its CILFIT decision.35 Arguably, already de lege lata the criterion of “reasonable doubt” leaves the national courts of the Member States a sufficient degree of flexibility in order to make sure that the number of cases referred to the CJEU does not get out of hand. If the national courts prove to be too reluctant to refer questions to the CJEU, one might consider, de lege ferenda, introducing the possibility of an appeal to the CJEU against the national court’s decision not to refer the case, as is found in numerous national legislations. As regards the number of cases to be decided by the CJEU, a reform would have to redraft the existing procedure laid down in Article 267 TFEU in view of the functional similarity of the CJEU with the US Supreme Court discussed above. Contrary to the CJEU, which presently is under an obligation to decide each question referred to it as long as the reference is admissible, which gives the 35

ECJ, case 283/81, ECR 1982 at 3415, ECLI:EU:C:1982:335 – CILFIT.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

80

Thomas Dreier

CJEU at least some leeway to reject references, the US Supreme Court has the power to refuse to sit on a particular case (to “deny certiorari”). This enables the US Supreme Court to discard cases of minor importance as well as not to take cases which are based either on exceptional facts or on facts which do not allow the US Supreme Court to decide a certain issue “in one go.” The flexibility which therefore exists for the US Supreme Court would, if applied to the CJEU, have a number of beneficial effects.36 To cite just one example: whether or not an Italian dental practice pays to the national collective management organization a remuneration for the playing of music to the limited number of its clients in the waiting area of the practice is a matter of hardly any concern to the European copyright system. Of course, an effect on the Single Market can always be construed by pointing to the fact that the music played was not only by national authors but also by authors in other EU Member States. However, it is apparent that this is only a weak link which can well be disregarded. In addition, the facts of the case were so limited that, while giving an answer to the particular set of facts of the case at bar, the CJEU opened up a Pandora’s box of questions on how to treat other, similar yet different cases, such as the playing of music in a rehabilitation center (both in the waiting rooms and a training room),37 in spa establishments (television and radio to patients in their rooms)38 and in hotels (retransmission to hotel rooms),39 to list just those cases already decided by the CJEU. Similar questions pose themselves with regard to hospitals (are reception areas to be treated as hotel lobbies, waiting areas for patients as dental practices, and hospital rooms like hotel rooms?). Apart from disrupting well-established licensing practices in all EU Member States, and the waste of resources in court battles that could well be avoided, the decisions of the CJEU are a disaster in terms of transaction cost: after the SCF decision all existing contracts had to be renegotiated and eventually litigated in each single EU Member State. Moreover, as demonstrated by the high number of later cases on the notion of “communication to the public,” the facts of the SCF case were so special that the CJEU was not able to fully flesh out its newly designed concept of public communication. For example, the importance of the factor of “profit-making nature” – which played a decisive role as regards the outcome of the SCF decision, but was not even mentioned in later cases before resurfacing as “pursuit of financial gain” in the GS Media case – is still unclear.

36

37 38 39

The introduction of a certiorari for the CJEU has already been proposed, some decades ago: see T. Kennedy, First Steps Towards a European Certiorari? (1993) 18 European Law Review 121. CJEU, case C-117/15, ECLI:EU:C:2016:379 – Reha Training (public communication). CJEU, case C-351/12, ECLI:EU:C:2014:110 – OSA (public communication). CJEU, case C-641/15, ECLI:EU:C:2017:131 – Verwertungsgesellschaft Rundfunk (no public communication with regard to broadcasting organizations, since the price of a hotel room is not, like the price of a restaurant service, an entrance fee specifically requested in return for a communication to the public).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

Copyright and the CJEU: Some Structural Deficits

81

The Court would indeed have been in a much better position had it been able to dismiss the SCF case and wait for another case whose facts would have enabled it to define the notion of communication to the public once and for all. Similar remarks can be made regarding the decisions handed down by the CJEU in the cases of different ways of linking as well as even more generally in cases concerning indirect communication. The “great triad” of the GS Media, Stichting Brein (Filmspeler) and Stichting Brein (Pirate Bay) cases all concerned highly untypical cases of bad intermediaries who acted with the full intention of allowing copyright infringement, but they are treated as starting points for a more general theory of intermediary liability.40 Of course, all of this might be studied in more detail and in greater depth. However, the sketchy remarks regarding the SCF case and its successors lead to the conclusion that it might be a good idea to empower the CJEU – similarly to the US Supreme Court – to dismiss cases referred to it by national courts of the EU, if the CJEU is of the opinion that the case is not sufficiently important for the further development of EU copyright harmonization, and/or if the Court considers the facts of the case referred to it to be unsuitable for the clarification of a broader systematic area. Of course, so far, this proposal is only based on the analysis of the CJEU’s jurisprudence in the area of copyright, as seen from the perspective of the German understanding of what constitutes a systematic approach towards the legal system. From this, it follows that it has to be discovered whether the CJEU’s practice of arriving at decisions gives rise to the same, or similar, concerns in other areas of IP law, and ultimately in other areas of law as well. In addition, the view from the perspective of other Member States would have to be ascertained.

40

CJEU, cases C-160/15, ECLI:EU:C:2016:644 – GS Media, C-527/15, ECLI:EU:C:2017:300 – Stichting Brein (Filmspeler) and C-610/15, ECLI:EU:C:2017:456 – Stichting Brein (Pirate Bay).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.010

6 Global Intellectual Property: Transition and Coherence Through Rules of Interpretation Susy Frankel1

The transition of intellectual property into the World Trade Organization (WTO) trade regime, through the vehicle of the TRIPS Agreement,2 not only took intellectual property protection to a new level of protection and with greater reach globally, it also placed intellectual property in a forum where it would be subject to binding dispute resolution. As part of that process, the Vienna Convention on the Law of Treaties (Vienna Convention)3 would be applied to the interpretation of the TRIPS Agreement provisions in disputes brought to the WTO.4 The Vienna Convention applies to the interpretation of all treaties (and other treaty-related matters) both within and outside the WTO dispute settlement context. Therefore, even prior to the conclusion of the TRIPS Agreement and the formation of the WTO in 1995, the Vienna Convention should have been the mechanism through which international treaties about intellectual property were interpreted by members of other international organisations, particularly the World Intellectual Property Organization (WIPO), and also by national and regional courts and policymakers. However, up until the formation of the WTO the Vienna Convention was rarely discussed in connection with intellectual property, even in the academic literature. This may in part have been because the United States has never ratified the actual Vienna Convention, but it has explicitly agreed to apply the Vienna Convention rules to the WTO Agreements.5 Also, many interpreters choose to discuss the drafting 1

2

3

4

5

FRSNZ, Professor of Law, Chair in Intellectual Property and International Trade, Director of the New Zealand Centre of International Economic Law, Victoria University of Wellington. Agreement on Trade-Related Aspects of Intellectual Property Rights (opened for signature 15 April 1994, entered into force 1 January 1996) 1869 UNTS 299 (TRIPS Agreement). Vienna Convention on the Law of Treaties (opened for signature 23 May 1969, entered into force 27 January 1980) 1155 UNTS 331 (VCLT). Understanding on Rules and Procedures Governing the Settlement of Disputes (opened for signature 15 April 1994), Marrakesh Agreement Establishing the World Trade Organization, Annex 2 (1994) 1869 UNTS 401, (DSU), Art. 3.2. Ibid.

Downloaded from https://www.cambridge.org/core. 85 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

86

Susy Frankel

history of the TRIPS Agreement (and other treaties such as the Berne Convention) ahead of using the Vienna Convention rules, which relegate that drafting history as supplementary material to be turned to only after an analysis of ordinary meaning of the treaty provision in context and in light of the treaty’s object and purpose.6 This fundamental rule exists because a treaty should be interpreted to reflect the intentions of all of the parties, whereas drafting history often only reflects some parties’ or one party’s intentions. Particularly where the draft from which the negotiations developed was prepared by a hegemonic trading nation or region, it cannot be relied on as a source of all of the parties’ intentions.7 This tendency to rely on other methods of treaty interpretation is perhaps, in part, explained by the way in which intellectual property has been treated as a specialised area of international law, and thus operated in remarkable isolation from much international law for a long time.

6.1 more extensive protection and counter-norm development The transition of intellectual property to the trade regime not only exposed intellectual property directly to trade rules and theory, but also provided a new pathway for other international regimes to interact with intellectual property institutions and to provide differing perspectives on norm making. This interaction has emerged in at least two distinct but related forms. Some interactions have eventually resulted in the making of new norms, such as those found in the Convention on Biological Diversity (CBD)8 and its Nagoya Protocol9 regarding the relationship between indigenous peoples’ knowledge (including their intellectual property) and access to and benefit sharing of genetic resources. Other interactions with intellectual property have not formed norms, but are in the nature of monitoring and sometimes challenging norms, such as the World Health Organization’s comments about patents and their impact on health.10 Reactions to intellectual property law arise not just as a response to the effects of the TRIPS Agreement, on matters such as public health and educational resources, but also to the increasing overall protection function of post-TRIPS free trade agreements (FTAs), which have progressively 6 7

8

9

10

VCLT, Art. 32. Susy Frankel, Test Tubes for Global Intellectual Property Issues (Cambridge University Press, 2015), 65–66. Rio Convention on Biological Diversity 1760 UNTS 79; 31 ILM 818 (opened for signature 5 June 1992, entered into force 29 December 1993) (CBD). The Nagoya Protocol is a supplementary agreement to the CBD: Nagoya Protocol on Access to Genetic Resources and the Fair and Equitable Sharing of Benefits Arising from their Utilization to the Convention on Biological Diversity (UN Doc. UNEP/CBD/COP/DEC/X/ 1, adopted on 29 October 2010). According to the CBD website, ‘[i]ts objective is the fair and equitable sharing of benefits arising from the utilization of genetic resources, thereby contributing to the conservation and sustainable use of biodiversity’. See WHO, Public Health, Innovation, Intellectual Property and Trade, https://www.who.int/ phi/implementation/ip_trade/en/.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

Global IP and Rules of Interpretation

87

developed and increased the levels of intellectual property protection. As Annette Kur has said:11 The picture has thoroughly changed in the post-TRIPS world. First and most importantly, TRIPS has resulted in an unprecedented level of substantive harmonisation of IP law. International IP began to “reach behind the borders” by regulating matters that were previously considered as presenting primarily domestic concerns.

While TRIPS brought protection to an unprecedented level, even then there was not complete harmonisation of most norms; rather a remarkable degree of flexibility remained. Although initially many commentators seemed to underappreciate the many flexibilities, focusing on the outcomes as harmonisation or supra-national norms, others (including some national legislatures) saw the potential of the flexibilities.12 Flexibility in TRIPS was perhaps more evident in some areas than in others. The subject matter of patent law, where key terms, including “invention” and “inventive step”, were undefined at international level allowing different countries to create their own parameters and definitions for those terms,13 can be contrasted with the field of trade marks where the subject matter is well-defined and the scope for subject matter differences is relatively more, but still not completely, prescribed.14 The very point of post-TRIPS bilateral and plurilateral trade agreements has been to limit, or even to close, those flexibilities. Some such agreements, for example, require that patents be available for second uses of known compounds, thus tightening the application of inventive step and expanding subject matter of patents.15 Some hoped that the move of intellectual property to regimes other than trade may have enabled the development of norms that could temper intellectual

11

12

13 14

15

Hanns Ullrich, Reto M. Hilty, Matthias Lamping and Josef Drexl (eds.), TRIPS plus 20: From Trade Rules to Market Principles (Springer, 2016), 135. As one commentator has noted, for the large economies of the developed world TRIPS was never enough: Susan K. Sell, TRIPS Was Never Enough: Vertical Forum Shifting, FTAS, ACTA, and TTP (2011) 18 J. Intell. Prop. L. 447. Another commentator describes the use of these flexibilities as “legal innovations”: see Ruth L. Okediji, Legal Innovation in International Intellectual Property Relations: Revisiting Twenty-One Years of the TRIPS Agreement (2014) 36 U. Pa. J. Int’l L. 191. TRIPS, Art. 27.1. TRIPS, Art. 15. That is not to suggest that differences, for example in the scope of rights relating to matters broadly called unfair competition, do not exist among states. See United States–Mexico–Canada Agreement, Can.-Mex.-U.S. (signed 30 November 2018, not yet in force) (USMCA), https://ustr.gov/trade-agreements/free-trade-agreements/unitedstates-mexico-canada-agreement, Art. 20.36 (2); and Trans Pacific Partnership (TPP), Art. 18.37(2) (now suspended), in the Comprehensive and Progressive Trans Pacific Partnership (open for signature 8 March 2018, entered into force 30 December 2018) (CPTPP). For an overview of the suspended provisions, see New Zealand Ministry of Foreign Affairs and Trade, CPTPP vs TPP, https://www.mfat.govt.nz/en/trade/free-tradeagreements/free-trade-agreements-in-force/cptpp/understanding-cptpp/tpp-and-cptpp-the-dif ferences-explained.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

88

Susy Frankel

property’s TRIPS-related (and post-TRIPS) excesses.16 These moves have not been as successful as perhaps they might have been, had those other regimes had the capacity to provide binding and enforceable counter-norms to the evolving web of FTAs creating more extensive intellectual property rights. This could have occurred in two different ways. First, through deeper institutional engagement in policy development; and, second, through dispute settlement, which should involve consideration of other relevant norms as part of a Vienna Convention analysis.17 But this has not happened. For example, the CBD needed the Nagoya Protocol to operationalise its aspirational provision regarding indigenous peoples’ rights.18 Yet that Protocol remains optional and, in any event, the United States does not belong to the CBD, meaning that it has consistently vetoed any detailed discussion of the CBD’s relationship with the TRIPS Agreement in the WTO’s TRIPS Council.19 Although there has been much discussion about the intersection of human rights and intellectual property, there are few changes to the international regime that have led to binding norms which address that intersection. One development in that regard was the amendment of the TRIPS Agreement to allow for the export of pharmaceuticals, made under a compulsory licence, to meet the needs of the importing country in a health-related emergency.20 That amendment represents one of the few successful outcomes of the post-WTO Doha Round negotiation process. However, the complexity involved in complying with its detail raises questions about its ultimate impact and utility.21 Another change in response to human rights claims is the Marrakesh Treaty for rights for the visually impaired.22 At face value, it should perhaps be more straightforward to provide rights of access to books for the visually impaired, as to do otherwise is a form of discrimination.23 But even that treaty was extremely hard-fought for.

16

17

18 19

20

21

22

23

Laurence R. Helfer, Regime Shifting: The TRIPs Agreement and New Dynamics of International Intellectual Property Lawmaking (2004) 29 Yale J. Int’l L. 1. Susy Frankel, The WTO’s Application of ‘the Customary Rules of Interpretation of Public International Law’ to Intellectual Property (2006) 46 Va. J. Int’l L. 365. CBD, note 8 above, Art. 8(j). WTO, Nagoya Gives New Context to Old Views in Intellectual Property Council, 1 March 2011, https://www.wto.org/english/news_e/news11_e/trip_01mar11_e.htm. TRIPS Agreement, Art. 31 bis. See WTO Analytical Index, https://www.wto.org/english/res_e/ publications_e/ai17_e/trips_art31_bis_oth.pdf. Donald Harris, TRIPS After Fifteen Years: Success or Failure, as Measured by Compulsory Licensing (2011) 18 J. Intell. Prop. L. 367. The Marrakesh Treaty to Facilitate Access to Published Works for Persons who are Blind, Visually Impaired, or Otherwise Print Disabled (opened for signature 27 June 2013, entered into force 30 September 2016) WIPO Doc VIP/DC/8. See Lida Ayoubi, Copyright Pluralism and Human Rights of Visually Impaired Persons, in Susy Frankel (ed.), Is Intellectual Property Pluralism Functional? (Edward Elgar, 2019).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

Global IP and Rules of Interpretation

89

6.2 international reform and experimentation in national regimes Reforms may also be entrenched in the international system through successive FTAs before they are tested at national level. The utility of testing laws at the national level is possibly more relevant for intellectual property than for other international regimes.24 However, in some circumstances an international response may, of necessity, precede detailed norms or experiments in national regimes. One of the challenges of, for example, the WIPO IGC negotiations25 for the protection of traditional knowledge, traditional cultural expressions and genetic resources is the relatively few national regimes on which to draw in order to develop treaty norms.26 It is precisely because of the absence of a sufficient number of national regimes to create global norms that the problems of a lack of protection and misappropriation and misuse of traditional knowledge have become acute, and, therefore, an international response may be necessary in order to grow an effective level of minimum standards in national regimes. To be sure, this makes the process of international negotiations more challenging, not only because of the paucity of national experience in effective protection, but also because those who oppose national protection do so at the international level too.27 At the heart of the calls for the protection of traditional knowledge and the use of a treaty-making process is acknowledgement that the current intellectual property regime recognises only some aspects of innovation and only the intellectual contributions to knowledge of some cultures, but not others. Put differently, the object and purpose of international intellectual property speaks to innovation, but only embodies and provides minimum standards for a partial understanding of global innovation. In spite of the merits, or otherwise, of experimenting with regulation at national level, what works nationally or even regionally is not appropriate for all countries or regions because of their stage of development, type of industry or even location. The objectives of national laws do not automatically equate to objectives for all nations. As Annette Kur has closely examined, one size does not fit all.28 One consequence 24

25 26

27

28

For discussion about the value of national experiments in international copyright law-making, see Graeme B. Dinwoodie, A New Copyright Order: Why National Courts Should Create Global Norms (2000) 149 U. Pa. L. Rev. 469, 514–515. See WIPO, Intergovernmental Committee (IGC), https://www.wipo.int/tk/en/igc/. Although as the negotiations continue more and more national and regional regimes have emerged, such as Traditional Knowledge Act 2013 (Cook Islands) and the Swakopmund Protocol of the African Regional Intellectual Property Organisation (signed 9 August 2010, entered into force 11 May 2015). For discussion of the United States’ opposition to the protection of traditional knowledge, see Susy Frankel, ‘Ka Mate Ka Mate’ and the Protection of Traditional Knowledge, in Rochelle Dreyfuss and Jane Ginsburg (eds.), Intellectual Property at the Edge (Cambridge University Press, 2014), 193–214. Annette Kur and Vytautas Mizaras (eds.), Structure of Intellectual Property Law: Can One Size Fit All? (Edward Elgar, 2011).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

90

Susy Frankel

of this is that when norms are internationalised their rationales (including their object and purpose), which are based on a particular locality and its economic conditions, are lost. Further, the connections to the corpus of intellectual property law (as innovation incentives), beyond the motivations for individual laws, become remote. One problem that Rochelle Dreyfuss and I have identified is that innovation incentives are gradually eroded.29 Another is that the overall framework of the object and purpose of intellectual property, as part of a global trade regime, is not often discussed in any real way that reflects relevant trade principles, such as comparative advantage and removing barriers to trade. Rather, single-interest-based issues are lobbied for in ways that are tantamount to protectionism and erecting barriers to trade.30 This is particularly pernicious when it is done in a framework that purports to be an extension of a trade-related regime. In part, this occurs because of the framework problem that allows increased levels of protection without any enforceable check on whether those protections remain consistent with the overall object and purpose of intellectual property. Articles 7 and 8, which are structural provisions of the TRIPS Agreement, could be used to have “maximum rules”,31 that is, to limit increased protection: providing higher levels of protection ought to have limits consistent with the wording of the TRIPS Agreement, in light of its object and purpose, its structure and the benefits that can be expected from it. Also, the ability to provide higher levels of protection does not equate to an unchecked right to coerce higher levels of protection through FTAs. Rather, the optional nature of the ability to increase protection is itself a purpose of the TRIPS Agreement that should be maintained.32

6.3 developing norms multilaterally and the role of the vienna convention When new norms of intellectual property are developed multilaterally, competing interests result in considerable debate. Any compromise that is ultimately reached likely has some kind of relationship to the competing goals of those involved in the system. The addition of Article 31 bis of TRIPS, which allowed export of 29

30

31

32

Rochelle C. Dreyfuss and Susy Frankel, From Incentive to Commodity to Asset: How International Law is Reconceptualizing Intellectual Property (2015) 36 Mich. J. Int’l. L. 557. As intellectual property can be used to stop the free movement of goods it is inherently a trade barrier, even if in part an internationally accepted type of barrier. Annette Kur, International Norm-Making in the Field of Intellectual Property: A Shift towards Maximum Rules? (2009) 1 WIPO J. 27. Susy Frankel, Challenging TRIPS-Plus Agreements: The Potential Utility of Non-Violation Disputes (2009) 12 JIEL 1023, 1030. The structural provisions of the TRIPS Agreement are those found in section 1 “General Provisions”, including non-discrimination (national treatment and MFN), and Articles 7 and 8, respectively objectives and principles of the agreement. See ibid.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

Global IP and Rules of Interpretation

91

pharmaceuticals made under a compulsory licence to certain countries that do not have the manufacturing capacity to benefit from compulsory licensing themselves, is an example of a detailed compromise. It also took into account the interests of producers and users, as Article 7 of the TRIPS Agreement indicates intellectual property should do. Thus, the resulting pharmaceuticals must be identifiable in ways that mean they can be distinguished from the same pharmaceuticals made for other markets. These sorts of compromises, however, do not always result in achieving any real change in practice. The burdens of labelling the pharmaceuticals and other administrative requirements have meant that the use of Article 31 bis is overly constrained, and indications are that the mechanism developed will not be used a second time by some, and, indeed, a first time by others.33 The role of the Vienna Convention in this sort of negotiated compromise is at a practical level non-existent. The engagement of intellectual property law with the Vienna Convention has arisen largely in response to its application as part of WTO dispute settlement. Yet, one might easily conclude that testing the design of a treaty against the principles by which it would be interpreted would be a worthwhile exercise for policy and law makers. In addition to the benefit of analysing whether the intention of the parties is expressed effectively, awareness of the rules of interpretation can also improve the understanding that international intellectual property norms are part of a greater body of international law. Further, the role of object and purpose could be used to analyse the relationship between negotiated norms and the rationales for them. This approach could result in the structuring of negotiations in ways other than as a compromise reached by a wrangle between competing interests. Consideration of the Vienna Convention key rules might also save the intellectual property community from creating treaties with unusual and controversial frameworks, such as the WIPO Copyright Treaty.34 That treaty includes agreed statements about the interpretation of certain articles in it.35 The agreed statements cannot be read as the terms of the treaty because they are not. If they were, they would be Articles. The agreed statements in Vienna Convention terms might be context, as Vienna provides that context includes “any agreement relating to the treaty which was made between all the parties in connection with the conclusion of the treaty”.36 This means they do not have the weight in international law that some may hope they do. While different parties will continue to argue about the meaning of the agreed statements, an understanding of their status in Vienna Convention terms may have led to something more usable and reflective of general international law treaty practices. 33 34

35 36

Harris, note 21 above. WIPO Copyright Treaty (opened for signature 20 December 1996, entered into force 6 March 2002) 36 ILM 65 (WCT). Ibid.; the agreed statements appear at the end of the WCT. VCLT, Art. 31.2(a).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

92

Susy Frankel

6.4 the coherence of difference The plethora of rights, and variations among them, that exist around the world is perhaps the inevitable consequence of a framework of minimum standards which allows for differences. As noted above, some provisions of TRIPS enable more flexibility than others. For example, while there may be little flexibility in the requirement that certain things are protected as trade marks, there are many different interpretations of what trade mark protection for shapes should look like and how protective it should be, given the functional aspects of shape.37 Differences in laws are not necessarily without a coherent underpinning. Being able to calibrate laws to meet local needs is part of the framework of TRIPS and, therefore, part of its object and purpose. Differences in patentability criteria, for example, represent different tools for countries to adjust their laws to reflect changes in policy and even comparative advantage, at particular points in time. The change in Canada’s utility doctrine to require a credible promise could arguably have been developed as an appropriate response to the evergreening of patents that resulted in limited competition in Canada in parts of the pharmaceuticals market. Equally, Canada’s changes to that law, subsequent to the investment dispute that had targeted those common law developments, could be looked at as a response to unintended effects of the law.38 One of the features of international intellectual property norms is that they are minimum standards and that the framework provides the ability for countries to provide more extensive protection.39 Claims about innovation incentives are utilised in the negotiation of various FTAs in which more extensive intellectual property protection has developed. While that is not surprising, the dialogue of innovation incentives is often used without adequate evidence to support such claims. In doing so, the notion of more intellectual property protection has been treated as a substitute for the comparative advantage approach that frames the debate about trade in goods and services. Comparative advantage in trade functions differently from intellectual property and their relationship is continuously evolving. Treaty interpretation is a method of analysis which can bring together trade and intellectual property-related purposes. As I have said elsewhere: Another difference between GATT and GATS, and the TRIPS Agreement is that the trade benefit of lowering tariffs is a philosophy that is clear in the GATT context. Any reduction in tariffs or other trade barriers is ‘good’ for international trade . . . The TRIPS Agreement is not a tariff lowering agreement but a minimum 37

38

39

Annette Kur, Too Common, Too Splendid, or ‘Just Right’? Trade Mark Protection for Product Shapes in the Light of CJEU Case Law, Max Planck Institute for Innovation & Competition Research Paper No. 14-17, 3 November 2014, http://dx.doi.org/10.2139/ssrn.2526124. Rochelle Dreyfuss and Susy Frankel, Reconceptualizing ISDS: When Is IP an Investment and How Much Can States Regulate It? (2018) 21 Vand. J. Ent. & Tech. L. 377, 385. TRIPS, Art. 1.1.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

Global IP and Rules of Interpretation

93

standard of protection. The lowering of barriers in one aspect of international trade does not, of necessity, require a corresponding increase in intellectual property protection. While the limit for lowering barriers might be complete removal, this alone is not a justification that can be used to suggest that the extent of intellectual property protection should be limitless.40

If the Vienna Convention is used to produce an agreed understanding of the object and purpose of intellectual property, then, among other matters, a more effective application of the objectives and principles in Articles 7 and 8 of TRIPS could emerge. In addition, this also could create a way to address the cumulative effect of norm pushing in bilateral FTAs.41 Instead, in the few TRIPS disputes that have been adjudicated at the WTO, object and purpose has not been effectively analysed. Indeed, it has mostly been sidelined. One consequence of a lack of analysis of intellectual property’s rationales as part of the objectives of TRIPS is that where domestic policy drivers are relevant to the challenged law, that policy has not even been treated as relevant. This has been noted by commentators discussing both the panel decision on the exception in US copyright law, permitting playing of music in certain retail establishments,42 and that about Canada’s approach to stockpiling generic medicines.43 The refusal of panels to engage with the underlying domestic purpose ignores that TRIPS makes national policy imperatives the basis for creation of specific limitations and exceptions through the three-step test for limitations and exceptions in copyright and patent.44 Also, the panel reports do not refer to the TRIPS Agreement preamble (in Vienna Convention terms, context45), which notes that the parties enter the agreement “Recognizing the underlying public policy objectives of national systems for the protection of intellectual property, including developmental and technological objectives”.

6.5 when flexibility can result in protectionism and incoherence The framework of minimum standards enables international norms to be complied with in different ways. The United States and the EU use FTAs to limit these 40 41 42

43

44

45

Frankel, note 17 above, 376. Frankel, note 32 above. WTO Panel Report, United States – Section 110(5) of the US Copyright Act, WT/DS160/R (15 June 2000). See Graeme B. Dinwoodie and Rochelle C. Dreyfuss, A Neofederalist Vision of TRIPS: The Resilience of the International Intellectual Property Regime (Oxford University Press, 2012), 117–118. WTO Panel Report, Canada – Patent Protection of Pharmaceutical Products, WT/DS114/R (17 March 2000). See, e.g., Jane C. Ginsburg, Toward Supranational Copyright Law? The WTO Panel Decision and the “Three-Step Test” for Copyright Exceptions (2001) 187 Revue Internationale du Droit d’Auteur 3, 5 (noting that the Panel “declined to undertake the evaluation of public policy”). VCLT, Art. 31(2)(a).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

94

Susy Frankel

differences. This approach can create problems if it is used as a form of protectionism rather than to enhance innovation. For example, while patent term or copyright extension may be a good fit in one economy, it does not follow that such extensions of term are a good fit in places that import the products subject to those extensions. The profit made in foreign markets is rarely, if ever, needed to recoup investment in the place of innovation, or to prevent any unfair competition. In such situations the rationale for extending term of protection is to maintain profits abroad. It may be inevitable that countries accept a certain level of protection which is out of step with their needs and level of development.46 However, that is somewhat different from an overall approach that seeks harmonisation regardless of economic differences. That sort of harmonisation can create incoherence in innovation policy. While flexibility can allow for appropriate adjustments to intellectual property law and policy, it can also result in incoherence. Such incoherence is demonstrable in the EU development to allow for export of medicines, including biosimilars (the generic versions of biological pharmaceuticals), during the time that the patented versions are subject to supplementary protection certificates (SPC), which are term extensions for certain pharmaceutical patents.47 The EU regulation allows for making and storing otherwise infringing medicinal products for the purpose of export.48 This waiver allows EU generics to compete in the United States and other foreign markets even before they can compete in the EU market. Because the EU is bound by the TRIPS Agreement, this waiver must meet the requirements of the three-step test found in Article 30.49 Ironically, the restrictive interpretation of that test is a result of the EU’s own policy pursued at the WTO through dispute settlement.50 One might say the EU has changed its mind over time and has recalibrated its more extensive protection regime according to its regional needs. But this is not persuasive, as the earlier EU policy, in framing the parameters of the three-step test at the WTO, is not something that the EU argued reflected its policy at the time of the dispute. Rather, the EU’s approach in the dispute was a direct attack on Canada’s use of patent law to make generic pharmaceuticals available earlier, especially in developing country destinations. The EU policy is, from an intellectual property perspective, somewhat contradictory. While protection of its own generic 46 47

48

49

50

See discussion in Frankel, note 7 above, chapter 8. For an explanation of the EU system, see Zaneta Pacud, Patents, SPC and Data Exclusivity at the Service of Legal Protection for Pharmaceuticals, in Susy Frankel (ed.), The Object and Purpose of Intellectual Property (Edward Elgar, 2019). Proposal for a Regulation of the European Parliament and of the Council amending Regulation (EC) No. 469/2009 concerning the supplementary protection certificate for medicinal products, https://data.consilium.europa.eu/doc/document/ST-6638-2019-INIT/en/pdf. SPCs are a more extensive form of protection than the requirement of patent term under the TRIPS Agreement, Art. 33. Extension is therefore permitted under the TRIPS Agreement, Art. 1.1, but is within the scope of TRIPS and so its provisions apply. See also Max Planck Institute for Innovation and Competition, Study on the Legal Aspects of Supplementary Protection Certificates in the EU – Final Report (European Union, 2018), 3.2. See Panel Reports (notes 42 and 43 above).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

Global IP and Rules of Interpretation

95

pharmaceutical industries explains the change in the EU’s approach, this reflects protectionism rather than the object and purpose of international intellectual property. In this chapter, I do not suggest that greater use of rules of interpretation is a panacea to the troubles of intellectual property. One mechanism cannot fix all problems. However, a theoretically consistent underpinning to the purpose of norm development should inevitably turn the interpreter’s eye to object and purpose. In other words, the interpreter asks ‘why do we have the rule and what should it achieve?’ Absent this mechanism, incoherence is not only likely but will continue to enable hegemonic power rather than rules-based systems to govern the international intellectual property landscape. As Annette Kur would have it, minimum standards need maximum rules.51 Short of a moratorium on increased standards or those standards backfiring on those who created them,52 insistence on adherence to object and purpose, even a trade-related one, can achieve some coherence.

51 52

Kur, note 31 above. Susy Frankel, It’s Raining Carrots: The Trajectory of Increased Intellectual Property Protection, in Gustavo Ghidini, Hanns Ullrich and Peter Drahos (eds.), Kritika: Essays on Intellectual Property, vol. ii (Edward Elgar, 2017), 159–186.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.012

7 Article 20 of the TRIPS Agreement: Up in Smoke? Daniel J. Gervais1

World Trade Organization (WTO) members must be able to regulate tobacco as they see fit. Was that argument sufficient for Australia to win the case filed against it by a number of WTO members contesting the compatibility of its plain-packaging legislation (which prohibited non-word marks entirely and limited the use, format and size of word marks on cigarette packs) with two major WTO instruments, including of course the TRIPS Agreement?2 Let us split this braid question in the next few pages by answering three more focused questions. First, is tobacco a product that requires its own rules in trade law? Second, what does the joint report by the Australia “panels”3 and the Appellate Body tell us about the interpretation of Article 20 of the TRIPS Agreement? Third, did the panels and the Appellate Body help develop the idea that WTO norms should not be read in “clinical isolation” from non-WTO norms of public international law?

7.1 tobacco I am no fan of tobacco. As Susy Frankel and I wrote in an article published as the WTO case was getting underway:4 1

2

3

4

Milton R. Underwood Chair in Law, Vanderbilt Law School. The Author acted as an adviser for Honduras (Appellate Body hearing only). Agreement on Trade-Related Aspects of Intellectual Property Rights, Annex 1C of the Marrakesh Agreement Establishing the World Trade Organization, signed in Marrakesh, Morocco on 15 April 1994. The cases were filed by (in order of filing): Honduras, the Dominican Republic, Cuba and Indonesia. A joint report (884 pages plus annexes) was issued for all the disputes: Australia – Certain Measures Concerning Trademarks, Geographical Indications and Other Plain Packaging Requirements Applicable to Tobacco Products and Packaging: Reports of the Panels, WTO documents WT/DS435/R, WT/DS441/R, WT/ DS458/R, WT/DS467/R, 28 June 2018. The other WTO instrument (not discussed in this short chapter) is the Agreement on Technical Barriers to Trade. The cases brought by different complainants were heard together and a single report was issued but technically by different panels (one for each case). Hence this chapter will use the plural “panels”. The “Request for Consultations by Honduras” was filed on 4 April 2012. See WTO doc. WT/ DS435/1, 10 April 2012.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms 96 of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

Article 20 of the TRIPS Agreement: Up in Smoke?

97

There is a significant body of evidence that smoking is very bad for human health and life. Cigarettes contain toxins that are known to cause cancer. People continue to smoke even though they know that smoking can kill them and may be bad for the health of those around them. One, if not the main, reason for this is that nicotine is addictive. The World Health Organization (WHO) has described this as an “epidemic of epic proportions”. Cigarettes are probably the only consumer product sold where over half its users will die as a result of its use.5

As legal scholars, however, we must endeavour to keep a cool head when discussing even difficult normative issues. What I can surmise is that the issue of exactly how “bad” tobacco is will be relevant in future cases, for reasons I outline below, in deciding the fate of the WTO dispute settlement reports.6 One could argue, for instance, that there are other “lifestyle” habits and products that kill many of their users. CBS reported that “in 2000, poor diet including obesity and physical inactivity caused 400,000 U.S. deaths – more than 16 per cent of all deaths and the No. 2 killer. That compares with 435,000 for tobacco, or 18 per cent, as the top underlying killer.”7 Of course, one can (and should) stop smoking, but one cannot stop eating. A better example might be another completely “unnecessary” product (for survival), namely alcohol. Alcohol reportedly killed 35,823 Americans in 2017.8 Indeed, alcohol and tobacco are often regulated under similar provisions, for example by limiting the age at which they can be purchased.9 Not surprisingly perhaps, the popular press often links these two “vices”.10 But it is not just the popular press, as experts now demand that plain packaging be applied to alcohol.11 Indeed, in a brief analysis of arguments concerning the plain packaging of alcohol, an observer of the UK packaging industry noted, “[a]lcohol is just one of a number of categories being targeted by health lobbyists at the moment, with sugar and soft drinks also in the 5

6

7

8

9

10

11

Susy Frankel and Daniel Gervais, Plain Packaging and the Interpretation of the TRIPS Agreement (2014) 46 VJ TL 1149, at 1157–1158 (footnotes omitted). I readily admit that I am not qualified to discuss the expert evidence submitted to the WTO panels as to the extent to which plain packaging reduces overall smoking or youth smoking in particular. Bootie Cosgrove-Mather, Americans Eat Themselves to Death, CBS News, 9 March 2004, https://www.cbsnews.com/news/americans-eat-themselves-to-death/. Trust for America’s Health/Well Being Trust, Pain in the Nation Update: While Deaths from Alcohol, Drugs, and Suicide Slowed Slightly in 2017, Rates are Still at Historic Highs (March 2019), https://www.tfah.org/wp-content/uploads/2019/03/TFAH-2019-PainNationUpdateBrief-06 .pdf. For example, in the United States, by the Bureau of Alcohol, Tobacco, Firearms and Explosives. Nancy Clanton, How Many Cigarettes are in a Bottle of Wine?, The Atlanta JournalConstitution, 30 March 2019, https://www.ajc.com/lifestyles/health/how-many-cigarettesare-bottle-wine/Jt7zvGO41KYqnlGnVY7GoK/. See Williams Hughes, The WTO’s “Decision” on Plain Packaging for Cigarettes: Will Alcohol be Next?, Lexology, 16 June 2017, https://www.lexology.com/library/detail.aspx?g=0434c9066372-42af-a156-2811847a05bb.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

98

Daniel J. Gervais

firing line. However, the noise around the dangers alcohol poses appears to be growing.”12 An Australian research team noted, along similar lines: Alcohol is a legal and social drug available for sale and consumption Australia-wide. It carries a comparable health risk to cigarettes when the long term, short term and health statistics are taken into consideration, yet the policy around health warnings is vastly different.13

A public health advocate might argue in this context that trade law must yield in the face of measures such as plain packaging of tobacco (and alcohol), and possibly yield in the same way to any major public health concern. Yet, WTO law has rules that have applied to tobacco (and other highly harmful products such as asbestos) in past cases.14 As is noted in the above-mentioned joint article,15 the panel in Thailand – Restrictions on Importation of and Internal Taxes on Cigarettes, for example, found against Thailand, despite support from the World Health Organization (WHO) for some of the measures taken by the Thai government.16 National treatment obligations did not simply vanish at the Dispute Settlement Body – even in the case of cigarettes. Be that as it may, the question of how tobacco’s “badness” compares to other products is bound to come up if plain-packaging requirements or similar measures are imposed on other products.

7.2 article 20 justification test Article 20 was first examined in an early WTO Dispute Settlement panel report. In the Indonesia – Certain Measures Affecting the Automobile Industry case,17 the United States claimed, inter alia, on the basis of Article 20 that Indonesia had violated its TRIPS Agreement obligations by imposing an obligation on certain mark owners to use marks associated with its National Car Programme. Essentially, because participating in the National Car Programme was seen as voluntary – and hence, trademark owners arguably “accepted” use of the imposed Indonesian mark – the encroachment upon trademark owners’ rights was considered not to violate Article 20. 12

13

14

15 16

17

Simon Creasy, Is Plain the Future? Category Focus – Alcoholic Drinks, Packaging News, 1 September 2017, https://www.packagingnews.co.uk/news/markets/plain-future-category-focusalcoholic-drinks-01-09-2017. Australian National University, student research project, Would Plain Packaging for Alcohol Communicate Health Risk Factors to Youth?, https://cpas.anu.edu.au/study/student-projects/ would-plain-packaging-alcohol-communicate-health-risk-factors-youth. See Appellate Body Report, United States – Measures Affecting the Production and Sale of Clove Cigarettes, WTO doc. WT/DS406/AB/R, 4 April 2012. Frankel and Gervais, note 5 above, at 1152. GATT Panel report, Thailand – Restrictions on Importation of and Internal Taxes on Cigarettes, paras. 50–57, 78–80. Available at https://www.wto.org/english/tratop_e/dispu_e/gatt_e/90cigart .pdf. Indonesia – Certain Measures Affecting the Automobile Industry, Panel report, WTO doc. WT/ DS54/R, 2 July 1998.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

Article 20 of the TRIPS Agreement: Up in Smoke?

99

Article 20 was at the heart of Australian – Tobacco Plain Packaging. Australia raised several arguments to oppose the application of Article 20 to its plain-packaging legislation, including that a ban is not an impediment because it is, in a way, more than an encumbrance. This puzzling interpretation of Article 20 was firmly rejected by the panels: [T]he [plain packaging] measures prevent a trademark owner from extracting economic value from any design features of its trademark through its use in the course of trade. In principle, therefore, the [plain packaging] measures’ prohibitions on the use of figurative trademarks on tobacco retail packaging and products, as well as of the figurative and stylized elements of composite and word marks, are far-reaching in terms of the trademark owner’s expected possibilities to extract economic value from the use of such features.18

The plain-packaging measure at issue was found to be an encumbrance. This also applied to the special form required for word marks. Indeed, use in a special form is specifically mentioned as covered by Article 20. Was it an encumbrance by special requirement? A “special requirement” is one that applies to the “use of a trademark in the course of trade”.19 The panels found that use covered by the Australian legislation was use in the course of commercial activities and thus rejected Australia’s argument that use relevant to Article 20 was “limited to the use of a trademark for the specific purpose of distinguishing the goods or services of one undertaking from those of other undertakings”.20 The next issue was whether the special requirements encumbered the use of marks “unjustifiably”. There never was any doubt that public health is a valid policy area to justify a regulatory intervention, that is, one in which the adoption of a policy measure can be justified.21 Public health is reflected, inter alia, in Article 8.1 of the TRIPS Agreement, and in the Doha Declaration on TRIPS and Public Health.22 The legal question here is different, namely whether invoking the fact that an impugned measure is in the area of public health suffices to rebut a claim of unjustifiability. According to the Appellate Body, the “concept of justifiability thus indicates that it connotes something that is fair and capable of being reasonably explained”,23 adding that the panel was correct in finding that “encumbrances on

18

19 20 21 22

23

Australia plain-packaging Panel report, note 2 above, para. 7.2569 (emphasis added). Australia did not appeal the panels’ report so these findings stand, at least as a formal matter. Ibid., para. 7.2231. Ibid., paras. 7.2282 and 7.2284. See Frankel and Gervais, note 5 above. Ibid., at 1152 and 1204. Declaration on the TRIPS Agreement and Public Health, WTO doc. WT/MIN(01)/DEC/2, 14 November 2001. Australia – Certain Measures Concerning Trademarks, Geographical Indications and Other Plain Packaging Requirements Applicable to Tobacco Products and Packaging: Reports of the Appellate Body, WTO docs. WT/DS435/AB/R and WT/DS441/AB/R, 9 June 2020, para. 6.645.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

100

Daniel J. Gervais

the use of trademarks by special requirements under Article 20 may also be imposed in pursuit of public health objectives”.24 The panels found unequivocally for Australia on this point, noting, first, that it was the complainants’ duty to prove that the encumbrance was unjustified,25 even though acknowledging that a respondent member, “as the Member having taken the measures, may be in a privileged position to explain, in response to a possible prima facie demonstration of violation, why a specific encumbrance is justifiable”.26 This position is reconcilable with the typical approach that the burden to establish all the elements included in an obligation provision is on the complaining party.27 Having thus shifted the burden to complainants, the panels had to find that the complainants had made a prima facie case, before asking the respondent (Australia) to rebut. As just noted, whether the level of justification should be in some way proportional to the nature of the encumbrance is certainly a plausible question to ask. The Appellate Body noted in that connection that “due account should also be taken of the nature and extent of the encumbrance at issue”.28 The panels devised a multi-factor test that facially included some degree of proportionality, including “the nature and extent of the encumbrance arising from the special requirements at issue, the reasons for which these requirements are applied, and whether these reasons sufficiently support them”.29 The panels could have weighed more thoroughly whether a measure they described as “extreme” should be matched by strong evidence that the plain-packaging measures were working, for example by there being a measurable drop in smoking after the introduction of the measure, duly controlling for other factors. As noted above, the panels instead chose to impose on complainants the burden of proving that the measures were not working. The Appellate Body agreed, noting that “the complainant has to demonstrate that a policy objective pursued by a Member imposing special requirements does not sufficiently support the encumbrances”.30 Interestingly, the panels did not include as part of the test the need to examine the availability of alternative measures that could provide an equivalent contribution.31 The Appellate Body also agreed on this point, noting that “such an examination is not a necessary inquiry under Article 20 of the TRIPS Agreement”.32 24 25 26 27

28 29 30 31 32

Ibid., para. 6.649. Australia plain-packaging Panel report, note 2 above, paras. 7.2164–7.2165. Ibid., para. 7.2167. See “Necessity Tests” in the WTO – Note by the Secretariat, WTO doc. S/WPDR/W/27/Add.1, 18 January 2011, para. 6. In the co-authored article, Professor Frankel and I took a different position on this point. But this is a minor point; very few GATT or WTO cases have turned primarily on a burden of proof question. Australia plain-packaging Appellate Body report, note 23 above, para. 6.651. Australia plain-packaging Panel report, note 2 above, para. 7.2441. Australia plain-packaging Appellate Body report, note 23 above, para. 6.659. Australia plain-packaging Panel report, note 2 above, para. 7.2601. Australia plain-packaging Appellate Body report, note 23 above, para. 6.653.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

Article 20 of the TRIPS Agreement: Up in Smoke?

101

By adopting a multifactor test that focuses essentially on the existence of a rational explanation to support the measure, the panels arguably discarded a number of approaches which are more typical of a TRIPS context. For example, in considering exceptions to patent rights allowed under Article 30, the key question was not simply why a government would want to allow for the production and stockpiling of generic medicines during the period of patent protection. In the context of the debate on access to medicines both leading to and following the adoption of the Doha Declaration, the reason why governments wanted to allow compulsory licensing was never the core issue. The question was about the balance of rights and obligations of intellectual property owners. The Appellate Body concluded hat “the term ‘unjustifiably’ in Article 20 of the TRIPS Agreement reflects the degree of regulatory autonomy that Members enjoy in imposing encumbrances”33 and a “certain degree of discretion”.34 The panels and Appellate Body could also have looked more deeply into the negotiating history of Article 20as per act 32 of the Vienna Convention. Absent “unjustifiably” in the first sentence, the second sentence, which qualifies the first, could be in contradiction with the first. The second sentence was designed, at least from the point of view of a number of demandeurs, to address requirements that international brands be shown in combination with domestic brands. The latter part of the second sentence is an attempt to limit (and thus arguably render “justifiable”) the specific measure at issue, namely a “requirement prescribing the use of the trademark identifying the undertaking producing the goods or services along with, but without linking it to, the trademark”. There are several known tests in trade law according to which a WTO member can justify a measure not otherwise compatible with its commitments under WTO instruments, including several shades of the so-called necessity test (the WTO itself refers now to “necessity tests”.) The scope of “outs” other than necessity is sometimes less clear, such as the current debates on whether invoking national security provides an automatic, unconditional exception for any measure is a correct interpretation of GATT, Article XXI and TRIPS, Article 73. This matter was addressed in a recent dispute, and its application in intellectual property was discussed by a panel which rejected its application to a failure to comply with Article 61 of TRIPS, in what is probably the first time the national security defence failed in a WTO case.35 In Australian - Tobacco Plain Packaging, the Appellate Body explained that the “degree 33 34 35

Australia plain-packaging Appellate Body report, note 23 above, para. 6.647. Ibid., para. 6.659. In Russia–Ukraine the panel, in discussing the national security exception, noted that it demanded “that the measures at issue meet a minimum requirement of plausibility in relation to the proffered essential security interests, i.e. that they are not implausible as measures protective of these interests” (para. 7.138) (emphasis supplied). A related question – that I cannot develop here – is the possible link between the rational connection and plausibility tests. See, regarding the TRIPS Article 61 case, Saudi Arabia – Measures Concerning the Protection of Intellectual Property Rights: Report of the Panel, WTO doc. WT/DS567/ R, 16 June 2020. The author appeared as an expert for Qatar in the Saudi Arabia case.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

102

Daniel J. Gervais

of connection between the encumbrance on the use of a trademark imposed and the objective pursued reflected through the term ‘unjustifiably’ is lower than it would have been had a term conveying the notion of ‘necessity’ been used in this provision”.36 The common thread in the application of exceptions and similar measures is that – except for a possible interpretation of the national security exception as an automatic carte blanche – panels and the Appellate Body typically require a particular level of connection for an exception to be validly invoked or applied by a member. The weakest such nexus is that of a “rational connection” between a measure and its policy objective. Arguably any measure designed to address the alleged public policy objective that is not otherwise frivolous, capricious or arbitrary will pass this “rational connection” test. This was the test suggested by Australia and a number of third parties in the plain-packaging case. The panels considered that test potentially applicable to Article 20, though sending conflicting signals on the matter, noted “the existence of any rational connection, no matter how tenuous, would [not] always sufficiently support the imposition of such encumbrance permissible under Article 20”.37 The Appellate Body cleared up the confusion. After noting that it “recognize[d] that the language that the Panel used in referring to the expected degree of contribution of the alternative measures in its analysis under Article 20 of the TRIPS Agreement was inconsistent”,38 it noted that “the term ‘unjustifiably’ should not be understood to require only the existence of some rational connection between encumbrances imposed on the use of a trademark and the reason for which they are imposed. Due account should also be taken of the nature and extent of the encumbrance at issue.”39 As I read the panels’ and Appellate Body reports, the test formulated by the panels and clarified by the Appellate Body does factor in proportionality; it is about the policy objective (public health was specially identified by the Appellate Body) pursued by the member imposing special requirements and whether it sufficiently supports the encumbrance, bearing in mind that it is the complainant’s burden to show that the policy objective is insufficient. What is particularly interesting in the panels’ reasoning is that it does not put limits on what could constitute a valid reason or policy objective. Under Articles 7 and 8, many more socio-economic interests would seem to be valid reasons for a policymaker to adopt certain measures. Affirming that in some cases a rational connection suffices to pass the Article 20 justifiability test was not a foregone conclusion. Indeed, Canada in its third-party brief noted, “Australia’s proposed test of requiring a complainant to establish that

36 37 38 39

Australia plain-packaging Appellate Body report, note 23 above, para. 6.647. Australia plain-packaging Panel report, note 2 above, para. 7.2829. Australia plain-packaging Appellate Body report, note 23 above, para. para. 6.693. Ibid., para. 6.651.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

Article 20 of the TRIPS Agreement: Up in Smoke?

103

there is no rational connection between the requirement and a legitimate public policy objective in order to demonstrate that the requirement is ‘unjustifiable’ would establish an extremely onerous burden to establish a violation of Article 20”.40 Canada had suggested that another test in the family of necessity tests, known as the “least-restrictive measure” test, was not the right one either.41 Canada had offered an alternative test consisting of four factors: (i) the legitimacy of the objective; (ii) the existence of a rational connection between the measure and the legitimate objective; (iii) the contribution of the measure to the requirements of said objective; and (iv) the extent of the encumbrance on the use of marks.42 This test would have meant a win for Australia on the first two factors. The third factor would have required a fair weighing of the evidence of the contribution of plain packaging to a reduction in smoking, a matter on which this chapter does not opine. The fourth and final factor would have introduced a more convincing involvement of proportionality; given the importance of the objective, Australia would likely have won this factor, but such an approach would be more in line with reports of panels in previous TRIPS and other WTO disputes. Indeed, in at least two WTO dispute-settlement cases, the proportionality of a measure in relation to its objective was considered.43 Ultimately, neither the panels nor the Appellate Body fully addressed the question of the threshold, finding, in their analysis of the evidence – including a most detailed dissection of statistical and econometric data in Appendix E to the panels’ report – that the complainants had not “demonstrated that Australia has acted beyond the bounds of the latitude available to it under Article 20 to choose an appropriate policy intervention”.

7.3 clinical isolation? Yet another interesting aspect of the reports is the manner and extent to which Australia relied on its implementation of the Framework Convention on Tobacco Control (FCTC). Guidelines to the FCTC (not the actual text of the FCTC) suggest that countries should “consider adopting measures to restrict or prohibit the use of logos, colours, brand images or promotional information on packaging other than brand names and product names displayed in a standard colour and font style (plain packaging)”.44 The use of the FCTC and its Guidelines is indeed a striking aspect of the reports. 40 41 42 43

44

Australia plain-packaging Panel report, note 2 above, para. 7.2352. See ibid. See ibid., para. 7.2354. See Korea – Measures Affecting Imports of Fresh, Chilled and Frozen Beef, Appellate Body report, WTO docs. WT/DS161/AB/R and WT/DS169/AB/R, 11 December 2000, para. 179; European Communities – Measures Affecting Asbestos and Products Containing Asbestos, Panel report, WTO doc. WT/DS/135/R, 18 September 2000, para. 8.196. World Health Organization, Framework Convention on Tobacco Control, Guidelines for Implementation of Article 11, para. 46.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

104

Daniel J. Gervais

There is a considerable body of literature on the interplay between WTO and non-WTO norms in WTO dispute settlement. It is obvious that the FCTC has broad membership (182 parties at the time of writing), though it must be noted that three out of four of the initial complainants were not party to the Convention or had not yet ratified the Convention.45 Second, plain packaging is only a recommendation to “consider” that is contained in the Guidelines, and it is thus not part of treatybased obligations. To accept the FCTC Guidelines (entry into force in 2005) as, say, customary international law (if that is what the panels did) was certainly not an obvious step. One cannot say with certainty what the legal bridge is between the FCTC Guidelines and the justification test in the panels’ report. What seems clear, however, is that the panels took the Convention and its Guidelines on board with little, if any, discussion.46 This could open up to arguments about the direct application of various non-WTO norms in several other contexts. If FCTC guidelines are sufficient then perhaps other key global instruments with near-universal membership, including the Framework Convention on Climate Change (FCCC), should also suffice, unless public health (or even just tobacco control) are seen as special cases.47 The Appellate Body has repeatedly relied on the provisions of the Vienna Convention on the Law of Treaties as a primary source for interpretative guidance.48 The Appellate Body has also found that WTO Agreements should “not to be read in clinical isolation from public international law”.49 This principle, which I have referred to as the “non-clinical isolation doctrine”, was reflected in reports that relied on the case law of international tribunals, namely the International Court of Justice, the European Court of Human Rights and the Inter-American Court of Human Rights.50 This should not lead to the conclusion that forum shopping to develop

45

46

47

48

49

50

I leave aside for now the somewhat complex issue of the effect of a signature not followed by a ratification. The FCTC plays a central role in the panels’ discussion of the Technical Barriers to Trade (TBT) arguments made by the complainants, especially under TBT, Article 2.5. In the case of TRIPS, and specifically Article 20, it seems accepted as the justification for the plain-packaging measures. See Australia plain-packaging Panel report, note 2 above, para. 7.2596. The FCCC has 197 parties as of July 2020. See https://bit.ly/30MY9Tf (accessed 5 July 2020). Article 3(3), for example, provides that “Parties should take precautionary measures to anticipate, prevent or minimize the causes of climate change and mitigate its adverse effects”. For example, a country might subject compliance with trade liberalization objectives to the desire to follow the Paris Trajectory. Vienna Convention on the Law of Treaties, adopted 22 May 1969, opened for signature 23 May 1969, entered into force 27 January 1980, 1155 UNTS 331. See India – Patent Protection for Pharmaceutical and Agricultural Chemical Products, WTO doc. WT/DS50/AB/R, 19 December 1997, para. 46. United States – Standards for Reformulated and Conventional Gasoline, WTO doc. WT/DS2/ AB/R, 29 April 1996, Part III.B. Ibid., note 34; Japan – Taxes on Alcoholic Beverages, WTO doc. WT/DS8/AB/R, 4 October 1996, Part D, note 19.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

Article 20 of the TRIPS Agreement: Up in Smoke?

105

norms outside the WTO should necessarily result in using them in interpreting WTO instruments through dispute-settlement proceedings.51 Reliance by the Appellate Body on extrinsic (that is, non-WTO negotiated) norms had thus far been limited to the application of well-accepted principles of international law.52 The suggestion that another treaty should be used as a blueprint for the interpretation of, or to effect a reduction in the scope of stated obligations under, the TRIPS Agreement must be considered with utmost caution.53 Add to this the fact that the FCTC is a treaty to which some complainants had not adhered – and that the reference is really to guidelines – and the step taken by the Australia panels appears novel. True, as former WTO Director-General Pascal Lamy stated at a conference on IP and global public policy issues: The international intellectual property system cannot operate in isolation from broader public policy questions such as how to meet human needs for basic health, food and a clean environment.54

That being said, whether and how those external norms should inform the interpretation of TRIPS in keeping with the principle that “unbargained for” concessions are not part of the WTO norm set – and if such external rules were to apply, to what extent – is an issue on which further Appellate Body guidance would no doubt be useful. Unfortunately, the Appellate Body punted on this issue, noting that the FCTC Guidelines had merely provided “additional factual support” to bolster the panels’ conclusions.55

7.4 conclusion While the Australia reports adopted a convincing analytical path to decide the case, their approach to the justification test, and their discussion of a rational connection, the absence of which must be established by a complainant, basically means that Article 20 went up in smoke short of the demonstrable absence of a policy objective to support the measure or clearly discriminatory measure. Some might say that this is

51

52

53

54

55

For a human rights illustration, see L. R. Helfer, Forum Shopping for Human Rights (1999) 148 U. Pa. L. Rev. 285. See S. Frankel, The WTO’s Application of “the Customary Rules of Interpretation of Public International Law” to Intellectual Property (2005) 46 Va. J. Int’l L. 365. See ibid.; and T. Eres, The Limits of GATT Article XX: A Back Door for Human Rights? (2004) 35 Geo. J. Int’l L. 597, at 624–625. Pascal Lamy, “Strengthening Multilateral Cooperation on IP and Public Health”, speech at the WIPO Conference on Intellectual Property and Public Policy Issues, Geneva, 14 July 2009, available at http://www.wipo.int/meetings/en/2009/ip_gc_ge/presentations/lamy.html, accessed 19 May 2020. Australia plain-packaging Appellate Body report, note 23 above, para. 6.707.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

106

Daniel J. Gervais

fully justified here because big tobacco is involved. As a matter of intellectual property law as well as trade law, however, this raises questions for the future. Among them is whether a sui generis box will be created for tobacco controlrelated cases, as it arguably was for asbestos in trade law, and whether and how the test will be applied to other products. My sense is that, now that the test has been clarified by the Appellate Body, it will be interpreted liberally by future panels.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.013

8 Implementing Treaty Obligations at the National Level Sam Ricketson1

8.1 introduction Much of Annette Kur’s work – and that of the other contributors to this volume – has been concerned with the ways in which international obligations dealing with intellectual property rights can be transmitted into national legal systems in a coherent and relevant fashion. This work, so carefully executed and challenging in its range, prompts the present essay, which is a consideration of the scope that these international obligations allow to national policy and law makers in implementing them at the national level.

8.2 the particular character of multilateral norms At the outset, it should be noted that we are concerned here only with obligations contained in multilateral agreements, such as the Berne, Paris and Rome Conventions,2 the WIPO Copyright Treaty3 and the TRIPS Agreement4 (to list the most significant). These are instruments that are open to accession by any state prepared to comply with their provisions, but are not otherwise subject to other joining requirements, such as geographic location, political alignment or developmental status. Consequently, membership of each of these agreements is now extensive – nearly universal in the case of Berne and Paris – and this immediately suggests that it is likely that there will be less precision in the content and 1 2

3 4

Professor Emeritus, Melbourne Law School, University of Melbourne, Australia. Berne Convention for the Protection of Literary and Artistic Works 1886–1971 (‘Berne’); Paris Convention for the Protection of Industrial Property 1883–1967 (‘Paris’); Rome Convention for the Protection of Performers, Producers of Phonograms and Broadcasting Organizations 1961 (‘Rome’). WIPO Copyright Treaty 1996 (‘WCT’). Agreement on Trade Related Intellectual Property Rights 1994 (‘TRIPS Agreement’), annexed agreement to membership of the World Trade Organization.

Downloaded from https://www.cambridge.org/core. 107 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

Sam Ricketson

108

formulation of their provisions than might be expected in the case of more tailormade bilateral or regional agreements that have involved far fewer states in their negotiation and adoption. The history of the Berne and Paris Conventions supports this view: their initial success in attracting adherents was based on the adoption of the principle of national treatment which required no specific action at the national level other than the obligation to treat foreign claimants for protection in the same way as national claimants. More specific obligations in the initial Berne and Paris texts were of a minimal nature, although the priority system under Paris was critical in the obtaining of access to the benefits of national treatment. Both Conventions, however, were viewed as dynamic rather than static instruments, so the obligation to revise was built into each of them from the outset, and revision occurred at regular intervals up to 1958 (in the case of Paris) and 1967 (in the case of Berne). Over the course of these revisions, Berne came to incorporate a series of additional norms of protection that member countries were to apply to the protection of foreign authors. Paris, by contrast, was less successful in this endeavour, but has now been supplemented in significant ways by the provisions of the TRIPS Agreement, just as Berne itself has been extended by specific obligations in relation to online uses under the WCT. As a general matter, once multilateral agreements move beyond the simple base position of national treatment, the formulation of additional norms of protection that are to be applied inevitably becomes subject to protracted processes of negotiation and compromise between a multitude of parties, leading to the consequence that any treaty provision that is ultimately adopted will usually be something of a compromise – hence, of course, the attraction of some of the bilateral or regional approaches that have characterised the last two decades in the form of free trade agreements that provide for more sharply defined norms of protection that are Berne plus, Paris plus or TRIPS plus, as the case may be.

8.3 a hierarchy of norms? It is unsurprising then that there are relatively few instances in which clear and hardedged multilateral norms are to be found which provide national law makers with firm guideposts or fixed obstacles around which it is impossible to manoeuvre in implementation at the national level (at least so far as claimants for protection under the Convention in question are concerned). An obvious one is the term of protection for literary and artistic works under Berne,5 which allows for departures only in limited instances (cinematographic and photographic works and works of applied art6); another is the no-formalities rule which operated for nearly a century to deter

5 6

Berne Convention, Article 7(1). Berne Convention, Article 7(2) and (4).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

Implementing Treaty Obligations at the National Level

109

the USA from Berne adherence.7 The Paris Convention does not contain prescriptive norms of this character, apart from its system of priorities which is laid down in some detail,8 but specific obligations in relation to minimum terms of protection for patents, designs and trade marks are now to be found in the TRIPS Agreement.9 Other positive norms contained in Berne, WCT, WPPT, 10 TRIPS and, to a lesser extent, Paris, allow varying degrees of flexibility in the way members can implement them, aided by the malleable rules of treaty interpretation that apply under both customary international law and the Vienna Convention on the Law of Treaties (VCLT) (to say nothing of the fact that, until the TRIPS Agreement, there was no effective enforcement machinery at the international level). There is considerable ‘give’ here which is allowed to member countries, as well as benefits to national policy and law makers that are not so often present in bilateral arrangements (other than to the more powerful party who is setting the agenda with which the other side has to comply). Without being too dogmatic, the following hierarchy of multilateral IP norms can be discerned. The first two of these will cause no surprise, as their scope will long have been appreciated by those charged with the interpretation and implementation of treaties;11 the third, however, discloses possibilities that may not previously have been apparent.

8.3.1 Norms Which Although Prescriptive in Form Allow National Laws Considerable Latitude in Their Interpretation A norm which states that the term of protection for a work is the life of the author plus fifty years does not permit any variation, particularly when it is also expressly provided that the fifty-year term does not commence until the first day of the year after the author dies.12 The same may be said of the priority periods specified in Paris, or the more detailed procedural provisions to be found in registration treaties such as the Patent Co-operation Treaty or the Madrid Protocol on Trademarks. On the other hand, a norm providing, in the case of Berne, for the grant of an exclusive right to reproduce, publicly perform, broadcast or distribute a work,13 or, in the case of Paris, ‘effective protection against any act of competition contrary to honest

7

8 9 10 11

12 13

Berne Convention, Article 5(2). Nonetheless, considerable room still remains for manoeuvre around and within the boundaries of this requirement: see further J. Ginsburg, With Untired Spirits and Formal Constancy: Berne-Compatibility of Formal Declaratory Measures to Enhance Copyright Title-Searching, 28 Berkeley Tech. L.J. 1583 (2014), 1584–1586. Paris Convention, Articles 4A–4I. TRIPS Agreement, Articles 18 (trade marks), 26.3 (designs) and 33 (patents). WIPO Performances and Phonograms Treaty. This may include both legislators and courts (in countries with monist systems in which treaty obligations may be capable of direct implementation). Berne Convention, Article 7(5). Berne Convention, Articles 9, 11, 11bis and 11ter and WCT, Article 6.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

110

Sam Ricketson

practices in industrial or commercial matters’,14 allows for a considerable degree of latitude in implementation in national laws. At the macro level, these are ‘conventional rules’ that seek to embody an overall uniform standard of protection to be applied by Member States. At the micro level, however, some of the terms used in these treaty provisions require further interpretation. For instance, does the term ‘reproduce’ in Article 9(1), Berne, extend to temporary copies in RAM or versions in different dimensions, how close must such reproductions be to the original, and is a mental element required on the part of a person charged with having carried out the act in question without authorisation? In the case of public performance under Article 11, what are the limits to the concept of ‘public’ and where is the line between ‘private’ and ‘public’ to be drawn? More generally, do references to ‘authorising’ in the formulation of each of the Berne exclusive rights involve concepts of vicarious or contributory as well as direct liability? In the case of the Paris provision on unfair competition, are comparative advertising and the unauthorised use of confidential information, which are not included in the list provided in Article 10bis(3), acts of unfair competition ‘contrary to honest practice’? The rules of treaty interpretation may provide some assistance here, requiring regard to be had to ‘the terms of the treaty in their context and in the light of its object and purpose’.15 Included in ‘context’, along with the surrounding text, preamble and annexes, are collateral agreements made by the parties in relation to the conclusion of the treaty,16 together with interpretations that may be established through subsequent state practice.17 Supplementary means of interpretation involving reference to preparatory materials and the ‘circumstances of its conclusion’ may also be consulted ‘to determine the meaning when the interpretation according to article 31: (a) leaves the meaning ambiguous or obscure; or (b) leads to a result which is manifestly absurd or unreasonable’.18 In many instances, however, these interpretative tools provide no guidance one way or the other, which leaves national legislators with considerable freedom to determine for themselves how a particular treaty term is to be interpreted and applied. This may, indeed, be an outcome that was contemplated by the parties negotiating the treaty provision in question, in that consensus at this level of detail would otherwise have been impossible or impracticable, with the consequence that the terminology adopted is deliberately open-ended to allow for a variety of national responses that will still fall within its scope. Harmonisation at this level may therefore allow for a ‘broad church’ of national responses that can reflect local conditions and interests without falling into non-compliance with the relevant treaty provision. 14 15

16 17 18

Paris Convention, Article 10bis. VCLT, Article 31.1 (although the rules here do not apply to Berne or Paris as pre-VCLT treaties, it is generally agreed that Articles 31 and 32 codify the customary rules of treaty interpretation that should be applied with respect to such treaties). VCLT, Article 31.2(a). VCLT, Article 31.3(a). VCLT, Article 32.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

Implementing Treaty Obligations at the National Level

111

8.3.2 Express Rules of Referral In other instances, the referral back to national laws is explicit, rather than implicit in the language chosen, with such matters being left to those laws to determine, at their discretion or within limits that are set in the particular treaty provision. The most extreme instances are to be found in the provisions of the Paris Convention requiring member countries to protect industrial designs, service marks and trade names respectively.19 These obligations go beyond what may be available as a matter of national treatment – which may be nothing at all, if there is no protection for such subject matter under national law – but otherwise leave apparently untrammelled latitude to laws as to how they give effect to these obligations (there may be some limits to this discretion, which are discussed below). Under the Berne Convention, rules of this kind are to be found in relation to exceptions and limitations which are also framed in permissive, rather than mandatory, terms.20 The most notable instance of such a rule of referral is to be found in the three-step test in relation to limitations and exceptions to reproduction rights under Article 9(2). No exceptions are either specified or mandated, but it is left to national policy and law makers to determine for themselves which exceptions are appropriate, having regard to the conditions laid down in the three steps.21 As we all know, this template for exceptions and limitations has been extended to other exclusive rights under the WCT,22 WPPT23 and TRIPS Agreement (where modified versions have been applied also to trade marks, designs and patents24).

8.3.3 Rules Where Compliance Depends Upon Outcomes Rather Than the Means Adopted This represents the most unusual case that came into prominence at the time of the negotiation of the WCT and WPPT in relation to the exclusive right of communication to the public. Unlike the exclusive rights of reproduction, public performance and the like under the Berne Convention, where it may be supposed that the Member State in question must accord something that is recognisably a reproduction, public performance, or similar right, it is argued that Article 8 of the WCT does 19 20

21

22 23 24

Paris Convention, Articles 5quinquies, 6sexies and 8. Berne Convention, Articles 9, 10 and 10bis (with the exception of Article 10(1) with respect to quotation). See further Christophe Geiger, Daniel Gervais and Martin Senftleben, The Three-Step-Test Revisited: How to Use the Test's Flexibility in National Copyright Law, 29 Am. U. Int’l L. Rev. 581 (2014). At the same time, there may be arguments that, in the USA, the concept of fair use has enabled the recognition of exceptions that extend beyond the limits set by the three-step test – a topic for another day that lies outside the scope of this essay. WCT, Article 10. WPPT, Article 16. TRIPS Agreement, Articles 13 (copyright), 17 (trade marks), 26(2) (designs) and 30 (patents).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

112

Sam Ricketson

not require the grant of a right of communication to the public formulated in similar terms to that Article, so long as the outcomes encompassed by that Article are achieved. On its face, the Article seems reasonably clear: countries should provide for a right of communication to the public of works by wire or wireless means, including a right of making the work available in such a way that members of the public may gain access from a place and at a time individually chosen by them. While the concept of ‘public’ has given rise to considerable controversy in national and regional courts since its adoption, the right so formulated appears apt to cover both active and passive forms of communication (‘push’ and ‘pull’ technologies), extending beyond the patchwork of exclusive rights required to be protected under Articles 11bis(1) and 11ter(1) of Berne. This is certainly how it has been implemented, using almost the same language, in the EU and in many other countries, including my own (Australia).25 But there was a problem of characterisation that became apparent in the preparations for the 1996 Diplomatic Conferences, where some countries clearly took the view that their existing formulations of exclusive rights, namely rights of public performance and distribution, including digital transmissions (in the case of the USA), already achieved the same outcomes.26 However, the need to align the language of national laws with that of the international norm is deftly avoided by the argument that the ‘technology neutral’ formulation of Article 8 embodies an ‘umbrella solution’ which downplays the question of characterisation so long as the ultimate effect of national law is to provide the protection required by the provision. In the circumstances, this was doubtless a convenient solution which avoided the need for changes in US law while permitting the treaty negotiations to move forward. However, it poses some important questions, one specific to the WCT and the other more general: (i) Its legal basis is relatively slight, if resort is had to the general rules of treaty interpretation reflected in the VCLT. In the case of other Articles of the WCT where uncertainty as to their meaning or scope arose, resort was had to the device of ‘agreed statements’ which can be properly added to the context of the provision in question for the purposes of interpretation under Article 31.2(a) of the VCLT. No such agreed statement was adopted in relation to the question of characterisation under Article 8,27 and the strongest basis for it is to be found in the following uncontested statement by the chief US delegate during 25

26

27

Directive 2001/29/EC of the European Parliament and of the Council of 22 May 2001 on the harmonisation of certain aspects of copyright and related rights in the information society, Article 3(1) and (2); Copyright Act 1968 (Australia), ss. 31(1)(a)(i), (b)(iii) and 10(1) (definition of ‘communicate’). See further J. Ginsburg and E. Treppoz, International Copyright Law: U.S. and E.U. Perspectives (Edward Elgar, 2015), 319–330. The Agreed Statement to Article 8, WCT that was adopted relates to the ‘mere provision of physical facilities for enabling or making a communication’.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

Implementing Treaty Obligations at the National Level

the discussions on Article 10 (which became Article 8) in Main Committee I of the Conference: 301. Mr. KUSHAN (United States of America) expressed support for Article 10 of Draft Treaty No. 1 and Articles 11 and 18 of Draft Treaty No. 2, concerning the rights of communication to the public and making available to the public, which were key to the ability of owners of rights to protect themselves in the digital environment. He stressed the understanding – which had never been questioned during the preparatory work and would certainly not be questioned by any Delegation participating in the Diplomatic Conference – that those rights might be implemented in national legislation through application of any particular exclusive right, also other than the right of communication to the public or the right of making available to the public, or combination of exclusive rights, as long as the acts described in those Articles were covered by such rights.28

To place this on the same level as an agreed statement appended to the text of the provision is probably to stretch things too far, lacking even the force of a comment or statement included in a general report adopted at the end of the Conference (there was no such general report produced at the end of the 1996 Diplomatic Conference, unlike previous revision conferences of Berne where such reports play an important interpretative role29). Nonetheless, the statement by the US delegate that this was an ‘understanding – which had never been questioned during the preparatory work’ of the Conference appears accurate. Certainly, these documents reveal a diversity of national approaches to the protection of interactive online communications, including extensions of existing broadcasting and distribution rights to cover these activities. But the conceptualisation and articulation of the ‘umbrella solution’ itself is found in the WIPO supporting documents prepared by Dr Mihaly Ficsor in his role as the responsible senior WIPO official and later expounded in his 2002 treatise.30 None of the records of preparatory meetings at which this issue was discussed indicate that there was any disagreement with the ‘umbrella solution’ (as indicated by the US delegate at the later diplomatic conference), but neither do they expressly adopt it in any formal sense. The ‘solution’ undoubtedly is an elegant one that has allowed readier adherence

28 29

30

Records of the Diplomatic Conference, 675, para. 301 (Minutes of Main Committee I). See, in particular, the authoritative general reports of Louis Renault at the Berlin Revision Conference 1908 and that of Savante Bergström at the Stockholm Revision Conference 1967. M. Ficsor, The Law of Copyright and the Internet: The 1996 WIPO Treaties, Their Interpretation and Implementation (Oxford University Press, 2002) (‘Ficsor’).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

113

114

Sam Ricketson

to both the WCT and WPPT by countries that formulate the exclusive rights in different terms,31 although, as noted above, most countries have done so with formulations that accord in form as well as in substance with the language of Article 8. Nonetheless, so far as the process of treaty interpretation is concerned, it calls for the elevation of the statement of the US delegate quoted above to the status of an agreed statement for the purposes of Article 31.2(a), VCLT, while reference to preparatory materials as permitted under Article 32 supports, at most, a negative rather than a positive inference, that is, that no delegates were opposed to such an interpretation and, perhaps more positively, that the essence of it is to be found in the supporting WIPO documentation. Subsequent and uncontested state practice might also be called in aid, having regard to Article 31.3(b) of the VCLT, in the sense that there has been no objection to accessions to the WCT by states that adopt the umbrella solution in their formulation of the rights required to be protected under Article 8. (ii) At the more general level, while the interpretation of Article 8 discussed above may arguably be justified by reference to the general rules of treaty interpretation, Dr Ficsor has articulated a wider principle that has application, at least in relation to the Berne Convention: it is in harmony with an age-old practice followed by countries of the Berne Union in the application of the various rights granted by the Convention (a practice the compatibility of which with the Berne Convention has never been questioned), namely that the legal characterization of a right – the choice of the applicable right – is frequently not the same under national laws as under the Convention.32

While the use of the words ‘age-old’ may be overstating the antiquity or prevalence of this practice, instances of its earlier occurrence can be cited, for example national laws where public performance rights include both public performance and public recitation rights which are the subject of separate Articles under Berne,33 or where rights of adaptation include translation rights which are treated separately under Berne.34 Prior to the WCT, it might have been argued that such a principle applied only in relation to rights of a similar character which

31

32 33 34

Note, however, doubts that may now arise as to whether or not US law is actually in compliance: see Ginsburg and Treppoz, International Copyright Law, note 26 above. Ficsor, note 30 above, 497. Berne Convention, Articles 11 (public performance) and 11ter (public recitation). Berne Convention, Articles 8 (translation) and 12 (adaptation, arrangement and other alteration).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

Implementing Treaty Obligations at the National Level

115

Dr Ficsor describes as ‘copy-related rights’ and ‘non-copy rights’ and labels as the principle of ‘freedom of relative legal characterization’.35 In the age of dissemination over digital networks, however, both sets of rights may be implicated, for example in streaming (involving the making of transient copies) and downloading (involving the making of more permanent copies), therefore making the application of a more far-reaching principle of legal characterisation ‘across the traditional borders between copy-related rights and non-copy related rights . . . hardly surprising, and . . . fully justifiable and legitimate’.36 If this is, indeed, an appropriate principle of treaty interpretation, at least so far as the WCT and Berne are concerned, it is nonetheless a new, or at least a newly recognised, one that finds explicit exemplification, to date, only in relation to Article 8 of the WCT. If applicable more broadly, it raises several questions which go to the very heart of the transition between international obligations and national implementation. These are considered briefly in the following section.

8.4 is legal characterisation a principle that can be readily applied outside the specific case of communication rights? In the case of communication rights, as Dr Ficsor carefully notes, the elements of what were to be protected were reasonably clearly identified and agreed upon in the text that was adopted: the relevant acts were the making available to the public; this was to include wired as well as wireless communications or a combination of both; it should cover all cases of making available regardless of the purpose behind the act, for example to hear or listen or make copies; it should apply regardless of the time and place at which members of the public received the work; the rights provided should be ‘exclusive’ (subject to whatever restrictions might be applicable under Articles such as Articles 9(2) and 11bis(2) of Berne and Article 13 of TRIPS); and should cover gaps presently arising in international norms, namely under Berne.37 The language moreover was deliberately technology neutral. Can the same kind of analysis be so readily applied to other exclusive rights provided for under Berne, the WCT or WPPT? In the case of exclusive translation rights under Article 8, Berne, no objection will, presumably, be taken to the fact that, under Australian law, for example, these are included within the adaptation right, so long as it is clear that

35 36 37

Ficsor, note 30 above, 499. Ibid. M. Ficsor, International Harmonization of the Protection and Management of Copyright and Neighbouring Rights, in Naples Book Forum, WIPO Publication No. 751(E) (1996), 137–138, quoted by the same author at Fiscor, note 30 above, 208.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

116

Sam Ricketson

they cover translations from one language to another.38 Again, with reproduction rights under Article 9, there will be no complaint if this is called a right against copying rather than reproduction, provided this extends to copies in ‘any manner or form’ (although the word ‘copying’ in itself may suggest only copies in the same form). Arguments about nomenclature should not mask matters of substance. However, a precondition to the principle of characterisation, whether ‘relative’ or ‘full’, must be that, as Dr Ficsor argues, the elements of what are to be protected are clearly identified in the international norm under consideration and capable of embodiment in the implementing measure at national level. Thus, to give a contrary example, it was never sufficient for common law members of Berne to assert compliance with the moral rights obligations of Article 6bis by reference to the incomplete protections available under common law actions of defamation or passing off or statute that did not meet the full requirements of that Article with respect to the content of the rights of attribution and integrity or the matter of duration.39 In the same way, suggested extensions of reproduction rights to include rights of distribution and importation by way of implication do not seem supported, either by the actual language of Article 9 or by reference to the preparatory materials of any earlier Berne revisions.40 These instances, therefore, suggest a limited scope for the application of the principle of characterisation to exclusive rights other than the communication right under Article 8, WCT.

8.5 application outside the case of exclusive rights under berne, wct, etc. As it is framed in a positive sense, namely as to the exclusive rights to be protected, there appears to be limited room, if any, for the application of the principle in other aspects of these agreements, such as duration standards or prohibitions on 38

39

40

Australia, Copyright Act 1968 (Cth), ss. 31(1)(a)(vi) and 10(1). To similar effect, see US Copyright Act 1976, s. 106(2). Under other national laws, such as that of France, there is no separate adaptation right as this is subsumed within the reproduction right. This recognition of Berne non-compliance was nonetheless slow in coming in a number of common law countries such as the UK and Australia that had acceded to the 1928 Rome Act of Berne with its obligation to protect moral rights for the first time: implementing legislation to give full effect to such rights was not adopted in these countries until 1988 (UK) and 2000 (Australia). Some may say also that the USA has not complied adequately in this respect, following its adherence to Berne in 1989, although testimony by the then Director General of WIPO, Dr Bogsch, to the US Congress sought to argue that existing statutory and common law protections under US law did have this effect and that nothing further was required. See further S. Ricketson and J. Ginsburg, International Copyright and Neighbouring Rights: The Berne Convention and Beyond (Oxford University Press, 2nd ed., 2005), vol. i, paras. 10.02–10.14; E. Adeney, The Moral Rights of Authors and Performers: An International and Comparative Analysis (Oxford University Press, 2006), Ch. 6. As suggested in some of the original WIPO proposals for a possible Berne Protocol, in the theory of ‘implied recognition’: see WIPO Document BCP/CE/III/2-111, 5–7, paras. 15–23, as extracted in Ficsor, note 30 above, 211–213.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

Implementing Treaty Obligations at the National Level

117

formalities, which are themselves expressed in positive terms, or in the case of exceptions that must fit within certain parameters but otherwise remain within the discretion of the individual Member States. These provisions contain within themselves the answers as to how they are to be interpreted and applied or otherwise indicate the matters that are left for national determination, making it unnecessary to refer to any wider principle of characterisation.

8.6 application to the interpretation of provisions in other ip treaties, such as paris There may be some scope for the application of the principle here. At the same time, it may not be capable of bridging the gap between the way the international norm is formulated and the way such matters are characterised at the national level. (i) Undisclosed information and comparative advertising: Does Article 10bis require protection of undisclosed information or the practice of comparative advertising as part of its general obligation to provide ‘effective protection against unfair competition’? Some national laws, such as those in common law countries, provide strong protection for undisclosed information where its misuse or unauthorised disclosure constitutes a breach of equitable obligations of confidence.41 However, common law conceptions of unfair competition, as embodied in actions for passing off and statutory restraints on misleading trading, would regard such protections as something distinct from unfair competition.42 Such laws would likewise make provision for comparative advertising activities under their registered trade marks laws (often by way of a defence against infringement rather than as act of unfair competition in itself ).43 It is therefore an open question whether protection against such activities is expressly required under Article 10bis, particularly when regard is had to the inclusive examples of unfair competition provided in Article 10bis(3), which appear directed to activities involving some form of deception, confusion or disparagement. In the case of undisclosed information, the inclusion of a requirement to provide protection against such actions as part of Article 10bis by virtue of Article 39(1) of the TRIPS Agreement may also suggest that Article 10bis otherwise does not require this. Invoking 41

42

43

For a comprehensive account of this action in the UK, see T. Aplin et al., Gurry on Breach of Confidence: The Protection of Confidential Information (Oxford University Press, 2012). See further C. Wadlow, The Law of Passing-Off: Unfair Competition by Misrepresentation (Sweet & Maxwell, 4th ed., 2011). In the case of Australia, see Trade Marks Act 1995 (Cth), s. 122(1)(d), which provides for a defence for comparative advertising in an action for infringement of a registered trade mark.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

118

Sam Ricketson

the Ficsor principle of characterisation here does not help either, as the listed instances in Article 10bis(3) can be said simply to articulate the outer limits of the obligation under that Article and not to require the protection of anything outside these boundaries. (ii) Industrial designs: This may provide the best potential candidate for application of the characterisation principle, albeit in a limiting rather than expansionary way. Article 5quinquies requires simply that ‘Industrial designs shall be protected in all countries of the Union’. On its face, this appears to leave national laws complete flexibility in its implementation, and this was certainly the view of Professor Bodenhausen, one of the leading commentators on the Paris Convention: Nothing is said about the means of providing such protection, so that countries may comply with that provision not only through special legislation for the protection of designs, but also through the grant of such protection, for example, in their laws on copyright or their provisions against unfair competition.44

This open-ended interpretation may go too far in suggesting that unfair competition laws may suffice for the purposes of compliance. Support for artistic copyright protection may well be supported by reference to attempts to protect works of applied art under both the Paris and Berne Conventions,45 and could be rationalised as an instance of application of the principle of ‘relative characterisation’. But both sui generis designs laws, whether requiring registration or not, and artistic copyright do have as their object the protection of designs as such and the protection provided under both kinds of laws is broadly the same, although the nature of the exclusive rights accorded and the term of protection under each may differ considerably. By contrast, under unfair competition, protection of the design arises only incidentally because of some kind of unfair competitive act, usually involving deception or confusion, and would not necessarily meet the general requirement under Article 5quinquies to protect industrial designs.46 Accordingly, it is suggested that it would not be possible for a country that accorded only unfair competition protection to

44

45

46

G. H. C. Bodenhausen, Guide to the Application of the Paris Convention for the Protection of Industrial Property, WIPO Publication No. 611(E) (1969), 86. See further S. Ricketson, The Paris Convention for the Protection of Industrial Property: A Commentary (Oxford University Press, 2015) (‘Ricketson’), Ch. 11, in particular at 11.30 ff. Indeed, Bodenhausen appears to acknowledge this in the section that follows the paragraph quoted above.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

Implementing Treaty Obligations at the National Level

comply with Article 5quinquies, and certainly not by invoking the broader Ficsor principle of legal characterisation. As noted above, the elegance of that principle in the case of communication rights under Article 8 of the WCT is to be found in the fact that that Article specifies with some precision the outcomes of the protection to be achieved through the according of exclusive rights (however these are framed). In the case of Article 5quinquies, the only outcome specified is the protection of industrial designs and it is submitted that this obligation could not be satisfied through the application of incidental protection which may only be available in limited circumstances, that is, where there is also some form of unfair competition involved. (iii) Service marks: These are also the subject of a simple provision in Article 6sexies, added in the Lisbon Revision of 1958, which requires Union countries ‘to undertake to protect service marks’, but without any requirement to provide for registration of such marks. The absence of any requirement to register appears to have arisen from the objection of one Union member (Australia) that, at the time, believed there were constitutional uncertainties as to its capacity to protect service marks under a registration system under federal law.47 Article 6sexies otherwise leaves service marks untied to the other requirements of Paris with respect to trade marks (priorities, restrictions on registration, and so on), and the absence of any registration requirement or any prohibition on requiring use of the mark for the purposes of protection clearly contemplates that protection of service marks through common law actions such as passing off will suffice for the purposes of compliance with the Article. It is hardly necessary, therefore, to invoke any principle of characterisation to justify such an interpretation of Article 6sexies, although the object of protection must clearly remain something that it is identifiably a mark for services. In practice, of course, almost all Paris Union countries, including Australia (since 1978), assimilate protection for marks for goods and services within one registration regime.48 (iv) Trade names: The obligation under Article 8 likewise requires protection of trade names without any requirement of registration, and whether or not the name forms part of a trade mark. Similar comments apply here as in the case of service marks: the obligation can be satisfied through any kind of registered or unregistered regime, whether separately from trade mark protection or not. The wording of the Article clearly contemplates this, without need to rely on any broader principle of legal characterisation. 47 48

See further Ricketson, note 45 above, paras. 12.29–12.30. A result that is contemplated, though not necessarily required, under the TRIPS Agreement.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

119

120

Sam Ricketson

8.7 concluding remarks The purpose of this chapter has been to examine the interpretative scope that is allowed to national laws in implementing their obligations under the multilateral IP Conventions. While there are hard edges that apply to some of the international norms contained in these instruments – of which provisions relating to terms of protection are the clearest examples – most provide differing degrees of flexibility for implementation at the national level, with the most expansive being exemplified by the umbrella solution in relation to Article 8, WCT. In consequence, there is a flexible, if shifting, interface between international obligation and national implementation that allows account to be taken of local circumstances and legal traditions. At the same time, this has also fostered the development of a coherent system of international protection for intellectual property rights in which there has been a steady and incremental harmonisation of norms. While this may be far from complete – and is certainly far removed from earlier universalist conceptions of intellectual property rights – transition and coherence are integral features of this dynamic and evolving system.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.014

9 Multiple and Overlapping Transitions in IP Marco Ricolfi1

9.1 introduction IP research does not often focus on transitions in legal regimes. This is hardly surprising. Usually lawyers, in analysing and interpreting legal rules, deal with the present and, sometimes, also with the future, by thinking up, discussing and advancing policy recommendations. Of course, in both cases some consideration of the past may be included. The extent of this historical input may vary, from scant to extensive, depending on the circumstances and the tastes of the scholar. Even so, the transition which led to the current situation or which is supposed to take us to the advocated legal reform is rarely given much more than cursory consideration. Thus, thinking about transitions in IP may hold some promise of opening up new vistas. There is a specific reason why we might learn from fresh consideration of IP transitions. When an IP legal regime is radically transformed, we may more than occasionally be taken by surprise. To make the point, I will consider two examples of recent transitions: the impact on IP of the advent of the internet; and the adoption of TRIPS. In retrospect, in both instances, the extent to which I was wrong-footed by the form change eventually took compared with my initial understanding and expectations now seems extraordinary. Here I talk about my own misunderstandings and misapprehensions, rather than those from which many other IP scholars suffered (my impression, however, is that I may find myself in good company). In this chapter, I will conclude by considering what I missed from the picture; and whether it is possible to use the lessons learned hereby to avoid, at least to a certain extent, incurring the same – or similar – mistakes in decoding subsequent IP transitions, including the one which would appear to be underway right now.

1

Chair of IP Law, Torino Law School.

Downloaded from https://www.cambridge.org/core. 121 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

122

Marco Ricolfi

9.2 two misunderstood transitions Let me try, then, to remind myself what was, in a nutshell, my original understanding of these two phenomena, both of which made for quite a grand upset of the previous legal regime. (a) Concerning the impact of the internet on IP, my idea was that the emergence of the digital environment had transformed the technical and social basis of creativity and of technological innovation. As far as copyright was concerned, the main feature I saw at the horizon was disintermediation.2 At long last, creators would be able to abandon the long route, where they had relied on the firm arm of a business to reach their public, and gain a chance to take the short route of a direct contact with end-users. The book and record industries were undergoing a process which announced the beginning of the end of the factory.3 The old business model whereby end-users paid businesses, and creators got a slice of the consideration, was increasingly proving outdated. In its place, two new modes of dissemination were emerging: outright gift by creators to end-users; and hybrid models, by and large imitating the successful operational mode of free software (free digital copies and payment for collateral goods and services). This development promised to have a significant impact on an essential feature of copyright: exclusivity. While exclusive legal title in the past was essential as an incentive both to create and to invest in the dissemination of creations, for creators choosing the short route it would no longer serve this purpose.4 Exclusivity could and would survive, but for different ends, in particular to enable creators to prevent unwanted commercial exploitation of their works by unwelcome third parties. (b) As far as TRIPS was concerned, my idea was that the adoption of this international compact would exacerbate the protectionist leanings of the global IP regime, which in most legal systems had become visible starting from the middle of the last century.5 2

3

4

5

See ex multis Marco Ricolfi, Individual and Collective Management of Copyright in a Digital Environment, in Paul Torremans (ed.), Copyright Law: A Handbook of Contemporary Research (Edward Elgar, 2008), 282 at 292–293. I first made this point back in 1996 in Marco Ricolfi, Da Chicago al Ciberspazio, in G. Clerico and S. Rizzello (eds.), Diritto ed economia della proprieta` intellettuale (Cedam, 1998), 83. When innovation and creativity are based on digital network-driven cooperation among a large number of contributors, and participants may be contributing ‘small grains’ of their time and attention, as typically happens in peer production and sharing, then the question of incentives itself becomes trivial, as has been shown convincingly: see Yochai Benkler, Coase’s Penguin, or, Linux and ‘The Nature of the Firm’, 112 Yale L.J. 369 (2002), at 376 ff. See Marco Ricolfi, Is There an Antitrust Antidote Against IP Overprotection within TRIPs?, 10 Marquette Intellectual Property Law Review 305–367 (2006); Marco Ricolfi, La tutela della proprietà intellettuale: fra incentivo all’innovazione e scambio ineguale, Riv. dir. ind., I,

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

Multiple and Overlapping Transitions in IP

123

In both cases I was wrong (at least to a large extent). On the first count, the internet, rather than announcing the dawn of a new, disintermediated environment of creativity, contributed to the emergence of new forms of re-intermediation, particularly social media platforms. In turn, if TRIPS indeed reinforced the previous protectionist trends, in retrospect it may be seen as a rather mild and benign arrangement when compared to the blatant overreach performed and planned by the IP sections of mega-regional agreements. Before asking myself what I had twice missed, and why this happened, a few words on the notion of transition adopted in this context may be in order.

9.3 on the notion of transition My goal here is not to try to think up a fully-fledged theory of transitions, or even of transitions in IP. My intention is much humbler: to clarify the meaning of and the conditions under which I use in this specific context the notion of transition, while trying to figure out what may go amiss in the effort of identifying the underlying driving forces and the implications of a transition from one legal IP regime to the next one. In a way, in setting this goal, I have already taken a position. Here the word ‘transition’ refers to a variation of legal regimes, to the process which leads from a starting point in legal rules to an outcome which is markedly different. The new regime which replaces the old one may be a different and novel institutional setting. In an even broader perspective, the notion of transition may encompass the change concerning systemic determinants, such as the general economic, technological and societal setting in which the old and new rules are placed. Also, in the perspective I have chosen here, the emphasis is placed both on the process which determines the outcome, the outcome itself and the larger, middle- to long-term implications of the change. It would also appear that the notion of transition, as employed here, has further determinants. In this setting, the concept has to be understood as indicating a phenomenon different from mere change and development. On the one hand, change and development hint at organic growth, at a process whereby the beginning and origin of a given legal regime still retain their hold into the later stages. In turn, the notion of transition has a component of discontinuity, the extent of which we may appreciate by comparing it to the (even more radical) idea of revolution.

511 (2002). In saying this I obviously was following well-established views: see among the many John Braithwaite and Peter Drahos, Global Business Regulation (Cambridge University Press, 2000), 56 ff.; Jerome H. Reichman, Legal Hybrids Between the Patent and Copyright Paradigms, 94 Columbia L. Rev. 2432 (1994); and The Globalization of Private Knowledge Goods and the Privatization of Global Public Goods, 7 Journal of Int. Ec. Law 279 (2004) (with Keith E. Maskus). Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

124

Marco Ricolfi

The idea of revolution is a comparatively recent concept.6 Originally it referred to the revolving of stars, ‘asters’, a process which entails an eternal return to the starting point. On the opposite, the three canonical modern revolutions – Copernicus, Darwin and Freud – take us to a point of non-return. The past is erased, rather than preserved. Now, transitions need not be as radical as revolutions; however, they share with them an essential characteristic: at the end of the process, we cannot move back to the starting point. As Hegel saw with characteristic acumen, after the split (die Trennung) there is no way to return to the undifferentiated harmony of the entire (des Ganzen); rather the two parts, before and after the split, come together into the Aufhebung, the German expression which includes both overcoming and deletion.7

9.4 missed in transition? Now, coming back to the question I left dangling earlier, what is it that one may ever so frequently miss in transitions – or, more specifically, that I lost in the two IP transitions I briefly sketched out above? If one is ever to learn a little something from past mistakes, I should pause here briefly to inquire what led me astray in connection with the internet and disintermediation once, and with TRIPS again. On reflection, it now seems to me that in both cases I missed out on a development which was taking place in parallel to the phenomenon I was studying. With the benefit of hindsight, we might now see that this development was bound to combine with the ongoing process in such a way as to lead to ultimate consequences markedly different from the ones which, at first glance, seemed to be on the horizon. More specifically: (a) As a technological development, initially the internet has, in fact, been relocating technological resources away from traditional industry into the hands of end-users.8 This has meant, indeed, decentralization: away from businesses and towards the direct nexus of creators and their public I just alluded to as ‘the short route’. However, experience teaches us that often communication technology waves start decentralized, as was the case with newspapers and radios, but end up being recentralized. Something of the kind would appear to have taken place with the internet. The wind changed when recentralization caught up: disintermediated creators were again taken up into the folds of a new 6

7

8

As remarked by current literature: see Juan Carlos De Martin, prefazione to Luciano Floridi, La rivoluzione dell’informazione (Codice edizioni, 2012), vii–xii. In this regard the idea of transition partakes of the features of transformation, in the meaning advanced by the great Karl Polanyi: Karl Polanyi, The Great Transformation: The Political and Economic Origins of our Time (1944) (Beacon Press, 1957). As noted by Yochai Benkler, Sharing Nicely: On Shareable Goods and the Emergence of Sharing as a Modality of Economic Production, 114 Yale L.J. 272 (2004), at 277.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

Multiple and Overlapping Transitions in IP

kind of industry, this time social media platforms.9 The implications of this shift for IP are huge: for the moment it is sufficient to say that whatever may happen to the role played by exclusivity as an incentive turns out to be less significant than I imagined, as de facto control of data and information is to a large extent decreasing the importance of the role played by formal titles of IP protection.10 (b) TRIPS was one of the creations of a world where trade institutions were fashioned after the multilateral model adopted at the end of World War II. This model has in the meantime come under stress, to a point where we end up looking back to TRIPS with some nostalgia and regret. True, compliance with TRIPS was not a piece of cake for the less powerful. Still, it was the result of a multilateral balancing exercise. Free trade agreements (FTAs) and, even more so, mega-regional agreements usually include an IP chapter. The imbalance is much bigger here than in TRIPS. Nations which control a large chunk of IP assets now have a chance to exert their power much more directly and forcefully. To the extent we conceive of IP as a balancing mechanism among multiple interests (of creators and innovators, other rightholders, users and subsequent generations of creators and innovators),11 the ratcheting up of the TRIPS niveau of protection by FTAs and mega-regionals raises concerns: in the meantime the scales have all too often been tipped to the benefit of rightholders. The reasons for the shift from multilateralism to bilateralism, which is eroding the role played by TRIPS, are multifarious. They may nevertheless be summed up by a simile advanced in a recent analysis: in the current landscape, international relations tend to take the shape of huband-spoke rather than the form of the multi-node framework.12 Against this background, we may come to realize that after all TRIPS had a silver lining which should not be neglected: it was the outcome of a multilateral balancing exercise rather than of unilateral exertion of power. 9

10

11

12

To my chagrin and shame, I should add that all the references to new intermediaries which appeared in the research I published in the field at the time were the result of (most appropriate) suggestions by two scholars, Federico Morando and Gustavo Ghidini, rather than of my own reasoning. This point is recurring: see Julie Cohen, Law for the Platform Economy, 51 U.C. Davis L. Rev. 133 (2017), at 148. For additional quotations supporting this position see Marco Ricolfi, The New Paradigm of Creativity and Innovation and its Corollaries for the Law of Obligations, in Peter Drahos, Gustavo Ghidini and Hanns Ullrich (eds.), Kritika: Essays on Intellectual Property, vol. i (Edward Elgar, 2015), 134–205, at 139 ff. The point is forcefully made by Hanns Ullrich, Select from Within the System: The European Patent with Unitary Effect, Max Planck Institute for Intellectual Property and Competition Law Research Paper No. 12-11, at 2–4. Pierluigi Fagan, Verso un mondo multipolare: Il gioco di tutti i giochi nell’era Trump (Fazi editore, 2017).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

125

126

Marco Ricolfi

9.4.1 The Omitted Overlapping Transitions To cut a long story short, what I seem to have missed in my understanding of the two transitions was in one case the technological shift under way, and in the other the restructuring of power relationships. The role played by technology and power: this is what was missing.13 Looking back, this may sound surprising. Here we are not talking about missing details; it is the macro-picture which was deficient, in that it missed an essential dimension – and defective, in that it obscured the ultimate direction of march of the ongoing process. On the other hand, these omissions are in a way understandable, if not excusable. While IP was undergoing one transition, a second, overlapping transition, respectively rooted in technology and power shifts, was under way. To tell the truth, this second, overlapping transition was in no way hidden; on the contrary, it was very visible, at least to the clear sighted. However, it was not in my area of direct vision; rather, it lay well below my radar screen. Which may, again, be understandable: as a lawyer, I tend to concentrate on law, rather than on technology and power. However, the result was a flagrant mistake in perspective.

9.5 on learning from past mistakes (and looking for the overlapping transition) Do we have a chance to learn from past mistakes? Normally the reply is in the affirmative; I am not so sure this is true in this connection, though, for reasons which I will elaborate at the end of the chapter. Nevertheless, we may still explore whether the lessons learned may help us out in the future, or, rather, in the present. Which, by the way, is a rather interesting moment, as everything in IP seems once more to be changing – or, rather, to stay in this frame of reference, ‘transitioning’. This prompts the following questions: is it possible to make a more educated guess as to the meaning and implications of the current transition in IP, on the basis of the hypothesis that the current legal transition also is likely to be intertwined with one (or more) separate but parallel transitions? If so, I should also ask, at which level are these overlapping transitions likely to be located (technology, power or some other driver)? I am sure that this question is rather a tall order. But, as the proof of the pudding is in the eating, maybe I should give it a try. Most of us would agree that the rate of change in the IP landscape is accelerating again. Of course, there are competing views on the direction of the current, rapid change. All of these are based on the idea that the digital infrastructure is taking us to the next stage, where the determinants that were not yet prevailing, say, a decade ago

13

The two are not separate: see Langdon Winner, Do Artifacts have Politics?, 109 Daedalus 121 (1980).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

Multiple and Overlapping Transitions in IP

127

are big data, machine learning, the internet of things (IoT) and artificial intelligence (AI). In a nutshell: the trappings of the algorithmic society.14 There are multiple ways to conceptualize the current transition and its meaning. Here I will (very tentatively) select two views, keeping in mind that the idea is to explore them as a preliminary step for identifying possible overlapping transitions which, while liable to be overlooked once more (as shown above), might, if correctly diagnosed, still prove essential for understanding the phenomena that are playing out around us. The first view deals with the impact on IP of the end of scarcity; the second focuses on the perils of abundance.

9.6 two views: the end of scarcity and the perils of abundance No question that Mark Lemley’s IP in a World Without Scarcity15 is one of the most notable contributions from one of the most remarkable American scholars. His main point is that digitization is just a beginning – ‘ce n’est qu’un début’, as Daniel CohnBendit used to say more or less half a century ago. In Lemley’s account, what I earlier dubbed as the end of the factory in publishing and music is seen as only a beginning; the ‘decoupling of informational content from manufacture and distribution’ is a process expanding to more or less proximate areas, from digital to 3D printing and bio-printing all the way to robots. In this perspective, where information in the form of bits interfaces at times with self-replicating genetic material, exclusivity is no longer a pillar of IP. This is so not only because exclusivity is no longer as important an incentive as it was in the past;16 the reason for this is that the main feature of this king’s cloth, enforceability, has been stripped, leaving the king naked. If we accept this view, the sweep of the transformation – or, in our parlance here, transition – we are presented with is mind-boggling. Originally the digital revolution had an impact on maybe 10 per cent of GDP: music, publishing, and, yes, advertising. But the impact Lemley refers to is perhaps three-quarters of the economy. Yet, in my view, the normative implications of this diagnosis are quite restrained and, all in all, laissez faire. In this novel context, the insistence on overprotection by business advocates is described as rather futile (which it is); a few caveats are advanced on issues such as forcing liability on platforms. The impression is that, in welcoming the partial demise of that particular regulatory tool which we describe as IP, Lemley ends up suggesting we should not rush to re-regulating too early.

14

15 16

For a forceful account see Jack Balkin, The Three Laws of Robotics in the Age of Big Data, 78 Ohio State L.J. 1217 (2017). 90 N.Y.U.L. Rev. 460 (2015). As suggested earlier in 9.2 and 9.4.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

128

Marco Ricolfi

I characterized the other view of the current transition as concerned with the perils of abundance. Lemley’s account, visionary as it is, does not yet place big data, IoT, algorithms, deep machine learning or AI (even though all this is to some extent factored into the section concerning the rise of robots) centre stage. In the meantime, these new inhabitants of the world of creativity and innovation have been explored intensively. What is striking in all these areas is the decrease of the role played by formal titles of IP protection. In the account provided by many scholars coming from quite different backgrounds and traditions,17 it would seem that many of the new players, particularly digital platforms (from Google, Apple, Facebook and Microsoft to Baidu, Alibaba and Tencent), forge ahead much more on the strength of de facto control over information resources than of legal title over them. Big data are stored away in the cloud, possibly a proprietary cloud. The algorithms which govern them as a way to analyse, control and steer behaviour are typically stashed away in black boxes.18 It is commonly accepted that this environment has seen the birth of information giants: rather than disappearing, intermediaries have taken the (not so novel) shape of platforms, often dominating their environment or ecosystem.19

9.7 normative implications Clearly the two views of the current transition I have tentatively sketched are not mutually exclusive. The end of scarcity may be seen as the outset of the age of abundance, rather than its negation. It seems to me, however, that the normative implications of the two narratives are rather different. Those who underline that in the current environment data, tools and AI are accumulated in few hands usually see what can be described as the ‘perils of abundance’, namely the dangers of de facto control of creativity and innovation by few players, and accordingly favour forms of re-regulation; while scholars focusing on scarcity and its end would seem to support a more hands-off approach. But then, what is meant here by re-regulation: why, when, how, by whom? As to the why: regulatory intervention postulates some form of market failure. One market failure which seems quite plausible in this context is the threat to sequential, downstream innovation. If few players are in a position to harness the tools I just 17

18

19

See Giulia Schneider, European Intellectual Property and Data Protection in the DigitalAlgorithmic Economy: A Role Reversal(?), 13 Journal of Intellectual Property Law & Practice 229–237 (2018), https://doi.org/10.1093/jiplp/jpx213; Cohen, Law for the Platform Economy, note 10 above, 148; Josef Drexl, Designing Competitive Markets for Industrial Data – Between Propertisation and Access, Max Planck Institute for Innovation and Competition Research Paper No. 16-13, at 29 ff. Peter Georg Picht and Benedikt Freund, Competition (Law) in the Era of Algorithms, Max Planck Institute for Innovation and Competition Research Paper No. 18-10, available at https:// ssrn.com/abstract=3180550 or http://dx.doi.org/10.2139/ssrn.3180550. The World’s Most Valuable Resource is no Longer Oil, but Data, The Economist, 6 May 2017.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

Multiple and Overlapping Transitions in IP

129

mentioned to observe the process of creation and innovation by unrelated third parties with what amounts to God’s eyes, it is conceivable that many a creative effort will be either promptly imitated, stopped or even gobbled up. Which is hardly surprising: information is a powerful competitive advantage. Also, accumulation of inordinate economic, social and political power may amount to a failure, albeit of a more general, institutional nature: as Senator Sherman clearly stated at the time of the adoption of the first antitrust legislation in the USA in 1890, what may be at stake is not just the functioning of competitive markets but the survival of democratic institutions,20 and, we may add today, of the democratic process itself.21 As to the when, lawyers’ usual reaction is that legislatures should not step in too early, when the facts on the ground have not yet taken a clear shape. Here a sense of urgency may lead in the opposite direction. In turn the question of regulatory mode, the how, allows a variety of approaches. In the USA, the interested parties seem to have made up their mind: as selfregulation (over State intervention) is the preferred mode, the chosen ground is the field of ethical debate rather than of outright political discussion. Opponents of the dominance of digital platforms seem to be doubtful about both options;22 they have a number of difficulties in devising alternatives, though. When J. K. Galbraith wrote his book The New Industrial State in 1967, it was clear both that the industrialmilitary complex required the opposing forces of powerful countervailing powers and that these were to be found in trade unions, consumer associations and democratic forces. Now, opponents have no clear view, nor can they, on the crucial issue of identifying which interests and forces are plausible candidates to become countervailing powers to the present powers that be. True, if the problem is de facto control of informational resources, ideally the solution of necessity has to be sought in ways to open up access to the same, foster the creation and preservation of informational commons and combine openness and interoperability with decentralization. All of which are probably worthy and even laudable ideas, theoretically, that nonetheless ring hollow unless social actors are identified – or more aptly rise up – who want to embrace them.

20

21

22

Senator Sherman said in a powerful speech made on the floor of the US Congress during the passage of the 1890 Act bearing his name, ‘If the concentrated powers of this combination are intrusted in a single man, it is a kingly prerogative, incompatible with our form of government’: see D. Millon, The Sherman Act and the Balance of Power, in E. T. Sullivan (ed.), The Political Economy of the Sherman Act: The First One Hundred Years (Oxford, 1991), 111. By this I mean to refer also to freedom of expression: see Josef Drexl, Economic Efficiency versus Democracy: On the Potential Role of Competition Policy in Regulating Digital Markets in Times of Post-Truth Politics, Max Planck Institute for Innovation and Competition Research Paper No. 16-16. See Brett Frischmann and Evan Selinger, Re-engineering Humanity (Cambridge University Press, 2018), at 52 ff., 136 ff. and 269 ff.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

130

Marco Ricolfi

9.8 in quest of the current overlapping transition Luckily, my task here is not to advance even tentative answers to these quite intractable questions. Rather, the object of the current effort is a different one (if by no means easier). Here we are not trying to figure out which of the two narratives I just referred to is right, but what dimension may still be missing in both of them. In other words, the question is whether, assuming these two narratives give, jointly or severally, a plausible account of the transformation of IP which is under way, there is also a parallel, overlapping transition which, while hidden from plain view, we should detect and factor into the analysis if we are to make sense of what is really going on. What I am asking here amounts to inquiring whether the incorporation into the equation of the variables which come from a different field may help us this time to avoid or attenuate the fallacies admittedly incurred in the past, in the effort of understanding the impact of the internet on IP and the silver lining of TRIPS. My hunch here is that the reply is in the affirmative: there is an overlapping transition which is under way. Of course, I do not know much of that other, overlapping transition. However, I am deeply impressed by a number of features which are playing out while the current IP transition unfolds. All these seem to turn around the issue I left out when discussing the missing coordinates of re-regulation: that is, the question concerning the who should regulate. Dealing with this issue we face a genuine geopolitical question. Few of the relevant regulatory issues raised in the current IP transition may be dealt with at the national or regional level. It is conceivable that access and transparency issues raised by big data may be tackled at the regional level, where, in fact, the EU is starting to resort to the tools provided by antitrust and data protection rules.23 I doubt, however, that the same regional level may be considered relevant in connection with the issues raised by algorithmic governance, the IoT and AI. One of the reasons why regional lawmakers are hardly up to the task is, as I just hinted, exquisitely geopolitical. Even if we confine our analysis to AI, we must take into account that any meaningful regulation of the phenomenon must be global, as AI is by definition deployable everywhere. But at the same time we cannot forget that over half of the investment in unmanned vehicles – terrestrial or airborne – and more generally in AI, is military (or, more exactly, concerns objects and acts which are amenable to dual, civilian and military uses). So, while regulation of AI (and algorithmic governance) has no chance of succeeding at a level which is not global, there is also no chance that the competing world powers, be they two or more, are going to give in to solutions which they find unacceptable, except via negotiations akin to the ones which in 1991 led to the strategic nuclear disarmament treaty (START).24 23 24

For a sceptical view, however, see Drexl, note 17 above, at 53 ff. While I subscribe to the approach chosen by Graeme Dinwoodie, Rochelle Dreyfuss and Annette Kur, The Law Applicable to Secondary Liability in Intellectual Property Infringement Cases, 42 NYU JILP 201 (2009), advocating at 225 ff. a global standard to deal with intermediary

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

Multiple and Overlapping Transitions in IP

131

9.9 three cheers for anti-disciplinary research The interesting thing, which leads me to my final remarks, is that of course I may be totally wrong. This is not only for the usual reason, the cognitive and intellectual limitations of any particular scholar (particularly one who admits to having had it wrong so many times as I proved to my chagrin at the start of this chapter). The reason is more specific, and it has to do with the disciplinary limitations which any legal scholar is subject to. How could we possibly be astute enough? The fact is that legal scholarship is rooted in a very specific tradition. The mission of legal academics is to teach. As their students aim to become legal professionals – be it attorneys, judges, counsel within public or private organizations – or fill jobs at the intersection of law and some other vocation, the subject matter of their endeavour is the law. Usually their understanding is not only confined to the law, as lawyers must by definition develop abilities to grasp the non-legal side of the questions that they will be asked to contribute to clarifying and eventually solving. It is nevertheless rare that a specific lawyer masters the fundamentals of one, or more than one, discipline other than law. This limitation is probably sound: as Goethe said, ‘in der Beschränkung zeigt sich erst der Meister’. However sound this approach may be under ordinary circumstances, it is bound to fail us when we are confronted with questions of the nature and order of magnitude that characterize the issues I have rather clumsily tried to deal with here (which in a way explains why in facing them I harbour a feeling of inappropriateness, or even of risking making a fool of myself ). Except it turns out that dealing with issues such as the ones at hand is not so much a matter of choice as of necessity: these are exactly the issues that are forced on us by the fraught times we are living in. As I have said, I do not know whether the tentative framing of the questions I have asked and the initial steps I have taken in the attempt to work out tentative replies have any plausibility at all. I submit, however, that the road undertaken has some merit. Here we have been dealing with IP transitions: the impact of the internet on IP, the assessment of the role played by TRIPS in the progressive ratcheting up of IP protection, the role played by IP in an increasingly algorithmic society. The tentative conclusion I seem to have reached is that it is not particularly helpful to deal with only the legal dimensions of such transitions, if we do not attune our ear to detect and explore the parallel and overlapping dimensions which accompany them at different, mostly non-legal levels. This additional layer of change may from time to time be determined by factors such as technology, power or geopolitics. Lawyers will have a hard time if they fail to address this second layer of inquiry, and are likely to be wrong-footed if their analysis is confined to the legal surface. Except that

liability in IP violations, I suspect that the difficulties involved in the process are very hard to overcome both in the area they deal with and, to an even larger extent, in connection with the issue discussed in the text. Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

132

Marco Ricolfi

normally lawyers are not in a position to address the additional layers, whatever their disciplinary coordinates may be. There is a variety of possible solutions to this stalemate. The most obvious one is to invest in interdisciplinary research. Cooperation between teams which include lawyers, rather than being confined to them, is on the increase. Law and economics, whatever its limitations may be, has helped us to move on. Empirical evidence is gradually being taken up by lawyers. Computer scientists and life science experts are increasingly interacting with lawyers. However, even interdisciplinary projects have their limitations. The most obvious one is that the very research question which must be at the centre of any research project tends to be defined starting from one specific disciplinary basis, and by adding on to it the dimensions coming from other disciplines. My hunch is that this approach, which we may call interdisciplinary, multidisciplinary or transdisciplinary, is bound to fail when the definition of the research question originates from only one field of research and additional perspectives from other fields are just added to or superimposed on it. There is an alternative, though, which is to go the whole hog and adopt an anti-disciplinary approach. Which means that research questions should be framed not from one specific disciplinary angle, but, from the onset, from a plurality of different disciplinary perspectives. This approach cannot be undertaken by a single scholar. It requires an appropriate institutional setting. Researchers from different backgrounds must work together under the same roof. Then, and only then, may rather intractable issues, like the ones I have tried to deal with here – which revealed that some IP transitions may not be figured out unless an overlapping transition is simultaneously addressed – have any hope of being appropriately dealt with.25 Of course this approach entails multiple questions: funding and freedom of research come first, but these cannot be addressed here.26

25

26

The impact of 5G telephony and of the emergence of quantum computers could be mentioned in considering an additional dimension of complexity which should be incorporated in any future analysis of IP transitions. What I may add here, however, is that much of what I say in this final paragraph was at the basis of the birth, and is part of the ongoing work, of the Nexa Center for Internet & Society, Politecnico di Torino (https://nexa.polito.it/), which I co-founded with Professor J. C. De Martin over ten years ago and which I currently co-direct with him.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.015

10 Transition and Continuity in the Private International Law of Intellectual Property Christian Heinze1

10.1 introduction Academically, Annette Kur is widely known as an intellectual property scholar. She has published extensively in the field, in particular on trade mark, design and unfair competition law, and her contributions are without doubt among the most fundamental writings on these subjects in the past thirty years. Against her achievements in IP, it may sometimes be less visible that Annette Kur is also a conflicts lawyer.2 The following contribution will try to sketch transition and continuity in the private international law of intellectual property since the 1990s – and the important contributions Annette Kur made in the different stages. To begin, however, I would like to express my personal and deeply felt gratitude: Annette Kur has always striven to invite, support, guide and give credit to the work of younger scholars. I profited immensely from working with her, and I am grateful for this privilege.

10.2 cautious steps towards a transition from strict territoriality to some universality in the 1990s Since the origins of modern intellectual property law, the private international law in this field has been dominated by the principles of lex loci protectionis3 and – on 1 2

3

Professor of civil law and intellectual property, Leibniz University Hanover. Apparently, she regarded international civil procedure as an ‘esoteric’ matter ‘difficult to access’, until she developed an interest while looking into the proposals for a Hague Judgment Convention; see A. Kur, Optionen für ein internationales Gerichtsstands- und Vollstreckungsübereinkommen im Bereich des Immaterialgüterrechts, in E. Keller, C. Plassmann and A. von Falck (eds.), Festschrift fu¨r Winfried Tilmann zum 65. Geburtstag (Carl Heymanns Verlag, 2003), 827–842 at 827. See J. Basedow in: European Max Planck Group on Conflict of Laws in Intellectual Property, Conflict of Laws in Intellectual Property: The CLIP Principles and Commentary (Oxford University Press, 2013), Article 3:102.C01-C02 with further references. For the (historically) broad understanding of exclusive jurisdiction of the protecting State for disputes over

Downloaded from https://www.cambridge.org/core. 133 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

134

Christian Heinze

the level of substantive law – territoriality.4 Both lex protectionis and territoriality are so firmly established that there is hardly any debate without reference to them. Because of the principle of territoriality and of the peculiarities of intellectual property, the conflict rules in this field (in a broad sense, including the rules of jurisdiction) were often considered as separate and distinct from general private international law. Accepted notions of general conflicts doctrine, such as party autonomy for (contractual and non-contractual) obligations, general jurisdiction for (in principle) all disputes at the defendant’s domicile, or restricted interpretation of exclusive jurisdiction rules were rejected or at least restricted for intellectual property disputes. Instead, many actors, in particular from practice, claimed a special place for intellectual property in conflicts doctrine, arguing for exclusion from general private international law instruments, or at least for special rules. This concept – strict territoriality and lex protectionis both in choice of law and even in jurisdiction,5 coupled with a perception of conflict of laws in intellectual property being a distinct and separate matter from general conflict of laws doctrine – came under pressure at the end of the 1990s. First, the rise of the internet and the internet-based distribution of protected works, signs or even inventions challenged the traditional notion that each dispute has to (and can) be split up along State boundaries for choice of law and even jurisdiction. While territoriality may have worked in a scenario of five to ten jurisdictions involved, it had to fail in practice where content was available online for more than a hundred jurisdictions. In such a

4

5

intellectual property rights, see the references in C. Heinze and E. Roffael, Internationale Zuständigkeit für Entscheidungen über die Gültigkeit ausländischer Immaterialgüterrechte, [2006] GRUR Int 787 at 787, fn. 2–6. For ownership of non-registered rights, in particular copyright, the application of the lex protectionis principle is sometimes doubted; see Basedow (this note), Article 3:102.N02 with further references. On territoriality, see Case C-9/93 IHT v. Ideal Standard ECLI:EU:C:1994:261 at [22] (trade mark law); Case C-192/04 Lagardère v. SPRE ECLI:EU:C:2005:475 at [46] (copyright); Case C523/10 Wintersteiger AG v. Products 4U Sondermaschinenbau GmbH ECLI:EU:C:2012:220 at [25] (trade mark law); Case C-173/11 Football Dataco v. Sportradar ECLI:EU:C:2012:642 at [27] (database right); Case C-441/13 Hejduk v. EnergieAgentur NRW ECLI:EU:C:2015:28 at [22] (copyright). On lex protectionis, see recital 26 Rome II: ‘universally acknowledged principle of the lex loci protectionis’. For an extensive analysis, see L. Lundstedt, Territoriality in Intellectual Property Law (doctoral thesis in private law, Stockholm University, 2016). For the traditional belief that disputes concerning intellectual property rights have to be litigated in the State of protection, see J. Kohler, Urheberrecht an Schriftwerken und Verlagsrecht (Ferdinand Enke, 1907), 393; further references in Heinze and Roffael, note 3 above. In England, it has been discussed since the 2000s whether claims based on (infringement of ) foreign intellectual property rights are justiciable in England: Lucasfilm v. Ainsworth [2009] EWCA Civ 1328 at [174]–[183], overruled by Lucasfilm v. Ainsworth [2011] UKSC 39 at [105]–[110]. For a similar discussion in other common law jurisdictions see TS Production LLC v. Drew Pictures Pty Ltd [2008] FCAFC 194 at [16]–[17] (Australia); Gallo Africa v. Sting Music [2010] ZASCA 96 at [19] (South Africa). See also Voda v. Cordis, 476 F 3d 887 at 898–904 (Fed Cir 2007) (US) (Federal courts cannot exercise supplemental jurisdiction over foreign patent rights); more liberal in the field of copyright, Boosey v. Walt Disney, 145 F 3d 481 at 493 (1998) (US) (reversing a forum non conveniens dismissal of actions based on foreign copyright).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

The Private International Law of IP

135

situation, the cost of ascertaining the applicable law and its content for all jurisdictions potentially involved becomes prohibitive, leading in practice to a concentration on a few select jurisdictions to which the plaintiff limits his or her claim to protection. While this pragmatic solution makes the lex protectionis principle workable in multistate cases, it does not solve all problems. For example, in the case of injunctive relief it may not always be possible to limit the plaintiff’s remedy along national boundaries. It therefore does not come as a surprise that the critics of territoriality and the country of protection principle – which existed well before the internet – became more vocal as the use of the internet became common.6 A second challenge to lex protectionis and territoriality was the application of general private international law doctrines to intellectual property. In the 1990s, courts and litigators in different European countries began to use the potential of the Brussels Convention to secure cross-border injunctions. Under the Brussels Convention (and also the Brussels Regulation 2001), exclusive jurisdiction of the State of protection (or registration) is only recognised ‘in proceedings concerned with the registration or validity of patents, trade marks, designs, or other similar rights required to be deposited or registered’.7 For all other disputes, the general rules of jurisdiction apply, which allow a consolidation of actions for several States of protection at the defendant’s domicile, at the place where one of several defendants is domiciled, at the place on which the parties agreed in a jurisdiction agreement, and possibly also at the place where the harmful act was committed (i.e. the place of the event giving rise to the damage).8 On the level of choice of law, the inclusion of intellectual property contracts in the Rome I Convention (now Rome I Regulation) promoted a further departure from the traditional focus on State of protection and territoriality.9 Not the lex protectionis, but rather the law chosen by the parties or the law of the State where the party effecting the characteristic performance has its habitual residence governs the contractual relationship between the parties, even if

6

7

8 9

See the discussion in WIPO in the 1990s, http://www.wipo.int/meetings/en/details.jsp?meet ing_id=4243. Article 16(4) Brussels Convention, Article 22(4) Brussels I Regulation 2001. The text in the current text of Article 24(4) Brussels Ibis Regulation has been amended to codify the GAT/LuK decision – see note 19 below. See Articles 4(1), 8 No. 1, 25 and 7 No. 2 of the Brussels Ibis Regulation 2012. As a consequence of a sentence in the preparatory Report on the Convention on the law applicable to contractual obligations by Mario Giuliano and Paul Lagarde, OJ C 282, 31.10.1980, p. 1 (‘First, since the Convention is concerned only with the law applicable to contractual obligations, property rights and intellectual property are not covered by these provisions’), it was not entirely clear whether and to what extent the Rome Convention would apply to contracts concerning intellectual property rights. The phrase was widely understood as excluding only non-contractual and proprietary aspects of intellectual property rights from the Rome Convention, while contractual issues were governed by it; see European Max Planck Group for Conflict of Laws in Intellectual Property (CLIP), Intellectual Property and the Reform of Private International Law – Sparks from a Difficult Relationship, [2007] IPRax 284 at 288.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

136

Christian Heinze

it involves intellectual property rights in different jurisdictions.10 While the possibility to choose the law governing the contractual obligations may have been accepted prior to the Rome Convention, some courts tended to prefer an application of the lex protectionis also to contractual issues where no choice of law had been concluded, even if it involved separating the applicable law to a license contract depending on the intellectual property right concerned.11 A third development around the turn of the millennium was the negotiations which started in 1992 on a general Hague Convention on jurisdiction, recognition and enforcement of judgments in civil and commercial matters.12 While the exact contours of this Convention remained heavily disputed until its eventual failure in 2001/02,13 it seems plausible to assume that its adoption could have served as a catalyst for consolidation of intellectual property litigation for several States of protection in a single forum, just as the Brussels Convention had for Europe. Such a consolidation of cross-border actions in a single jurisdiction would have affected not only jurisdictional, but also choice of law issues: only a single law, the lex fori, would have governed all procedural matters, and only one regime of choice of law rules would have been applied. Moreover, a court faced with a multi-country litigation is likely to find ways to make the application of different laws under the lex protectionis principle manageable. For example, the court may require the parties to help in determining the applicable foreign laws, or simply apply (directly, or more likely indirectly) notions familiar from its own law even in the context of a foreign law. Annette Kur was among the first scholars to realise the impact of all three developments. As early as 1996, she discussed the protection of names and signs in cyberspace and saw a necessity to act at the international level.14 She returned to this subject in several papers that highlight the ‘Territorialität versus Globalität’ conflict.15 She also discussed the possible impact of a global (general) Hague 10 11

12 13

14

15

See Articles 3(1), 4(2) Rome I Regulation and Articles 3(1), 4(1), (2) Rome Convention. For an example, see Oberlandesgericht Düsseldorf 4.8.1961, Case 2 U 66/61, Tubenverschluss [1962] GRUR Ausl 256 at 257. In other cases, courts applied the law of the transferor/licensor or transferee/licensee; see A. Metzger in: European Max Planck Group on Conflict of Laws in Intellectual Property, Conflict of Laws in Intellectual Property – The CLIP Principles and Commentary, note 3 above, Article 3:502.N04–N08 with references. See https://www.hcch.net/en/instruments/conventions/publications1/?dtid=35&cid=98. For details, see Kur, Optionen für ein internationales Gerichtsstands- und Vollstreckungsübereinkommen im Bereich des Immaterialgüterrechts, note 2 above, at 827–828. Untitled case note, [1996] Computer und Recht 590; see also her work on the WIPO proposals for trade mark conflicts on the internet, e.g. in Die WIPO-Vorschläge zum Schutz notorisch bekannter und berühmter Marken, [1999] GRUR 866 and Die WIPO-Empfehlungen zur Benutzung von Marken im Internet, [2001] GRUR Int 961. [2000] Wettbewerb in Recht und Praxis 935; Trademarks Conflicts on the Internet: Territoriality Redefined?, in J. Basedow, J. Drexl, A. Kur and A. Metzger (eds.), Intellectual Property in the Conflict of Laws (Mohr Siebeck, 2005), 175–193. See the Joint Recommendation Concerning Provisions on the Protection of Marks, and Other Industrial Property Rights in Signs, on the Internet (with Explanatory Notes), adopted by the

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

The Private International Law of IP

137

Convention on intellectual property where she clearly saw the parallels to the Brussels regime and argued for a limited role of exclusive jurisdiction for validity disputes.16 After the broad concept of a general Hague Convention had failed in 2001/02, she participated in conferences and groups that laid the foundation for the later work of larger groups producing principles that could serve as the basis of a European or even international instrument on conflict of laws in intellectual property. In fact, she was one of the few people who participated in groups on both sides of the Atlantic that later produced a comprehensive set of principles, with comments and notes, on private international law for intellectual property.17

10.3 the affirmation of the traditional doctrine in the 2000s After the millennium had turned with hopes for a more flexible approach to conflict of laws in intellectual property, the traditional notions of territoriality and lex protectionis were affirmed in the early 2000s by courts and legislators. A first step towards a strict understanding of lex protectionis was the adoption of the Rome II Regulation in 2005. While the inclusion of a specific conflict rule for intellectual property in Article 8 Rome II was justified, the exclusion of any form of party autonomy in Article 8(3) Rome II goes too far. Such a strict understanding of territoriality excludes the parties’ choice of law even in matters such as remedies where a limited choice may be acceptable. It even complicates an ex post choice of law in the course of a pending infringement dispute that can be useful to reduce the cost and complexity of the litigation by limiting it to one (or a few) applicable laws.18

16

17

18

Assembly of the Paris Union for the Protection of Industrial Property and the General Assembly of the World Intellectual Property Organization (WIPO), 24 September to 3 October 2001, https://www.wipo.int/publications/en/details.jsp?id=345&plang=EN. Immaterialgüterrechte in einem weltweiten Vollstreckungs- und Gerichtsstandsübereinkommen – Auf der Suche nach dem Ausweg aus der Sackgasse, [2001] GRUR Int 908 at 910–913; International Hague Convention on Jurisdiction and Foreign Judgments: A Way Forward for IP?, [2002] EIPR 175; Optionen für ein internationales Gerichtsstands- und Vollstreckungsübereinkommen im Bereich des Immaterialgüterrechts, note 2 above. She participated in the working group in the Munich Max Planck Institute for Intellectual Property, Competition and Tax Law which was founded in 2002 (see A. Drexl and A. Kur (eds.), Intellectual Property and Private International Law – Heading for the Future (Sweet & Maxwell, 2005)) and in the later joint working group of the Munich Institute and the Hamburg Max Planck Institute for Comparative and International Private Law which was founded in late 2004. This work led to the adoption of the European Max Planck Group on Conflict of Laws in Intellectual Property (CLIP) (Conflict of Laws in Intellectual Property: The CLIP Principles and Commentary, note 3 above). She was also advisor to the project of the American Law Institute which led to the adoption of the ALI Principles, Intellectual Property: Principles Governing Jurisdiction, Choice of Law, and Judgments in Transnational Disputes (American Law Institute, 2008). In my opinion, the exclusion of choice of law for non-contractual obligations under Article 8(3) Rome II can be avoided if the parties conclude a contractual agreement (e.g. a license contract or a settlement agreement) with retroactive effect. This renders the conduct lawful ex post, so

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

138

Christian Heinze

While the exclusion of party autonomy for non-contractual matters in the Rome II Regulation was a setback, the jurisprudence of the Court of Justice of the European Union (CJEU) in the GAT/LuK and Roche Nederland cases was more than that. In GAT/LuK, the CJEU decided that exclusive jurisdiction for validity disputes on registered IP rights (Article 22(4) Brussels I Regulation 2001) covered also scenarios where the invalidity was raised as a mere defence to an infringement action.19 As a consequence, any infringement action outside the State of protection risked an ‘invalidity’ defence (even after the proceedings had started) which would either exclude the jurisdiction of the courts of any State other than the State of protection, or at least require a stay of proceedings for the validity dispute to be decided by the courts of the country of protection. In Roche Nederland, the CJEU denied a consolidation of parallel infringement actions against defendants from the same corporate group if these actions were based on parallel, but separate national patents in different countries.20 In a similar vein, the US Court of Appeals for the Federal Circuit held that US Federal courts cannot exercise supplemental jurisdiction over foreign patent claims.21 The affirmation of lex protectionis and territoriality did not come as a surprise. Both principles fulfil important functions in safeguarding the regulatory (economic, social, etc.) policies of the State granting the intellectual property right. In the context of registered rights, they are also justified by the fact that the registration is also governed by the lex protectionis. If the existence, scope and limits of intellectual property rights reflect regulatory interests of the State granting these rights, it is legitimate to respect these interests within this State’s borders, but not beyond. Moreover, it was difficult to arrive at a better alternative. While possible alternatives had been intensely discussed in academic writing, there was no solution apparent which could avoid the problems of lex protectionis without raising other issues.22

19

20

21

22

that there is no delict; see C. Heinze, Article 8, in M. Würdinger (ed.), juris PraxisKommentar BGB, vol. vi, Internationales Privatrecht und UN-Kaufrecht (8th ed.) (C. H. Beck, 2017), Article 8, para 51. For contracts, including license contracts, choice of law is possible under Article 3 Rome I Regulation. Parties should, however, avoid in their agreement any reference to choice of law for a non-contractual agreement. Case C-4/03 Gesellschaft für Antriebstechnik mbH & Co. KG v. Lamellen und Kupplungsbau Beteiligungs KG ECLI:EU:C:2006:457 at [25]. Case C-539/03 Roche Nederland BV v. Frederick Primus und Milton Goldenberg ECLI:EU: C:2006:458 at [33]. Voda v. Cordis, 476 F 3d 887 at 898–904 (Fed Cir 2007) (US). It is uncertain whether this decision excludes only supplemental jurisdiction of the Federal courts or jurisdiction of US courts (including state courts) over foreign patent rights in general; see M. Schauwecker, Extraterritoriale Patentverletzungsjurisdiktion: Die internationale Zusta¨ndigkeit der Gerichte ausserhalb des Patenterteilungsstaates fu¨r Verletzungsverfahren (Carl Heymanns Verlag, 2009), 345–348. For details, see J. Drexl, Internationales Immaterialgüterrecht, in F. J. Säcker, R. Rixecker, H. Oetker and B. Limperg (eds.), Mu¨nchener Kommentar zum Bu¨rgerlichen Gesetzbuch, vol. xii (C. H. Beck, 7th ed., 2018), paras 337–344.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

The Private International Law of IP

139

However, recognising that territoriality and lex protectionis play an important role in the private international law of intellectual property does not mean that these concepts must be absolute. In particular in the field of jurisdiction and international civil procedure, other considerations come into play that justify granting jurisdiction to courts outside the State of protection. In the discussions during the 2000s, Annette Kur elaborated on her earlier studies and identified both the value and the limits of the principles of lex protectionis and territoriality. She recognised the role that both principles have to play regarding issues such as existence, validity, scope and infringement of intellectual property rights. But she also identified countervailing principles such as (procedural and transactional) efficiency, party autonomy, legal security and foreseeability or fairness and equity, including the protection of weaker parties, which justify both granting jurisdiction to and applying the law of States other than the country of protection.23 Her writings24 helped to pave the way to a more nuanced and detailed approach to conflict of laws in intellectual property which was eventually implemented in different sets of principles adopted in the USA, Europe and East Asia towards the end of the 2000s.25 The general thrust of 23

24

25

A. Kur, Are there any Common European Principles of Private International Law with regard to Intellectual Property?, in S. Leible and A. Ohly (eds.), Intellectual Property and Private International Law (Oxford University Press, 2009), 1 at 6–7. For example her paper on the GAT/LuK and Roche Nederland decisions, A Farewell to CrossBorder Injunctions? The ECJ Decisions GAT v. LuK and Roche Nederland v. Primus and Goldenberg, 37 IIC 844 (2006); see also the joint and parallel publications of the CLIP Group, Intellectual Property and the Reform of Private International Law – Sparks from a Difficult Relationship, [2007] IPRax 284, Exclusive Jurisdiction and Cross-border IP (Patent) Infringement: Suggestions for Amendment of the Brussels I Regulation, [2007] EIPR 194, Comments on the European Commission’s Proposal for a Regulation on the Law Applicable to Contractual Obligations (Rome I) of 15 December 2005 and the European Parliament Committee on Legal Affairs’ Draft Report on the Proposal of 22 August 2006, 38 IIC 471 (2007), or on the law applicable to secondary liability in intellectual property cases, G. Dinwoodie, R. Dreyfuss and A. Kur, The Law Applicable to Secondary Liability in Intellectual Property Cases, 42 N.Y.U. J Int’L L. & Pol 201 (2009). The American Law Institute (ALI), Intellectual Property: Principles Governing Jurisdiction, Choice of Law, and Judgments in Transnational Disputes, note 16 above; (Japanese) Transparency Proposal on Jurisdiction, Choice of Law, Recognition and Enforcement of Foreign Judgments in Intellectual Property; (Japanese-Korean) Principles of Private International Law on Intellectual Property Rights (Joint Proposal Drafted by Members of the Private International Law Association of Korea and Japan (Waseda University Global COE Project); European Max Planck Group on Conflict of Laws in Intellectual Property (CLIP), Conflict of Laws in Intellectual Property: The CLIP Principles and Commentary, note 3 above. For earlier proposals see J. Ginsburg and R. Dreyfuss, Draft Convention on Jurisdiction and Recognition of Judgments in Intellectual Property Matters, 77 Chi.-Kent L. Rev. 1065 (2002); see also J. A. L. Sterling, International Codification of Copyright Law: Possibilities and Imperatives, 33 IIC 270 (2002). For a comparative analysis of the ALI and CLIP Principles, see A. Kur and B. Ubertazzi, The ALI Principles and the CLIP Project: A Comparison, in S. Bariatti (ed.), Litigating Intellectual Property Rights Disputes Cross-border: EU Regulations, ALI Principles, CLIP Project (CEDAM, 2010), 89–147; for a summary of the CLIP Principles see A. Kur, The CLIP Principles – Summary of the Project, 12 The quarterly review of Corporation Law and Society 202 (2011) and A. Kur,

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

140

Christian Heinze

these principles, with differences in detail, was similar26: while the important role of territoriality and lex protectionis is recognised, all principles try to give appropriate weight to countervailing principles which limit the role of the protecting State and its law. The principles and academic proposals do not try to abolish the principles of lex protectionis and territoriality, but rather limit their effects in an appropriate way,27 either on the level of jurisdiction, of choice of law or even of substantive law, by requiring certain minimum effects in the protecting State or by reducing28 the extraterritorial application of intellectual property laws. These principles were not without impact on court practice: not only did the UK Supreme Court refer to the academic projects when it adopted the important Lucasfilm v. Ainsworth judgment,29 which permitted jurisdiction for copyright infringement disputes outside the State of protection; the impact of Annette Kur’s writings and of the CLIP Principles to which she made important contributions was also reflected in different opinions of Advocate Generals at the CJEU: in at least seven opinions – interestingly all on issues of jurisdiction – different Advocate Generals cited the CLIP Principles or works which discussed these principles.30 As the CJEU does not cite doctrine, it can only be speculated how far it may have been influenced by the CLIP Principles. It is noteworthy, though, that the Court tended to relax its strict understanding of territoriality in later decisions on jurisdiction and gave greater leeway to consolidation of actions or preliminary measures

26

27

28

29 30

Die Ergebnisse des CLIP-Projekts – zugleich eine Einführung in die deutsche Fassung der Principles on Conflict of Laws in Intellectual Property, [2012] GRUR Int 857. See A. Kur, Are There Any Common European Principles of Private International Law with Regard to Intellectual Property?, in S. Leible and A. Ohly (eds.), Intellectual Property and Private International Law (Oxford University Press, 2009), 1–14 at 13: different principles ‘are not so far apart’. See J. Drexl, Internationales Immaterialgüterrecht, in Mu¨nchener Kommentar zum Bu¨rgerlichen Gesetzbuch, note 22 above, para 344. As a contrary development, it is also conceivable that the extraterritorial reach of court orders is expanded to avoid the problem of applying foreign intellectual property law. Lucasfilm v. Ainsworth [2011] UKSC 39 at [94]. Opinion of Advocate General Trstenjak delivered on 12 April 2011, Case C-145/10 Eva-Maria Painer v. Standard VerlagsGmbH and Others ECLI:EU:C:2011:239 in fn. 31; Opinion of Advocate General Cruz Villalón delivered on 16 February 2012, Case C-523/10 Wintersteiger AG v. Products 4U Sondermaschinenbau GmbH ECLI:EU:C:2012:90 in fn. 7 (citing an article of C. Heinze on the CLIP Principles); Opinion of Advocate General Cruz Villalón delivered on 29 March 2012, Case C-616/10 Solvay SA v. Honeywell Fluorine Products Europe BV and Others ECLI:EU:C:2012:193 in fn. 24 (citing a group publication by CLIP); Opinion of Advocate General Jääskinen delivered on 13 June 2013, Case C-170/12 Peter Pinckney v. KDG Mediatech AG ECLI:EU:C:2013:400 in fn. 53; Opinion of Advocate General Cruz Villalón delivered on 11 September 2014, Case C-441/13 Pez Hejduk v. EnergieAgentur.NRW GmbH ECLI:EU:C:2014:2212 in fn. 26; Opinion of Advocate General Wathelet delivered on 9 November 2016, Case C-618/15 Concurrence Sàrl v. Samsung Electronics France SAS and Amazon Services Europe Sàrl ECLI:EU:C:2016:843 in fn. 3; Opinion of Advocate General Szpunar delivered on 20 June 2018, Case C-379/17 Società Immobiliare Al Bosco Srl ECLI:EU: C:2018:472 in fn. 21 (citing an article of P. de Miguel Asensio on the CLIP Principles).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

The Private International Law of IP

141

outside the State of protection.31 Unfortunately, the European legislator amended the ground of exclusive jurisdiction in Article 24(4) Brussels Ibis Regulation in 2012 to codify the unfortunate GAT/LuK decision and thereby petrified the development in this field. It can only be hoped that this mistake will soon be corrected.

10.4 where now from here? the new hague conventions, eu unitary rights, frand litigation It remains to be asked whether the current state of affairs is likely to last for the foreseeable future, or whether we might expect further development towards more universalism, and, if so, in which way. On the level of academic projects, it may be mentioned that a committee of the International Law Association is working on intellectual property and private international law.32 This work aims at drafting, on the basis of comparative studies of pre-existing projects, ILA Guidelines on jurisdiction, applicable law, and recognition and enforcement. On the level of international law, the Hague Conference on Private International Law – after the failure of a general Judgments Convention – scaled down its project and focused on international cases involving exclusive choice of court agreements where consensus was easier to reach. The fruit of this work is the Hague Convention of 30 June 2005 on Choice of Court Agreements.33 The Convention has only limited impact on intellectual property disputes as it applies only to exclusive choice of court agreements concluded in civil or commercial matters. The Hague Choice of Court Convention is therefore only relevant for disputes between parties of a contract concerning intellectual property rights which includes an exclusive choice of court agreement (with further exceptions concerning validity disputes), while ‘pure’ infringement disputes fall outside its scope.34 In spite of its limited scope, it underlines an important point which has been made already by the different 31

32

33 34

E.g. in Case C-145/10 Eva-Maria Painer v. Standard VerlagsGmbH and Others ECLI:EU: C:2011:798 at [81] (difference in legal basis between the actions brought against the various defendants, does not, in itself, preclude the consolidation of these actions); Case C-616/10 Solvay SA v. Honeywell Fluorine Products Europe BV and Others ECLI:EU:C:2012:445 at [27], [50] (no application of Roche Nederland if same IP national right is enforced, no application of GAT/LuK to provisional measures); Joined Cases C-24/16 and C-26/16 Nintendo Co. Ltd v. BigBen Interactive GmbH and BigBen Interactive SA ECLI:EU:C:2017:724 at [49], [51]– [52] (consolidation of actions based on EU unitary rights possible); Case C-341/16 Hanssen Beleggingen BV v. Tanja Prast-Knipping ECLI:EU:C:2017:738 at [35] (no application of GAT/ LuK to ownership disputes). International Law Association Sydney Conference (2018), Intellectual Property and Private International Law – Fourth Committee Report, http://www.ila-hq.org/images/ILA/ DraftReports/DraftReport_IntellectualProperty.pdf. Available at https://www.hcch.net/en/instruments/conventions/full-text/?cid=98. See M. Dogauchi and T. Hartley, Convention of 30 June 2005 on Choice of Courts Agreements – Explanatory Report at [33] et seq., available at https://assets.hcch.net/upload/ expl37final.pdf. There are further limitations concerning issues of validity of registered rights; see Article 2(2) lit. n, o Hague Choice of Court Convention 2005.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

142

Christian Heinze

academic projects: party autonomy in the form of a choice of court agreement may justify jurisdiction outside the State of protection, and such jurisdiction is not necessarily excluded if any issue of validity of the intellectual property right is raised during proceedings.35 A more general approach was discussed in preparation of the recent Hague Convention on the Recognition and Enforcement of Foreign Judgments in Civil or Commercial Matters.36 This Convention does not include any direct rules on jurisdiction, but rather establishes in its Articles 5 and 6 ‘Bases for Recognition and Enforcement’. Judgments rendered by courts in one of these ‘bases’ ‘shall be recognised’ – if no exception applies (e.g. no proper service or public policy violation) – in all Contracting States (Article 4(1) Hague Convention).37 After extensive discussion, it has been decided that intellectual property shall be excluded from the scope of the final Convention.38 Even if intellectual property had been included in the Convention, the acceptable ‘base of jurisdiction’ for intellectual property disputes would have been small: Article 5(3) of the Draft Convention 201839 provided that the general bases of jurisdiction (such as jurisdiction at the defendant’s habitual residence) did not apply to intellectual property disputes. Rather, the only acceptable fora which would have permitted recognition and enforcement under the Draft Convention 2018 for disputes concerning validity and infringement of intellectual property rights were the courts of the State where the right was registered or for which protection is claimed. Therefore, the Draft Convention 2018 would have returned to the traditional approach (which has been abandoned in Europe) that only the courts of the State of registration or protection can rule, not only on validity, but also on infringement of intellectual property rights. While such a narrow rule would probably not find Annette Kur’s support,40 there was some consolation: in the proposed texts of the acceptable bases of jurisdiction, there was 35

36 37

38

39

40

For details of the approach of the Convention to dealing with preliminary matters such as the validity of an intellectual property right, see Article 10 Hague Choice of Court Convention and Dogauchi and Hartley,note 34 above at [198] et seq. Available at https://www.hcch.net/en/instruments/conventions/full-text/?cid=137. For an analysis and its potential impact on EU law, see the study on the Draft Convention for the European Parliament, available at http://www.europarl.europa.eu/RegData/etudes/STUD/ 2018/604954/IPOL_STU(2018)604954_EN.pdf. Article 2(2) lit. (m) Hague Convention of 2 July 2019 on the Recognition and Enforcement of Foreign Judgments in Civil or Commercial Matters. For the discussion, see Treatment of Intellectual Property-Related Judgments under the November 2017 draft Convention – background document of May 2018, available at https://www.hcch.net/de/projects/legislative-pro jects/judgments/special-commission. Available at https://www.hcch.net/en/projects/legislative-projects/judgments/specialcommission. Arguably, such a small step is better than no harmonisation on the international level, as the Draft Convention permits national or supranational rules more favourable to recognition: Article 16 (with an unfortunate exception for disputes falling in the scope of exclusive jurisdiction under Article 6, which covers also decisions on the registration or validity of registered intellectual property rights, Article 6 lit. a).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

The Private International Law of IP

143

a proviso proposed which limited the scope of the jurisdiction of the State of protection,41 that seemed to be inspired by Article 2:202 CLIP Principles. A further field where we find a trend to universalism is EU unitary rights, in particular EU trade marks and designs. While it is clear that the law applicable to such rights is a single law if the issue is governed by the respective unitary rights regulation, it has long been unclear which law applies to issues not covered by the unitary rights regulations. The text of the respective regulations, Article 129(2) EU Trade Mark Regulation and in particular Article 8(2) Rome II Regulation,42 seemed to suggest a mosaic approach in multi-State infringement scenarios. However, the CJEU has now paved the way to a consolidation both on the level of jurisdiction and on the level of the applicable law,43 even if the exact contours of this way are yet to be explored. A final, most recent example where the traditional conflict between territoriality and lex protectionis or universalism has been discussed is jurisdiction for (worldwide) FRAND decisions. In spite of the fact that the underlying standard-essential patents are territorial in nature, UK courts came to the conclusion that they could decide on the terms of a worldwide FRAND license and that even, in the circumstances, only a global licence would be FRAND.44 It has been argued that this constitutes a new form of extraterritoriality: in contrast to traditional notions of extraterritorial application of intellectual property law where national courts try to police conduct abroad which is unlawful under domestic laws, in the global FRAND decisions national 41

42

43

44

‘[U]nless the defendant has not acted in that State to initiate or further the infringement, or their activity cannot reasonably be seen as having been targeted at that State’. Article 8(2) Rome II reads: ‘In the case of a non-contractual obligation arising from an infringement of a unitary Community intellectual property right, the law applicable shall, for any question that is not governed by the relevant Community instrument, be the law of the country in which the act of infringement was committed. This could be read as referring to the law of each individual Member State in which an act of infringement was committed, leading to a mosaic of different applicable laws for all issues not covered by the Regulation (in particular remedies).’ Joined Cases C-24/16 and C-26/16 Nintendo Co. Ltd v. BigBen Interactive GmbH and BigBen Interactive SA ECLI:EU:C:2017:724 at [49], [51]–[52] (consolidation of actions based on EU unitary rights possible); [103]–[104] (‘overall assessment of that defendant’s conduct in order to determine the place where the initial act of infringement at the origin of that conduct was committed or threatened’; ‘single connecting factor linked to the place where the act of infringement at the origin of several acts alleged against a defendant was committed or threatened’). For a detailed analysis see A. Kur, Durchsetzung gemeinschaftsweiter Schutzrechte: Internationale Zuständigkeit und anwendbares Recht Zugleich Anmerkung zu den Entscheidungen EuGH, Rs. C-360/12 – Coty und EuGH, Rs. C-479/12 – Gautzsch, [2014] GRUR Int 749. Unwired Planet International Ltd v. Huawei Technologies Co. Ltd [2018] EWCA Civ 2344 at [129]; Unwired Planet International Ltd. v. Huawei Technologies (UK) Co. Ltd. [2020] UKSC 37 at [63], [67]; see also Oberlandesgericht Düsseldorf 17.11.2016, Case 15 U 66/15 at [21] (juris database); Landgericht Mannheim 8.1.2016, Case 7 O 96/14 at [119] (juris database); Landgericht Düsseldorf 31.3.2016, Case 4a O 73/14 at [226] (juris database); Landgericht Düsseldorf 8.1.2019, Case 4c O 12/17 at [353] (juris database): requiring to obtain a worldwide licence including a certain market is only FRAND if it is customary in the industry to always co-licence the respective market.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

144

Christian Heinze

courts ‘have begun to use disputes over domestic patent rights as vehicles for shaping the global business arrangements of private parties absent any violation of national law’.45

10.5 conclusion I hope this sketch shows that conflict of laws in intellectual property has been oscillating between transition (towards more universality) and continuity (of territoriality and lex protectionis) in the past thirty years, and that Annette Kur’s contributions have been an important part of this development. While both the technology and the legal issues on a granular level will keep changing, the fundamental issues are likely to remain the same: how territorial can and should intellectual property law be, and how much space can and should be allowed for universalism? In times when one might feel a certain trend towards renationalisation (particularly outside the world of intellectual property), it seems important not to forget the advantages that (some) universality may bring. I hope that Annette Kur will keep contributing to this debate for many years to come.

45

J. Contreras, The New Extraterritoriality: FRAND Royalties, Anti-Suit Injunctions and the Global Race to the Bottom in Disputes over Standards-Essential Patents, 25 BU J. Sci. & Tech. L. 251–290 at 254(2019), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3339378, at 2.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.016

11 From Nintendo Wii to Perfumes, Driving a BMW Car: A Tale of Transition to the Wrong Kind of Coherence Paul Torremans1

11.1 introduction The Regulations on the EU trade mark2 and the Community design3 contain specific provisions on international jurisdiction. These provisions are clearly inspired by the provisions of the Brussels I Regulation,4 but differ from them in some respects. The Brussels I provisions deal with any kind of tort, not merely with IP infringement. Could that explain the slightly different language that is used, and is there a coherent approach to all IP infringement cases, irrespective of whether they are based on unitary rights or not? Private international law in relation to EU trade marks and Community designs5 has only recently been the subject of preliminary questions to the Court of Justice of the European Union (CJEU), in Nintendo6 and Bayerische Motoren Werke.7 The Bundesgerichtshof has also looked at this area in the Parfummarken decision.8 These judgments are innovative and they raise the question of whether the CJEU is developing a new approach to IP rights and whether this approach can be applied 1

2

3 4

5

6

7 8

Prof. Dr. Paul L. C. Torremans, Professor of Intellectual Property Law, School of Law, Faculty of Social Sciences, University of Nottingham. Regulation (EU) 2017/1001 of the European Parliament and of the Council on the European Union trade mark [2017] OJ L 154/1. Council Regulation (EC) No. 6/2002 on Community designs [2002] OJ L 003/1. Regulation (EU) No. 1215/2012 of the European Parliament and of the Council on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters [2012] OJ L 351/1. See in detail James Fawcett and Paul Torremans, IP and Private International Law (2nd ed.) (Oxford University Press) (2011), chapter 7. Joined cases C-24/16 and C-25/16 Nintendo Co. Ltd v. BigBen Interactive GmbH and BigBen Interactive SA ECLI:EU:C:2017:724. Case C-433/16 Bayerische Motoren Werke AG v. Acacia Srl ECLI:EU:C:2017:550. Case Parfummarken, Bundesgerichtshof, 9 November 2017, I ZR 164/16, http://juris .bundesgerichtshof.de/cgi-bin/rechtsprechung/document.py?Gericht=bgh&Art=en&nr= 80153&pos=0&anz=1.

Downloaded from https://www.cambridge.org/core. 145 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

146

Paul Torremans

outside the context of the EU trade mark and the Community designs. Is this a complete transition, namely a transition to a differently shaped coherence? The parallelism proposed by the Court between the rules of jurisdiction and the rules of applicable law also deserves some explanation. Are we seeing a new form of coherence here, this time between jurisdiction and choice of law?

11.2 the starting point It is not surprising that the specific rules of jurisdiction first of all connect to the domicile of the defendant. Private international law is often aiming for legal certainty and predictability, which also drives the CJEU’s approach, and this connecting factor meets that target well. Legal actions concerning infringement and threatened infringement of EU trade marks or Community designs, legal actions to establish non-infringement and counterclaims for revocation or for annulment may be brought before the courts of the Member State where the defendant is domiciled.9 This also fits well within the framework of a single intellectual right for the entire European Union. One court that is competent for the whole case is logical in such a system and the best balance is then achieved by giving the defendant, who is forced to respond to the claim, the advantage of being able to do so in his or her most familiar court. In the absence of a domicile in a Member State, the Regulations seek to achieve the same balance by connecting to the Member State where the defendant has an establishment. If the defendant is neither domiciled nor established in any of the Member States, the connecting factor moves from the person of the defendant to the business link between the unitary intellectual property right and the Union. A court in the Union system specifically created for this purpose is preferred to a court in a third country where, for example, the defendant is domiciled. And in order to find the (proper) national court in the absence of a supranational court at the level of the Union, a forum actoris is created. Thus, the claimant’s domicile is taken and, failing that, the establishment that the plaintiff has in a Member State.10 And in the case of absolute lack of residence or establishment in the Union, the Member State where the European Union Intellectual Property Office (EUIPO) has its seat, namely Spain, is the final alternative.11 Even if there are tweaks in here that are unitary right specific, the approach runs parallel to that in the Brussels I Regulation and therefore demonstrates a coherent approach to IP rights. The Regulations also have a connecting rule that is based on the facts of the infringement. With the exception of legal actions to establish non-infringement, it is therefore possible to use the place where the infringement took place or threatens to 9 10 11

Article 125(1) Regulation (EU) 2017/1001 and Article 82(1) Regulation (EG) 6/2002. Article 125(2) Regulation (EU) 2017/1001 and Article 82(2) Regulation (EG) 6/2002. Article 125(3) Regulation (EU) 2017/1001 and Article 82(3) Regulation (EG) 6/2002.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

From Wii to Perfume and BMWs: A Tale of Transition

147

take place as a connecting factor and this then leads to the jurisdiction of the courts of the Member State concerned.12 This is a clear difference from the standard in Article 7(2) of the Brussels I Regulation, which refers both to the place of the infringement and to the place where the damage occurs. The Brussels I Regulation, as interpreted by the CJEU since the Shevill case,13 limits the jurisdiction of the national court in the latter case to the local damage, but that clearly does not fit in with the concept of unity of both the EU trade mark and the Community design. After all, the idea of a single intellectual property right for the whole Union also requires that one court can deal with the whole matter. Hence the preference for the location of the infringement and the exclusion of the place where the damage occurs. That is the logic that follows from the Coty14 judgment of the CJEU. However, the concept of the location of the infringement may have to be viewed in the specific context of particular intellectual rights. The distinction between a harmful event and damage does indeed make sense in a general tort context as in the Brussels I Regulation, because they can occur in another country. This is different with territorial rights such as intellectual property rights. After all, there can only be an infringement when there is both a harmful event and damage, and the principle of territoriality requires that these two elements occur in the same country, even if indirect effect can also be produced in another country. One can then refer to the whole with one reference: the place of the infringement. That reference then refers both to the harmful event and to the damage. Perhaps the emphasis on the harmful event is therefore wrong and flows from a misunderstanding of the specific nature of intellectual rights.15 The legislature may thus have envisaged a coherent approach to IP despite the difference in wording.

11.3 an exclusive jurisdiction system The specific jurisdiction rules for the EU trade mark and the Community designs exclude, for matters within their scope, the jurisdiction rules of the Brussels I Regulation. This became very clear in Bayerische Motoren Werke AG v. Acacia Srl.16 The Italian producer of wheels, Acacia Srl, who made wheels very similar to wheels on which BMW had registered a design, filed a legal action to establish noninfringement at its forum actoris, the court in Naples.17 But because BMW as the 12 13

14 15

16 17

Article 125(5) Regulation (EU) 2017/1001 and Article 82(5) Regulation (EG) 6/2002. Case C-68/93 Fiona Shevill, Ixora Trading Inc., Chequepoint SARL and Chequepoint International Ltd v. Presse Alliance SA ECLI:EU:C:1995:61. Case C-360/12 Coty Germany v. First Note Perfumes NV ECLI:EU:C:2014:1318. A. Kur, Abolishing Infringement Jurisdiction for EU Marks? – The Perfume Marks Decision by the German Federal Court of Justice, IIC 49, 452–465 (2018). Above, note 7. This court is known to favour the producers of spare parts and to have a restrictive view of design protection on this point.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

Paul Torremans

148

defendant is domiciled in a Member State, the CJEU ruled that the case had to be brought before the court of the defendant’s domicile in application of the specific rule of jurisdiction, in this case before the German court. Article 82(1) of the Regulation on Community designs is mandatory. Article 82 merely clarifies that the parties can still choose the competent court and voluntary appearance remains possible. Articles 25 and 26 of the Brussels I Regulation thus continue to play their part, including in relation to Community designs and claims covered by the specific jurisdiction rules of that Regulation.18 Acacia Srl therefore did not have the possibility to rely on Article 7(2) of the Brussels I Regulation to sustain the jurisdiction of the court in Naples.19 The latter provision cannot even be invoked to determine an abuse of a dominant position or unfair competition if this is linked to the legal claim for a declaration of non-infringement, and the determination can happen only once the claim for a declaration of non-infringement has been assigned.20 Article 125 of the EU Trade Mark Regulation is an exact copy of Article 87 of the Community Design Regulations. The same conclusion thus also applies to EU trade marks.21 One can therefore not mix up both sets of jurisdiction rules. From a business perspective national trade marks or designs under the Brussels I Regulation are not that different from unitary rights under their own rules. A coherent approach along the same lines is therefore expected, even if two separate sets of rules apply.

11.4 the place where the infringement took place or threatens to take place The jurisdiction rule based on the place where the infringement took place or threatens to take place is clearly the rule that deviates most from the Brussels I norm, and not only because the place where the damage occurs or threatens to take place is left out of consideration.22 Infringing acts can after all be carried out in more than one Member State. Let us, by way of example, look at a simplified overview of Article 9 of the EU Trade Mark Regulation. An EU trade mark gives the holder the exclusive right to prohibit any third party who has not obtained his or her consent from using a sign in the course of trade for goods and services where the sign is identical to the EU trade mark and is used for goods or services which are identical to those for which the EU trade mark has been registered. This applies inter alia to the affixing of the sign to the goods or to the packaging of those goods, the offering of goods under the sign and the use of the sign in advertisements. It is not at all unrealistic that the defendant, in a practical case, carries out this combination of infringements and is 18 19 20 21 22

Above, note 7, para. 42. Ibid., para. 46. Ibid., para. 52. See above note 5, chapter 7. Ibid.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

From Wii to Perfume and BMWs: A Tale of Transition

149

sued on that basis by the proprietor of the EU trade mark for infringement of the trade mark under Article 9 of the Regulation. The problem that arises is that each of those (alleged) infringing acts can take place in another Member State. A rather simple scenario unfolded in the Parfummarken case23 that was recently dealt with by the German Bundesgerichtshof. Trade mark-protected perfumes were offered by an Italian company to a German public on a website, set up in Italy, that was also available in German. Signs that were identical to the protected trade marks were thus used in one way or another in Italy and Germany (there is no need to go into the details here). The question that arises, however, is what that means for the jurisdiction of the courts in Germany and Italy if the proprietor of the mark wishes to rely on the rule that the court of the place where the infringement took place or threatens to take place has jurisdiction. One could opt for a distributive application of the jurisdiction rule based on each of the infringing acts in Article 9. The competent court must then be determined per infringing act. But that would also mean that the German court would be competent for certain infringements, while the Italian judge would be competent to judge other infringements. Such a fragmentation and the fact that what constitutes one infringement pattern will not be judged by one, but by different judges seems to be at odds with the idea of the EU trade mark as one intellectual right for the whole Union. But that criticism can, of course, be altered by stating that the EU trade mark holder can always bring the whole case before the court of the place of domicile or place of establishment of the defendant or, depending on the circumstances, the court of his or her place of domicile or place of establishment. It also seems to be a logical consequence of the fact that Article 9 splits the protection afforded by the exclusive right into protection against various infringing acts, each of which is sufficient in itself to lead to an infringement. Nevertheless, the Bundesgerichtshof rejected the distributive application of the jurisdiction rule. According to the Bundesgerichtshof, it is not necessary to refer to any allegedly infringing act, but the conduct of the defendant must be assessed in its entirety in order to determine the place where the original infringing act on which the alleged conduct is based was carried out by the defendant or where it is likely to be carried out. In the Parfummarken case that means the setting up of, or the decision to set up, the process by which the perfumes protected under trade mark law were offered to the public via the website of the defendant. However, there must be a positive action by the defendant (or a threat of active action).24 All other infringing acts are ancillary in this logic and follow from the setting up of the website on which the perfumes are offered. That setting up of the website is thus the original infringing act if one considers the behaviour of the defendant as a 23 24

Above, note 8. Ibid., para. 27.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

150

Paul Torremans

whole. The fact that the website is accessible in German25 from Germany26 and that perfumes may be supplied to a German public in Germany only gives rise to infringing acts that in the behavioural pattern of the defendant in its entirety follow on from, or stem from, the setting up of the website. With regard to jurisdiction, those infringing acts are then left out of consideration and the entire infringement case against the defendant is linked to the place where the original infringing act took place. This leads to the jurisdiction of one judge, in this case the Italian judge, because the defendant set up and operated the website in Italy. The great advantage of this approach is that any fragmentation of competence is avoided and all grounds of jurisdiction in the EU trade mark Regulation now lead to the result that the whole case will be dealt with by one judge. This is also fully in line with the unitary nature of the EU trade mark. There is also no question of a conflict with the principle of territoriality, because the entire European Union is the territorial unit with regard to the EU trade mark.27 The approach taken by the Bundesgerichtshof, the refusal to refer the case to the CJEU and the parallel that is drawn with the applicable law in Article 8(2) of the Rome II Regulation, despite the fact that in relation to jurisdiction and choice of law very different factors are of importance, make a very critical analysis of the Parfummarken judgment possible.28 Suffice it to highlight here, as a starting point, that the ‘central place’ will quite often be the defendant’s seat so that Article 125(5) loses most of its meaning. Advocate General Szpunar took that criticism to heart in his conclusion in AMS Neve. He specifically rejected the link with choice of law and the Rome II Regulation and he also rejected the exclusive reliance on the criterion of the place where the website was activated. Instead he proposed a tailormade approach for unitary rights that respects the IP principle of territoriality and that, for example, gives jurisdiction on the basis of the fact that a website targeted clients in that country. That local impact on the unitary right as a result of the targeting is sufficient to allow the local judge to deal with the infringement that results from the use of the website in that jurisdiction. Fortunately, the CJEU agreed and went for the targetting approach.29 But if one follows the logic of the Bundesgerichtshof’s decision, if only to see where this might ultimately lead, one finds that, since the Regulation in question contains exactly the same provision, this approach applies not only to the EU trade mark, but also to Community designs. Parfummarken was concerned with the infringement of trade mark law and in particular of EU trade marks, but the

25 26 27 28 29

Ibid., paras. 17 and 29. Ibid., para. 31. See AMS Neve Ltd v. Heritage Audio [2018] EWCA Civ 86, now pending as case C-172/18. A. Kur, above, note 15. Opinion of the A-G and judgment of the CJEU in case C‑172/18 AMS Neve Ltd, et al. v. Heritage Audio SL, et al. ECLI:EU:C:2019:276 and ECLI:EU:C:2019:674.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

From Wii to Perfume and BMWs: A Tale of Transition

151

Bundesgerichtshof bases its approach entirely30 on Nintendo v. BigBen.31 This decision by the CJEU has its roots in Community design law. The nice thing is that that decision deals with applicable law in the design context and the perceived parallelism with choice of law was a major concern in Parfummarken. It makes sense therefore to look in more detail at the CJEU’s choice of law analysis.

11.5 the applicable law Nintendo v. BigBen was not brought before the German court on the basis of the place of the infringing act. But the analysis of the CJEU focusses very much on the applicable law. This brings us in an IP context to Article 8(2) of the Rome II Regulation:32 2. In the case of a non-contractual obligation arising from an infringement of a unitary Community IP right, the law applicable shall, for any question that is not governed by the relevant Community instrument, be the law of the country in which the act of infringement was committed.

In the present case, we were dealing with the sanctions that could be imposed, for which the Regulation provided no rule, relegating applicable law to Article 8(2).33 It is striking and perhaps not at all a coincidence that the connecting factor for the law applicable to unitary rights is the same as that for jurisdiction and based on the place of the infringing act. According to Nintendo, BigBen had committed infringing acts in several Member States. German customers could order remote controls and accessories from BigBen Germany that were compatible with Nintendo’s Wii game consoles. These remote controls and accessories were made by BigBen France in France and they were also delivered to German (and other) customers from there. With regard to the applicable law, two solutions were possible. One can, on the one hand, determine the applicable law in a distributive way for each of the infringements. That would then lead to French or German law, depending on the infringing act and the place where it was carried out. On the other hand, it is possible to analyze BigBen’s infringement pattern as a whole and to identify one applicable law for the whole pattern, for example on the basis of the strongest link between the pattern of infringing acts and a certain legal system. The Court clearly rejected the distributive application of different laws and applied a single national law to the whole case. In order to determine that one national law, the conduct of the defendant must be assessed in its entirety in order to 30 31 32

33

Above, note 8, paras. 30, 31 and 34. Above, note 6. Regulation (EC) 864/2007 of the European Parliament and of the Council on the law applicable to non-contractual obligations (Rome II) [2007] OJ L199/40. Above, note 6.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

152

Paul Torremans

determine the place where the original infringing act on which the conduct that is complained of is based was committed or is threatened to be committed.34 The connecting factor is therefore the place of the original infringing act and it is determined on the basis of an analysis of the behaviour of the defendant in its entirety. According to the Court, this approach flows from the unitary nature of the Community design (or the EU trade mark) and it allows the competent court to determine the applicable law in a simple way, whilst making the outcome of the process predictable.35 BigBen used electronic commerce and offered on its website, intended for consumers in different Member States, products for sale that infringe Community designs. The harmful event or conduct then consists of the offering for sale of allegedly infringing products, in particular by placing an offer to sell on its website. The place where the event giving rise to the damage occurred is therefore the place where the economic operator operates the procedure for placing the offer to sell on its website. All further infringing acts, such as the production and delivery of the products by BigBen France, are ancillary and result from that original infringing act. The fact that the same connecting factor is used for both jurisdiction and applicable law is of course also an advantage. The Bundesgerichtshof also referred to this as an advantage when it applied the logic and approach of the Court without hesitation, not only with regard to trade mark law, but also with regard to the jurisdiction of the court.36

11.6 an evaluation It is undeniable that a solution that leads to the application of one law charms by its simplicity, certainly in the context of a unitary right that applies throughout the European Union. It also fits perfectly into the logic of private international law.37 But the underlying substantial IP right functions on a totally different basis. Each infringing act gives rise to an infringement in itself. There is no causal or necessary connection between the various infringing acts. They are not components of a more complex infringing act in a legal sense, even if some of the infringing acts often occur together in practice. The logic is therefore that for each infringing act the applicable law is determined separately, especially if the acts are carried out by separate legal entities, as in the BigBen group.

34 35 36

37

Ibid., para. 103. Ibid., para. 104. Above, note 8, para. 35. See also the application of the BigBen doctrine by the court in Milan in Diesel v. Zara (http://ipkitten.blogspot.com/2018/07/milan-court-applies-nintendo-v-big-ben.html). See Article 3:601 CLIP Principles.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

From Wii to Perfume and BMWs: A Tale of Transition

153

From an IP perspective, the new approach of the Court is therefore a small revolution. But perhaps it is true that here the unitary character of IP prevails. It is then not only acceptable, but also desirable that one law applies to the whole case and therefore also to all infringing acts.

11.7 conclusion This is essentially the first time the CJEU had a crack at working out an approach to unitary rights, but the way in which the Court goes about these cases could lead one to think that a new approach is proposed and that this approach has the potential to shake up generally accepted principles in the world of IP. With regard to both jurisdiction and applicable law, the Court opts for a unitary link that leads to one court and one applicable law. The Court postulates the dominance of the original infringing act and all other aspects then become ancillary. As a result, all these other aspects and infringing acts are no longer (autonomously) connected to a jurisdiction and an applicable law. This approach is unusual in the context of national IP rights, but it is defensible in the context of the EU trade mark and the Community design. The unitary line that is already strongly felt in the Regulations is extended through private international law and there is a logic in the tight concept of one intellectual law, one competent court and one applicable law. How that one competent court is determined remains more problematic, though, as it can oblige a claimant to go to a faraway jurisdiction where the infringement chain started, but which is not at all the market it wants to protect. But perhaps it should be borne in mind that the Court only proposes this approach for unitary rights. Given the emphasis on the unitary nature of these intellectual rights and the fact that this argument is invoked as the main justification of the approach, it seems unlikely that the Court would also like to apply the approach outside the context of the unitary rights. After all, that would also jeopardize the entire concept of territoriality, or at least seriously limit it. Suffice it to think of the idea that a violation of a parallel national IP right could be ancillary to other infringements of another parallel IP right, both in terms of jurisdiction and applicable law. That is probably a step too far. But the CJEU puts all this in such general terms38 that one cannot rule out that we see here the start of a broader transition that could apply to all IP rights. That is not the transition that we need. It relies too much on the logic of private international law and not enough on the logic of IP, even if in terms of logic unitary rights come closer.

38

See e.g. joined cases C-24/16 and C-25/16 Nintendo Co. Ltd v. BigBen Interactive GmbH and BigBen Interactive SA ECLI:EU:C:2017:724, paragraph 111.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

154

Paul Torremans

Admittedly, it would create coherence in the approach to unitary and non-unitary IP rights, but maybe we need a different form of coherence. Maybe we need coherence between the IP approach with separate infringing acts and a corresponding private international law approach. That offers better guarantees for a correct choice of the best-placed court and an applicable law that reflects the specificities of the local IP law and the underlying culture. And, of course, maybe the different wording of the unitary rights Regulations merely reflects the specific nature of IP, rather than a departure from the general tort rule. So maybe there was all along more coherence than expected and less need for this apparent transition.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.017

12 Transition through Automation Henning Grosse Ruse-Khan1

Quite fitting with one of the themes for this book, the first major project on which I had the privilege to work with Annette Kur was called ‘Intellectual Property Rights in Transition’ (IPR in Transition); its outcome was subsequently published as Intellectual Property Rights in a Fair World Trade System: Proposals for Reform of TRIPS (EE Publishing, 2011). Based on our earlier discussions about key shortcomings within the international IP system, Annette and I focused on the concept of ceilings: the idea was to challenge the commonly held view that international IP law is only about setting minimum standards and can be further expanded without limits (or ceilings).2 While I am not sure what role our work played in this, a range of subsequent developments in international IP law show a transition to a system where IP treaties are not only about guaranteeing exclusive rights and other protections,3 and where IP protection is conceptualized as an element within a broader set of international norms that can counter attempts for limitless expansions of IP protection and enforcement.4 1

2

3

4

Reader in International and European Intellectual Property Law, King’s College, University of Cambridge. I thank Graeme Dinwoodie and Oliver Englisch for their comments on earlier drafts. See H. Grosse Ruse-Khan and A. Kur, Enough is Enough – The Notion of Binding Ceilings in International Intellectual Property Protection (8 December 2008), Max Planck Institute for Intellectual Property, Competition & Tax Law Research Paper Series No. 09-01, https://ssrn .com/abstract=1326429. See, e.g., Marrakesh Treaty to Facilitate Access to Published Works for Persons Who Are Blind, Visually Impaired or Otherwise Print Disabled, adopted on 27 June 2013 at Marrakesh (entered into force 30 September 2016). Consider, for example, the movement that led to the eventual rejection of the AntiCounterfeiting Trade Agreement (ACTA) by the European Parliament – which was primarily driven by fears that overreaching IP enforcement norms could interfere with privacy, freedom of expression and information, and general notions of due process and proportionality. See, e.g., European Academics Opinion on ACTA, at https://www.statewatch.org/media/docu ments/news/2011/jul/acta-academics-opinion.pdf.

Downloaded from https://www.cambridge.org/core. 157 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

158

Henning Grosse Ruse-Khan

The IPR in Transition project reflected well the reasons why I admire and respect Annette: as a scholar who is not only extremely well versed in technical details of essentially all areas of IP law, but equally open to new perspectives and fresh ideas – which she is probably best placed to give an analytical reality check. Not in terms of destructive criticism, but rather as a critical engagement with an excellent intuition on what is feasible, and worth considering further. With this spirit in mind, this chapter discusses another form of transition – focused on technological change in the form of automated decision-making employed by entities mediating content in global networks – and offers some initial thoughts on the challenges this poses for IP law. Technological change that affects society and/or our environment necessarily has an impact on law as one of the principal tools for regulating human interaction and our relations with the wider world. In that regard, technology has, over and over again, transformed human society and, by the need to respond to and accommodate new technologies, the laws made by humans. IP law in particular is in constant transition through the dynamic development of technology. Changes in technology have always had a specifically profound effect on IP law because (1) the subject matter protected by IP is often technology based or at least technology related; (2) the production, dissemination, consumption and (commercial) exploitation of IPprotected subject matter is often based on technology; and (3) as I want to argue here – nowadays, technology increasingly shapes the norm-setting, implementation and adjudication of IP law, at least in digital networks which are at the centre of this chapter. The starting points for this argument are the well-understood features of the digital, network environment and their effects on copyright law in particular. First, the digital aspect: against the background of fairly low thresholds for copyright eligibility, and the fact that any access or use of content which can be represented in electronic form in the digital environment necessarily involves copying (which in principle implicates exclusive rights), copyright is essentially always involved when digital content is used. Second, the network aspect: the internet as a decentralized, global network of interconnected electronic devices has exponentially amplified opportunities for citizens across the world to access, share and consume content. The decentralized nature of networks like the internet requires entities that mediate communications and other interactions between users – in other words, intermediaries. It is the use of automated content-protection technologies by some of these intermediaries, namely platforms that bring together internet users in order to create, copy and share content online, that serves as my example for how technology effectuates transition and change in social and legal norms linked to IP. Market actors like Facebook, YouTube, Instagram and Twitter live on users ‘freely’ sharing content via their platforms in exchange for the data that users generate in the process (which can be sold on to others, including those running targeted ads on the platform that the targeted user then has to watch). The content Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

Transition through Automation

159

users create and share is frequently based on, or at least involves elements of, appropriation of existing (copyright-protected) content. User creativity and ‘usergenerated content’ – for example in the form of memes, mash-ups, playalongs and supercuts – tends to be expressed through repurposing, transforming or otherwise modifying or adding on existing content: a form of ‘remix culture’ that somewhat resembles sampling in 1990s hip-hop. These types of remixes can become quite popular, occasionally bringing fame to their user-creators when shared widely on a platform. From a copyright perspective, quite a bit of that remixed content is likely to be infringing under most domestic laws – although much will still depend on the respective quantitative or qualitative thresholds for infringement and on any defences or exceptions, such as fair use concepts or specific exceptions for the purpose of parody, pastiche, satire, quotation, news reporting, education, research and so on.5 Hence, platforms that allow such content to be shared run the risk of meeting traditional accessory liability tests, or of otherwise being held liable for infringing acts of their users – unless of course they can rely on ‘safe harbours’ that exempt liability for intermediaries under certain conditions. While details differ and not all major jurisdictions offer specific safe harbours, these safeguards from liability converge around a ‘notice and takedown’ model whereby intermediaries have to act expeditiously when they become (or should be) aware of infringing acts of their users (often via a notification).6 In order to avoid liability, platforms are taking steps to respond to infringement notifications by removing or blocking access to allegedly infringing content.7 Increasingly, however, platforms also rely on filtering systems and other automated content-protection tools that check, usually in real time, any materials uploaded or otherwise shared by users against reference files in huge databases for content submitted by those claiming rights in that content.8 As the specific 5

6

7

8

For a discussion from a US perspective, see M. Sag, Internet Safe Harbors and the Transformation of Copyright Law, 93(2) Notre Dame L. Rev. 499 (2017), at 518–520. See, e.g., Art. 14 of Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 on certain legal aspects of information society services, in particular electronic commerce, in the Internal Market (hereafter E-Commerce Directive); 17 USC 512(c) (2018) (US) (introduced by the Digital Millennium Copyright Act 1998). For an empirical account of the notice and takedown system in the USA, see J. Urban, J. Karaganis and B. Schofield, Notice and Takedown in Everyday Practice, UC Berkeley Public Law Research Paper No. 2755628 (22 March 2017), https://ssrn.com/abstract=2755628; see also K. Erickson and M. Kretschmer, ‘This Video is Unavailable’: Analyzing Copyright Takedown of User-Generated Content on YouTube, 9(1) JIPITEC 75 (2018). For a discussion on the EU system, see C. Angelopolous, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer, 2016) and M. Husovec, Injunctions Against Intermediaries in the European Union: Accountable But Not Liable? (Cambridge University Press, 2017). A general description of automated enforcement tools by one of the major providers (Audible Magic) is at https://www.audiblemagic.com/compliance-service/#how-it-works. Specific content-protection tools – other than the one by YouTube discussed in detail below – are, for example, discussed here (https://variety.com/2018/digital/news/instagram-block-copyright-

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

160

Henning Grosse Ruse-Khan

mechanisms and technical tools used by platforms such as Google, YouTube, Facebook and Instagram differ, this discussion focuses on one of the most sophisticated and arguably most profitable filtering systems: YouTube’s ‘Content ID’.9 Once Content ID’s automated content recognition tool detects a match, the entity/entities claiming rights over that content is/are notified. That entity can then decide (case by case or in the form of setting general rules) over the further fate of the user content uploaded or shared. It can of course be blocked – but, given the popularity of usergenerated content, which often gets thousands (and occasionally millions)10 of views, content owners may find it much more economically interesting to monetize the user content, for example by running ads against it.11 Given that safe harbours do not require upload filters and actually can discourage proactive monitoring,12 why would platforms invest up to US$100 million13 in the development, maintenance and further improvement of these sophisticated contentprotection tools? It may in part be that they prefer to cooperate with content owners in order to avoid risky litigation,14 and appear as a trustworthy partner for content

9

10

11

12

13

14

infringement-video-1202692290/ – regarding Instagram) and here (https://variety.com/2017/ digital/news/facebook-rights-manager-copyright-videos-third-party-1202578122/ – concerning Facebook). To understand the basics of automated enforcement tool Content ID, which, as per YouTube’s owner Google reported in 2018, now deals with 98 per cent of all copyright issues on YouTube, see ‘How Google Fights Piracy’, 7 November 2018, at 21 and 25, https://www.blog.google/ outreach-initiatives/public-policy/protecting-what-we-love-about-internet-our-efforts-stoponline-piracy/, and this short (YouTube) video: https://support.google.com/youtube/answer/ 2797370?hl=en-GB. For a legal analysis, see the authors discussed below and also S. Jacques, K. Garstka, M. Hviid and J. Street, Automated Anti-Piracy Systems as Copyright Enforcement Mechanism: A Need to Consider Cultural Diversity, 40(4) European Intellectual Property Review 218 (2018). See, e.g., https://www.youtube.com/watch?v=l4HpWQmEXrM&feature=youtu.be, a video of a surprise marriage proposal using an indie pop song as background that, due to the video’s popularity, has been listened to by 14 million viewers. See also the ‘Harlem Shake’ example, discussed in Michael Soha and Zachary J. McDowell, Monetizing a Meme: YouTube, Content ID, and the Harlem Shake, Social Media & Society, January–March 2016, 1–12. YouTube, for example, explains that since January 2014, Content ID claims have ‘outnumbered copyright takedowns by more than 50 to 1’ (see https://support.google.com/youtube/ answer/7002106?hl=en-GB). In 2017, 90 per cent of Content ID claims imposed automated monetization by running ads against an upload that matched with claimed content – this in turn has resulted in YouTube paying out more than US$6 billion in total (and over US$1.8 billion from October 2017 to September 2018) to right holders who have monetized the use of their content on the platform – see ‘How Google Fights Piracy’, note 9 above, at 23. Active monitoring may not only breach data protection or privacy rules and cannot be demanded from service providers under Art. 15 of the E-Commerce Directive – it also may lead to liability for inaction once a platform becomes aware of infringing content. According to Google’s own 2018 report, it has invested more than US$100 million into building Content ID – see ‘How Google Fights Piracy’, note 9 above. The experience of operators like Napster or Grokster (see the US Supreme Court decision in MGM Studios, Inc. v. Grokster, Ltd., 545 US 913 (2005)) will have told platform operators that a full collision course with content owners is not helpful in guaranteeing the survival of their

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

Transition through Automation

161

owners – or that they aim to anticipate legislative15 (or jurisprudential)16 change which requires them to be more proactive. As a result, and in order to rather ‘err on the side of caution’, there is likely to be a tendency to over-enforce copyright claims. Most importantly, however, employing content protection tools that allow monetization of user-generated content also creates quite a lucrative market for platforms, as usually they will be able to claim a (significant) share in the advertising revenue (or other form of monetization) generated by sharing the user’s content.17 In addition, large-scale availability of user content attracts more users and keeps existing users on the platform – generating even more revenues, including via shares in monetization activities by those claiming rights in user content. In other words: the remix culture of users is turned into a profit enterprise for platforms and content owners,18 which collude to exploit the user’s ‘digital labour’19 involved in creating and sharing content. Since platforms generally ask all their users for a worldwide, royalty-free and transferable licence to use any element of the content a user might own,20 one might say that from a technical-legal perspective, everyone should be happy. But apart from the colluded exploitation of user content and user data in which platforms and content owners engage (of which most users are most likely unaware), one might wonder about the type of copyright (or rather, ‘content protection’) norms that are imposed and automatically enforced through the platform. How does this form of private ordering align with our domestic copyright regimes, and the normativity embedded therein? In particular, is the widespread monetization of user-generated content supported by copyright principles as we know them? One might think that exploiting users’ content through monetization has no legal basis if the use of content falls under a copyright exception, or is otherwise non-infringing. But even if user-generated content does

15

16

17

18

19 20

own business model. That can be better achieved by avoiding too much confrontation with right holders. See, e.g., the approach adopted in Art. 17 of Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on copyright and related rights in the Digital Single Market and amending Directives 96/9/EC and 2001/29/EC, OJ L 130, 17 May 2019 (‘Digital Single Market’ or ‘DSM’ Directive), discussed further below. See, e.g., the ‘Pirate Bay’ judgment of the CJEU, Stichting Brein v. Ziggo BV and XS4All Internet BV, Case C-610/15 (14 June 2017) (extending primary liability for copyright infringement to certain hosting platforms). In the case of YouTube, that appears to be around 40 to 50 per cent of the money generated through monetising content claimed by those eligible for Content ID – see Soha and McDowell, note 10 above. Consider the telling examples provided by Stephen Witt, When Your YouTube Video Becomes a Corporate Profit Center, Los Angeles Times, 27 June 2015 – online at https://www.latimes .com/opinion/op-ed/la-oe-0628-witt-youtube-copyright-20150628-story.html, as well as the Harlem Shake meme discussed by Soha and McDowell, note 10 above. See the discussion in Soha and McDowell, note 10 above, at 6. See, e.g., the ‘Licence to YouTube’ and ‘Licence to Others’ conditions in YouTube’s Terms of Service, at https://www.youtube.com/static?gl=GB&template=terms (as updated on 31 March 2020).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

162

Henning Grosse Ruse-Khan

involve a bit of copyright-infringing content (e.g. a one-minute extract of a recorded song that serves as background in a fifteen-minute video), it is not immediately clear that the user content as a whole (the entire fifteen-minute video that otherwise does not infringe) should be monetized exclusively by the platform and the copyright owner.21 In this short chapter, I want to sketch a ‘brave new world’ of automated content protection based on the example of YouTube’s Content ID system – although this is just one (albeit a quite sophisticated and complex) case of a broader trend towards parallel legal universes that occasionally reflect state-made IP laws, but often diverge from them in significant ways. My focus is on the transition we are facing as globally operating online platforms set their own rules (or ‘community guidelines’), implement them through their codes and user interfaces and enforce them through algorithms. In constitutional terms, platforms (partially in concert with those who claim rights in the content users’ upload or share) set norms, apply and enforce them, and also act as quasi-adjudicators over disputes. This concentration of legislative, executive and judicative power is not only in the hands of those who have an evident self-interest in how this power is exercised. It also often has global reach, and in any case transcends national borders. Its efficacy and impact of prescribing how we experience audio-visual content easily surpasses international and national IP norm-setting. Even the most detailed rules in FTAs require transposition into domestic IP laws and, crucially, enforcement ‘on the ground’ through domestic offices and courts. In the world of automated content-protection tools, however, everyone only gets to access, experience and use what passes through filters and other algorithmic gatekeepers, implementing the normative decisions they embed and execute.22 One may counter that the reach of platform normativity is at best inter partes and voluntary – and hence cannot be compared with the inter omnes effects of binding legal rules, set out in domestic IP statutes and their international treaty counterparts. Surely, no one is forced to use a platform like YouTube, Facebook or 21

22

This, however, appears to be the current practice on YouTube, i.e. a tiny amount of copyrighted content if claimed under Content ID generally prevents the user from monetizing her/ his videos, regardless how irrelevant that copyrighted material is for the success of the user’s video as a whole – see ‘What kind of content can I monetise?’, https://support.google.com/ youtube/answer/2490020?hl=en-GB and the blog discussion at https://support.google.com/you tube/thread/1281991 where YouTube admits that it is ‘clearly not ideal’ that a copyright owner is ‘getting all of the creator’s revenue for a few seconds of a video’. An exception exists for certain eligible cover videos – see https://support.google.com/youtube/answer/3301938. As Niva Elkin-Koren has aptly put this: ‘Algorithmic copyright enforcement . . . has changed copyright default: if copyrighted materials were once available unless proven to be infringing, today materials that are detected by algorithms are removed from public circulation unless explicitly authorized by the right holder.’ See N. Elkin-Koren, Fair Use By Design, 64 UCLA L. Rev. 1082 (2017), at 1093. In addition, the norms and potential biases embedded in automated enforcement tools are likely to frame user experiences and gradually change user behaviour. See generally D. Burk, Algorithmic Fair Use, 86 U. Chi. L. Rev. 283 (2019).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

Transition through Automation

163

Instagram – and we are also not compelled to do online searches via Google. As private individuals choosing to use the services offered to us by private market actors, we are in well-known territory where (with some occasional limits) party autonomy and freedom of contract reign. Even if it can be proven that users have validly agreed to the rules of the platform,23 it appears doubtful that notions of party autonomy and freedom of contract were designed to serve as foundations for creating private content protection regimes with global reach, paired with immediate and automated enforcement. Furthermore, the argument that users are free not to use the platform neglects the fact that many of these platforms nowadays function as essential facilities in accessing and communicating content online.24 Their role as facilitating freedom of speech and information has been recognized by courts, which take this function into account when determining platform liability, for example for direct and indirect copyright infringements.25 If platforms are privileged based on their public-interest function, why should they not also have to be held responsible for actually fulfilling this function? I will come back to this key issue of accountability and responsibility below – after setting out in some greater detail how systems such as Content ID operate. First of all, Content ID is not for everyone. Without giving away any specific thresholds, YouTube indicates that ‘to be approved, you must own exclusive rights to a substantial body of original material that is frequently uploaded by the YouTube user community’.26 The platform makes clear that it reserves to itself the final say on what enforcement tool it decides to provide to whom27 – pointing also to more basic (i.e. less automated and scalable) options such as its ‘Content Verification Programme’,28 23

24

25

26

27 28

The Tribunal de Grande Instance de Paris, for example, held that a blanket licence to use all user-generated content on Twitter, including future works, in Twitter’s general terms of service breaches various author-protection rules of the French Copyright Code and hence is invalid – see Tribunal de Grande Instance, Décision du 07 août 2018, 1/4 social No. RG 14/07300 – http://entreprises.claisse-associes.com/wp-content/uploads/2018/08/TGI-Paris-7-ao%C3%BBt2018-UFC-Twitter.pdf. See, e.g., the discussion of the essential role of private platforms with regard to free speech: Kate Klonick, The New Governors: The People, Rules, and Processes Governing Online Speech, 131 Harv. L. Rev. 1598 (2018). For a competition law analysis (with a focus on big data), see I. Graef, EU Competition Law, Data Protection and Online Platforms: Data as Essential Facility (Kluwer, 2016). See CJEU judgments in Case C‑360/10, SABAM v. Netlog (16 February 2012), paras. 48–50 and Case C‑314/12, UPC Telekabel Wien v. Constantin Film Verleih (27 March 2014), paras. 55–57; and in particular the decision of the German Federal Supreme Court (BGH) in Vorschaubilder III (I ZR 11/16 – 21 September 2017), paras. 60–62. See ‘How Content ID Works’ – https://support.google.com/youtube/answer/2797370 (emphasis added). See ‘Qualify for Content ID’ – https://support.google.com/youtube/answer/1311402. See ‘Content Verification Programme’ (CVP) – which YouTube describes as ‘designed especially for copyright-holding companies to issue multiple removal requests’ and hence equally not accessible for everyone, and again subject to an application to be accepted by YouTube –https://support.google.com/youtube/answer/6005923. A key difference to Content

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

164

Henning Grosse Ruse-Khan

a ‘Copyright Match Tool’29 and a simple copyright notification web form.30 YouTube apparently tries to offer the right type of tool to copyright owners with different needs, mainly based on the amount of their content shared on YouTube. A more cynical view would point out that (copy)rights and effective protection are provided only to major players – not individual creators with a handful of selfcreated videos. In any case, leaving it to the platform to decide who gets efficient and automated filters and who needs to engage in individual searches is quite problematic. The more these automated tools develop into the key mechanisms for enforcing copyright online, the more important is a non-discriminatory approach to eligibility. The most powerful tools will offer the most favourable protection, while not being eligible may at some point in the future mean that there is no effective protection available. Of course, in the ‘real’ world, a large-scale copyright owner will usually be better off than an individual author in terms of enforcing rights based on IP laws and through the judicial system – but at least the same rights and remedies are available to all. Offering differentially effective enforcement tools means to administer selective justice – in this case with no convincing reason for such a discriminatory approach. Such an approach is even more problematic where less powerful and economically weaker parties are likely the ones that are stuck with inefficient enforcement tools – while scalable and automated systems with significant potential for abuse (as shown below) are reserved for major copyright holders. Even if smaller copyright owners have the option to sign up with a third-party service provider that administers Content ID for smaller players, that creates additional costs and will not usually be a realistic option for an individual creator. Secondly, not all content is meant to be protected under Content ID. But there is no mechanism to check whether the (presumably low) thresholds for copyright protection of the respective lex loci protectionis (or in fact those of any domestic law) are in fact met. Seemingly operating under the premise that all video and sound recordings – however short – are protected under neighbouring rights and that any use not covered by an exception necessarily infringes,31 there is no attempt to check

29

30

31

ID appears to be that CVP relies essentially on active searches by content owners, whereas Content ID filters any upload against content submitted to its reference databases, and conducts a legacy scan of all existing content on YouTube, searching for matches. The Copyright Match Tool (CMT) can be used to individually search for full re-uploads (rather than just partial matches) of one’s own videos on other YouTube channels, and operates on a ‘first to file’ basis where only (near to) full matches with subsequent video uploads can be claimed. It is currently only available as a pilot for a limited number of YouTube channels – see https://support.google.com/youtube/answer/7648743. See generally ‘Copyright management tools’ – https://support.google.com/youtube/answer/ 9245819. A normatively flawed proposition in light of the original intent of such neighbouring rights to protect (significant) investment into activities that mediate and disseminate copyrighted works – but one which has been supported by the Opinion of the Advocate General to the CJEU in Moses Pelham GmbH v. Ralf Hütter (Case C‑476/17), 12 December 2018. On 29 July 2019, the

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

Transition through Automation

165

submitted content for subject matter or threshold eligibility. Nor does Content ID verify whether the amount used in a claimed video would constitute infringement under the law of the country where protection is claimed. Instead, YouTube emphasizes that eligible Content ID users should only claim content for which they actually hold exclusive rights in the territories claimed. Its web pages explain, for example, that content licensed under a Creative Commons (CC) licence, ‘public domain footage’ or material ‘used under fair use principles’ is ‘ineligible for use in or as a reference [file]’.32 While it warns not to abuse the system and threatens to act if it becomes aware of such abuse, YouTube eventually leaves it to the content owner to ensure that no wrongful claiming occurs;33 and to resolve ownership and other disputes over any claims made.34 In its further explanations on eligible content, YouTube simply stresses that the reference content provided must be ‘sufficiently distinct’ (hence excluding remasters, soundbeds or production loops) and individualized (i.e. excluding compilations, DJ mixes or mash-ups).35 Anecdotal evidence of all kinds of false positives in claiming content not copyright protected, not owned by the entity claiming it, or clearly within the ambit of copyright exceptions (in particular fair use) shows that leaving it to content owners to ‘selfregulate’ and simply warn them not to overclaim has, at least from a normative perspective, not worked well in the past.36 Third, the scope of protection and ‘remedies’ offered by the platform vis-à-vis user uploads that match claimed content focus on common options (called ‘policies’37 in the case of Content ID), such as: (1) monetizing (allowing the matching upload to be viewed and displaying advertisements with it); (2) tracking (merely collecting statistics about views); and (3) blocking (i.e. not allowing the matching upload to be viewed).38

32 33

34 35 36 37

38

CJEU judgment in this case confirmed that ‘the reproduction by a user of a sound sample, even if very short, taken from a phonogram must, in principle, be regarded as a reproduction “in part” of that phonogram’ and that therefore ‘such a reproduction falls within the exclusive right granted to the phonogram producer’. See ‘Content eligible for Content ID’ –https://support.google.com/youtube/answer/2605065. YouTube explains that as it automatically generates claims against those uploading matching content, the Content ID user is ‘responsible for avoiding incorrect results’, such as ‘claims that interfere with authorised uses of content’. It, however, also warns that ‘YouTube takes action to address cases of abuse and error, including disabling specific reference files . . . and releasing all associated claims, disabling Content ID or even terminating YouTube partnership’. See ‘Content eligible for Content ID’, note 32 above. See ‘Clean up incorrect claims’ – https://support.google.com/youtube/answer/4352063. See ‘Content eligible for Content ID’, note 32 above. See Sag, note 5 above, at 544–554. For a comprehensive database on policies set for audio content on YouTube, see the ‘Music Policy Directory’ at https://www.youtube.com/music_policies. See ‘Policy and claims basics’ – https://support.google.com/youtube/answer/107383.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

166

Henning Grosse Ruse-Khan

Content owners can distinguish between how their own content and matching uploads are treated,39 customize policies for individual countries,40 or simply rely on YouTube’s default policy, which is to monetize throughout the world.41 As this also pays out for the platform, the system is unsurprisingly geared towards monetization and towards the use of automated tools like Content ID more generally. According to YouTube, Content ID use ‘has outnumbered copyright takedowns by more than 50 to 1’.42 To incentivize content owners to use this system, YouTube provides them with sophisticated tools to tailor algorithmic enforcement to their individual needs.43 Enforcement means that in principle any match with content claimed by an owner will trigger a Content ID claim,44 without a de minimis threshold for finding ‘infringement’.45 YouTube, however, does ask Content ID users not to ‘overclaim’ and to ‘clean up mistaken claims’.46 It also requires them to ‘[v]alidate potentially invalid reference content’ – suggesting that it employs further checks reviewing in particular whether content claimed is actually eligible for Content ID, including whether it incorporates public domain footage.47 The extent to which these checks occur, and whether they are based on automated or human review, nevertheless remains unclear. Fourth, while platforms generally offer mechanisms for users to dispute a claim by a content owner, dispute resolution is often without any involvement of independent third parties. From the outset, the onus is on the user whose uploaded content has been automatically blocked. YouTube, for example, explains to its users how to dispute a Content ID claim, in particular if they can argue for ‘fair use’ and/or their content being in ‘the public domain’.48 An interesting side-note is that YouTube 39 40 41 42

43

44 45

46

47 48

See ‘Upload and match policies’ – https://support.google.com/youtube/answer/107129. See ‘Create a custom policy’ – https://support.google.com/youtube/answer/106964. See ‘Set default policies’ – https://support.google.com/youtube/answer/3369992. See ‘The difference between copyright takedowns and Content ID claims’ – https://support .google.com/youtube/answer/7002106. For example, to customize a policy in a way that it allows short uploads from fans (which of course may function as promotional tools, and facilitate goodwill – see https://support.google .com/youtube/answer/107129), or to ‘whitelist’ (i.e. exempt) certain channels run by partners or promoters from Content ID claims (see https://support.google.com/youtube/answer/6070344). Since an uploaded video may well be caught by multiple claims, including from different content owners, YouTube also offers mechanisms for prioritizing the more ‘restrictive’ policies (i.e. blocking prevails over tracking – see https://support.google.com/youtube/answer/6300781), to resolve ‘asset ownership conflicts’ and to distribute shares in monetization (see https://support .google.com/youtube/answer/3013321). See ‘Choose effective references’ – https://support.google.com/youtube/answer/4389910. While YouTube does seem to set a minimum length for matches at around thirty seconds, that can be manually set at a lower threshold by those eligible for Content ID. See ‘Clean up incorrect claims’ – https://support.google.com/youtube/answer/4352063. For example, one has to exclude content one does not own from a reference file submitted for identifying matching uploads (https://support.google.com/youtube/answer/4389910). See https://support.google.com/youtube/answer/6013183. See ‘Dispute a Content ID claim’ – https://support.google.com/youtube/answer/2797454. In a November 2018 report, YouTube indicates that less than 1 per cent of all Content ID claims are disputed – see ‘How Google Fights Piracy, note 9 above, at 28.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

Transition through Automation

167

appears to apply the US concept of fair use throughout its platform49 – rather than distinguishing based on the applicable exceptions and other limits in the country for which Content ID protection is sought. If it was to serve as a meaningful defence on the platform (akin to how section 107 of the 1976 Copyright Act has been applied by US courts), this arguably would lead to a more effective ‘globalisation’ of fair use than any of the advocates for rolling the concept out internationally could have hoped for. The problem, however, is that on Content ID, users can’t rely on fair use (or any other defences) effectively. If the user disputes a claim, it is then entirely up to the content owner whether it ‘releases’ or ‘upholds’ the claim. In the latter case, the user normally will have a right to appeal – which is in principle no more than another plea to the content owner to reconsider its claim over the upload.50 An appeal to the content owner, however, does require the latter to request a copyright takedown, based (again regardless of the country for which protection is claimed) on the Digital Millennium Copyright Act 1998 (DMCA), if it wishes to continue to uphold its claim.51 That in turn not only leads to a copyright strike (by YouTube) for the user,52 but also sets the DMCA machinery in motion. The user then has to submit a counter-notification under section 512(g) DMCA if she or he insists on the content to be released.53 In order to keep the disputed content offline, under section 512(g)(2)(C) DMCA, the content owner would in principle have to initiate court proceedings against the user. But ‘to save time for right holders’, YouTube conducts an initial screening as to whether counter-notices ‘provide a sound rationale for reinstatement’.54 For a significant majority of counter-notices, which apparently fail this test,55 the disputed claim will be finally resolved in favour of the content owner – without the user ever having been given the opportunity for independent, let alone judicial, review. One might argue that at least for those cases which have passed YouTube’s initial screening, eventually the matter will be decided by judicial authorities and that the onus is on content owners to initiate such proceedings. This probably is true for USbased users (assuming that they are willing to submit a counter-notification and later

49

50 51 52 53

54 55

See, e.g., the YouTube page targeting a German audience and explaining fair use: ‘Was ist “Fair Use” (angemessene Verwendung)’ – https://www.youtube.com/yt/about/copyright/fairuse/#yt-copyright-protection. See ‘Dispute a Content ID claim’, note 48 above. Ibid. Ibid. For the user, receiving just one ‘copyright strike’ serves as a warning and affects the user’s ability to monetize his or her own content. A user who has received three strikes will have their account terminated and all uploaded videos removed, and will not be able to create new channels on YouTube – see https://support.google.com/youtube/answer/2814000. Users therefore put their YouTube presence on the line if they appeal, and then have to issue a counternotification (thereby risking a lawsuit in the USA – see below) in order to counter the strike. See ‘How Google Fights Piracy’, note 9 above, at 31. YouTube rejects more than two-thirds of counter-notifications at this level – see ibid.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

168

Henning Grosse Ruse-Khan

risk a lawsuit)56 – but will not work as well or as easily for users outside the USA, who apparently are still subject to the DMCA-based process if they wish to retain their uploads on YouTube.57 Under section 512(g)(3)(D) of the US Copyright Act, the contents of any counter-notification must include a statement whereby the user consents to the jurisdiction of the US Federal District Court of her/his permanent residence; or – if the user’s address is outside the USA – to the jurisdiction of ‘any judicial district in which the service provider may be found’.58 In the case of YouTube, this arguably entails consent to be sued in California, where YouTube has its headquarters. Submitting a counter-notification hence means, for a user from outside the USA, that he or she consents to the jurisdiction of a US court, and arguably risks being sued as well as receiving a default judgment. It is apparent that this will prevent most non-US based users from filing counter-notices – which are necessary to overcome the content owner’s rejection of an appeal against any Content ID claim. The fact that YouTube occasionally will support fair use litigation in favour of users59 does little to correct the bias inherent in the Content ID system, which runs on simply accepting a content owner’s claim. In summary then, YouTube is affording content owners a powerful automated and scalable tool to act against any upload which matches with content they claim to own. As a private market actor which benefits significantly from monetization of its users’ content, YouTube has a strong economic interest in its Content ID system being widely used – and little incentive to ensure that the system reflects the traditional limits of copyright protection in terms of subject matter, thresholds for protection and infringement, and exceptions. It is, then, not surprising that YouTube leaves it essentially to content owners to stick to the limits of copyright law when operationalizing Content ID against its users. Whether or not these limits are actually adhered to, enforcement of content owners’ policies is automatic and built into the platform. This usually means that a user has to take legal action, arguing that, under the applicable national law, for example: (1) the claimed content is not copyright protected; (2) the amount utilized does not amount to an infringement; or (3) an exception or other defence applies. So, the burden is on users to take matters into their own hands to enforce their ‘rights’ or other mandatory copyright rules that work in their favour. The problem of course is that, by and large, copyright laws do not conceptualize the various limits to copyright protection as ‘users’ rights’, and they are not 56

57 58

59

Empirical evidence, however, suggests that counter-notifications are ‘a dead letter – impractical and rarely used’ – see Urban et al., note 7 above, at 44. See ‘Counter notification basics’ at https://support.google.com/youtube/answer/2807684. See section 512(g)(3)(D) DMCA – whereby the user will also ‘accept service of process from the person who provided notification under subsection (c)(1)(C) [of the DMCA] or an agent of such person’. See ‘How Google Fights Piracy’, note 9 above, at 32 and ‘YouTube’s Fair Use Protection’ – https://www.youtube.com/intl/en-GB/yt/about/copyright/fair-use/#yt-copyright-protection.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

Transition through Automation

169

necessarily contract or technological protection mechanism (TPM) proof either.60 Arguably, in the world prior to automated enforcement, there was no need for this since the user would normally61 just use content and then copyright owners might need to take action to enforce their rights. Automated content protection through the platform turns this on its head – which requires us to rethink ways to properly conceptualize limits to copyright in this context. In an environment where those who assert ownership over content determine what is protected and the platform defines how, and where these determinations are directly enforced through algorithms, society may wish to consider new and effective tools that re-enforce the public domain and other basic parameters within copyright law. Coming back to the theme of IP in transition, the tools to reinforce these basic parameters would ideally be directly built into the AI mechanisms platforms employ – ‘fair use by design’, as Niva Elkin Koren has called it.62 Others, however, such as Dan Burk, have been rather critical of that approach, warning that ‘the design values embedded in algorithms will inevitably become embedded in public behavior and consciousness’, so that ‘algorithmic fair use’ exports and disseminates programmers’ conscious or unconscious biases or predisposed (training) datasets to all users, thereby ‘progressively altering the fair use standard it attempts to embody’.63 Without addressing these concerns, Article 17(7) of the Digital Single Market Directive (DSM) requires EU Member States to ensure that non-infringing user content remains ‘available’ on the platform. In particular, users must ‘be able to rely on’ the quotation, criticism and review, as well as the caricature, parody and pastiche exception.64 While a lot will depend on how exactly EU Member States implement Article 17(7) DSM, it seems to focus on preventing non-infringing content being blocked or removed. That alone is not enough, however. As the Content ID example shows, nowadays blocking or removing content is the rare exception and 60

61 62 63 64

There are, of course, exceptions: on a proactive approach to copyright exceptions as user’s rights, see the Canadian Supreme Court judgment in CCH Canadian Ltd v. Law Society of Upper Canada [2004] 1 SCR 339, 2004 SCC 13. Mandatory exceptions which cannot be overridden by contract can be found in ss. 29, 29A, 30, 30A and 32 of the UK Copyright, Designs and Patents Act 1988 (CDPA) and in, for example, the EU Software and Database Directives. Also, as discussed below, Art. 17(7) of the Digital Single Market Directive aims to ensure the applicability of certain exceptions in the context of content removal and blocking online. With the exception of situations where TPMs are deployed. Elkin-Koren, note 22 above, at 1193. See Burk, note 22 above, at 2, 9–17; see also the examples in Sag, note 5 above, at 562–563. In full, Art. 17(7) states: ‘The cooperation between online content-sharing service providers and rightholders shall not result in the prevention of the availability of works or other subject matter uploaded by users, which do not infringe copyright and related rights, including where such works or other subject matter are covered by an exception or limitation. Member States shall ensure that users in each Member State are able to rely on any of the following existing exceptions or limitations when uploading and making available content generated by users on online content-sharing services: (a) quotation, criticism, review; (b) use for the purpose of caricature, parody or pastiche.’ (emphasis added).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

170

Henning Grosse Ruse-Khan

monetization the rule. EU Member States should consider this when implementing the DSM. Since the ability to rely on the listed copyright exceptions arguably implies an option to invoke them also against attempts to monetize user content that amounts to a parody, quotation, criticism or review, one would hope that Article 17(7) will be implemented to cover monetization. It has also been argued that (at least by US standards, under section 512(f ) of the Copyright Act) after Universal v. Lenz,65 ‘consideration of fair use requires human review’ – even if the Ninth Circuit did not explicitly say so.66 Outside the notice and takedown system, however, US law does not impose such requirement on privately ordered content-protection tools.67 Article 17(9) DSM on the other hand does require in such cases that ‘decisions to disable access to or remove uploaded content shall be subject to human review’ – but could be read narrowly not to apply to disputes over monetization. Extrapolating from the debates about incorporating fair use into machine learning, a key question will be whether future automated copyright tools can be trained to understand and apply the boundaries and limits of our copyright laws, or at least the core normative principles. What machinecomprehendible proxies can we feed these tools in order for them to recognize works of authorship, and originality; to distinguish between idea and expression; to apply other thresholds for infringements (such as a substantiality test); or to filter out other non-infringing uses, for example based on the more specific, enumerated exceptions in Europe? Without digging deep into the literature on machine learning here,68 one will have little difficulty imagining that the outcomes of such an exercise are very likely to change our copyright norms, perhaps modifying the underlying value judgements and eventually how copyright protection is socially constructed and experienced along the way. The key then is not whether copyright by design is possible (it is already here), or whether it will change copyright protection (it will), but what kind of change we consider desirable, or at least acceptable. That in turn essentially depends on who designs automated copyright protection, and whether there is oversight and accountability within that process. The example of YouTube’s Content ID perhaps primarily shows an overall drive away from blocking towards monetization of user content as default embedded in and facilitated by the system. At best, monetization can be seen as automated licensing of the copyrighted material users integrate into their videos and other created content. In that way, monetizing appears more user-friendly than large-scale takedowns when it comes to user content that infringes. But remember that no one checks whether the content claimed is actually copyrighted, and whether its use is actually infringing. Unless a valid licensing agreement exists, such

65 66 67 68

Lenz v. Universal Music Corp., 815 F.3d 1145, 1154 (9th Cir. 2016). Sag, note 5 above, 533. Ibid., at 544. Instead, see the extensive discussion in Burk, note 22 above.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

Transition through Automation

171

monetization of user-generated content arguably amounts to copyright infringement by the platform – as the user’s content will almost always contain a recording which qualifies for protection under neighbouring rights. Questions about the validity of the licensing aspects in the terms of service agreements aside, excluding the user from any revenues generated through monetizing seems grossly unfair and exploitative. Even where user-generated content involves copyright-infringing material, that may well have minor significance for the popularity of that content, and in any case will usually not be the sole element that drives monetization. Exploiting the user’s ‘digital labour’69 without compensation in this way effectively allows the platform and content owners to cash in on any original or public domain elements of the content uploaded by the user. Such outcomes should not be surprising when system design is by private market actors primarily driven by profit motives. If we think it is important that automated systems also reflect other interests, the stakeholders representing these must have a say and hence get to participate in the ‘quasi-legislative’ process of system design. Of course, this brings us back to the initial question of the role of intermediaries. Do we consider platforms such as YouTube as private parties, public utilities, essential facilities, human rights duty bearers or facilitators (or even guarantors) of freedom of expression online? If they are essentially private entities, then party autonomy, freedom of choice and contract should govern the extent to which users get to influence automated contentprotection tools – perhaps with a few mandatory contract law protections in favour of the weaker party.70 Taking account of the transnational (often global) and (close to) inter omnes effects of norm-setting by platforms described above, one may be inclined to categorize them in a way which takes these effects and their functions into account as well. These (and other) perspectives will determine follow-on questions of accountability and legal responsibility of platforms.71 In my view, a more proactive approach is in principle justified by the effects and functions described above. Legislation should require automated content protection to be more transparent and comprehensible, for example by demanding that a machine-generated outcome is accompanied by explanations showing to those affected in an easy-to-understand way that (and how) copyright eligibility, infringing use, and exceptions and limitations have been considered.72 Platform regulation

69 70

71

72

See the discussion in Soha and McDowell, note 10 above, at 6. These can be in the form of mandatory (contract-proof ) exceptions, as for example in ss. 29 (4B), 29A(5), 30(1ZA) and 30A(2) CDPA. See also Maayan Perel and Niva Elkin-Koren, Accountability in Algorithmic Copyright Enforcement, 19 Stan. Tech. L. Rev. 473 (2016), at 481. This could be linked to a good-faith belief requirement that the claimed user content is actually infringing – akin to the approach in the Lenz judgment (note 65 above) – which should offer, as a minimum, an explanation to the user as to why her/his upload is claimed to be infringing,

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

172

Henning Grosse Ruse-Khan

could also adopt a more rigid standard whereby the platform and those who use its tools to enforce their copyrights show that they have undertaken a due diligence analysis of whether the use of their content as claimed actually amounts to an infringement, under the law of the country where protection is claimed. The platform as principal designer of the system could also be asked to show what exactly it does to prevent abuse (or more specifically, avoid protection beyond the core normative limits of applicable copyright laws) by those claiming rights in user content. Perhaps most importantly, we should consider more effective and fair remedies. Users must be able to contest these matters in meaningful dispute settlement proceedings offered by the platform, which (1) must involve an independent and impartial review – for example by an ombudsperson for platform users; and (2) allow users to challenge claims – importantly, including those involving monetization – in front of courts with jurisdiction over the protection claimed by a content owner. In case a user contests a claim, the onus should be on the content owner to show a prima facie case of copyright or neighbouring right infringement – away from the current default where the ‘process accepts the rightsholder’s claim of infringement as correct until proven otherwise’.73 If automated enforcement allows allegedly infringing use to be stopped just on the basis of a claim, the decision to automatically enforce must be taken with care, and be reasoned and comprehensible to users – which for the time being might well mean that there needs to be a ‘human in the loop’. Article 17(9) DSM addresses some of these demands,74 but focuses on blocking and removing content – hence to a large extent disregarding exploitation of user content through monetization.75 Finally, akin to a duty to indemnify for

73 74

75

and what can be done to challenge this effectively in front of an impartial and independent third party, such as an ombudsman for platform users. In terms of additional transparency and fairness for (business) users of platforms, see also Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services – which, as per Art. 2(1), however, does not apply to consumers and other non-commercial end-users. Sag, note 5 above, at 555. While a full discussion of Art. 17(9) is not possible here, this provision calls for an ‘effective and expeditious complaint and redress mechanism’ available to users in cases of disputes over ‘disabling access to, or removal of’ their content. It asks right holders to ‘duly justify the reasons’ for their blocking and/or removal requests, and demands that ‘decisions to disable access or to remove uploaded content shall be subject to human review’. In its last sentence, however, Art. 17(9) does impose an obligation on EU Member States to ‘ensure that users have access to a court or another relevant judicial authority to assert the use of an exception or limitation to copyright and related rights’ (emphasis added). Assertion of an exception in this sense would arguably cover cases where the user complains about being subjected to monetization of his or her content that benefits from a copyright exception. One hopes that EU Member States will implement this provision in such a way that not only disputes over blocking or removing content where users invoke copyright exceptions must end up in front of courts – but also those about monetization of content that falls under copyright exceptions.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

Transition through Automation

173

damages suffered because of a request for preliminary (injunctive) relief or related court orders, one might even consider strict liability of those who decide to automatically enforce, should a court or other independent third party later find that no infringement existed. Overall, the idea is to adopt measures that account for, and rebalance, the fundamental shift which is inherent in algorithmic enforcement where our ability to experience content online is predetermined by what automation allows us to see and hear.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.019

13 Eye, Robot: Artificial Intelligence and Trade Mark Registers Dev S. Gangjee1

13.1 introduction: reading trade mark registers The edifice of trade mark registration exists primarily to provide useful information. Registers tell us who owns what. They signal the existence of exclusive property rights associated with commercial signs, thereby allowing other traders to plan around that information. These signals exist in ever increasing numbers. According to the World Intellectual Property Organization (WIPO), an estimated 9.11 million new trade mark applications were filed worldwide in 2017, while in the same year there were an estimated 43.2 million active trade mark registrations at 138 offices worldwide.2 Until recently, it was axiomatic that registers for marks were directed at human readers – an applicant for a trade mark, trade mark registry examiners, vigilant competitors, employees of search and watching agencies as well as the occasional judge. This list now has a new entrant. What are the implications for the registered trade mark ecosystem, when algorithms begin to efficiently and comprehensively read trade mark registers? This is an especially apt question for this volume, since the colleague we are honouring – Professor Annette Kur – has a long-standing interest in registrationbased protection regimes. My personal favourite, where scholarly rigour meets forensic triaging on a pan-European scale, is the insightful Max Planck Study which influenced the reform of EU trade mark law.3 Annette’s scholarship on registered

1

2 3

Faculty of Law, University of Oxford. I am grateful to Peter Lambert and Peter Keyngnaert (Clarivate) as well as Lee Curtis (HGF) for background information and comments. WIPO, World Intellectual Property Indicators 2018 (WIPO, 2018), 88, 95. Max Planck Institute, Study on the Overall Functioning of the European Trade Mark System (2011). Equally admirable is the mastery demonstrated (with Martin Senftleben) in European Trade Mark Law: A Commentary (Oxford University Press, 2017).

Downloaded from https://www.cambridge.org/core. 174 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

Artificial Intelligence and Trade Mark Registers

175

trade mark and design protection guides us through some of the most complex and contested debates in these fields.4 However, the influence of artificial intelligence (AI) on trade mark registration is more subtle than its impact on patent or copyright law. The creative and inventive domains of IP have to contend with seemingly existential challenges: whether increasingly autonomous computer software ought to be considered an inventor or author and whether the corresponding outputs should be recognised as protectable subject matter.5 In the trade mark context, machine learning has developed to the point where AI algorithms can readily assess the similarity between marks as well as goods and services, flagging up potential conflicts. At first glance, this seems like an enhancement which merely allows users of registries to do what they already do, but better. The legal implications of an AI algorithm which ‘reads’ a trade mark register, replicating or entirely replacing human judgement for certain stages of analysis, has not yet been considered in any detail. Machine learning algorithms are involved in identifying potentially conflicting prior rights when selecting a mark, during clearance checks by private service providers or by trade mark registries themselves. The speed and comprehensiveness of coverage, as well as the increasingly routine application of AI, is potentially game changing. This chapter outlines these recent developments and considers some of their implications.

13.2 ai and trade marks: setting the scene AI is commonly used as an ‘umbrella term to cover a set of complementary techniques that have developed from statistics, computer science and cognitive psychology’.6 According to a White House report, it conventionally refers to ‘a computerized system that exhibits behaviour that is commonly thought of as requiring intelligence. Others define AI as a system capable of rationally solving complex problems or taking appropriate actions to achieve its goals in whatever real world

4

5

6

For representative works by Anette Kur, see Not Prior in Time but Superior in Right – How Trademark Registrations Can be Affected by Third Party Interests in a Sign (2013) 44 IIC 790; Probleme im Überschneidungsbereich von Marken und Designs (2017) MarkenR 185 (considering trade mark and design protection overlaps); Gemeinfreiheit und Markenschutz (2017) GRUR 1082 (analysing the Vigeland decision which facilitates the free use of public cultural symbols, thereby strengthening the public domain); Ersatzteilfreiheit zwischen Marken- und Designrecht (2016) GRUR 20 (exploring whether the exceptions relating to spare parts need to be better co-ordinated across trade mark and design law). See e.g. R. Abbott, I Think, Therefore I Invent: Creative Computers and the Future of Patent Law (2016) 57(4) Boston College L. Rev. 1079; WIPO, Technology Trends 2019: Artificial Intelligence (WIPO, 2019); A. Ramalho, Will Robots Rule the (Artistic) World? A Proposed Model for the Legal Status of Creations by Artificial Intelligence Systems (2017) 21 Journal of Internet Law 1; J. Grimmelmann, Copyright for Literate Robots (2016) 101 Iowa L. Rev. 657. W. Hall and J. Presenti, Growing the Artificial Intelligence Industry in the UK (HM Government, Independent Report, 2017), 4.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

176

Dev S. Gangjee

circumstances it encounters’.7 The goal is for software to replicate intelligent behaviour, and remarkable progress has been made on so-called ‘Narrow AI, which addresses specific application areas such as playing strategic games, language translation, self-driving vehicles, and image recognition’.8 Recent successes are in part attributable to the advent of big data and improved computing power. Improved data flows form the basis for better-quality machine learning, which ‘is the technology that allows systems to learn directly from examples, data, and experience’.9 This technology allows software algorithms to learn from data (or examples), drawing statistical inferences and identifying patterns, rather than by following preprogrammed rules. Machine learning is iterative, so that when an algorithm is exposed to new data, it can adapt. These systems are already ubiquitous. ‘Many people now interact with machine learning-driven systems on a daily basis: in image recognition systems, such as those used to tag photos on social media; in voice recognition systems, such as those used by virtual personal assistants; and in recommender systems, such as those used by online retailers’.10 While image recognition technologies are considered below, recommendation systems in the online retail context hint at the broader implication for trade mark infringement.11 A platform’s virtual assistant or AI-powered recommendation system – think Amazon or eBay – might respond to a search query by suggesting an infringing product, based purely on statistical correlations relating to past searches on that platform. Alternatively, the system might treat the trade mark in a search query generically, as shorthand for a product class and offer the products of competitors within that class. Should the platform or online service provider be held liable if an autonomous AI system is making retail recommendations that infringe trade mark rights? Online grocery retailers or supermarkets already recommend substitutes if the desired product is not available and may offer a competing brand in the process. Is this helpful to consumers and competition enhancing, or damaging to trade mark owners’ legitimate interests? Conversely, when virtual assistants shop on our behalf, does the ‘average consumer’ hypothetical construct, characterised by imperfect recollection and the inability to make side-by-side comparisons, still apply to AI shoppers when assessing infringement?12 7

8 9

10 11

12

US Executive Office of the President, Preparing for the Future of Artificial Intelligence (October 2016), 6. Ibid., 7. Royal Society, Machine Learning: The Power and Promise of Computers that Learn by Example (April 2017), 16. Ibid. For infringement issues, see L. Curtis and R. Platts, AI is Coming and it Will Change Trade Mark Law (2017) Managing IP 9; L. Curtis and R. Platts, Alexa, ‘What’s the Impact of AI on Trade Mark Law?’ (2019) Managing IP 43; M Grynberg, ‘AI and the “Death of Trademark”’ (2019-2020) 108 Kentucky Law Journal 199. On the average consumer, see G. B. Dinwoodie and D. Gangjee, The Image of the Consumer in European Trade Mark Law, in D. Leczykiewicz and S. Weatherill (eds.), The Image(s) of the Consumer in EU Law (Hart, 2016), 339.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

Artificial Intelligence and Trade Mark Registers

177

The ongoing transformation of retail services forms the backdrop for this set of questions. As enticing as they are, we must leave a more detailed consideration for another day. As an initial response, parallels might be drawn with keyword advertising case law, where AI algorithms have offered competing products in response to search terms consisting of trade marks. The answers to infringement questions may turn on how the results of the search are presented to consumers, as opposed to how the AI internally processes the trade mark. Where product recommendations are provided with suitable clarifications and qualifications, they should be permitted. Where they are misleading or ambiguous, they are likely to be infringing.13 Putting infringement to one side, the task for this chapter is to highlight the remarkable inroads that AI is making into the everyday processes of trade mark registration, as reflected in its adoption by two very significant institutional actors: (i) trade mark registries, as well as (ii) trade mark clearing, searching and watching agencies.14 The developments described in this chapter are intended to augment existing practices relating to trade mark search, examination and watching. Commercial search and watching agencies have been at the forefront of technological developments. Most of the technology described below is designed to identify prior conflicting rights, as evidenced by similar marks for similar goods. Searching, for the purposes of clearing a potential new mark, is described as: [T]he critical legal step in the process of selecting a new mark. The search enables a trade mark lawyer to determine whether a mark is available for use and likely to be registrable. For a business to launch a new product or service without first conducting a search is to flirt with commercial disaster. A search is necessary because, simply stated, trade mark rights are granted on a first-come, first-served basis.15

Searches therefore anticipate objections by the registry or oppositions based on prior rights, allowing applicants to assess the risks of proceeding with a registration. Watching agencies provide a commercial service for clients with successfully registered trade marks, whereby the agencies monitor new trade mark applications by third parties across selected jurisdictions and flag up potential conflicts. The clients can then decide whether or not to oppose them.16 Both these processes are primarily concerned with registry-level conflicts.

13 14

15 16

See e.g. Cosmetic Warriors and Lush v. Amazon.co.uk and Amazon EU [2014] EWHC 181 (Ch). On the significance of bureaucratic practices such as search and classification in constituting the very objects of trade mark protection, see J. Bellido and Hyo Yoon Kang, In Search of a Trade Mark: Search Practices and Bureaucratic Poetics (2016) 25(2) Griffith L. Rev. 147; J. Bellido, Towards a History of Trade Mark Watching [2015] IPQ 130. G. A. Gundersen, Trademark Searching and Clearance (INTA Guide, 2018), 1–2. Bellido, Towards a History, note 14 above.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

178

Dev S. Gangjee

13.3 searching for similarity In 2018, WIPO convened a review of the experiences of intellectual property offices that had experimented with AI algorithms to increase efficiency and reduce costs.17 While AI is supporting major IP Offices in a number of specific (e.g. classifying patents according to relevant technology groups for examination purposes) and general ways (e.g. chatbots as part of help desk services, to assist applicants with queries), certain trade mark specific applications have been identified.18 13.3.1 Goods and Services Classification First, AI algorithms are being used to automatically recommend classes for goods and services contained in trade mark applications. Along with the sign being claimed as a mark, the trade mark application also indicates the goods and services identified by that sign.19 This reflects commercial practice: the Nike ‘swoosh’ logo is applied to athletic footwear, sports clothing and accessories. The international reference point is the Nice Classification system, maintained by WIPO, which consists of a list of headings of thirty-four classes of goods and eleven classes of services as well as an alphabetic list of goods or services in each class.20 Historically, there has been no mandatory form or prescribed terminology for specifying these goods and services. Applicants are free to choose, based on their own commercial preferences. However, the terms selected by an applicant are subsequently slotted into the relevant classification taxonomy adopted by the trade mark registry. Thus an applicant selling remote-controlled aerial vehicles might specify ‘toy drones’ on the form but will have to identify Class 28 (‘Games and Playthings’) of the Nice Classification as the relevant class. Accuracy is important here. One of the primary purposes of bureaucratic classification is to enable efficient searching by registries and third parties for conflicting prior marks in relation to identical or similar goods. In order to avoid rejections based on inaccurate terminology or mistaken classifications, the Intellectual Property Office of Singapore (IPOS) has developed an International Classification of Goods and Services (ICGS) Autochecker software 17

18

19

20

See generally the Meeting of Intellectual Property Offices (IPOS) on ICT Strategies and Artificial Intelligence (AI) for IP Administration, 23–25 May 2018, Geneva (WIPO/IP/ITAI/ GE/18). The responses to a survey relating to AI usage is contained in WIPO Secretariat, Original Replies from IPOs in English, French or Spanish, 29 March 2018 (WIPO/IP/ITAI/GE/ 18/2 REV). See WIPO, Index of AI initiatives in IP Offices, at: https://www.wipo.int/about-ip/en/artificial_ intelligence/search.jsp. Case C-307/10, Chartered Institute of Patent Attorneys v. Registrar of Trade Marks (IP TRANSLATOR) ECLI:EU:C:2011:784 (AG Bot), [1]: ‘The two essential components of the registration of a trade mark are (a) the sign and (b) the goods and services which that sign is to designate. Each of those components makes it possible to define the precise subject-matter of the protection conferred by the registered trade mark on its proprietor.’ See: https://www.wipo.int/classifications/nice/en/.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

Artificial Intelligence and Trade Mark Registers

179

tool, which relies on a natural-language processing AI system.21 Applicants can verify their list of selected product terms against the thousands of pre-approved terms in the IPOS database. The software identifies misspellings, duplicated items, terms not found within the database of approved terms and items listed under the wrong class heading. TM Class is another widely used tool which helps applicants to identify appropriate terms and corresponding classes from within a consolidated preapproved list provided by participating registries.22 It additionally provides a hierarchical taxonomy of pre-approved terms, thereby enabling applicants to locate their preferred product description within broader or more narrow fields by moving up or down the hierarchy. China is presently developing a ‘Standard Goods System’, which clusters existing terminology relating to goods into groups, based on their similarity, so as to ‘establish the Goods Relation Dictionary. With this dictionary, the system automatically allocate[s] newly-supplied goods into the respective . . . group. For goods supplied for the first time, a mother goods [sic] would be designated to begin a group.’23 Germany and Japan have also invested in developing such systems,24 while WIPO is working on AI that will be used to predict the most relevant Nice classifications, improving on the former text-search matching model.25 These tools reduce examination times and rejection rates based on incorrect classifications, leading to cost savings for registries and users.

13.3.2 Identifying Similar Marks: Semantic and Image Searches Assessing the similarity of marks is at the heart of legal tests for: (i) relative grounds of opposition, which allow the trade mark proprietor to oppose the registration of a similar subsequent mark (a registry level conflict); or (ii) trade mark infringement, where that proprietor objects to a similar or identical sign being used in the marketplace by a third party (a real world conflict). Registry level conflicts are the primary focus of this chapter, although the underlying technology is relevant for policing infringement as well. The most widely used legal test considers whether the similarity of marks, when combined with the similarity of goods or services, is likely to cause confusion for relevant consumers of those products.26 Marks are assessed in terms of their visual, aural/phonetic or conceptual similarity, also referred to as sight, 21

22 23

24

25

26

The instructions are available at: https://www.ipos.gov.sg/docs/default-source/resources-library/ trade-marks/resources/autochecker-user-guide.pdf. See http://euipo.europa.eu/ec2/. WIPO, Summary of the Replies to the Note on Applications of AI to IPO Administration, 8 February 2018 (WIPO/IP/ITAI/GE/18/1), [23]. Japan Patent Office, Outline of JPO's Activities for Using AI, 29 May 2018 (WIPO/IP/ITAI/GE/ 18/P9). Patrick Fiévet, Artificial Intelligence applied to IPC and NICE Classifications, 25 May 2018 (WIPO/IP/ITAI/GE/18/P18). I. Fhima and D. S. Gangjee, The Confusion Test in European Trade Mark Law (Oxford University Press, 2019).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

180

Dev S. Gangjee

sound and meaning analysis.27 For complex or composite marks, which combine words and/or figurative elements, the comparison should consider each mark as a whole while also recognising the distinctive and dominant elements that consumers would notice. Algorithms for assessing the similarity of marks can be distinguished based on the types of mark being compared. Relatively straightforward computerised text searches have been available for several decades.28 In the past, these search systems employed ‘text-based retrieval technology . . . [which] look[s] for trade marks that match some or all words in a query string text’.29 Text search has improved over the years to incorporate phonetic analogies, synonyms and permutations of letters so that slightly modified marks are also returned in the search results.30 Recent advances have expanded the scope of similarity searching across three dimensions. As regards the first of these, algorithms are being developed to assess the conceptual similarity between marks, on the basis of shared or even oppositional meanings. A simple text search will not flag up the semantic similarity between ‘H2O’ and ‘water’. On the other hand, signs which look textually or visually similar may relate to different concepts. Homophones sound similar but have different meanings (steel v. steal) while homographs are spelled the same, but the context clarifies the difference in meaning (bass being a type of fish or the lowest frequencies in music). Capitalisation alone can produce significant changes: compare ‘Polish’ (the nationality) with ‘polish’ (for furniture or shoes). Therefore search technology based on semantic or conceptual similarity considers synonyms or antonyms, comparable words in another language with similar meanings and ‘lexical relations’ (PINK LADY v. LADY IN ROSE).31 The second domain of similarity comparison, which has seen significant improvements in recent years, is image search. The technology has advanced to the point where a user can directly upload an image in a recognised file format such as JPG, PNG, GIF or TIFF and search for similar images within the relevant registry database. Both WIPO and the European Union Intellectual Property Office

27

28

29

30

31

USPTO, Trade Mark Manual of Examining Procedure (April 2016) at §1207.01(b) (Similarity of the Marks); EUIPO, Guidelines for Examination of European Union Trade Marks, Part C, Section 2, Chapter 4 – Comparison of Signs (v 1.1, October 2017). The history of computerisation and the pivotal role of the watching agency Compumark is detailed in Bellido, Towards a History, note 14 above. F. Mohd Anuara, R. Setchia and Y.-K. Lai, A Conceptual Model of Trademark Retrieval based on Conceptual Similarity, (2013) 22 Procedia Computer Science 450, 451. C. J. Fall and C. Giraud-Carrier, Searching Trademark Databases for Verbal Similarities (2005) 27(2) World Patent Information 135. For an application of text-search technology measuring the distance between word marks and identifying conflicts, in order to assess whether viable word marks are being depleted, see B. Beebe and J. C. Fromer, Are We Running Out of Trademarks? An Empirical Study of Trademark Depletion and Congestion (2018) 131(4) Harvard L. Rev. 945. Anuara et al., note 29 above, 453.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

Artificial Intelligence and Trade Mark Registers

181

(EUIPO) offer this facility.32 National offices are actively incorporating this technology into the internal registry examination process.33 A moment’s reflection reveals some of the information-processing challenges that need to be overcome. Figurative marks and logos ‘are designed to have visual impact . . . consisting of multiple homogeneous elements, which may be closed regions, lines, or areas of texture. They may represent a given type of object (such as a dog or car) in stylised form, or consist purely of abstract patterns. They may be coloured or monochrome.’34 Human observers consider shape to be the single most important feature of an image but image structure (the layout of individual image elements) and semantic interpretation (the image of a tree evoking trees) are also relevant. Then there is the matter of identifying what counts as an image element – how granular does it get and why?35 It rapidly becomes apparent that there are several, oftentimes subjectively prioritised, parameters according to which similarity might be assessed. One attempt to respond to these challenges is found in the creation of the International Classification of the Figurative Elements of Marks, also referred to as the Vienna Classification, administered by WIPO.36 As matters presently stand, figurative marks are manually indexed by trade mark examiners, with codes or keywords being assigned to them.37 The most widely adopted of these coding systems is the Vienna Classification, a hierarchical system proceeding from the general to the particular. It consists of categories, divisions and sections, each of which has been assigned a number. Figurative elements classified within a section are referred to by three numbers: ‘the first, which may be any number from 1 to 29, denotes the category; the second, which may be any number from 1 to 19, the division; and the third, which may be any number from 1 to 30, the section. For instance, the representation of “a little girl eating” [02.05.03] belongs to Category 2 (Human beings), Division 5 (Children), Main Section 3 (Girls).’38 The result is over 1,300 different taxonomic categories. Matching the Vienna codes of a new

32

33

34

35 36

37

38

WIPO facilitates images searches in its Global Brand Database: https://www.wipo.int/branddb/ en/. The EUIPO consolidates registration information via its TMview database, which combines the register of EU-wide European Union Trade Marks with that of twenty-seven national EU Member States as well as the WIPO database: https://www.tmdn.org/tmview/welcome. WIPO/IP/ITAI/GE/18/1, note 22 above, refers to image search initiatives by the IPOs of Australia, Chile, China, Japan, Norway and Singapore as well as the EUIPO and WIPO. J. Schietse, J. P. Eakins and R. C. Veltkamp, Practice and Challenges in Trademark Image Retrieval, in Proceedings of the 6th ACM international Conference on Image and video Retrieval (2007), 1, 1. Ibid. See the Vienna Agreement Establishing an International Classification of the Figurative Elements of Marks, 1973. The Agreement entered into force on 9 August 1985. The current (eighth) edition has been available online since 2017 and in force from 1 January 2018. See: https://www.wipo.int/classifications/vienna/en/. For an example of a code-based system, see the USPTO Design Search Code Manual, available at: http://tess2.uspto.gov/tmdb/dscm/index.htm. See https://www.wipo.int/classifications/vienna/en/preface.html.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

182

Dev S. Gangjee

application with those already in the registry database generates a list of similar figurative marks. However, not all trade mark registries use the Vienna Classification system – it presently has thirty-four contracting parties – while the allocation of codes (or keywords in some systems) to figurative elements inevitably involves some subjectivity, with the attendant risk of leaving gaps.39 This background helps to explain why recent advances in content-based image retrieval systems, which can be used to directly compare a target image to other images in a database, have been welcomed by trade mark registries and professionals. The search results list (hit list) is prioritised by an AI-assisted process so that only the closest matches and most relevant images are presented. More tightly focused lists of potential conflicts save reviewing time and therefore reduce costs for trade mark examiners, trade mark attorneys or agents, paralegals, legal practitioners and ultimately clients. Image recognition systems directly analyse the colours, shapes and textures of images rather than relying on representative keywords or codes.40 Research by WIPO indicates that image searches to date have been more effective in relation to simple geometric shapes, with Vienna Classification searches as the background comparator for measuring effectiveness.41 However, there is room for improvement when it comes to complex shapes or logos, combining both figurative elements and text. Nevertheless, the ability to search by directly uploading an image is a significant, potentially tectonic, shift, especially as the technology continuously improves. Thus WIPO’s new AI-based image comparison service innovates ‘by using deep machine learning to identify combinations of concepts – such as an apple, an eagle, a tree, a crown, a car, a star – within an image to find similar marks that have previously been registered’.42 Meanwhile the United States Patent and Trademark Office (USPTO) has built on its extensive experience with manually coding figurative images to train AI algorithms: ‘A six-digit numerical design search code is assigned to each design element of a trade mark, such as a depiction of a star 39

40

41

42

WIPO, Future Development of the Vienna Classification: Questionnaire Results (3 April 2019), 3. Although most participating registries were satisfied with the system, the following issues were raised: ‘a number of Offices reported difficulties with the classification of colours. Codes that are over-used, as well as codes that are not used enough and are simply ignored also lead to an incomplete classification. Modernisation . . . by adding new codes for commonly used figurative elements and for keeping pace with new technologies was also highlighted.’ For technical background, see A. Alaei, P. P. Roy and U. Pal, Logo and Seal Based Administrative Document Image Retrieval: A Survey, (2016) 22 Computer Science Review 47; C. A. Perez et al., Trademark Image Retrieval Using a Combination of Deep Convolutional Neural Networks, in International Joint Conference on Neural Networks (IJCNN) (IEEE, 2018). Christophe Mazence, Machine Learning Applied to Trademarks Classification and Search, 29 May 2018 (WIPO/IP/ITAI/GE/18/P17). WIPO Launches State-of-the-Art Artificial Intelligence-Based Image Search Tool for Brands, 1 April 2019 (PR/2019/831), at: https://www.wipo.int/pressroom/en/articles/2019/article_0005 .html.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

Artificial Intelligence and Trade Mark Registers

183

(01.01.03) or flower (05.05.25). Using years of images with corresponding examinerannotated design codes, we are able to train deep learning systems that can predict design codes of a new trade mark image.’43 The USPTO also uses neural networks to retrieve and store features of mark images, that can then be compared, via an image similarity measure, to other marks’ features. Instead of developing their own systems, some registries use commercially developed image search solutions. For instance, IP Australia has an Image Search option within its Australian Trade Mark Search feature, to search for existing trade mark images based on a given image. It uses the commercially available TrademarkVision Image Recognition software for this purpose.44 The EUIPO uses this system as well. TrademarkVision is now a part of Clarivate Analytics, which had previously acquired a leading trade mark clearance and protection company CompuMark.45 The software incorporates a deep learningbased reverse image search, similar to the facial recognition algorithms of Facebook and Google but applied to figurative marks and logos. At the time of writing, the algorithm generates a list of similar figurative marks, which can be filtered by jurisdiction, class of products, status (live or dead), specific goods or services and the owner’s name. The system also facilitates proactive monitoring, emailing alerts when similar marks are subsequently discovered.46 TrademarkVision is thus a relatively new entrant in the trade mark search/clearance and watching ecosystems. There are others as well: ‘TrademarkVision, MikeTM Suite and LawPanel’s Aila are some of the existing AI-powered solutions. With various specialties, they essentially aim to search and monitor words, phrases, and images using various government trademark offices’ databases of marks and the internet.’47 We return to the implications for search and watching agencies in the next section of this chapter. The third aspect which is being explored could be described as putting the pieces together or developing a gestalt understanding. Attempts are underway to develop AI algorithms that can combine different measures of similarity – the words and images in the two complex marks being compared – to arrive at an integrative assessment. The goal is to mimic the assessment of a human examiner who must synthesise visual, aural and conceptual similarity to arrive at an overall conclusion on whether

43

44

45

46 47

USPTO, Emerging Technologies in USPTO Business Solutions, 25 May 2018 (WIPO/IP/ITAI/ GE/18/P5). Robert Bollard, IP Australia Artificial Intelligence Initiatives, 29 May 2018 (WIPO/IP/ITAI/GE/ 18/P7). Clarivate Analytics Snaps Up TrademarkVision as New Research Offerings Eyed, World Trademark Review, 31 October 2018, at: https://www.worldtrademarkreview.com/ip-offices/ clarivate-analytics-snaps-trademarkvision-new-research-offerings-eyed. See: https://trademarkvision.compumark.com/solutions/trademarks/. How AI Impacts Trademarks, Trademark Times, 19–23 May 2018 (Managing Intellectual Property: Seattle 2018 INTA Annual Meeting).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

184

Dev S. Gangjee

the marks conflict.48 Datasets of past examiners’ decisions are a resource for training algorithms to assign different ‘weights’ to different measures of similarity, depending on the type of mark. Another approach presents the relevant measures of similarity in parallel – for example image or pixel similarity, text similarity, automated content similarity and manual similarity – to avoid their significance being diminished or compressed into a single metric.49

13.3.3 Assisting with Examination The techniques described above, for assessing the similarity of goods or services as well as marks, are assisting registries in processing trade mark applications more efficaciously. For example, the Japanese Patent Office uses an AI-assisted goods and services similarity assessment tool so that its examiners can identify the closest match between a new application and previously registered (approved) terms to describe goods and services. This speeds up the process for checking whether the correct product classes have been indicated in the application, as well as the field of prior marks with which to compare the new application.50 IP Australia has been developing a ‘Smart Assessment Toolkit’ for examiners, using machine learning models to flag up potential issues. This includes identifying prior similar word marks for similar goods or services to prioritise conflicting marks. It also incorporates a form of distinctiveness assessment. For example, the candidate sign may be considered descriptive for the goods selected or contain a common or generic term.51 The software uses a combination of natural language processors and datasets of historic reports by examiners for learning purposes when evaluating registrability. Meanwhile IPOS has been experimenting with using ‘machine learning to automatically measure the distinctiveness of a given word mark and also to suggest evidence for the measurement (Trade Marks Distinctiveness Checker). This [initial steer] helps officers speed up the examination step of distinctiveness and thus reduces turnaround time. The automatic measurement of this task can be also used by applicants, in order to reduce rejection rate due to [non-]distinctive word marks.’52

48

49

50 51

52

R. Setchi and F. Mohd Anuar, Multi-Faceted Assessment of Trademark Similarity, (2016) 65 Expert Systems with Applications 16; see also Perez et al., note 40 above (assessing the combined visual and conceptual similarities of images). I. Mosseri, M. Rusanovsky and G. Oren, TradeMarker – Artificial Intelligence Based Trademarks Similarity Search Engine, in C. Stephanidis (ed.), HCI International 2019 – Posters, Communications in Computer and Information Science, vol. 1034 (Springer, 2019), 97. WIPO/IP/ITAI/GE/18/P9, note 24 above. At https://www.govtechreview.com.au/content/gov-tech/news/trademark-applications-stream lined-by-machine-learning-1008018169. WIPO/IP/ITAI/GE/18/1, note 23 above, at [44].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

Artificial Intelligence and Trade Mark Registers

185

Besides assisting examination, IPOs are developing user assistance tools designed with self-filers in mind. To take one example, in May 2018, IP Australia launched Trade Mark Assist, an online tool that assists users wishing to check whether their chosen sign is a good candidate for registration.53 The tool indicates whether prior similar marks exist on the Australian register and whether non-distinctiveness objections might apply to the sign.

13.4 implications for trade mark law An AI-assisted search results in a relevancy ranked list of potential conflicts along with a risk profile, often presented statistically (e.g. a 71 per cent similarity match). This assists registry officials and professional advisers when making recommendations. Experience to date therefore suggests that AI algorithms are intended to augment human judgement – to effectively sift through ever-increasing volumes of registration data – and not to replace it. Emphasising the continuing need for nuanced human evaluations, a CompuMark report observes: ‘While AI and neural networks will play an expanding role in CompuMark solutions . . . they are intended to complement, not replace, human analysts.’54 As the founder of LawPanel puts it: ‘AI will speed up legal research, but it will not replace advice formulation . . . [since it] only works on repetitive tasks in a very tightly-defined domain’.55 Machine learning technology can comprehensively filter the ever-growing numbers of trade mark applications and registrations, displaying the most relevant list of results for human experts to assess. Yet despite the projection of ‘(enhanced) business as usual’, is a more fundamental set of realignments taking place? A general question, beyond the remit of this chapter but one of interest to decision makers across many legal and regulatory fields, is the extent to which algorithmic decision making differs from human decision making.56 That depends on the nature of the analysis: where the data for a machine learning approach is derived from judicial content analysis – past decisions by human tribunals where factors are coded and correlations derived – the algorithm will behave like the human decision maker it is modelled after, warts 53 54

55

56

At https://trademarks.business.gov.au/assist/welcome. CompuMark White Paper, Artificial Intelligence, Human Expertise: How Technology and Trademark Experts Work Together to Meet Today’s IP Challenges (2018), 5. T. Lince, ‘No Imminent AI Apocalypse’ – Tech Expert Rejects Predictions of Mass Job Losses in Trademark Industry, 1 February 2018, World Trademark Review, at: https://www .worldtrademarkreview.com/brand-management/no-imminent-ai-apocalypse-tech-expertrejects-predictions-mass-job-losses. See e.g. European Parliamentary Research Services, Understanding Algorithmic DecisionMaking: Opportunities and Challenges (Report PE 624.261, March 2019); Centre for Data Ethics and Innovation, Interim Report: Review into Bias in Algorithmic Decision-Making (25 July 2019), at: https://www.gov.uk/government/publications/interim-reports-from-thecentre-for-data-ethics-and-innovation/interim-report-review-into-bias-in-algorithmic-decisionmaking.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

186

Dev S. Gangjee

and all. On the other hand, when it comes to assessing semantic similarity, algorithms may also stretch beyond the capacities of human analysts and identify alternatives from within a dictionary or thesaurus which might escape human assessors, or, using a historic database of goods and services, identify product similarities that humans would be unlikely to perceive. At this point, the algorithm may be outperforming a human trade mark examiner. Any resulting match and associated risk indicator may be erring on the side of caution, since real-world examiners would not pick up on those similarities. Unless of course registries are using similar algorithms(!). Algorithms may also find patterns which are ‘valid’ but not causally related in a meaningful way to the rules of trade mark law. Peter Keyngnaert provides an example of an irrelevant pattern – where a new company X files fifty new trade mark applications on (say) 14 August 2019 but all contain an error (a reference to class 54 instead of 34) and are rejected, the algorithm may identify a high risk of rejection associated with company X, or a high risk of rejection associated with 14 August.57 One implication is that the existing gap between the likelihood of confusion test for relative grounds analysis and infringement analysis will continue to expand. An attenuated version of the likelihood of confusion test in trade mark law is applied upstream, at the initial stages of trade mark clearance. Conflict analysis conducted by algorithms focuses on similarity between the marks and goods as they appear on the register. However, the broader multifactor likelihood of confusion test for infringement incorporates additional elements. In the USA, although the test varies across judicial circuits, significant factors include: the similarity of the marks; the similarity between goods or services; evidence of actual confusion; the strength of the plaintiff’s mark (in the form of inherent distinctiveness or acquired repute); the defendant’s intent (including bad faith); and consumer sophistication.58 In the EU, the assessment proceeds in three interdependent and cumulative stages:59 (i) similarity of the marks; (ii) similarity of goods or services; and (iii) does this lead to a likelihood of confusion? Under the third stage, the (hypothetical) average consumer for the overlapping goods or services needs to be constructed, her levels of attention and sophistication established, her linguistic attributes clarified and evidence of market conditions (for example, whether products are likely to be sold over the counter, with professional assistance or via websites) identified. Evidence of actual confusion may also be submitted, in the form of witness statements, trade evidence or consumer surveys. Once these additional factors are incorporated, it may emerge that the similarity between two marks derives from themes common to the product sector (both incorporate the colour green and the image of a tree on environmental 57

58

59

I am extremely grateful to Peter Keyngnaert, Manager Research Scientist at CompuMark, for his patient and helpful guidance. A representative approach is: Polaroid Corp v. Polarad Electronics Corp, 287 F 2d 492 (2d Cir 1961). For details, see Fhima and Gangjee, The Confusion Test, note 26 above.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

Artificial Intelligence and Trade Mark Registers

187

products). Consumers might conclude the marks appear similar but they would not be confused about the commercial origin of the respective products, because that similarity could be discounted (motifs common to the sector).60 Or a common class heading – such as Class 28 for Games and Playthings – may actually resolve into two very different subordinate categories of products: soft toys aimed at toddlers versus role-playing board games for a teenage market. Some of these additional factors – such as constructing the notional consumer perspective – are relevant for relative grounds analysis as well, which is why leading search or watching agencies clarify that human analysts, with the experience to draw these distinctions, remain integral to the services that they offer. Today conflict analysis for registry clearance purposes is already a thinner, more formalistic version of conflict analysis for real-world infringement purposes, where context can be taken into account.61 AI-enhanced similarity searching may serve to further attenuate registry-level conflict analysis. Will the test effectively shrink to just two factors (marks and products) in the commercially significant clearance or registry opposition context?62 The reactions of a real-world consumer, so often alluded to in trade mark doctrine, may be muted even further as a result. While human expertise continues to assess the conflicts in results lists generated by algorithms, for risk-averse commercial clients it is extremely tempting to be guided by clearly defined percentages of similarity. Alongside the attenuation of the multifactor test, a related issue is the increasing legal significance of bureaucratic classification of goods and services by registries. It has been repeatedly emphasised that bureaucratic classification systems such as Nice do not have a bearing on the similarity of goods and services analysis when rights conflict. EU trade mark law confirms that goods and services cannot be assumed to be similar from the fact that they are in the same class and, vice versa, it cannot be assumed from the fact that they are in different classes that they are dissimilar.63 However, similarity within the Nice Classification hierarchy is seemingly considered relevant for risk analysis by algorithms. One is left to wonder whether the algorithmic assessment of similarity 60

61

62

63

As one US decision memorably puts it: Paula Payne Prods Co v. Johnson’s Publ’g Co, 473 F 2d 901, 902, 177 USPQ 76, 77 (CCPA 1973) (‘[T]he question is not whether people will confuse the marks, but rather whether the marks will confuse people into believing that the goods they identify emanate from the same source’). As recognised in Case C-533/06 O2 Holdings Ltd v. Hutchison 3G Ltd EU:C:2008:339, [66]. See also Specsavers International Healthcare Ltd v. Asda Stores Ltd [2012] EWCA Civ 24, [78]– [88] (Kitchin LJ). Empirical research already suggests that, for (human) tribunals, the similarity of marks and of products may be dispositive. This leads to a ‘stampeding’ of the remaining factors to align them with the more important ones, thereby shaping the outcome of the confusion test. See B. Beebe, An Empirical Study of the Multifactor Test for Trade Mark Infringement (2006) 94 California L. Rev. 1581; I. Simon Fhima and C. Denvir, An Empirical Analysis of the Likelihood of Confusion Factors in European Trade Mark Law (2015) 46 IIC 310. See Art. 33(7) of Regulation (EU) 2017/1001 of the European Parliament and of the Council of 14 June 2017 on the European Union trade mark, OJ L 154 16.6.2017.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

188

Dev S. Gangjee

will make its way across to infringement analysis, as evidence in a real-world dispute. It has been suggested that businesses might find it attractive if an AI solution were to replace judicial determination, or at least provide a preliminary assessment of confusion even in an infringement context.64 A second set of concerns relates to the quality of information contained in trade mark registers. In recent years, attention has focused on the problem of clutter or deadwood – the existence of marks on the register that are partly or wholly unused by their owners.65 It is not uncommon for trade mark applicants to claim broad swathes of goods and services on which they never intend to use the mark.66 Once such marks are registered, they occupy space on the register as prior rights in relation to these goods and services. Emphasising that registered trade marks need to be used in order to be maintained, the US Registrar of Trademarks recently summarised the problem at a Congressional Hearing in 2019: The register itself provides notice to applicants, other trademark owners, and our examining attorneys of the registrant’s claim of ownership in a mark and allows them to search the register to determine the availability of marks for registration in the United States. The register is a valuable tool in making business decisions, and its accuracy is paramount. When businesses are selecting names for new products, they turn to the register to figure out whether their chosen mark is available for their use and registration. But, for the register to be useful, it must accurately reflect marks that are in use in the United States for the goods and services identified in the registrations. If the register is filled with marks that are not in use, or features registrations obtained by improper means, it makes trademark clearance more difficult, time-consuming and expensive. An inaccurate register also leads to expensive opposition and cancellation proceedings, or federal court litigation, to correct inaccurate registrations and to enforce rights. And, in turn, it may cause companies to alter business decisions, often at significant cost.67

64

65

66

67

See the discussion during a WIPO and UKIPO conference in 2019, reported at: https://www .worldipreview.com/news/ai-as-judges-and-patent-destroying-tools-panel-discussion-18237. G. von Graevenitz, R. Ashmead and C. Greenhalgh, Cluttering and Non-Use of Trade Marks in Europe (UK IPO Report 2015/48); USPTO, Post Registration Proof of Use Pilot Final Report (25 Aug 2015); US House Committee on the Judiciary Hearings, Counterfeits and Cluttering: Emerging Threats to the Integrity of the Trademark System and the Impact on American Consumers and Businesses (18 July 2019), at: https://judiciary.house.gov/calendar/eventsingle .aspx?EventID=2271. The Court of Justice of the European Union (CJEU) has held that such an application might be in bad faith where there is either (i) a dishonest intention to undermine third party rights; or (ii) it represents an intention to obtain exclusive rights that extend beyond the legitimate functions of a trade mark. Contrast the more robust approach taken by the Advocate General with the Court’s final decision: Sky plc & Ors v. Skykick & Ors (C-371/18) ECLI:EU:C:2019:864 (AG); ECLI:EU:C:2020:45 (CJEU). Statement of Mary Boney Denison, Commissioner for Trademarks, USPTO, US House Committee on the Judiciary Hearings, note 65 above, at 2.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

Artificial Intelligence and Trade Mark Registers

189

Although these unused categories are vulnerable to a claim for invalidation, this requires costly and time-consuming real-world investigations into the commercial offerings of trade mark proprietors, followed by legal proceedings to revoke the mark. So long as the register unhelpfully generates false positives, AI algorithms will take formally valid registration at face value. Risk profiles are generated based on potentially vulnerable prior rights that won’t survive scrutiny. Conversely, AI algorithms are presently unable to identify conflicts on the basis of trade mark dilution, which is a broader form of infringement protection for marks with a reputation.68 It involves the use of an identical or similar mark in connection with non-competing or dissimilar goods or services, where the use by the junior user will generate a link in the minds of consumers to the reputed mark. No confusion need be established. Instead, as a result of that link, the reputed mark might be blurred or tarnished. Alternatively, the junior user might obtain a free ride or unfair advantage, through a form of image transference, based on the mental link between the two signs.69 The crucial point is that for reputed marks, protection extends unpredictably (at least for now) beyond similar goods and services. Trade mark law also recognises unregistered prior rights which may generate conflicts, such as common law trade marks protected via the tort of passing off, trade names protected under unfair competition or copyright-protected images from which trade mark logos are derived. These exist outside of trade mark registers or indeed any form of registration in some cases.70 Therefore while the machine learning technology of today offers clearance assistance for one very significant layer of conflicts – where a similar mark on similar goods would be assessed under the likelihood of confusion test – its limitations must be borne in mind. Finally, the technology offers the potential to support defensive strategies to narrow the scope of trade mark rights. Visual search now offers users submitting a complex or composite image an in-built editing tool for close cropping a region of interest within the image.71 It is possible to use this to claim that certain elements within an image are widely used by other trade mark owners (such as an image of a cow on dairy products) and are therefore unlikely to be considered the distinctive elements of the mark by consumers.72 They should be discounted from the 68

69

70

71 72

R. Burrell and M. Handler, Dilution and Trademark Registration (2008) 17 Transnational Law and Contemporary Problems 713. These theories of harm may not rest on empirically or doctrinally valid foundations and are contested. See B. Beebe, R. Germano, C. J. Sprigman, J. H. Steckel, Testing for Trademark Dilution in Court and the Lab (2019) 86(3) University of Chicago L. Rev. 611; M. Handler, What Can Harm the Reputation of a Trademark? A Critical Re-evaluation of Dilution by Tarnishment (2016) 106 Trademark Reporter 639. R. Burrell and M. Handler The Intersection between Registered and Unregistered Trade Marks (2007) 35(3) Federal L. Rev. 375. The WIPO Global Brands Database image search has this functionality. EUIPO Guidelines, note 27 above, Part B, Section 4, Chapter 3, [8] (Commonplace Figurative Elements).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

190

Dev S. Gangjee

similarity analysis. These are just some of the implications of the comprehensive, large-scale algorithmic reading of trade mark registers that is available today.

13.5 conclusion The algorithmic reading of registers is engineered for completeness, certainty and convenience. As the technology improves, the similarity of marks and goods analysis is a useful resource to augment human expertise. Trade mark registry officials, analysts and lawyers can focus on what is important – analysing relevant results in greater detail – as opposed to spending time on searching for those results through trial and error. Yet the risk is that the genie may escape the bottle. Algorithms designed to produce a heuristically helpful upstream snapshot of conflict risks, based on two dimensions of similarity, may unintentionally edge out the more complex multifactor test in a wider range of situations, including trade mark infringement analysis. The seductive appeal of the all-seeing algorithm should be resisted, since a range of conflicts beyond likelihood of confusion (similar marks on similar goods) is also possible. The seduction is especially powerful since similarity-matching algorithms can seemingly keep up with the high-pressure hosepipe of new trade mark applications, bringing a degree of consistency, comprehensiveness and ‘objectivity’ to the intrinsically subjective task of assessing similarity. Their limits should be borne in mind.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.020

14 Patent Protection of Inventions Involving Artificial Intelligence Stefan Luginbuehl1

14.1 introduction The Fourth Industrial Revolution has created a variety of new technologies. It seems to be undisputed that they will not only be an important economic driver but also have a great impact in basically every field of technology, on the economy and society at large, and on the relevant underlying laws.2 The development and control of these technologies is thus directly linked to future technological and military supremacy, and is the main reason for the trade dispute between the US and China. The Internet of Things (“IoT”) is one such technology. It is a network formed of billions of connected objects and sensors that gather data at home, the workplace and other places and communicate with each other. In other words, it is a network that is collecting and exchanging data. The collection of this “big data”, the storage thereof3, its extraction and finally its use for new devices and systems (in particular via Artificial Intelligence (AI)), are major drivers for this kind of innovation. Common standards will ensure the interconnectedness of these technologies.4 AI is already used to help create products in the transportation and telecommunications, as well as life sciences (for example pharmaceuticals, medical devices and biotech) sectors, and it is expected that in the near future it will also be used more 1

2

3

4

Ph.D., lawyer, Head of Department European Legal Affairs, European Patent Office, Munich, Germany. I would like to thank Dr Lukas Bühler, LLM and Dr Dieter Stauder, formerly Prof. associé Université Robert Schuman, Strasbourg, for their highly valuable input on the issue. I am expressing my personal views. See Georg Borges, Rechtliche Rahmenbedingungen für autonome Systeme (2018) 71 NJW 977; Jan Phillip Rektorschek, Industrie 4.0 und künstliche Intelligenz – Risiko oder Chance für den gewerblichen Rechtsschutz? (2017) 108 Mitt. 438, at 438; see also Sidharth Chauhan, Artificial Intelligence – A Competition Law Perspective (2019) 40(3) ECLR 139. With regard to a possible trade secret protection of data generated and managed within the framewok of Big Data, see Cristina Sappa, How Data Protection Fits with the Algorithmic Society via Two Intellectual Property Rights – A Comparative Analysis (2019) GRUR Int. 135. Cf. Rektorschek, supra note 2, 441.

Downloaded from https://www.cambridge.org/core. 191 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

192

Stefan Luginbuehl

extensively in the finance, energy, nutrition, chemistry and manufacturing sectors.5 However, it should also be noted that a great part of AI-related research goes into digital marketing, surveillance and robot weapons,6 which raises, not least, human rights and ethical concerns. AI is systems based on algorithms that perform certain predetermined tasks in a structured way and are designed to perform cognitive tasks on the basis of deep learning and neural networks7 – just like a human trying to solve a problem AI will try to maximize the chances of reaching a particular result by steadily improving its performance on the basis of learning from experience or external data.8 This definition makes it clear that AI as such is not a new phenomenon.9 In the past few decades, such inventions were usually qualified as ‘computer implemented inventions’ in Europe, for which there is an established practice to grant patent protection.10 Nevertheless, it comes as no surprise that industries, in order to preserve their R&D investments, are looking for appropriate intellectual property protection for this new technology, in particular via copyrights11 and patents. On the one hand, users are seeking patent protection for AI itself or for AI-based inventions. Studies have revealed that, before the European Patent Office (EPO) alone, patent applications relating to smart connected objects – including AI – grew by 54 per cent between 2014 and 2016.12 Nevertheless, companies are often reluctant to disclose the role of AI when filing a patent application.13

5 6

7

8

9

10 11

12

13

Machine Yearning: AI and Patents, “Patent Strategy,” Managing IP, February 2019, 6f. Cf. quote of Boi Faltings, in: WIPO Technology Trends 2019, Artificial Intelligence, “Key Issues Arising from AI and Policy Responses,” 124. Cf. Yann LeCun, Yoshua Bengio and Geoffrey Hinton, Deep Learning, Nature, vol. 521, 28 May 2015. Cf. Florian De Rouck, Moral Rights and AI Environments: The Unique Bond between Intelligent Agents and their Creations (2019) 14 JIPLP 299; Ronny Hauck and Baltasar Cevc, Patentschutz für Systeme Künstlicher Intelligenz? (2019) 11(2) ZGE/IPJ 135, at 141. See WIPO Technology Trends 2019, Artificial Intelligence, “A Short History of AI,” supra note 6, 19. See in particular G 3/08 (Programs for computers), OJ EPO 2011, 10. Cf. with regard to copyright protection of AI, for example, Anne Lauber-Rönsberg, Autonome “Schöpfung” – Urheberschaft und Schutzfähigkeit (2019) 3 GRUR 244; Tim W. Dornis, Der Schutz künstlicher Kreativität im Immaterialgüterrecht (2019) 12 GRUR 1252; De Rouck, supra note 8; Jani Ihalainen, Computer Creativity: Artificial Intelligence and Copyright (2018) 13 JIPLP 724; Pratap Devarapalli, Machine Learning to Machine Owning: Redefining the Copyright Ownership from the Perspective of Australian, US, UK and EU Law (2018) 40 EIPR 722; Frank Hartmann and Matthias Prinz, Immaterialgüterrechtlicher Schutz von Systemen Künstlicher Intelligenz (2018) WRP 1431. Patents and the Fourth Industrial Revolution, EPO study, 2017. See also WIPO Technology Trends 2019, Artificial Intelligence, “Evolution of AI Patent Applications and Scientific Publications,” supra note 6, 38ff. Machine Yearning, supra note 5, 7.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

Patent Protection of Inventions Involving AI

193

On the other hand, although it seems that AI systems have not yet developed sufficiently in order to create their own patentable inventions,14 it must reasonably be expected that in the future AI will autonomously establish inventions for which patent protection will be sought. Once this is possible, such AI systems will challenge the paradigms of the existing patent system and its legal basis, in particular as regards patentable subject matter, the inventive step and disclosure requirements, and with regard to inventorship.

14.2 patentability and patentability requirements of ai and ai-based inventions 14.2.1 Patentable Subject Matter The EPO grants European patents for any inventions, in all fields of technology.15 ‘Invention’ implies a requirement of technical character fulfilled by an invention as claimed in order to be patentable.16 As regards the patentability of AI and AI-based inventions, a distinction should be made between AI systems and their products. The Protection of AI Systems To a great extent, AI systems are mathematical methods which, under Article 52(2) (a) of the European Patent Convention (EPC)17 and most national patent laws, are not regarded as patentable. According to Article 52(3) EPC, patentability of mathematical methods and computer programs as provided for in Article 52(2) EPC is excluded to the extent that a European patent application relates to such subject matter or activities as such. However, if a claimed invention is directed either to a method involving the use of technical means (e.g. a computer) or to a device, and if its subject matter has a technical character as a whole, then the invention is not excluded from patentability and can be patentable if the general patentability requirements, in particular novelty and inventive step, are met.18 The patent protection of AI systems follows these general lines of protection of mathematical methods. However, it seems open still with regard to patent protection of AI systems whether an invention includes a technical character if it solves the 14

15 16 17

18

Yann Ménière and Heli Pihlajamaa, Künstliche Intelligenz in der Praxis des EPA (2019) GRUR 332, at 333; Machine Yearning, supra note 5, 6. Art. 52(1) European Patent Convention (EPC); Art. 27.1 TRIPS Agreement. T 931/95, T 1173/97, OJ EPO 1999, 609 and T 935/97. Convention on the Grant of European Patents of 5 October 1973. Currently the Convention has 38 contracting states, including the 27 EU member states and 11 non-EU member states, including Norway, Switzerland, Turkey and the United Kingdom. “Brexit” has no impact on the status of the United Kingdom in the European Patent Organisation. Guidelines for Examination in the European Patent Office, November 2019, Part G – Chapter II-3.3, available at www.epo.org/law-practice/legal-texts/guidelines.html.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

194

Stefan Luginbuehl

technical problem of making a computer program adaptive so that it autonomously develops a technical solution. This should normally be the case. Furthermore, it is noteworthy that there are calls from industry to make the algorithms which play a crucial role in AI patentable as such, not least because more and more start-up companies19 and SMEs are making AI-based inventions. In Europe, this would require an amendment of the EPC and of most national patent laws. In 2000, at the Diplomatic Conference for the revision of the EPC, removal of the exclusion of computer programs from patentability was discussed, but due to certain fears of possible excessive upstream patenting there was no agreement on the issue and it was decided not to amend the EPC.20 Protection of Products Established by AI Systems Products established by AI systems could, for example, include new therapeutic or diagnostic methods developed on the basis of an evaluation of scientific literature, or new mechanical tools modelled on the basis of an evaluation and combination of existing tools in other technical areas. Such products would be patentable before the EPO within the legal boundaries of the EPC.

14.2.2 Inventive Step Inventive step is the central point of any examination of whether a patent can be granted for an invention. An invention is considered to involve an inventive step if it provides a technical solution to a technical problem and, having regard to the state of the art, it is not obvious to a person skilled in the art.21 As regards computer implemented inventions, it is regularly the case that technical and non-technical features appear in a claim. When assessing the inventive step of such mixed-type inventions, all those features which contribute to the technical character of the invention are taken into account. This also includes the features which, when taken in isolation, are non-technical, but do, in the context of the invention, contribute to producing a technical effect serving a technical purpose, thereby contributing to the technical character of the invention.22 This approach also applies to AI-based mixed-type inventions. The EPO recommends including within the patent specification as much information about the technical effect as possible. This requires a clear identification of the technical features and 19

20 21 22

In 2016, start-up companies invested 6–9 billion US dollars in AI; see Kay Firth-Butterfield and Yoon Chae, Artificial Intelligence Collides with Patent Law, White Paper, World Economic Forum, April 2018, 5. Cf. Conference Resolution, Special Edition No. 1, OJ EPO 2001, 50. Article 56 EPC. Guidelines for Examination in the EPO, November 2019, Part G – Chapter VII-5.4, supra note 18; Stefan Steinbrener in: Singer/Stauder/Luginbühl (eds.), Europa¨isches Patentu¨bereinkommen (Wolters Kluwer, 8th ed., 2019), Art. 52 N 20.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

Patent Protection of Inventions Involving AI

195

the corresponding technical effect of the invention.23 At the conference organised by the EPO in 2018 on patenting AI, it was therefore suggested that in fields such as autonomous vehicles and healthcare, AI might be claimed as a tool embedded in a larger claim and defined as a functional feature providing a partial solution. In such cases, the claim as a whole should be assessed for inventive step, and not just the functionally defined AI tool.24 In order to show the technical effect of an AI-based invention, the application could also include objectifiable parameters and a corresponding meaningful test.25 AI systems should regularly involve an inventive step, as the common computer implemented inventions will not provide the required technical solutions to the technical problems at stake. AI systems are thus comparable to molecular biology, where the development of biological tools – such as the PCR method in the early days of this technology – did not give rise to doubts about being inventive. However, over time the determination of inventive step will be reduced to advantages vis-à-vis public AI systems with the increasing state of the art of AI. In addition, it can be expected that in certain fields of technology nothing will be invented anymore without the use of (possibly standardized) AI. This will have an impact on the determination of the person skilled in the art, and of the obviousness of an invention for such a person. Already there are algorithms that gather information from different industries and determine if there is a technical solution in one field that could be applied to a technical problem in another area.26 Currently, the EPO qualifies the person skilled in the art as follows: The person skilled in the art should be presumed to be a skilled practitioner in the relevant field of technology, who is possessed of average knowledge and ability and is aware of what was common general knowledge in the art at the relevant date. He should also be presumed to have had access to everything in the “state of the art”, in particular the documents cited in the search report, and to have had at his disposal the means and capacity for routine work and experimentation which are normal for the field of technology in question. If the problem prompts the person skilled in the art to seek its solution in another technical field, the specialist in that field is the person qualified to solve the problem. The skilled person is involved in constant development in his technical field. He may be expected to look for suggestions in neighbouring and general technical fields . . . or even in remote technical fields, if prompted to do so . . . Assessment of whether the solution involves an inventive step must therefore be based on that specialist’s knowledge and ability . . . There may be instances where it is more appropriate to think in terms of a group of persons, e.g. a 23

24 25

26

Cf. Conference summary of the EPO conference of 30 May 2018 on patenting artificial intelligence, 4, available at www.epo.org. Conference summary of the EPO conference of 30 May 2018, supra note 23, 6. See Joel Nägerl, Frank Steinbach and Benedikt Neuburger, Artificial Intelligence: A Game Changer for the Patent System, IAM, 25 September 2018, https://www.iam-media.com/artificialintelligence-game-changer-patent-system. Machine Yearning, supra note 5, 6.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

196

Stefan Luginbuehl

research or production team, rather than a single person . . . It should be borne in mind that the skilled person has the same level of skill for assessing inventive step and sufficient disclosure . . .27

Against this general definition, a person skilled in the art should be qualified as being aware of possible AI which is seeking solutions in other technical fields beyond the current expectations28 that will be used to develop subject-matter for which patent protection is sought (“a skilled human using a machine”29). This will have to be proven at the relevant moment. After the person skilled in the art has been determined, it must be decided whether the claimed invention was obvious to that person at the date of filing of the application. The term obvious means that which does not go beyond the normal progress of technology but merely follows plainly or logically from the prior art, namely something which does not involve the exercise of any skill or ability beyond that to be expected of the person skilled in the art.30 Obviousness is thus determined objectively and does not take into account how the invention was established (i.e. with or without the use of AI).31 Nevertheless, the question arises as to whether the general use of AI to create inventions would not lead to a “raising of the bar”, namely whether in such cases a fair and balanced approach to determining obviousness would not require that the person skilled in the art was qualified with a higher degree of specialist knowledge by taking into account the skill and ability of the used AI.32 As a consequence, in certain areas of technology fewer patents would be granted for inventions. Once AI is able to create inventions autonomously, the criteria of obviousness will have to be reconsidered33 as most such inventions will go beyond the normal technological progress to be expected of a specialist human being. A possible approach could be to qualify the standard AI used in certain fields of technology as a person skilled in the art,34 and use it as a basis for determining whether something is obvious. This however could not reasonably be decided by a human 27 28 29 30 31

32

33

34

Guidelines for Examination in the EPO, November 2019, Part G – Chapter VII-3, supra note 18. See also Nägerl, Steinbach and Neuburger, supra note 25. Conference summary of the EPO conference of 30 May 2018, supra note 23, 8. Guidelines for Examination in the EPO, November 2019, Part G – Chapter VII-4, supra note 18. Sven Hetmank and Anne Lauber-Rönsberg, Künstliche Intelligenz – Herausforderungen für das Immaterialgüterrecht (2018) GRUR 574, at 576. See also Hetmank and Lauber-Rönsberg, supra note 31, 580f; Joel Nägerl, Bendikt Neuburger and Frank Steinbach, Künstliche Intelligenz: Paradigmenwechsel im Patentsystem (2019) GRUR 336, at 338; Kemal Bengi and Christopher Heath, Patents and Artificial Intelligence Inventions in: Heath/Kamperman Sanders/Moerland (eds.), Intellectual Property Law and the Fourth Industrial Revolution (Wolters Kluwer, 2020), 142. See also Firth-Butterfield and Chae, supra note 19, 10; Roman Konertz and Raoul Schönhof, Erfindungen durch Computer und künstliche Intelligenz – eine aktuelle Herausforderung für das Patentrecht? (2018) 10(4) ZGE/IPJ 379, at 403. See also Ryan Abbott, I Think, Therefore I Invent: Creative Computers and the Future of Patent Law (2016) 57 BCL Rev. 1079, at 1124.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

Patent Protection of Inventions Involving AI

197

being but again only by AI, thus raising the question of whether the inventive step requirement would not have to be abolished after all. 14.2.3 Disclosure It is a basic substantive requirement for a patent35 that the application discloses the invention in a manner sufficiently clear and complete for it to be carried out by a person skilled in the art.36 This principle of protection in exchange for disclosure ensures that the technical knowledge embodied in the invention becomes publicly available and can in general be used by anyone after the expiration of the patent right. In order to fulfil the disclosure requirement before the EPO it is necessary that the invention is described not only in terms of its structure but also in terms of its function, unless the functions of the various parts are immediately apparent.37 In the case of AI-based inventions, the description of the function is more important than the description of the structure. However, AI often performs its processes in a “black box” which humans cannot understand to the fullest extent due to the different approach that machines take towards solving a problem. This makes it very difficult for an applicant to fulfil the disclosure requirement if he or she is not able to explain how the invention works so that the technical solution can be reworked on the basis of the specifications in the patent. Thus the applicant will often be forced to define the function from a different perspective, such as from a network structure or a functional perspective, and by precisely describing the inputs and outputs of the AIbased invention.38 This will also ensure that the protection conferred by the patent extends to the products directly obtained by the patented process.39 The difficulties surrounding the disclosure of such inventions will furthermore be reflected at the enforcement level – it could be very difficult to prove a possible infringement if, on the basis of the applicable law, a patentee is required to prove that a potentially infringing product was established on the basis of the protected process.

14.3 inventorship of ai or ai-based inventions It is a generally recognized principle that inventorship is limited to human beings,40 as they are considered to be the only possible creators that can come up with an 35

36 37 38

39 40

Article 29.1. TRIPS Agreement; Rudolf Teschemacher in: Singer/Stauder/Luginbühl (eds.), Europa¨isches Patentu¨bereinkommen, supra note 22, Art. 83 N 2. Article 83 EPC. Guidelines for Examination in the EPO, November 2019, Part F – Chapter III-1, supra note 18. Cf. Nägerl, Steinbach and Neuburger, supra note 25; Bengi and Heath, Supra note 32, 137. Ménière and Pihlajamaa, supra note 14, 335. Article 64(2) EPC. On 25 November 2019, the EPO refused applications EP 18 275 163 and EP 18 275 174 in which an AI system (“DABUS”) was designated as the inventor. It held that the applications do not

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

198

Stefan Luginbuehl

invention.41 The inventor (or his/her successor) is the one that has the right to a patent,42 and not least the European patent application has to designate the inventor.43 In addition, many states provide that employees have a right for an appropriate reward by their employer in the event that they devise an invention during their employment. Industries have thus requested a possible change to the paradigm and have called for a more flexible approach with regard to inventions made by AI. Less problematic, as regards the current qualification that an inventor must be a natural person, are calls for having the user, namely the programmer, developer or implementer,44 qualified as the inventor of inventions created by AI. This would also provoke solutions for employee inventions and appropriate awards, in particular if AI developed by an employed developer becomes very successful. More challenging are proposals to allow legal persons, inventors and/or owners of the AI to be qualified as inventors of inventions created by their AI.45 Proposals to qualify AI as an electronic person46 or to simply qualify AI as a possible inventor47 go in a similar direction. The existing laws do not cover these proposals and would thus have to be amended. Until then it is to be expected that applicants will continue to indicate on the patent application the name of an employee as inventor if AI created

41

42 43 44

45 46 47

meet the legal requirements of the EPC that an inventor designated in the application has to be a human being, and not a machine. The decisions were appealed and are currently pending before the EPO Boards of Appeal (See also Martin Stierle, Artificial Intelligence Designated as Inventor. An Analysis of the Recent EPO Case Law (2020) GRUR lnt. 918). The USPTO and the UKIPO refused applications indicating DABUS as inventor on the basis of similar grounds (see decision of the USPTO at https://www.uspto.gov/sites/default/files/documents/16524350_ 22apr2020.pdf and decision of the UKIPO at https://www.ipo.gov.uk/p-challenge-decisionresults/o74119.pdf). The appeal against the decision of the UKIPO was dismissed by the UK High Court (Patents), [2020] EWHC 2412 (Pat). See for example Tobias Bremi and Dieter Stauder in: Singer/Stauder/Luginbühl (eds.), Europa¨isches Patentu¨bereinkommen, supra note 22, Art. 60 N 4; Rudolf Krasser and Christoph Ann, Patentrecht (C. H. Beck, 7th ed., 2016), §1 N 25; Thomas Heinz Meitinger, Erfinderlose Erfindungen durch Know-How einer Organisation und Erfinderprinzip: kein Widerspruch (2017) 108 Mitt. 149, at 149; decision of the Swiss Federal Patent Court (Bundespatentgericht) of 6 December 2013, O2012_001, N 27; Yeda Research and Development Co Ltd v. Rhone-Poulenc Rorer International Holdings Inc. and others [2007] UKHL 43, N 20; decision of the German Federal Supreme Court (Bundesgerichtshof, BGH) of 10 November 1970, X ZR 54/67; Abbott, supra note 34, 1092; Rektorschek, supra note 2, 442. Art. 60(1) EPC. Art. 81 EPC. Cf. Konertz and Schönhof, supra note 33, 401; Hetmank and Lauber-Rönsberg, supra note 31, 581; Nägerl, Steinbach and Neuburger, supra note 25; Conference summary of the EPO conference of 30 May 2018, supra note 23, 8. See also decision of the UK High Court (Patents), Supra note 40, N52(2). Nägerl, Steinbach and Neuburger, supra note 25. Cf. Rektorschek, supra note 2, 442. Cf. Abbott, supra note 34, 1114; Firth-Butterfield and Chae, supra note 19, 10; cf. also Ménière and Pihlajamaa, supra note 14, 336.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

Patent Protection of Inventions Involving AI

199

a part or all of the invention for which they seek patent protection,48 in order to avoid rejection of their application.49

14.4 conclusion For the time being, AI systems will be particularly useful for developing solutions to technical problems on the basis of solutions to problems in other technical fields which will go beyond the general expectation of a skilled person in the art. They will provide results that no human would think of or would ever have expected. Nevertheless, it will be humans who will have to discover that there is a concrete result which can be used to solve a technical problem and which can then be commercialized. Inventors will thus make use of these solutions evolved by AI systems and develop them further through experiments and with their own mind and logic. Inventors and possibly even AI itself will combine solutions provided by several different kinds of AI. It will consequently become an indispensable tool for improving inventions. The existing general principles as regards the patentability of mathematical methods and computer implemented inventions, and the applicable patentability criteria of inventive step and disclosure, seem to be sufficient to provide for a balanced solution for the protection of AI-based inventions in Europe. This will no longer be the case once AI systems are able to autonomously create inventions. In order to take account of this new reality it will be necessary to either adjust the existing patent law related criteria, or establish a new sui generis intellectual property right for inventions and creations made by AI. However, from a political perspective it would be very difficult to realise such a new sui generis right and – as with molecular biology when it emerged as a new technological field for which patent protection was sought – it can thus be expected that patent practice and jurisprudence will find a way to deal with the challenges posed by AI.

48 49

See also Abbott, supra note 34, 1088, 1097. Articles 90(3), (5) and 81 in conjunction with Rules 19(1) and 60(1) EPC for the European patent procedure.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.021

15 Automated Profiling in New Media and Entertainment Markets: What to Protect, and How? Megan Richardson1

15.1 the threat How should a legal system treat advances of new technologies and practices that may end up reshaping human identities and social interactions? In this chapter, I consider practices of firms drawing on automated decision-making technologies to construct profiles of individuals based on their personal data and using these to develop new products and services, frame individual preferences, and foster consumption in new media and entertainment markets. Still at an early stage, the practices are becoming a basic business model of the digital economy despite all the argued social costs of removing distinctions between private and public life, appropriating valuable data resources away from individuals who generate the resources, and treating human identity as subject to effective control of the machine and its masters. Should the weight of an intellectual property system geared to the protection of intellectual assets, in this context valuable data, be extended to such activities, subject to specified limits and exceptions? Or should the practices be left to operate outside the rubric of legal regulation, at least for a while, adopting a ‘sand-box’ approach, effectively leaving it to firms to regulate themselves in exploiting the available data resources and in the process accumulating intellectual assets? Or is there a way in between, offering scope for effective legal regulation of social costs associated with misuses of personal data as and when these costs emerge, but otherwise allowing practices to operate without legal constraint?

1

Professor, Melbourne Law School, The University of Melbourne. With grateful thanks to Damian Collier and Kobi Leins for helpful comments and advice.

Downloaded from https://www.cambridge.org/core. 200 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

Automated Profiling in New Media and Entertainment

201

15.2 potential solutions Well before these practices became fully operational, Reto Hilty and Annette Kur suggested that unfair competition doctrines in civil law jurisdictions offered a way in between, allowing judges to address social costs associated with new technologies and innovative practices as and when they arise, with the possibility that they might become stepping stones to full-scale intellectual property rights down the line.2 Primarily, they were concerned with economic costs. But in a passing comment Kur drew a parallel with portrait rights which were designed to address problems of unauthorised circulation of portraits that arose with the rise of photography. A specific provision in the German copyright statute, dating back to 1907,3 was over the course of the twentieth century expanded to offer broad protection to a right of personality in the face of a wider set of incursions, now ingrained in the German postwar Constitution (or ‘Basic Law’).4 That personal and economic costs may nevertheless be intertwined is shown by Kur’s example of the Marlene Dietrich case.5 There, the daughter and heir of the famous German actress relied on the rights to dignity and freedom in the Basic Law to successfully maintain a claim for damages for unauthorised commercial exploitation of her name and image in an advertisement, claiming infringement of her right of personality in an explicitly commercial sense. As Kur put it, ‘[b]y acknowledging that the financial interest vested in the publication, which appeared after Marlene Dietrich’s death, arose independently of the actress’s personal interest to be free from interference concerning her dignity and individual freedom, the Federal Supreme Court finally gave its blessing to the concept of a full-fledged exclusive personality right comprising moral as well as economic aspects’.6 There is much that common law jurisdictions can learn from the civil law experience. The United Kingdom offered its own minimalist version of a portrait right in the Fine Arts Copyright Act 1862,7 allowing limited derogation from the principle that copyright belonged to the author of a portrait for an individual who 2

3

4 5

6 7

Reto Hilty, The Law Against Unfair Competition and its Interfaces, in R. Hilty and F. Henning-Bodewig, eds., Law Against Unfair Competition – Towards a New Paradigm in Europe (Springer, 2007), 1, 27ff; Annette Kur, What to Protect, and How? Unfair Competition, Intellectual Property, or Protection Sui Generis, in N. Lee, G. Westkamp, A. Kur and A. Ohly, eds., Intellectual Property, Unfair Competition and Publicity (Edward Elgar, 2014), 11, 27–31. See s 22ff Kunsturhebergesetz (KUG) and generally Ansgar Ohly, German Law, in Huw Beverley-Smith, Ansgar Ohly and Agnes Lucas-Schloetter, Privacy, Property and Personality Civil Law Perspectives on Commercial Appropriation (Cambridge University Press, 2005), ch 4. Kur, above n 2, 22–23. Bundesgerichtshof (BGH), (2000) GRUR 709 – Marlene Dietrich; BGH, (2000) GRUR 715 – Der Blaue Engel; BVerfG 1971 NJW 1645. Kur, above n 2, 23. Fine Arts Copyright Act 1862, c 68, s 1.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

202

Megan Richardson

commissioned the portrait. But this is barely renamed a ‘right to privacy of certain photographs and films’ in the latest iteration.8 And the UK still does not have a fullscale right of personality although a more limited right to privacy/private life is provided for by the Human Rights Act 1998, giving effect to the European Convention on Human Rights, Art 8.9 On the other hand, as Kur also points out, British judges have granted protection under doctrines such as breach of confidence and passing off against acts which would, ‘according to the continental European perception, amount to unfair competition’ and/or ‘to infringement of particular exclusive rights, like the personality right’.10 So, for example, in Douglas v. Hello!,11 celebrity actors Michael Douglas and Catherine Zeta Jones successfully claimed damages for unauthorised publication of surreptitiously obtained wedding photographs in Hello! magazine on the basis of breach of confidence, notwithstanding they had ‘sold’ most of their rights to another magazine (in the tendering process rejecting Hello!) – a claim that was presented as one of breach of privacy under Art 8 of the European Convention, but had at least a partial economic dimension, offering protection to their professional as well as personal reputations. Likewise, in Fenty v. Arcadia Group Brands Ltd, the tort of passing off was successfully relied on to object to the unauthorised use of celebrity musician and fashionista Rihanna Fenty’s portrait on a t-shirt, on the basis that there was a ‘misrepresentation, implied in all the circumstances, that she had endorsed the t-shirt’.12 In some states in the United States, legislatures and/or courts have moved even further in the combined operation of various privacy torts alongside the right of publicity to protect the more obvious indicia of personality, such as name and/or likeness. A notable example is California, home of many of the digital economy companies which have been at the forefront of the automated profiling business models. In fact, California has a long history of framing its laws creatively to fashion privacy and publicity rights associated, at least in part, with its status as the home of Hollywood. For instance, in the 1931 case of Melvin v. Reid,13 the plaintiff, Gabrielle Darley Melvin, succeeded in obtaining damages based on a privacy tort framed in terms of a ‘right to live one's life in seclusion, without being subjected to unwarranted and undesired publicity’ following commercial exploitation of her life story identifying her by name in the defendant’s biopic film The Red Kimono.14 The Californian court cited Samuel Warren and Louis Brandeis’ article on ‘The Right to Privacy’ in the 1890 Harvard Law Review,15 where these authors argued for a 8 9 10 11 12 13 14 15

Copyright, Designs and Patents Act 1988 (UK), s 85. Human Rights Act 1998, appending the European Convention on Human Rights, Rome, 1950. Kur, above n 2, 11, n 1. Douglas v. Hello! Ltd [2006] QB 125. Fenty v. Arcadia Group Brands Ltd (trading as Topshop) [2015] 1 WLR 3291, Kitchin LJ at 3303. Melvin v. Reid 112 Cal App 285 (1931). Ibid, 289. Samuel D. Warren and Louis D. Brandeis, The Right to Privacy (1890) 4 Harvard Law Review 193.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

Automated Profiling in New Media and Entertainment

203

common law right to privacy extending to ‘invasion either by the too enterprising press, the photographer, or the possessor of any other modern device for recording or reproducing scenes or sounds’,16 as well as the right to happiness in the California Constitution.17 The plaintiff’s alternative claims based on property were rejected by the court on the basis that ‘no authorities sustained such a property right in the story of one’s life’.18 But the substantial damages sought and received for breach of the privacy tort might suggest that the interests concerned were at least partly economic.19 The case of Melvin v. Reid predated the US Supreme Court’s reliance on the First Amendment right of freedom of speech and the press to establish a significant constitutional constraint on the ability to claim invasion of privacy against media and other defendants engaged in ‘newsworthy’ publications in cases from the 1960s.20 But if the publisher’s interests are judged as merely commercial there is (currently) less prospect of asserting broad constitutional immunity – and the range of laws that might be drawn on in California in support of the plaintiff’s rights in her name, likeness and other indicia of personality against commercial exploitation has expanded considerably since that case to include statutory and common law rights of publicity (albeit with limitations for news uses).21 So in Fraley v. Facebook,22 in 2011, Facebook’s use of the plaintiffs’ names and likenesses in exploiting their ‘likes’ for Facebook’s commercial advertising purposes was held sufficiently likely to amount to a breach of their right of publicity, and violation of California’s Unfair Competition Law,23 for the claims not to be struck out. And Facebook’s argument that its use was news rather than commercial was rejected in light of statements of senior executives about the commercial value of Facebook friends’ endorsements. Further, in the 2012 case of Monge v. Maya Magazines,24 copyright protection was successfully claimed in wedding portraits published without the subjects’ authority in the defendant magazine. Rejecting the magazine’s argument of privileged

16 17

18 19 20

21

22 23

24

Ibid, 206. See Jessica Lake, The Face that Launched a Thousand Lawsuits (Yale University Press, 2016), ch 6. Melvin v. Reid, above n 13, 292. Lake, above n 17, 205. See Melville B. Nimmer, The Right to Speak from Times to Time: First Amendment Theory Applied to Libel and Misapplied to Privacy (1968) 56 California Law Review 935; Daniel J. Solove and Paul M. Schwartz, Information Privacy Law (Kluwer, 6th ed., 2011), 146. See California Civil Code § 3344; Midler v. Ford Motor Co, 849 F2d 460 (9th Cir, 1988); see generally Graeme Dinwoodie and Megan Richardson, Publicity Right, Personality Right, or Just Confusion?, in Megan Richardson and Sam Ricketson, eds., Research Handbook on Intellectual Property in Media and Entertainment Law (Edward Elgar, 2017), ch 16; David Tan, the Commercial Exploitation of Fame (Cambridge University Press, 2017). Fraley v. Facebook, Inc, 830 F Supp 2d 785 (ND Cal, 2011). Specifically, that without adequate notice, the practice was an ‘unlawful, unfair or fraudulent business act or practice’ under the California Business and Professional Code § 17200 et seq. Monge v. Maya Magazines, Inc, 688 F 3d 1164 (9th Cir, 2012).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

204

Megan Richardson

newsworthy publication, the court held that, without a transformative element to the publication this did not fall within the fair use defence to infringement.25 Taken together, these cases suggest some scope for plaintiffs to claim control over key indicia of their personalities (name, likeness, and so on), with respect to a range of media and entertainment uses.

15.3 developing approaches Recent developments around what is quite commonly called ‘data privacy’ law have contributed substantially to this conclusion. For reasons of space, I offer only a handful of examples. First, an apparently minor but nevertheless intriguing example is the California case of In re Vizio, Inc, Consumer Privacy Litigation.26 Plaintiffs objected to the smart television company’s tracking of the viewing habits of some 11 million customers, using a built-in spy device employing image recognition technology, and selling this information to advertisers and media content providers for targeted advertising. A California court held that plaintiffs ‘plausibly alleged that Vizio's provision of – to quote its own prospectus – “highly specific viewing behavior data on a massive scale with great accuracy” amounts to the disclosure of personally identifiable information’ in breach of the Video Privacy Protection Act 1988,27 and further that ‘Vizio’s collection practices amount to a highly offensive intrusion [on seclusion]’, a privacy tort under California law.28 These were not the only claims against the smart television company. In separate proceedings, the Federal Trade Commission alleged unfair or deceptive practices under § 5 of the Federal Trade Commission Act, with the settlement terms including $2.2 million, a mandated privacy program and measures introduced to obtain consent from customers before collecting and sharing their data.29 The claims in the California case have now also been settled, with Vizio agreeing to pay $17 million as well as to provide more explicit notification of its data collection and use practices, including prominent displays on TV screens plus additional language in the device’s quick-start guide, coupled with deletion of all viewing data collected before February 2017.30 This is 25

26 27 28

29

30

See, critically, Pamela Samuelson, Protecting Privacy through Copyright Law?, in Marc Rotenberg, Jeramie Scott and Julia Horwitz, eds., Privacy in the Modern Age: The Search for Solutions (New Press, 2015), ch 22. In re Vizio, Inc, 238 F Supp 3d 1204 (CD Cal, 2017). Ibid, 1225. Ibid, 1233. Other claims including fraud, negligent misrepresentation and false advertising were dismissed with leave to amend. See Federal Trade Commission v. Vizio, Inc, Stipulated Order for Permanent Injunction and Monetary Judgment, 6 February 2017, available at https://www.ftc.gov/system/files/documents/ cases/170206_vizio_stipulated_proposed_order.pdf. See Sara Merken, Vizio Settles Claims It Collected, Sold Customer TV Viewing Habits, Bloomberg Law: Privacy & Data Security Law News, 9 October 2018, available at

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

Automated Profiling in New Media and Entertainment

205

only one of a number of class actions and investigations involving digital platforms’ handling of personal information – including class actions directed at Facebook in the wake of the Cambridge Analytica scandal citing privacy and other claims, at least one of which (In re Facebook, Inc, Consumer Privacy User Profile Litigation) has been allowed to proceed to trial despite Facebook’s efforts to have the claims struck out,31 as well as a $5 billion fine issued by the Federal Trade Commission in July 2019 following its investigation.32 More substantial and potentially far-reaching steps have been taken by way of legislation. For instance, in June 2018 the California Consumer Privacy Act 2018 was signed into law,33 providing, inter alia, that consumers have the right to request that ‘a business that collects a consumer’s personal information disclose to that consumer the categories and specific pieces of personal information the business has collected’, that ‘a business delete any personal information about the consumer which the business has collected from the consumer’,34 that their information be provided to them in a ‘portable format’, and that companies allow them to opt out of practices involving selling their data, including providing ‘a clear and conspicuous link on the business’s Internet homepage, titled “Do Not Sell My Personal Information,” that enables a consumer, or a person authorized by the consumer, to opt-out of the sale’.35 In addition, consumers have a right to bring proceedings in court in aid of the above rights.36 While targeted specifically at the Cambridge Analytica/Facebook scandal earlier in the year, the Act seems adapted also to address more minor scandals such as the Vizio breach. This may be only a first step towards more comprehensive ‘consumer privacy’ regulation across the United States. Peter M. Lefkowitz, chief privacy and digital risk officer at Citrix Systems and 2018 chairman of the International Association of Privacy Professionals, writing an opinion piece in the New York Times in June 2019, notes that ‘[a]t least 15 bills have been filed this year’ following the CCPA, adding that ‘[t]ime is running out for Congress to pass meaningful privacy legislation before the implementation of the California privacy law in 2020, the adoption of inconsistent laws in other states, and the distraction of the 2020 elections’.37

31 32

33

34 35 36 37

https://news.bloomberglaw.com/privacy-and-data-security/vizio-settles-claims-it-collected-soldcustomer-tv-viewing-habits. In re Facebook, Inc, 402 F Supp 3d 767 (ND Cal, 2019). See FTC Imposes $5 Billion Penalty and Sweeping New Privacy Restrictions on Facebook, Federal Trade Commission, 24 July 2019, available at https://www.ftc.gov/news-events/pressreleases/2019/07/ftc-imposes-5-billion-penalty-sweeping-new-privacy-restrictions. TITLE 1.81.5. California Consumer Privacy Act of 2018 [Cal Civ Code §§ 1798.100–1798.199] (CCPA). CCPA, § 1798.105 (a). CCPA, § 1798.135 (a). CCPA, § 1798.150 (a). Peter M. Lefkowitz, Why America Needs a Thoughtful Federal Privacy Law, New York Times, 25 June 2019, available at https://www.nytimes.com/2019/06/25/opinion/congress-privacy-law .html.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

206

Megan Richardson

A third important development is, of course, the EU’s General Data Protection Regulation 2016 (GDPR),38 in effect from May 2018. The GDPR draws directly on the ‘right to the protection of personal data’ in Art 8 of the EU Charter of Fundamental Rights 2000,39 specifying that ‘data must be processed fairly for specified purposes and on the basis of the consent of the person concerned or some other legitimate basis laid down by law’.40 This requirement finds an interesting parallel in Art 22(1) GDPR, which states that ‘[a] data subject shall have the right not to be subject to a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her’.41 Much will depend on how the phrases ‘based solely on automated processing’ and ‘significantly affects’ are construed.42 However, the exceptions provided for in Art 22(2) for ‘explicit consent’ by the data subject, entry into or performance of contractual terms, and profiling allowed under EU or Member State law, all subject to various requirements for measures to be adopted to safeguard ‘the data subject’s rights and freedoms and legitimate interests’,43 suggest that there will be a substantial focus on the quality of consent and protective measures/safeguards adopted.44 A wide range of provisions in the GDPR may be drawn on to address contemporary concerns about data processing in the digital context, including rigorous conditions for consent,45 requirements for 38

39 40 41

42

43

44

45

Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) 2016. Charter of Fundamental Rights of the European Union, 2000/C 364/01, Art 8(1). Ibid, Art 8(2). GDPR, Art 22(1). ‘Profiling’ is defined in GDPR Art 4 as ‘any form of automated processing of personal data consisting of the use of personal data to evaluate certain personal aspects relating to natural persons in particular to analyse or predict aspects concerning that natural person’s performance at work, economic situation, health, personal preferences, interests, reliability, behavior, location or movements’. Frederike Kaltheuner and Elettra Bietti, Data is Power: Towards Additional Guidance on Profiling and Automated Decision-making in the GDPR (2018) 2 Journal of Information Rights, Policy and Practice, available at https://jirpp.winchesteruniversitypress.org/articles/ abstract/10.21039/irpandp.v2i2.45/. For some guidance, see Article 28 Data Protection Working Party, Guidelines on Automated Individual Decision-making and Profiling, for the Purposes of Regulation 2016/679, 17/EN WP251rev.01, 3 October 2017, revised 6 February 2018, available at https://ec.europa.eu/newsroom/article29/item-detail.cfm?item_id=612053 . GDPR Arts 22(2)(a), (b) and (c), and Art 22(3). As to Arts 22(2)(a) and (c) (exceptions for explicit consent and contractual terms), Art 22(3) adds that the measures should include ‘at least the right to obtain human intervention on the part of the controller, to express his or her point of view and to contest the decision’. Note also that, under GDPR Art 22(4) and Art 9, further restrictions apply to processing of sensitive data, including as to racial or ethnic origin, political opinions, religious or philosophical beliefs, sex life or sexual orientation, and biometric or health data, including again a requirement that ‘suitable measures to safeguard the data subject’s rights and freedoms and legitimate interests are in place’. GDPR Arts 4 and 7.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

Automated Profiling in New Media and Entertainment

207

transparency,46 a right to data portability,47 and to rectification and erasure,48 and the ability to bring an action and obtain a remedy for breach.49 But it is very clear that the right not to be subjected to a decision based on automated processing under Art 22 is seen as a central plank of the GDPR’s protection of personality in the coming century.50

15.4 the way forward? In short, even as we approach our new algorithmic society, we see a variety of legal mechanisms being adopted, across different jurisdictions, that can be drawn on to constrain automated data processing, including data profiling practices. At minimum, they attempt to ensure a degree of transparency about the technologies and practices, coupled with some ability to exercise real control on the part of those affected and take cases to court where standards are breached. The question remains whether they will develop into more concerted protections against algorithmic control, geared to what English philosopher John Stuart Mill in 1859 called ‘the freedom of pursuing our own good in our own way’51 – arguing that ‘[h]uman nature is not a machine to be built on the basis of a blueprint, and set to do exactly the work prescribed for it; rather, it is a tree that needs to grow and develop itself on all sides, according to the tendency of the inward forces that make it a living thing’.52 In Europe, with its tradition of personality rights, we can expect to see continuing development of data protection along these lines, and not necessarily relying on the language of privacy now that data protection – the right to ‘informational selfdetermination’ as it is sometimes termed53 – is a recognised right.54 In the US we may see expanded consumer protection standards developed specifically for the processing of personal data, supplementing existing protections, and even at some stage potentially the development of fully tradeable property rights in personal data.55 As to other common law jurisdictions, my expectation is that practically speaking the focus will continue to be on the quality of consent and safeguarding

46 47 48 49 50 51

52 53 54 55

GDPR Arts 5(1)(a) and 12. GDPR Art 20. GDPR Arts 16 and 17. GDPR Art 79. Noted in Recitals 24, 60, 63, 70, 71, 72, 73 and 91. John Stuart Mill, On Liberty (1859), reprinted in Mary Warnock, ed., Utilitarianism – John Stuart Mill (William Collins, 1962), 138. Ibid, 188. Census Act Case (1983) 65 BVerfGE 1. Charter of Fundamental Rights, above n 39, Art 8. As argued, for instance, by Luigi Zingales and Guy Rolnik, A Way to Own Your Social-Media Data, New York Times, 1 July 2017, available at https://www.nytimes.com/2017/06/30/opinion/ social-data-google-facebook-europe.html.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

208

Megan Richardson

of interests.56 But, for as long as they can, policymakers in these jurisdictions will stick to the old order, using the language of ‘privacy’ rather than admitting there is a larger concern regarding the ability of individuals to fashion and maintain a sense of identity in the face of external pressures – as German sociologist Georg Simmel observed in 1903,57 resisting social determinism and seeking to keep a semblance of control over individuality.

56

57

See, for instance, Australian Competition & Consumer Commission, Digital Platforms Inquiry: Final Report, June 2019, available at https://www.accc.gov.au/system/files/Digital% 20platforms%20inquiry%20-%20final%20report.pdf, ch 7 (‘Digital Platforms and Consumers’). Georg Simmel, The Metropolis and Mental Life (1903), translated by Edward A. Shils, reprinted in Donald N. Levine, ed., On Individuality and Social Forms (University of Chicago Press, 1971), 324.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.022

part ii

Coherence

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

16 The (Lack of ) Coherence of Data Ownership with the Intellectual Property System Josef Drexl1

16.1 introduction In the European Union, policy makers and scholars have been discussing whether the modern data economy is in need of data ownership rights for quite some time. The European Commission has so far taken a regulatory approach, in particular considering a potential data producer’s right to promote access to machinegenerated data.2 In response to this, the Max Planck Institute for Innovation and Competition,3 and the author of this chapter as an individual scholar,4 have advised the Commission and other policy makers against new exclusive ownership rights in data. This chapter strives to deepen the debate from a more normative perspective, asking whether a potential data ownership right would fit into the overall intellectual property (IP) system. The author dedicates this chapter to Annette Kur, with whom it has been his great pleasure to cooperate on many occasions. The topic fits the research interests of Annette excellently, although data ownership has not particularly featured in her writing. She has always been at the forefront of IP scholarship that looks at the IP 1

2

3

4

Prof. Dr., LLM (UC Berkeley), Director of the Max Planck Institute for Innovation and Competition, Munich. See in particular Communication of the Commission of 10 January 2017 – Building a European Data Economy, COM(2017) 2 final, p. 13. Position Statement of the Max Planck Institute for Innovation and Competition of 26 April 2017 on the European Commission’s ‘Public Consultation on Building the European Data Economy’, available at https://www.ip.mpg.de/fileadmin/ipmpg/content/stellungnahmen/MPI_ Statement_Public_consultation_on_Building_the_EU_Data_Eco_28042017.pdf (accessed 1 July 2020). See, for example, Josef Drexl, Data Access and Control in the Era of Connected Devices, Study on Behalf of the European Consumer Organisation BEUC (2018), available at https://www .beuc.eu/publications/beuc-x-2018-121_data_access_and_control_in_the_area_of_connected_ devices.pdf (accessed 1 July 2020).

Downloaded from https://www.cambridge.org/core. 213 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

214

Josef Drexl

system as a whole with the interest in improving individual IP rights systems by learning from others. Moreover, she has identified herself as a critical IP scholar who often raised their voice to prevent excessive or dysfunctional IP legislation.5 Nevertheless, she has always acknowledged the benefits of an IP system that serves society as a whole and not only big business. From the latter perspective, data ownership presents a complex issue. The fact that ‘big business’ – not least the typically US-based internet platform operators – is anyhow empowered to control large volumes of data, including personal customer data, may indeed argue for rules that allocate data ownership to different persons to bring more balance into the system. This raises the question whether data ownership in personal data, as an extension or an additional prong of personal data protection law, should be supported,6 or whether, as considered by the European Commission, the users (consumers) of connected devices should be recognised as the holders of new data producer’s rights.7 This chapter will show that intellectual property principles do not support either of the two propositions. To the contrary, both propositions would conflict with fundamental intellectual property considerations.

16.2 data ownership as intellectual property Before the analysis can address the issue of coherence with intellectual property, a preliminary question is whether future data ownership can be considered intellectual property at all. For the purposes of this chapter, data ownership will be understood in the sense of an economic ownership right in data as intangible assets in the form of an exclusive right that enables the rightholder to appropriate the economic benefits from the use of these data. Accordingly, this chapter will not explore the availability and role of EU trade secrets protection in the data economy.8 5

6

7 8

Most recently, it was the author’s pleasure to work with Annette Kur to support the German Federal Ministry of Justice to introduce a repair clause into Germany’s design legislation. See Stellungnahme des Max-Planck-Instituts für Innovation und Wettbewerb zum Referentenentwurf (RefE) eines Gesetzes zur Stärkung des fairen Wettbewerbs vom 11 September 2018 (co-authored by Annette Kur, Frauke Henning-Bodewig, Reto M. Hilty and Josef Drexl) (16 October 2018), available at https://www.ip.mpg.de/fileadmin/ipmpg/content/stellungnahmen/Stellungnahme_ RefE_fairer_Wettbewerb.pdf (accessed 1 July 2020). Indeed, with this motivation, the French liberal think-tank Génération Libre proposed such data ownership legislation on the national level as part of the implementation of the EU General Data Protection Regulation in 2018. However, this proposal, which was taken up in a motion by a French MP, was rejected. See Génération Libre, Mes datas sont à moi – Pour une patrimonialité des données personnelles (authored by Isabelle Landreau, Gérard Peliks, Nicolas Binctin and Virginie Pez-Pérard), January 2018, 46–100 (Part 2 of the Report). See also the discussion of this proposal in Josef Drexl, Legal Challenges of the Changing Role of Personal and Non-Personal Data in the Data Economy, in Alberto De Franceschi and Reiner Schulze (eds.), Digital Revolution – New Challenges for Law (C. H. Beck/Nomos, 2019), paras 43–44. As considered by the European Commission, note 2 above. EU law only provides a legal framework against unlawful acquisition, use or disclosure of trade secrets and explicitly states that this protection is not designed to create any exclusive rights in

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

The Lack of Coherence of Data Ownership with IP

215

The reason for this is not only and specifically the lack of exclusivity of trade secrets protection, but the fact that data subjects cannot rely on trade secrets protection to appropriate the economic benefits of the use of personal data by third persons.9 The same holds true for firms as the users of connected devices where the data collected by such devices will only sometimes constitute trade secrets.10 Hence, to allocate the economic benefits of the use of personal data to the data subjects or of machinegenerated data to the users of connected devices, additional data ownership legislation would be required. In line with the above understanding, the following analysis identifies data ownership as intellectual property both with regard to its subject matter of protection and its scope of protection. Leaving aside the question whether the personal data protection rules of the General Data Protection Regulation (GDPR)11 qualify, or at least contain elements of IP protection,12 personal data as such can be considered potential subject matter of IP protection. The reason is that EU data protection law defines personal data as ‘information’, namely, as ‘any information relating to an identified or identifiable person’.13 This puts personal data on the semantic level where data conveys meaning. It does not depend on which signs (or languages) are used to convey such meaning. Similar to the description of an invention in a patent grant, such personal information fulfils all the requirements of an intangible (public) good. It is ‘non-rival’ as it can be used by an unlimited number of persons; and it is ‘nonexclusive’ in the sense that nobody can be excluded from its use once the information is made accessible to the public. Conversely, the same can be said about non-personal data, since it can only be distinguished from personal data on the semantic level of information. In contrast, it is less obvious that digitised data, whether they encode personal or non-personal data, can be considered an object of intellectual property as

9

10 11

12

13

the underlying information. See Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the protection of undisclosed know-how and business information (trade secrets) against their unlawful acquisition, use and disclosure, [2016] OJ L157/1, recital 16. For a positive evaluation of the trade secrets protection with regard to data generated by connected devices, see Drexl, note 4 above, 91–106. Yet personal data can constitute trade secrets of a firm that collects personal data of others. On the relationship of data protection rules with trade secrets protections in personal data of others, see in more detail Drexl, note 4 above, 100–102. Drexl, note 4 above, 95. Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/16/EC (General Data Protection Regulation), [2016] OJ L119/1. This has been rejected by Drexl, note 4 above, 59–67 (distinguishing between personal data protection as an expression of a personality right, on the one hand, and a future data ownership right in personal data as a property right, on the other hand). Art. 4(1) GDPR.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

216

Josef Drexl

distinguished from real property. The case of hacking illustrates this point. If somebody destroys data, for instance in the form of a computer program, by hacking the computer of another person, the data (computer program) will be lost for this person, while the copyright (as an IP right) in the computer program still continues to exist. Integrity presupposes existence of the asset in the real world, albeit in the form of digital data. Tort law allocates the damage claim for interference with a tangible asset to the owner of that asset. Whether property in digital data should be recognised for the purpose of protecting integrity is especially an issue in Germany, where private law in principle requires physical property as a basis for tort liability. While it is recognised that property in the physical carrier (the computer) suffices to justify a tort claim of the owner of this carrier, problems arise when data is downloaded in the cloud. In this latter case the person interested in compensation is the one who has downloaded the data in the cloud and not the owner of the server. The Court of Justice of the European Union (CJEU) has contributed to the debate on the recognition of data ownership in identifiable digital data in a different context, namely by recognising ‘property’ in the downloaded dataset embodying a permanently licensed computer program for the purpose of arguing exhaustion of the copyright in the program.14 This shows that property in ‘data’ can and should sometimes be legally separated from the intellectual property, namely the computer program as a copyrighted work. Hence, the law sometimes distinguishes between rights in mere, yet identifiable, digital data in the form of digits on the syntactic level and the subject matter of intellectual property on the semantic level. In the policy debate on data ownership, the European Commission seems to go in a similar direction by considering the introduction of a data producer’s right in raw, machine-generated, data.15 For such protection it would only matter that such data have been produced or processed by a connected technical device. This idea seems to be inspired by the enormous economic value of large datasets, whereby it is often the case that the economic value is realised through big data analytics by which new information is generated based on the empirical conclusions drawn from rather random correlations between different pieces of information. Yet, from a legal perspective, the attempt to limit protection to ‘raw’ or ‘machinegenerated’ data has to attract serious concerns. Using such concepts fails to identify with sufficient precision what the subject matter of protection should be. In the Commission’s considerations, it does not become sufficiently clear whether this concept indeed only refers to the syntactic level of data, as a sequence of signs, or whether it also extends to the semantic level of information. In fact, the Commission is particularly concerned that data ownership could interfere with EU data

14 15

C-128/11 UsedSoft ECLI:EU:C:2012:407, paras 42–46. European Commission, note 2 above, 8–9, and particularly 10 (‘Raw machine-generated data are not protected by existing intellectual property rights . . .’) and 12–13 (by considering a potential data producer’s right as a means to promote access to machine-generated data).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

The Lack of Coherence of Data Ownership with IP

217

protection law, and therefore argues that data collected by technical devices that is personal in nature should be anonymised in order to qualify as raw, machinegenerated data that could be accepted as the subject matter of a potential data producer’s right.16 Accordingly, the Commission argues for the need to limit protection to ‘non-personal’ machine-generated data. This indicates that the subject matter of protection cannot be identified without taking into account what meaning on the semantic level the machine-generated data is able to convey. This reflects the view in legal scholarship according to which it is not possible to limit data ownership to the syntactic level of using signs without producing adverse effects on access to information on the semantic level.17 Yet the major argument for qualifying the above-mentioned proposition for data ownership as intellectual property arises from the proposed scope of protection. The focus is not specifically – or exclusively – on how to protect data integrity and security, but on whether a person as the data owner should be vested with exclusive rights to control the use of data for the purpose of allocating the gains from economic exploitation to this rightholder. Such use will typically involve the copying or analysing of data without the need to depossess the data owner of the data. These forms of exploitation – involving the use and analysis of data on the semantic level – characterise data ownership as an intellectual property right. This is confirmed by the existing, similarly structured EU system of sui generis protection for databases, which may already provide rather broad or even excessive power to control and exclude access to data.18 16 17

18

Ibid, 13. See, in particular, Thomas Heymann, Recht an Daten – Warum Daten keiner eigentumsrechtlichen Logik folgen (2016) Computer und Recht 650, 653; Andreas Wiebe, Protection of Industrial Data – A New Property Right for the Digital Economy? (2016) Gewerblicher Rechtsschutz und Urheberrecht Internationaler Teil 877, 882 (both arguing that protection of data on the semantic level would restrict access to information, since digital data encode information on the semantic level). See also the more thorough discussion of Drexl, note 4 above, 44–46. The European Commission mentions existing sui generis protection of databases, but in the sense that such protection may not go far enough to provide protection. In contrast, the more recent review of the Database Directive 96/9/EC in 2018 and more recent scholarship assessing the effect of sui generis database protection on free flow of information in the modern data economy identify potentially very far-reaching negative effects on free flow of data. See Study in Support of the Evaluation of Directive 96/9/EC on the Legal Protection of Databases: Final Report (prepared for the European Commission by the Joint Institute for Innovation Policy and the Technopolis Group, together with Lionel Bently and Estelle Derclaye as external experts), 40 (confirming that the sui generis right may especially prevent access to data contained in solesource databases and noting that this has already given rise to a revival of the debate on whether a compulsory licensing system should be added to the regime). The Report especially relies on Leistner, who argued the need for a compulsory license system. See Matthias Leistner, Big Data and the EU Database Directive 96/9/EC: Current Law and Potential for Reform, in Sebastian Lohsse, Reiner Schulze and Dirk Staudenmayer (eds.), Trading Data in the Digital Economy: Legal Concepts and Tools (Nomos/Hart, 2017), 27. For a more recent analysis see Drexl, note 4 above, 67–85 (arguing that legislation on data access regimes should

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

218

Josef Drexl

16.3 economic justifications for ip protection Whether data ownership can be integrated in a coherent system with other IP rights depends on whether the objectives of such legislation can be justified by the standard grounds of IP protection. Intellectual property typically builds on the economic incentive theory. The incentive theory responds to the market failure arising from the public goods nature of the subject matter of protection. It is argued that, without legal exclusivity, markets lead to insufficient production of the intangible asset. However, the data economy does not seem to be characterised by underproduction of data. Rather, competition in the market fuels a very dynamic transition of whole industries to go digital.19 In addition, de facto control over the data firms collect, generate or process may already suffice as a basis to exclude others from the use of the data and to charge others for the use of data as a basis for workable data markets. However, this does not exclude other economic justifications for the introduction of data ownership. Data ownership may help data holders to invest in making their data available in secondary markets, since the commercialisation of data may often involve additional costs for guaranteeing the quality of data or their anonymisation,20 as well as costs for guaranteeing data security and secrecy. Yet data holders should already be able to recoup such costs by relying on their de facto control of the data and technical protection measures. The fact that data sharing and trading still seem underdeveloped may be explained by types of market failures, such as information asymmetries about the quality of the data21 and lack of data interoperability, that cannot be remedied by data ownership.22 Accordingly, economists also refrain from claiming that there is sufficient economic evidence for the need to introduce data ownership.23 In addition, the incentive theory would at best argue in favour of allocating the data ownership right to the economic actor who makes the investment in the digital

19 20

21

22 23

go further than a compulsory licensing system and generally prevail over sui generis database protection). See Drexl, note 4 above, 135 with further references. As argued by Wolfgang Kerber, Rights on Data: The EU Communication ‘Building a European Data Economy’ from an Economic Perspective, in Sebastian Lohsse, Reiner Schulze and Dirk Staudenmeyer (eds.), Trading Data in the Digital Economy: Legal Concepts and Tools (Nomos/Hart, 2017), 109, 117. Yet economists also argue that the so-called ‘information paradox’ is not a real problem in the data economy, since data holders can sufficiently circumscribe the quality of the data without making the data accessible to a potential data purchaser. See Wolfgang Kerber, A New (Intellectual) Property Right for Non-Personal Data? An Economic Analysis (2016) Gewerblicher Rechtsschutz und Urheberrecht Internationaler Teil 989, 994. Another way to solve the information problem would be to appoint an independent data trustee to assess the quality of the data for the purposes of the purchaser. In this sense Kerber, note 20 above, 121. Conclusion by Kerber, note 21 above, 994–995.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

The Lack of Coherence of Data Ownership with IP

219

business model. This would typically be the operator of an internet platform or the producer of a (smart) connected device.24 Such allocation of the right would hence further strengthen the existing de facto control of the firm collecting the data, without taking into account the colliding interests of data subjects in their personal data25 or the interests of the owners or long-term users of connected devices in access to the data these devices collect. This raises the question whether there are economic justifications of vesting data ownership rights in a different person. As mentioned above, the Commission itself argues in favour of allocating a data producer’s right to the owner or long-term user of a connected device in order to promote access to data collected by such device.26 Here, the right is meant to remedy a different market failure, namely data lock-ins. The Commission seems to hope that the data producer’s right will enable owners and long-term users to get access, or to provide other suppliers of goods and services with the possibility to get access, to those machine-generated data without the consent of the de facto data holder (typically the manufacturer of the device). Yet the data producer’s right as conceived by the Commission is not the appropriate means to remedy data lock-ins for several reasons of inconsistency with general principles of IP legislation. First, IP legislation provides power to the rightholder to prohibit the unauthorised use of the subject matter by others. But IP rights do not generally provide a positive right to use the subject matter; indeed such use may often be excluded for other legal grounds.27 In the case of data, the mere exclusive right to prohibit the use does not enable the rightholder to use the data or enable a third person to use the data as long as the manufacturer maintains de facto control over the data. If the data owner licensed the use of the data to a third person without being able to provide access to the data, the data owner would still remain unable to fulfil the contract. The exclusive right of use alone would only create some economic leverage, but no legal claim, against the de facto data holder to get

24

25

26 27

The latter is confirmed by the Commission Staff Working Document on the Free Flow of Data and Emerging Issues of the European Data Economy, 10 January 2017, SWD (2017) 2 final, 35 (considering the manufacturer of a connected device as the data producer based on the classical incentive theory). Still, the Commission Working Staff considers the owner or the long-term user as a co-producer of the data collected by a connected device. The latter seems to reflect the conclusion of the European Commission in the Communication on Building a European Data Economy, note 2 above, 13, where only the owner or long-term user is considered as the holder of a potential data producer’s right. On the coordination of personal data protection with an economic data ownership right in the personal data allocated to the firm collecting the data, see Drexl, note 4 above, 138 (arguing that cumulating two rights would in principle be possible, but that the right to withdraw consent under data protection law would torpedo the licensing of the data ownership right of the data producer, thereby undermining the economic utility of the data producer’s right). European Commission, note 2 above, 13. For instance, for lack of consumer safety. To bring a patented pharmaceutical to the market, the law also requires a marketing authorisation based on the safety and efficacy of the drug.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

220

Josef Drexl

access to the data.28 This shows that the owner or long-term user of such device is not in need of an IP right, but a data access or data portability right.29 Of course, nothing argues against including a data access right in legislation on a data producer’s right. But this would remain legislation that is in substance alien to intellectual property.30 Second, the economic rationale of unlocking data does not justify the introduction of an exclusive right to prohibit the use of the data by anybody else. In particular, it would immediately create additional costs based on the need to clear rights, especially when third persons seek access to the aggregated data controlled by the manufacturer of a connected device.31 While the Commission claims to promote free flow of data, the discussed data producer’s right would by definition limit free flow of data. Third, data ownership framed as intellectual property would also entail the right to waive, transfer or license the right. Therefore, the exclusive right would not protect the owner or long-term user against transferring the right for free in a case of unequal distribution of bargaining power.32 Manufacturers marketing connected devices could easily use standard contract terms to get the data producer’s rights of their customers assigned for free. At best, data lock-in is a problem of consumer law, and not IP law, and should be addressed by remedies that target the imbalance of bargaining power. In sum, the appropriate solution does not consist in an exclusive data producer’s right but in a non-waivable data access right.

16.4 ownership in personal data in particular As mentioned, recognition of data ownership is also advocated for personal data as a right of the data subjects.33 Yet such protection is explained by justice considerations rather than by any IP-like incentive theory. But application of such a theory is not excluded upfront. Indeed, it could be argued that recognition of a second economic 28

29 30

31 32 33

On the insufficiency of the exclusive data producer’s right to get access to the data, see more broadly Drexl, note 4 above, 141–142. For instance, according to the model of the data portability right of Art. 20 GDPR. Indeed, a right of access to the subject matter of protection cannot be found in any other IP legislation. The right of the author to get access to the physical copy of a copyright work, such as a painting (see, for instance, s. 25 German Copyright Act), is not a right of access to the intangible work as the subject matter of copyright protection. Rather, this right of the author limits the property right of the owner of the tangible copy to withhold that copy from others. See Position Statement of the Max Planck Institute, note 3 above, para 19. Ibid, para 18. See also Drexl, note 4 above, 143. See in particular the Study of Génération Libre, note 6 above. In the German literature, personal data protection or, in broader terms, personal autonomy, seems to be the starting point for arguing data ownership for individuals using connected devices; see Karl-Heinz Fezer, Data Property of the People – An Intrinsic Intellectual Property Law Sui Generis Regarding People’s Behavior-generated Informational Data (2017) Zeitschrift fu¨r Geistiges Eigentum 356. In favour of adding an economic prong to personal data protection to guarantee economic participation in the income generated from the economic exploitation of data, see also ArthurAxel Wandtke, Diskussion des ‘Warencharakters’ von Daten aus persönlichkeits- und urheberrechtlicher Sicht (2017) MultiMedia und Recht 6.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

The Lack of Coherence of Data Ownership with IP

221

pillar of personal data protection has the very function of creating incentives for data subjects to license their rights. From an economic perspective this could enhance the development of the data economy, which also has to rely on the use of valuable personal data.34 Yet such logic conflicts with the specific objectives of personal data protection law in the EU. The GDPR is not an instrument to promote the commercialisation of data, but to preserve the autonomy of the data subject. Hence, ‘consent’ in terms of the GDPR is not ‘consent’ as part of a contract to license the economic use of the subject matter of protection of IP law or even more a contract to transfer an IP right, but a permanent means of control guaranteed to the data subject.35 The difference with IP contract law principles arises most clearly from the right to withdraw consent at any time according to Article 7(3) GDPR and, subsequently, to claim erasure of that data according to Article 17(1)(b) GDPR.36 Hence, the GDPR should not be considered a nucleus of already existing EU law on data ownership in personal data, nor is such recognition a logical next step in the development of European data protection law. To the contrary, from an IP perspective an economic ownership right in personal data would need to include the transferability of the right or, at least, the capability of the data subject to permanently license the use of personal data to third persons.37 If the legislature implemented such property right in personal data on the national level, it could not set aside the non-waivable autonomy rights enacted under the EU GDPR under the principle of supremacy of EU law. Whether the European legislature would – at least in theory – be empowered to liberalise EU data protection law and adopt IP-style data ownership legislation for personal data on the EU level, depends on the question whether such legislation would be in conformity with the right to data protection under Article 8 of the EU Charter of Fundamental Rights.38

34

35

36

37

38

The problem is also that GDPR defines personal data so broadly that economic regulation of the data economy cannot ignore the use of personal data in the digital economy. On the definition of personal data and its application, especially in cases relating to big-data analyses, see Manon Oostveen, Identifiability and the Applicability of Data Protection to Big Data (2016) 6 International Data Privacy Law 299, 305–306. See also Drexl, note 4 above, 48. Yet data protection law based on the autonomy of the data subject does not prevent the data subject from giving consent for economic purposes. The data subject can also charge a price for the data, or, as typically happens, accept that personal data are used for getting access to a digital service without monetary consideration. The consent given in the framework of such a contract has to be legally distinguished from the consent in the sense of the GDPR. On the contract law implications of the withdrawal of the latter, see Drexl, note 4 above, 122–123. Following Luisa Specht, Das Verhältnis möglicher Datenrechte zum Datenschutzrecht (2017) Gerwerblicher Rechtsschutz und Urheberrecht Internationaler Teil 1040, 1043 (arguing that workable IP systems are in need of stable contracts). The lack of compatibility of an intellectual property right in personal data with personal data protection law is more broadly argued by Drexl, note 6 above, paras 53–56.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

222

Josef Drexl

16.5 data as an unfit subject matter of ip protection Finally, data also appears an unfit subject matter of IP protection, for two reasons. First, data would define the subject matter of protection too broadly. This concern arises from a comparison with other forms of IP protection. As seen in Section 16.2 above, data as the potential subject matter of IP protection would either have to be placed on the semantic level of information, or the attempt to limit protection to the raw, machine-generated data on the syntactic level would still limit access to the information encoded in raw data. Other IP rights, such as under patent law and copyright law in particular, also provide protection on the semantic level. Yet the difference is that data ownership legislation would recognise ownership in any encoded information without any additional substantive requirements, while the patentability requirements guarantee that not just any technical information can be patented, and the originality requirement prevents information as such from qualifying for copyright protection. To the contrary, patent law takes the interest of the public in access to information into account by publishing the invention and allocating the invention to the public domain after the expiry of the patent. Copyright law only protects the creative parts of works and, thereby, promotes the distribution of information by creating incentives to invest in media and publishing businesses. The conclusion from this is that data ownership should attract even closer scrutiny and caution than that the legislature would have been well advised to apply on the occasion of the adoption of sui generis database protection. In the latter context, the legislature has already committed the error of providing protection exclusively based on (economic) input coined as ‘investment’, without any requirements regarding the output in terms of quality requirements that serve society.39 The second argument refers to the need for legal certainty in the context of rightsclearing and licensing. The workability of an intellectual property system depends crucially on the ability of the public to identify existing rights and – to a lesser extent – rightholders, as well as the ability of rightholders to monitor the execution of licensing agreements and to identify infringements by any third party in the markets. This requires that the subject matter of protection is clearly identifiable. This may still be possible in the case of databases. Yet data as part of huge datasets, where the data can originate from very different sources and may have gone through various steps of big data analytics, may create tremendous problems of legal certainty for users of accessible data, while failing to enable the rightholders to control the use of their data in secondary markets. Registration of data ownership rights and technical protection measures are clearly insufficient against the backdrop of the 39

On the information policy of copyright and, in light of this, criticising sui generis data base protection and data ownership rights, see P. Bernt Hugenholtz, Data Property in the System of Intellectual Property, in Sebastian Lohsse, Reiner Schulze and Dirk Staudenmeyer (eds.), Trading Data in the Digital Economy: Legal Concepts and Tools (Nomos/Hart, 2017), 75.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

The Lack of Coherence of Data Ownership with IP

223

technical and economic reality of the digital economy, considering the huge volume of data and the particular use of data in the context of big data analytics. The latter would in particular require a decision on the legal regime of derivative data – similarly to the legal regime of ‘derivative works’ (adaptations) in copyright law – and the rights of different persons in such data, which in substance consist in empirical informational conclusions drawn from large amounts of pre-existing data. To reach the objective of guaranteeing allocation of the income from secondary markets to data owners, it would be necessary to both guarantee rights throughout long value chains of big data analyses and monitor the use of such rights in practice. While the first requirement could theoretically be implemented in substantive legal provisions, implementation and enforcement of such rules would be practically impossible in a digital economy where even computer experts cannot explain how AI-based computer programs reach their decisions.40

16.6 conclusion In sum, data ownership does not fit into the IP landscape. The legislature should refrain from creating an illegitimate child of IP law and confusing the privacy interests, as protected under existing EU data protection rules, with economic property interests. Rather, the legislature should concentrate on the elaboration and conditions of data access regimes to promote the economic development of the data economy.

40

With similar arguments, Specht, note 37 above, 1040. The problems relating to the status of derivative data are more thoroughly discussed by Drexl, note 4 above, 143–145.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.025

17 The Threefold Fictitiousness of Intellectual Property Alexander Peukert1

17.1 introduction To Annette Kur, the contributors to this Festschrift, and many other academics around the globe, intellectual property (IP) is a topic fascinating enough to merit lifelong attention. In this chapter – dedicated to Annette as a sign of deep gratitude and admiration for all her support and inspiration during my academic life – I claim that the fascination displayed by IP cannot be reduced to its economic and political significance or a peculiar predisposition of IP academics, such as a particular interest in innovation, the arts, or market communication. Instead, I suggest that IP is a theoretically, doctrinally (! coherence) and practically fascinating, eminently dynamic (! transition) body of law because it constitutes, as has often been remarked in passing, a “fiction”,2 and fictions are, in and beyond the law, powerful and stimulating imaginations.3 1

2

3

Professor of Civil and Commercial Law, Faculty of Law, Goethe University, Frankfurt am Main. Michel Foucault, The Essential Foucault 377, 382 (Paul Rabinow and Nikolas Rose eds., The New Press, 2003) (“fiction of the work”); James E. Penner, The Idea of Property in Law 118 et seq. (Clarendon Press, 1997) (“idiotic fiction that intellectual property constitutes property in ideas (patents) or expressions”; “in general it does no harm to speak of rights in ideas, or in manuscripts, or in marks, any more than it does to refer to one’s rights in one’s labour”); Peter Drahos, A Philosophy of Intellectual Property 67, 111, 151–56, 211 (Ashgate, 1996); Hugh Breakey, Properties of Copyright, in Concepts of Property 137, 152 et seq. (Helena R. Howe and Jonathan Griffiths eds., Cambridge University Press, 2013); Robert H. Rotstein, Beyond Metaphor: Copyright Infringement and the Fiction of the Work, 68 Chi.Kent L. Rev. 725 (1993); Brad Sherman and Lionel Bently, The Making of Modern Intellectual Property Law 28 (Cambridge University Press, 1999); Alain Pottage and Brad Sherman, Figures of Invention 4, 7 (Oxford University Press, 2010) (“Intangibility is a figment”). The contribution is based in part on Alexander Peukert, Fictitious Commodities: A Theory of Intellectual Property Inspired by Karl Polanyi’s “Great Transformation”, 29 Fordham Intellectual Property, Media & Entertainment Law Journal 1151 (2019).

Downloaded from https://www.cambridge.org/core. 224 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.026

The Threefold Fictitiousness of IP

225

With the claim that IP constitutes a fiction, I obviously do not deny that IP rights (IPRs), IP laws and the subject matter of these rights and laws exist. They all form part of social reality, but in a way that is much more complex than the standard account suggests, according to which IPRs allocate objects to owners, just like ownership allocates real property.4 This simplistic paradigm operates on the basis of three fictions, explored below, that establish the impression that there are IP objects that one can own, like a piece of cake or land.

17.2 information not produced for sale The first fiction concerns IP subject matter that was originally not produced for sale but created for other purposes. Karl Polanyi coined those commodities “fictitious” in contrast to “real” commodities. His examples for “fictitious” commodities were labor, land, and money. In his classic study The Great Transformation, he explains that and how those goods came to be treated as if they had been produced for sale, although they were either not produced at all (like land) or, if so, were not produced for sale (like labor).5 IPRs also apply to works, inventions, signs etc. that were brought about in nonmarket contexts and without the perspective of commercialization. Grace periods in patent law are precisely meant to allow for such a transformation.6 Copyright comes into existence automatically and therefore covers each and every artefact displaying a modicum of creativity, including countless works created for pleasure in completely private settings. If these works retain their non-commercial social status, their copyright protection does not attain relevance. Thanks to the Internet, however, works created in private nowadays often see the light of day and sometimes even spark great commercial success. Such a move from fan fiction to bestseller in itself reorganizes the communicative context from non-commercial to commercial.7 More visible and contested transformations concern academia and the artistic field. These spheres operate separately from the market on the basis of autonomous logics 4

5

6

7

For a recent example see CJEU Case C-310/17, ECLI:EU:C:2018:899 paras 32–46 – Levola Hengelo. Karl Polanyi, The Great Transformation – The Political and Economic Origins of Our Time 75–76 (Beacon Press, 2001) (“Labor is only another name for a human activity which goes with life itself, which in its turn is not produced for sale but for entirely different reasons, nor can that activity be detached from the rest of life, be stored or mobilized; land is only another name for nature, which is not produced by man; actual money, finally, is merely a token of purchasing power which, as a rule, is not produced at all, but comes into being through the mechanism of banking or state finance. None of them is produced for sale. The commodity description of labor, land, and money is entirely fictitious.”). 35 U.S.C. § 102(b)(1) (2018). More restrictive Art. 55(b) EPC (six months grace period for displays of the invention at an international exhibition). Except perhaps in the view of literary critics; Liz Bury, Fifty Shades of Pay: Erotica Yarn Sends EL James to Top Spot in Earnings List, The Guardian (Aug. 14, 2013), https://www .theguardian.com/books/2013/aug/14/el-james-highest-earning-author.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.026

226

Alexander Peukert

of truth and aesthetics and respective allocations of reputational gains and losses among academics and artists. If genuine academic writings and artworks originally created for their own sake (“l’art pour l’art”) are later marketed as products up for sale, their perception and evaluation change fundamentally. They are not valued anymore according to their truth, depth of thought, or aesthetic originality but rather according to their market success. Depending upon its frequency, such commodification can exhibit systemic effects that tend to supplant an open and reciprocal “republic of science” and an equally autonomous artistic field with profit-oriented transactions.8 In the area of trademark law, finally, ex-post commodification of noncommercial signs concerns artworks in the public domain and cultural icons that are later used as signs indicating the origin of a good or service.9 The European Free Trade Association (EFTA) Court recently confirmed that, in general, trademark protection is available and legitimate in such cases, unless there is a genuine and sufficiently serious threat of “misappropriation or desecration” of the respective work, in which case a trademark registration may be refused on the basis of the public policy/morality exception.10

17.3 commodifying communication In quantitative terms, these instances of ex-post commodification do not, however, justify labeling IP as “fictitious” across the board. Much, if not most, IP subject matter is originally produced for sale under conditions of the market and thus presents a real capitalist commodity in Polanyian terms.11 Suffice it to mention patented medicines and other technologies invented within private companies for 8

9

10

11

Academia: Michael Polanyi, The Republic of Science: Its Political and Economic Theory, 38 Minerva 1 (2000); Alexander Peukert, Das Verhältnis zwischen Urheberrecht und Wissenschaft: Auf die Perspektive kommt es an!, 4 J. Intell. Prop. Info. Tech. Elec. Comm. 142 (2013) para. 1; Academic Capitalism in the Age of Globalization (Brendan Cantwell and Ilkka Kauppinen eds., Johns Hopkins University Press, 2014). Art: Pierre Bourdieu, The Rules of Art: Genesis and Structure of the Literary Field (Susan Emanuel trans., Stanford University Press, 1996); Martin Senftleben, Copyright, Creators and Society’s Need for Autonomous Art – The Blessing and Curse of Monetary Incentives, in What if We could Reimagine Copyright? 20, 48 (Rebecca Giblin and Kimberlee Weatherall eds., Australian National University Press, 2017). Katya Assaf, The Dilution of Culture and the Law of Trademarks, 49 IDEA 1 (2008); Martin Senftleben, Free Signs and Free Use: How to Offer Room for Freedom of Expression within the Trademark System, in Research Handbook on Human Rights and Intellectual Property 354, 357 et seq. (Christophe Geiger ed., Edward Elgar, 2015). Case E-5/16, Municipality of Oslo, para. 102 (EFTA Court 2017) and Annette Kur, Gemeinfreiheit und Markenschutz – Bemerkungen zur Entscheidung des EFTA-Gerichtshofs im “Vigeland”-Fall, GRUR 2017, 1082. For US law see 15 U.S.C. §1052(a) (2016) and Matal v. Tam, 137 S. Ct. 1744 (2017). Bob Jessop, Knowledge as a Fictitious Commodity: Insights and Limits of Polanyian Analysis, in Reading Karl Polanyi for the 21st Century – Market Economy as a Political Project 115, 118–119 (Ay¸se Bu˘gra and Kaan A˘gartan eds., Palgrave Macmillan, 2007).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.026

The Threefold Fictitiousness of IP

227

commercial gain, proprietary software, entertainment products, phonograms, broadcasting signals, databases and other products protected by rights related to copyright, industrial designs, and signs created for use as trademarks. Industrial property law is even confined to the commercial context. Private uses of patented inventions, protected designs, trademarks, etc. are beyond the scope of these IPRs.12 Copyright does extend to the private sphere but often in the weaker form of a right to remuneration (liability rule).13 Thus, many, if not most, IPRs institutionalize markets for original commodities. This statement is, however, premature because it fails to recognize the peculiar communicative character of all inventions, works, and signs, irrespective of whether they have been created for non-commercial or commercial purposes. Communication is generally defined as a process in which one person imparts something that she knows to another.14 The creation and further use of inventions, works, and trademarks always involves such acts of communication. To bring about a new technical solution, a creative expression, or a distinctive sign, firstly requires an immense amount of personal knowledge on the part of a novice innovator that has to be acquired by learning about existing technologies, works, brands, etc.15 Secondly, and more importantly, the result of this preparatory act of communication is itself an artefact that communicates something. Most notably, texts, but all other categories of copyrightable works too, express information, be it a scientific theory, a story, or another visually or aurally perceivable “idea.”16 Immanuel Kant therefore characterized printed matter as a dynamic speech of the author (opera) and not as an objective thing (opus).17 The German Federal Constitutional Court also finds that a published work will “serve as a link to an artistic dialogue.”18 The same can be said of patentable inventions. For inventions are neither to be equated with a machine or other “dead” artefacts nor with a “relation between a person and an object.”19 Instead, inventions teach a person having ordinary skill in the art (PHOSITA) 12

13 14

15

16 17

18

19

Cf. Patentgesetz [PatG] [German Patent Act], Dec. 16, 1980, BGBl. I at 1, § 11(1); Community Design Regulation 6/2002, Art. 20(1)(a), 2002 OJ (L 3) 1, 6 (EU). Cf. Parliament and Council Directive 2001/29, Art. 5(2)(b), 2001 OJ (L 167) 1. Case C-273/00, Sieckmann v. Deutsches Patent- und Markenamt, Opinion of AG Colomer, 2002 E.C.R. I-11737, paras. 19–20; Niklas Luhmann, Die Gesellschaft der Gesellschaft 81 et seq. (Suhrkamp, 1997). KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) (“advances, once part of our shared knowledge, define a new threshold from which innovation starts once more”); see also dissenting opinions in Laboratory Corp. v. Metabolite, 548 U.S. 124 (2006) (Breyer, J., Stevens, J., and Souter, J., dissenting). Art. 9(2) TRIPS Agreement. Immanuel Kant, On the Unlawfulness of Reprinting (1785), Primary Sources on Copyright (1450–1900), http://www.copyrighthistory.org/record/d_1785 (last visited Jul. 02, 2020); Abraham Drassinower, What’s Wrong with Copying? 8, 16, 113 (Harvard University Press, 2015). BVerfG May 31, 2016, NJW 2016, 2247 para. 87 – “Metall auf Metall” (English version available at: http://www.bverfg.de/e/rs20160531_1bvr158513en.html, my emphasis). Contra Drassinower, supra note 17, at 64–65.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.026

228

Alexander Peukert

how to solve a particular problem by making use of natural resources and the laws of nature.20 In other words, an invention as claimed and described in a patent document communicates technical information (the message) from a sender (the inventor) to a recipient (the PHOSITA). Trademarks, finally, also “convey a message.”21 It is their very function to inform the public about the origin of a product and to create an attractive image. Thus, works, inventions, and trademarks are not static commodities produced for consumption but elements of dynamic communicative processes. They are derived from the state of the art and further conveyed to the public, whose members in turn rely on them as the basis for further innovation, creative expression, and competition. To treat IP as if it is a marketable good is fictitious because such commoditization ignores IP’s embeddedness in communication. What is more, the communicative significance of a given work, invention, or trademark for the public – and thus its use and, eventually, its exchange value – is not produced by the IPR holder. All that an author, inventor, or trademark owner can do is to impart her artistic, technical, or marketing message to the public and hope that some recipients will receive, understand, and find interest in it. Only in this case is the communication complete and a use value created. A book unread, a technical teaching ignored, or a trademark not perceived is worth just the paper on which it is printed. IP only enters the picture if there is an active recipient. The moment when an IP communication is successfully completed is, however, not only the moment when use and exchange value is created but also the moment in which a work or other IP subject matter becomes “common intellectual and cultural property.”22 This common property and sometimes even the meaning of a work or trademark is created by members of the public.23 It thus cannot be attributed to the author, inventor, or other IPR holder alone. The contrary rule of IPR ownership is thus based on a fiction.

17.4 commodifying artefacts and actions of non-owners One could still try to defend IP as a standard or “real” commodity in Polanyian terms by pointing to IP subject matter that does not communicate anything. For example, the rights “related” to copyright such as the rights of phonogram and film producers and of broadcasters do not attach to a message that a sender imparts to a recipient 20

21

22 23

BGH Mar. 03, 1969, GRUR 1969, 672 (673) – Rote Taube; BGH Jun. 30, 2015, GRUR 2015, 983 para. 27 – Flugzeugzustand (Erfindung als “Lehre zum planmäßigen Handeln unter Einsatz beherrschbarer Naturkräfte zur Erreichung eines kausal übersehbaren Erfolgs”); BGH Sep. 27, 2016, GRUR 2017, 261 para. 21 – Rezeptortyrosinkinase II (“Lehre zum technischen Handeln”). Matal v. Tam, 137 S. Ct. 1744, 1752 (2017) (“trademarks often consist[ed] of catchy phrases that convey a message”). Supra note 18. Regarding the consumer understanding of a trademark see Dev Saif Gangjee, Property in Brands, LSE Law, Soc’y and Econ. Working Papers, Jun. 13, 2013, at 1, 19, available at https:// papers.ssrn.com/sol3/papers.cfm?abstract_id=2249765 (unpaid labor of consumers).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.026

The Threefold Fictitiousness of IP

229

but to the medium or channel employed for the transmission of a piece of information, namely to fixations of sounds and moving images, and to broadcasting signals.24 If a phonogram, a film carrier, or a broadcasting signal produced for the entertainment market is not a “real” commodity, what else will qualify for this category? In addition, both this and other IP subject matter, including works, inventions, and trademarks, have for a long time been signified and regulated as “goods” that can be owned and traded on markets.25 Is this absolutely dominant practice not proof enough of the ontological adequacy of IP commodification?26 Well, not if one takes into account that mainstream IP theory itself characterizes IPRs as legal institutions creating “artificial scarcity” of otherwise “public goods.”27 For to treat public goods as if they were private commodities and to execute this transformation by legal means is exactly the kind of fictitiousness Polanyi had in mind.28

24

25

26

27

28

See Art. 14 TRIPS Agreement, WPPT, BTAP. Another example of this materialist approach concerns plant variety rights that attach to a “plant grouping within a single botanical taxon of the lowest known rank, which grouping, irrespective of whether the conditions for the grant of a breeder's right are fully met, can be defined by the expression of the characteristics resulting from a given genotype or combination of genotypes, distinguished from any other plant grouping by the expression of at least one of the said characteristics and considered as a unit with regard to its suitability for being propagated unchanged” (International Convention for the Protection of New Varieties of Plants (UPOV), Art. 1(vi), Mar. 19, 1991, 815 U.N.T.S. 89); Alexander Peukert, Kritik der Ontologie des Immaterialgu¨terrechts 68 et seq. (Mohr Siebeck, 2018). This is true even for critical observers. See, for example, Drahos, supra note 2, at 156 et seq., 212; Jessop, supra note 11, at 120 (“non-rival good”). But see Ugo Pagano, The Crisis of Intellectual Monopoly Capitalism, 38 Cambridge J.Econ. 1413 (2014) (“Knowledge is not an object defined in a limited physical space. The same item of knowledge can be encoded in multiple languages, using many different objects existing in a potentially infinite number of places. For this reason, the full-blown private ownership of knowledge means a global monopoly that limits the liberty of many individuals in multiple locations.”). In this sense Reinold Schmu¨cker, Was ist Kunst? Eine Grundlegung 267 (Klostermann, 2014); Maria E. Reicher, Wie aus Gedanken Dinge werden: Eine Philosophie der Artefakte, 61 Deutsche Zeitschrift fu¨r Philosophie 219, 227 et. seq. (2013); Andrew Chin, The Ontological Function of the Patent Document, 74 U. Pitt. L. Rev. 263 (2012). Mark A. Lemley, IP in a World Without Scarcity, 90 N.Y.U. L. Rev. 460 (2015) (“In effect, the point of IP laws is to take a public good that is naturally nonrivalrous and make it artificially scarce, allowing the owner to control how many copies of the good can be made and at what price.”). Maurizio Borghi, Writing Practices in the Privilege – and Intellectual Property – Systems, Social Science Research Network Working Series Papers (2003), https://case.edu/affil/sce/authorship/ Borghi.pdf, 3–4 (“No law of the market can convert an idea, or a poem, or a creation, into a scarce item . . . ideas . . . are not naturally commodities, but they are nonetheless treated as if they were commodities . . . No natural law of the market (no ‘invisible hand’) is capable of producing this fiction by itself. The fiction must be established as such.”); Jessop, supra note 11, at 120 (“knowledge is collectively produced and is not inherently scarce . . . it is made artificially scarce and access thereto depends on payment of rent”); Pagano, supra note 25, at 1414 (“commons were turned into exclusive private property”).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.026

230

Alexander Peukert

Yet even this observation does not exhaust the problem. The continued talk about “public” goods that, through a legal measure, miraculously turn into “private” goods neglects and obscures a transformation at a deeper level, namely the level of how we collectively conceive of reality and regulate human interaction accordingly. The worldview I allude to here is the view that IP “goods,” like works, inventions, brands etc., exist as distinct objects and that these “goods” can be allocated to certain owners in exclusion of all others. The respective ontology assumes that IP objects exist as abstract objects (types) independently from their instantiations (tokens) in books, products, digital files, and other physical or mental manifestations.29 I have shown elsewhere in detail30 that this ontology is implausible because the existence of allegedly abstract IP is always dependent upon the existence of at least one physical or mental “embodiment.” The dominant paradigm is also untenable from a legal perspective because law can legitimately only regulate behavior that relates to brute facts that humans are able to control. Abstract types exactly defy such control. I also show that the idea of the abstract IP object was the result of a quite recent historical process, in which signifiers like “the” book, work, or invention changed their meaning. Instead of referencing many distinct but sufficiently similar artefacts and actions, they henceforth signified abstract IP objects. Whereas early modern privileges and the first British patent and copyright statutes regulated exclusive rights to print a book or work a machine, today’s paradigm was only implemented in full by the French Revolutionary Acts of 1791 and 1793, which granted exclusive property rights in “ouvrages,” “idée nouvelle,” and “découvertes industrielle.” Since then, we have treated books, machines, other items with physical existence, and public performances as secondary “embodiments” of a primary, abstract “intellectual property.” The transformation at stake here concerns the dominant perception of the world. An idealized world of abstract objects superseded a realistic focus on artefacts and actions having brute, measurable existence. This fundamental shift occurred solely in our language and thinking. The brute facts of artefacts and actions (books, machines, performances) retained their physical existence. But they were signified and conceived of differently: not as artefacts and actions whose use or occurrence was regulated, but as exemplars of an IP object that belonged to someone else. This abstraction is fictitious in the sense that we speak of works, inventions, and other IP objects as of tangible commodities, where in fact IP objects only exist insofar and because we speak and regulate as if they exist

29

30

See Art. 2(1) Berne Convention; 17 U.S.C. § 101 (1976) (“embodied”); Code de la propriété intellectuelle, Art. L111–3 (Fr.) (“La propriété incorporelle définie par l’article L. 111–1 est indépendante de la propriété de l’objet matériel.”); Michael J. Madison, The End of the Work As We Know It, 19 J. Intell. Prop. L. 325, 333 (2012) (“The work subject to copyright is solely and purely an intangible thing.”); on designs and signs Arts. 15–16, 26(1) TRIPS Agreement. Peukert, supra note 24; on information as an object see Paul Duguid, The Aging of Information: From Particular to Particulate, 76 J. Hist. Ideas 347 (2015).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.026

The Threefold Fictitiousness of IP

231

as abstract “things” of value. In other words, IP objects only exist in our linguistic practice and collective imagination.31 From a legal realist perspective, IPRs are exclusive rights to prevent or authorize the reproduction and further use of certain Master Artefacts.32 And the only reason for the “wild”33 conceptual move from privileges to act to modern IPRs in abstract objects was the commodity function of IP. The emerging market for books and other innovative yet easily reproducible products required property rights in distinct abstract objects that represented the input of authors and inventors.34

17.5 concluding remarks In order to understand the striking dynamic and intellectual fascination of IP, it is necessary to pose and tackle the ontological question, in what ways does the subject matter of IPRs exist? Proponents of commodification assume that IPRs allocate objects, which in principle lend themselves to propertization like land and other goods, whereas their opponents perceive IP as an integral part of communication and thus society. According to the latter view, IP falls into Polanyi’s category of fictitious commodities. I agree with this qualification because IPRs in part attach to information that was not produced for sale, they partition communication into commodified pieces, and they are based on the obscure fiction of the abstract IP object. The theoretical and normative implications following from this ontological analysis35 are far reaching and eventually confirm the observation with which this article started: IP is a fascinating area of study to which Annette Kur will hopefully continue to contribute for many years to come.

31

32

33

34 35

loci classici for this legal realist approach: Alf Ross, Ophavsrettens grundbegreber, 58 Tidsskrift for Rettsvitenskap 321 (1945); Richard Rudner, The Ontological Status of the Esthetic Object, 10 Philosophy and Phenomenological Research 380 (1950). Arts. 11, 16, 26, 28 TRIPS Agreement (Members shall provide rightholders with exclusive rights to “prevent,” “prohibit” or “authorize” certain conduct); Peukert (supra note 24) at 49 et seq.; from the perspective of economics Michele Boldrin and David K. Levine, Intellectual Property and the Efficient Allocation of Social Surplus from Creation, 2 Review of Economic Research on Copyright Issues 45 (2005). Millar v. Taylor, [1769] 4 Burrow 2303, 2357 (Yates, J.); see Sherman and Bently, supra note 2, at 19 et seq. Millar v. Taylor, ibid. See Peukert (supra note 3).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.026

18 An Intellectual Property Structural Engineer Extraordinaire and Her Lifelong Quest for Coherence Peter K. Yu1

18.1 introduction Striking an appropriate balance in the intellectual property system has been the subject of a perennial scholarly and policy debate. Article 7 of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement) states explicitly that ‘[t]he protection and enforcement of intellectual property rights should contribute to . . . a balance of rights and obligations’. The preamble of the WIPO Copyright Treaty (WCT) also recognises ‘the need to maintain a balance between the rights of authors and the larger public interest, particularly education, research and access to information’.2 Yet, despite the dedicated and ceaseless efforts of both policymakers and commentators, the search for a balanced intellectual property system has remained unfinished. Finding this balance is imperative because overprotecting intellectual property rights is just as dangerous as underprotecting them. While the latter will take away the much-needed support for creativity and innovation, the former will impoverish society by reducing access to information, knowledge, and scientific productions. It is therefore no surprise that policymakers and commentators alike, Professor Annette Kur included, have devoted considerable time, energy, and effort to developing a balanced intellectual property system. In the past two decades, the rapid expansion of intellectual property rights, the proliferation of disruptive technologies, and the active negotiation of bilateral, regional, multilateral, and plurilateral agreements have sparked a new but related

1

2

Professor of Law, Professor of Communication, and Director, Center for Law and Intellectual Property, Texas A&M University. WIPO Copyright Treaty (adopted 20 December 1996, entered into force 6 March 2002) (1997) 36 ILM 65.

Downloaded from https://www.cambridge.org/core. 232 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

An IP Structural Engineer Extraordinaire

233

pursuit: the quest for coherence.3 Although this quest has not yet attracted as much scholarly and policy attention as the more longstanding search for balance, Professor Kur is among the rare few who have devoted a significant part of her distinguished career to fulfilling this quest.4 Recognising this important contribution to the intellectual property field, the present chapter explores her role as an intellectual property structural engineer extraordinaire and her lifelong quest for a coherent intellectual property system. It draws illustrations from two edited books, Intellectual Property Rights in a Fair World Trade System5 and The Structure of Intellectual Property Law.6 The chapter also touches on structural concepts that are essential to the development of intellectual property law, such as foundation, floor, ceiling, frame, space, architecture, and construction plan.

18.2 intellectual property rights in a fair world trade system Intellectual Property Rights in a Fair World Trade System: Proposals for Reform of TRIPS bore the fruit of a highly ambitious multi-year project led by Professor Kur and Professor Marianne Levin of Stockholm University in Sweden.7 Established as a collaborative effort between the Institute for Intellectual Property and Market Law (Institutet för Immaterialrätt och Marknadsrätt) at Stockholm University and the Max Planck Institute for Intellectual Property and Competition Law8 (MPI) in Munich, Germany, the Intellectual Property in Transition Project (IPT Project) aimed to remove the ‘structural bias’9 in the TRIPS-based intellectual property system – at both the domestic and international levels.10 As Professors Kur and Levin observed: 3

4

5

6

7 8

9

10

Peter K. Yu, International Enclosure, the Regime Complex, and Intellectual Property Schizophrenia (2007) Michigan State Law Review 1, 18–19. Another notable exception is the Gowers Review of Intellectual Property, which devoted considerable discussion to the need for UK intellectual property rights to be ‘internally and externally coherent’. Andrew Gowers, Gowers Review of Intellectual Property (HMSO 2006) 58–61. Annette Kur and Marianne Levin, Intellectual Property Rights in a Fair World Trade System: Proposals for Reform of TRIPS (Edward Elgar Publishing 2011). Annette Kur and Vytautas Mizaras (eds.), The Structure of Intellectual Property Law: Can One Size Fit All? (Edward Elgar Publishing 2011). Kur and Levin (note 5). Since January 2014, the MPI has been renamed the Max Planck Institute for Innovation and Competition. Annette Kur and Henning Grosse Ruse-Khan, Enough Is Enough – The Notion of Binding Ceilings in International Intellectual Property Protection in Kur and Levin (note 5) 377. Annette Kur and Marianne Levin, The IPT Project – Proposals to Reform the TRIPS Agreement in Gustavo Ghidini, Rudolph J. R. Peritz, and Marco Ricolfi (eds.), TRIPS and Developing Countries: Towards a New IP World Order? (Edward Elgar Publishing 2014).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

234

Peter K. Yu

[A] tendency to apply a one-sided and inflexible approach to [intellectual property] on the global as well as on the domestic level might distort the necessary equilibrium between the different interests involved, and thus become a source of systemic malfunctions and increasing discontent. The response to that . . . could only be found in promulgating an overarching ‘balancing instrument’ to which all [intellectual property] must conform.11

Featuring contributions from scholars affiliated with the IPT Project, this edited volume included three chapters written or co-written by Professor Kur. These chapters provided useful insights into how this intellectual property structural engineer extraordinaire would modify the system. They also showcased her tireless efforts towards fulfilling the quest for a coherent intellectual property system. In her chapter on limitations and exceptions, Professor Kur offered a trenchant critique of the prevailing interpretation of the three-step test in the TRIPS Agreement, the WCT, and countless TRIPS-plus bilateral, regional, and plurilateral agreements.12 Specifically, this chapter ‘explained why and how the interpretation by the WTO panels is not doing full justice to the inherent flexibilities of the test’.13 Providing a careful but convincing analysis and building upon the Declaration on a Balanced Interpretation of the ‘Three-Step Test’ in Copyright Law – a document developed jointly by the MPI and Queen Mary, University of London – this chapter underscored the need to reconsider ‘the structure and meaning of the three-step test in the framework of TRIPS’.14 In Professor Kur’s view, a more liberal interpretation of this test will be conducive to developing new limitations and exceptions that can help restore the balance in the existing intellectual property system. The insights in this chapter have since been extended to newer contexts, such as the Marrakesh Treaty to Facilitate Access to Published Works for Persons Who Are Blind, Visually Impaired or Otherwise Print Disabled.15 In the second chapter, Professor Kur and her former colleague, Henning Grosse Ruse-Khan (now a reader at the University of Cambridge), proposed to establish ‘binding ceilings’ in international intellectual property law.16 These mandatory limits sharply contrast with the international minimum standards established in leading international intellectual property agreements, such as the Paris Convention for the Protection of Industrial Property (Paris Convention), the Berne Convention for the Protection of Literary and Artistic Works (Berne

11 12

13 14 15

16

Annette Kur and Marianne Levin, preface in Kur and Levin (note 5) viii. Annette Kur, Limitations and Exceptions under the Three-Step-Test – How Much Room to Walk the Middle Ground? in Kur and Levin (note 5). Ibid. 208. Ibid. 250. Annette Kur, From Minimum Standards to Maximum Rules in Hanns Ullrich, Reto M. Hilty, Matthias Lamping, and Josef Drexl (eds.), TRIPS Plus 20: From Trade Rules to Market Principles (Springer 2016). Kur and Grosse Ruse-Khan (note 9).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

An IP Structural Engineer Extraordinaire

235

Convention), and the TRIPS Agreement. Joining other scholars who have made similar suggestions,17 Professor Kur and Dr Grosse Ruse-Khan explained: In addition to their main aim of calling a halt to the upward movement of [intellectual property] protection levels, ceilings might . . . help to improve the ability of TRIPS Members to appreciate the room available under international law for legislative measures limiting the availability and scope of [intellectual property] protection.18

In the third chapter, Professor Kur and Professor Jens Schovsbo of the University of Copenhagen in Denmark challenged the predominant exclusivity paradigm in intellectual property law.19 The chapter cautioned policymakers against ‘canonis[ing property rules] as the only regulatory model permitted’20 – or the best model for promoting creativity and innovation. Instead, the authors encouraged policymakers to consider liability rules as a viable alternative, underscoring the fact that these rules ‘are already part of the known [intellectual property right] fabric’.21 Building on these chapters and other contributions to the book, Professor Kur and her collaborators in the IPT Project concretised theoretical ideas into reform proposals. To emphasise the important role liability rules can play in the TRIPSbased intellectual property system, and to allow for further adjustments based on the structural variations in different bodies of intellectual property law, the participants of the IPT Project recommended the addition of a ‘balance of interests’ provision to the TRIPS Agreement. If adopted, this provision would require the consideration of the ‘extent [to which] the right holders and/or original creators are granted equitable and appropriate compensation under efficient and practicable mechanisms’ (Proposed Article 8a.2(b)(i)). At the intersection of intellectual property and private international law – an area in which Professor Kur, Professor Josef Drexl, and their MPI colleagues have made notable contributions22 – the participants of the IPT Project further proposed the 17

18 19

20 21 22

Graeme B. Dinwoodie, Federalized Functionalism: The Future of Design Protection in the European Union (1996) 24 AIPLA Quarterly Journal 611, 715, fn. 274; Rochelle Cooper Dreyfuss, TRIPS-Round II: Should Users Strike Back? (2004) 71 University of Chicago Law Review 21, 27; Ruth Okediji, Toward an International Fair Use Doctrine (2000) 39 Columbia Journal of Transnational Law 75, 168; Peter K. Yu, TRIPs and Its Discontents (2006) 10 Marquette Intellectual Property Law Review 369, 402. Kur and Grosse Ruse-Khan (note 9) 367. Annette Kur and Jens Schovsbo, Expropriation or Fair Game for All? The Gradual Dismantling of the IP Exclusivity Paradigm in Kur and Levin (note 5). Ibid. 445. Ibid. 444. Josef Drexl and Annette Kur (eds.), Intellectual Property and Private International Law (Hart Publishing 2005); Graeme B. Dinwoodie, Rochelle C. Dreyfuss, and Annette Kur, The Law Applicable to Secondary Liability in Intellectual Property Cases (2009) 42 New York University Journal of International Law and Politics 201; Annette Kur, Applicable Law: An Alternative Proposal for International Regulation – The Max-Planck Project on International Jurisdiction and Choice of Law (2005) 30 Brooklyn Journal of International Law 951.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

236

Peter K. Yu

introduction of a new TRIPS provision on ‘jurisdiction and choice of law in transborder conflicts in intellectual property matters’ (Proposed Article 41b). While this provision applies lex loci protectionis by default, it contains flexibility that allows for derogation from such application in special cases, such as those involving multiterritorial infringements in the digital environment. Taken together, all of these chapters and reform proposals aimed to restructure the TRIPS-based intellectual property system from the inside. In Professor Kur’s view, the existing system is unbalanced not because the notion of intellectual property rights is fundamentally flawed, but because of the system’s failure to attain a proper structure. By providing the much-needed adjustments, the book and the IPT Project helped retool this system to better promote creativity and innovation.

18.3 the structure of intellectual property law The Structure of Intellectual Property Law: Can One Size Fit All? showed a different side of Professor Kur as the intellectual property structural engineer extraordinaire, as well as a yet-to-be-explored aspect of her quest for a coherent intellectual property system. Marking the inaugural volume of the ATRIP Intellectual Property Series published by Edward Elgar Publishing,23 this edited volume brought together papers selected from the 2008 and 2009 annual congresses of the International Association for the Advancement of Teaching and Research in Intellectual Property (ATRIP). The first half of the collection was devoted to the exclusive rights granted by the existing intellectual property system, while the second half turned to limitations and exceptions, utilising illustrations from specific local contexts. As the ATRIP president from 2007 to 2009, Professor Kur organised the congress at the Justizpalast in Munich in July 2008.24 The event’s title was ‘Can One Size Fit All?’, which eventually became the book’s subtitle. Professor Vytautas Mizaras, the volume’s co-editor, served as the local host of the ATRIP congress at the University of Vilnius in Lithuania in September 2009. The title of that follow-up event was ‘Horizontal Issues in Intellectual Property Law: Uncovering the Matrix’. As Professors Kur and Mizaras recounted, ‘Whereas the first year focused on the commonalities in the structure of intellectual property . . . throughout the various disciplines, the second year addressed specific topics which were analysed in the same horizontal manner’.25 Unlike the first book, in which Professor Kur and her collaborators wore safety helmets, holding a construction plan and walking around the TRIPS site to propose structural modifications, this edited volume reflected Professor Kur’s considerable 23

24

25

Kur and Mizaras (note 6). On the launch of this book series, see Annette Kur and Jan Rosén, ATRIP and Publications (2013) 4 WIPO Journal 269, 271–273. On the 2008 and 2009 ATRIP congresses, see Peter K. Yu, Taking ATRIP Down Memory Lane (2013) 4 WIPO Journal 278, 288. Annette Kur and Vytautas Mizaras, foreword in Kur and Mizaras (note 6) ix.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

An IP Structural Engineer Extraordinaire

237

success in rallying scholars to explore key questions concerning the appropriate structure of the TRIPS-based intellectual property system. These questions include the following: Is there a coherent system underlying the entire field of intellectual property . . . or are the different constituent areas only loosely connected with each other? Have the foundations on which the different rights were originally grounded converged, or are they drifting apart? Does the fact that the law must react to a host of problems characterised by an ever-increasing degree of sophistication necessarily entail fragmentation ultimately leading to inconsistency, or are the principles informing legal choices sustainable enough to provide a solid and resilient frame? Vice versa – are those principles too rigid and inflexible, and therefore incapable of responding to novel challenges? To what extent, and where, should the system make room for more differentiation? Which tools does the law provide for the adjustment of protection to different needs and circumstances, and how much flexibility exists in employing those tools? Can One Size Fit All?26

Featuring contributors at various stages of their career, this ATRIP volume was as welcoming to junior scholars as Professor Kur’s other projects and conferences. The book not only showcased her leadership in the quest for a coherent intellectual property system, but also reminded the intellectual property community of how fortunate it was to have her as an ATRIP leader, setting examples for others. Also notable is Professor Kur’s brief introduction to the book, which analogised a unitary intellectual property regime to a tree, where one strong and massive trunk grows out of a meshwork of roots, then separates into different branches which sprout a large number of twigs, and finally end up in smaller and smaller sprigs which are too numerous to be counted.27

Emphasising the ‘one right approach’ that supports the development of ‘a coherent system underlying the entire field of intellectual property’,28 this powerful imagery pushed readers to reconceptualise different intellectual property laws as a single, coherent body of law. Such reconceptualisation, while arguably aspirational, provided a helpful thought experiment that put readers in the position of an intellectual property structural engineer. This thought experiment is particularly important at a time when intellectual property rights continue to expand rapidly in myriad directions, creating a mishmash of overlapping rights that offer multiple layers of protection to the same subject matter. As Professor Kur lamented, ‘While patent and copyright protection were originally based on quite distinct motives and concepts, patent law – in some parts of the world – has ventured out of the field of technology, while copyright long ago left 26 27 28

Annette Kur, Introduction in Kur and Mizaras (note 6) xi. Ibid. xii. Ibid. xi.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

238

Peter K. Yu

the sphere of purely artistic creations’.29 Worse still, ‘legal fields such as patent and copyright law [have] become increasingly compartmentalised: pharmaceutical patents have little in common with patents in “classical” technical fields such as machinery, and copyright in works of fine art is quite another discipline than copyright in news articles or computer programs’.30 To a large extent, the edited volume represented Professor Kur’s open invitation to join her quest for a coherent intellectual property system. This invitation is significant for three reasons. At the domestic or systemic level, greater coherence will provide clarity, certainty, and predictability not only to authors and inventors, but also to users and third parties (such as investors, distributors, and intermediaries). By properly demarcating legal boundaries, such a system will help minimise the complications created by ill-defined, inconsistently applied, or overlapping rights. Appropriately designed, such a system will also be better positioned to meet the continuous challenges posed by modern technological advances and the ubiquitous use of intellectual property goods and services. At the global level, a coherent intellectual property system will facilitate and enhance integration between the disparate national intellectual property systems across the world. Such integration will help minimise transaction costs while enabling individuals, businesses, and their respective countries to fully realise the promise of globalisation. Having a fully integrated international intellectual property system was a key aspiration not only of TRIPS negotiators, but also of the drafters of the Paris and Berne Conventions. Within the larger economic order, greater coherence between intellectual property standards and other regulatory standards can also be beneficial. Such coherence will reduce fragmentation, thereby benefiting countries with limited expertise and resources to navigate complex rules and standards.31 A coherent global intellectual property system will also alleviate the tensions and conflicts between different international regimes, including those relating to public health, human rights, biological diversity, and climate change.

18.4 a positive attitude towards restructuring In the past decade, I have had the good fortune of participating in a number of conferences and projects at the MPI and elsewhere, many of which allowed me to deeply engage Professor Kur on a number of intellectual property topics. Having been a strong advocate for using human rights to address challenges in the

29 30 31

Ibid. xii. Ibid. xi. Eyal Benvenisti and George W. Downs, The Empire’s New Clothes: Political Economy and the Fragmentation of International Law (2007) 60 Stanford Law Review 595.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

An IP Structural Engineer Extraordinaire

239

intellectual property field,32 I cannot help but notice her reluctance to reform the intellectual property system from the outside. To be sure, Professor Kur is open and holistically minded, and she ‘embrac[es] a pluralist approach to intellectual property’.33 She is also keenly aware of the important benefits provided by external norms, which may ‘function as a limit or ceiling on existing . . . protection’.34 In her chapter on binding ceilings in international intellectual property law, for example, she and Dr Grosse Ruse-Khan wrote: Various international treaties may confer rights or otherwise protect the interests of individuals or certain groupings within a society in a way which may necessitate, imply or at least allow limits on [intellectual property] protection. Such treaties can relate to diverse fields such as the protection of human rights, to biological or cultural diversity, access to knowledge and information as well as to environmental protection or climate change.35

As if these observations were not enough, Professor Kur’s home base at the MPI requires her to frequently address questions concerning the tensions and conflicts between intellectual property rights and fundamental rights protected by the European Convention of Human Rights and the Charter of Fundamental Rights of the European Union. In the past decade and a half, those tensions and conflicts have been central to a number of disputes before the European Court of Human Rights and the Court of Justice of the European Union.36 Notwithstanding her awareness of the importance and benefits of external safeguards, Professor Kur remains uncomfortable in bringing doctrines from other

32

33

34 35 36

Peter K. Yu, The Anatomy of the Human Rights Framework for Intellectual Property (2016) 69 SMU Law Review 37; Peter K. Yu, EU Economic Partnership Agreements and International Human Rights in Josef Drexl, Henning Grosse Ruse-Khan, and Souheir Nadde-Phlix (eds.), EU Bilateral Trade Agreements and Intellectual Property: For Better or Worse? (Springer 2014); Peter K. Yu, Intellectual Property and Human Rights 2.0 (2019) 53 University of Richmond Law Review 1375; Peter K. Yu, Intellectual Property and Human Rights in the Nonmultilateral Era (2012) 64 Florida Law Review 1045; Peter K. Yu, Reconceptualizing Intellectual Property Interests in a Human Rights Framework (2007) 40 U.C. Davis Law Review 1039. Graeme B. Dinwoodie and Annette Kur, Framing the International Intellectual Property System in Rochelle Cooper Dreyfuss and Elizabeth Siew-Kuan Ng (eds.), Framing Intellectual Property Law in the 21st Century: Integrating Incentives, Trade, Development, Culture, and Human Rights (Cambridge University Press 2018) 297. Kur and Grosse Ruse-Khan (note 9) 389. Ibid. 384. Henning Grosse Ruse-Khan, Overlaps and Conflict Norms in Human Rights Law: Approaches of European Courts to Address Intersections with Intellectual Property Rights in Christophe Geiger (ed.), Research Handbook on Human Rights and Intellectual Property (Edward Elgar Publishing 2015) 71–78; Laurence R. Helfer, The New Innovation Frontier? Intellectual Property and the European Court of Human Rights (2008) 49 Harvard International Law Journal 1; Tuomas Mylly, The Constitutionalization of the European Legal Order: Impact of Human Rights on Intellectual Property in the EU in Geiger 107–126.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

Peter K. Yu

240

international regimes to address problems in the intellectual property system. As she and Dr Grosse Ruse-Khan observed: [I]nternalising non-trade concerns into TRIPS appears to be a sensible way of ensuring greater coherence, security and predictability in the process of applying [intellectual property] rules. It may also provide more tailored solutions to potential conflicts and leave fewer open questions on how to resolve them. Also militating in favour of internalisation is the fact that according to experience, interests which have to be imported from other (external) rule-systems to become part of the recognised context for interpretation cannot compete effectively with detailed ‘codifications’ of [intellectual property] right holders’ (or more generally: trade) interests. If properly drafted and implemented, internalisation of ceilings could thus lead to a more balanced international system of economic regulation, and would support synchronisation between distinct areas of public international law.37

To Professor Kur, internal ceilings provide more effective structural adjustments than their external counterparts. Because the latter often implicate differing justifications, concepts, values, and policy preferences, they will not only undermine the internal coherence of the intellectual property system, but may also generate new questions and unintended consequences. In fact, Professor Kur is quite confident that internal adjustments can adequately address the problems in this system. Her preferred approach is evident in the TRIPS reform proposals advanced through her publications and the IPT Project. Her ATRIP volume with Professor Mizaras embraced a similar approach, which focused on exclusive rights and their limitations and exceptions, not the use of external safeguards. Taken together, these scholarly works and the IPT Project have revealed Professor Kur’s positive attitude – and, perhaps, even optimism – towards restructuring the existing intellectual property system. Even though she recognises the many problems in this system, she remains confident that solutions can be found inside. At a time when policymakers, commentators, and the public at large actively question the suitability, sustainability, and continued vitality of the existing intellectual property system, her confidence is both refreshing and inspiring.

18.5 conclusion As long as intellectual property rights continue to exist, there will always be debate on how these rights can be restructured to better meet economic and technological challenges. Because a large part of this debate concerns legal responses to these challenges, there will also be endless inquiries about whether the intellectual property system should be restructured from the inside or outside. Fortunately, Professor Kur’s lifelong work has provided us with useful guidance to these inquiries. As she rightly reminded us: 37

Kur and Grosse Ruse-Khan (note 9) 377.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

An IP Structural Engineer Extraordinaire

241

Granting intellectual property . . . rights is like setting a big ship on a straight course: While the course as such may, on the whole, lead to the aspired goal, corrections may be needed here and there in order to avoid dangerous or shallow waters.38

In guiding this ever-expanding ship, she draws advice from a deep passion for intellectual property law, a positive attitude towards addressing problems from the inside, and a willingness to engage with, and set examples for, younger scholars. To this intellectual property structural engineer extraordinaire, I congratulate her not only for having a long, rich, and fruitful career, but also for making a significant impact in the field of intellectual property law. If we get closer to developing a more balanced and more coherent intellectual property system, a lot of credit should go to Professor Annette Kur.

38

Annette Kur, Of Oceans, Islands, and Inland Water – How Much Room for Exceptions and Limitations under the Three-Step Test (2009) 8 Richmond Journal of Global Law and Business 287, 288.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.027

19 Open Yet Secret – Trading of Tangible Goods and Trade Secrets Nari Lee1

19.1 introduction Intellectual property law aims to promote innovation and creativity. If successful, it would lead to proliferation of innovative technologies and creative expression. As a result, intellectual property law does need to deal periodically with new technologies. New technologies first challenge categories of subject matters of intellectual property. As each subsystem of intellectual property follows a rationale that may uniquely fit its subject matter, having categorically defined subject matter eligibility rules provides certainty and efficiency in regulation. However, as commercial opportunities expand and develop without heed to legal categories, every new technological disruption may raise claims of under- or over-expansive protection. Trade secret protection seems to provide a perfect solution to this problem. Trade secrets could include any and all subject matter of intellectual property as well as other types of subject matter ineligible for intellectual property protection, such as raw data, information and knowledge.2 In new technologies, trade secret protection may indeed become a substitute for other types of intellectual property protection, especially when greater restrictions are imposed on traditional subject matters. When the US Supreme Court has imposed stricter subject matter requirements for patents in software and biogenetic technology,3 more businesses have migrated to 1

2

3

Professor of Intellectual Property, Hanken School of Economics, Finland. The author gratefully acknowledges financial support from Hanken Support Foundation and the Academy of Finland research project, FAME-IP. The work was first conceived during the IP Law Researchers in Asia Workshop (2017). I thank Prof. Irene Calboli for arranging that workshop. I also thank Prof. Lionel Bently and Dr. Henning Grosse Ruse-Khan for hosting my numerous visits to Cambridge University, and Prof. Masabumi Suzuki for hosting my research visit to Nagoya University, Japan in 2019. Jerome H. Reichman, Legal Hybrids Between the Patent and Copyright Paradigms, 94 Columbia Law Review 2432 (1994). Alice Corp. v. CLS Bank International, 134 S. Ct. 2347 (2014); AMP v. Myriad Genetics, 133 S. Ct. 2107 (2013).

Downloaded from https://www.cambridge.org/core. 242 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

Trading of Tangible Goods and Trade Secrets

243

protect data or algorithms directly using trade secret law.4 Moreover, as expressive works or methods may also be protected by trade secret law, works such as computer program codes produced by artificial intelligence (AI) or 3D printing CAD files as well as algorithms for AIs may be protected as trade secrets even though their protectability under copyright is more uncertain. Moreover, as regulations on public disclosure for public access to information treat trade secrets differently from other types of information,5 trade secret protection may be considered a versatile tool to avoid public scrutiny over sensitive information. In the context of international trade, trade secret protection is treated rather minimally, leaving much of the details to national legislators.6 Recently, national legislators have been quite active in the US, EU, Japan and China.7 Moreover, investment treaties as well as regional free trade agreements such as the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) include clauses on trade secret protection with detailed definitions of misappropriation which shall give rise to criminal sanctions.8 Against this backdrop, this chapter asks to what extent the international trade in tangible goods may be affected by trade secret protection. Trade secret law is

4

5

6

7

8

See, for example, Dan L. Burk, Patents as Data Aggregators in Personalized Medicine, 21 Boston University Journal of Science and Technology Law 233 (2015), at 242–245; Jacob S. Sherkow and Christopher Thomas Scott, The Pick-and-Shovel Play: Bioethics For GeneEditing Vector Patents, 97 North Carolina Law Review 1497 (2019). For example, the US Freedom of Information Act exempts trade secrets categorically from its scope: 5 USC § 552(b)(4) (2000). The Paris Convention for the Protection of Industrial Property, signed in Paris, France, on 20 March 1883 (hereinafter ‘Paris Convention’). The Paris Convention refers to unfair competition only in Art. 10bis. The TRIPS Agreement elaborates on this in Art. 39 by providing a definition of trade secrets, and formulating minimum standards of protection. See Agreement on Trade-Related Aspects of Intellectual Property Rights, 15 April 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, 1869 UNTS 299, 33 ILM 1197 (1994) (hereinafter ‘TRIPS Agreement’). The USA adopted a federal law on trade secrets in 2016: Defend Trade Secrets Act (DTSA) (Pub. L. 114–153, 130 Stat. 376, enacted 11 May 2016, codified at 18 USC § 1836, et seq.). The EU followed suit with Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the protection of undisclosed know-how and business information (trade secrets) against their unlawful acquisition, use and disclosure, OJ L 157, 15.6.2016, pp. 1–18 (hereinafter ‘EU Trade Secrets Directive’). It should have been transposed into national laws by June 2018. Likewise, in November 2017, China amended the ‘Anti-Unfair Competition Law of the People’s Republic of China’, including Art. 9 which deals with trade secrets misappropriation. The law came into effect on 1 January 2018 and now includes a definition of trade secrets as well as third party liability on employee misappropriation. The Japanese Unfair Competition Act, which includes both civil and criminal liability for trade secrets protection, was amended in 2015 and again in 2018 with a view to enhancing trade secrets protection. The Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) came into effect on 30 December 2018. While many provisions of Chapter 18 on intellectual property have been suspended by Art. 2 and Annex para. 7, the trade secret section (§18.78) has survived.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

244

Nari Lee

traditionally perceived to regulate unfair conduct.9 Using the EU Trade Secrets Directive as an example, this chapter explores how trade secret protection may turn from a liability rules regime into a property rules one when the marketing of ‘infringing goods’ is regarded as a misappropriation. This chapter first examines how intangible objects and tangible objects are conceptualized in intellectual property law and compares this regime to trade secret law. In comparison to intellectual property law, trade secret law does not have a strong concept of ‘intangible objects’. Thus, when such a concept is introduced, as seen in Articles 2(4) and 4(5) of the EU Trade Secrets Directive, it contests coherence within the law. This chapter concludes by recommending an explicit counterbalancing doctrine such as general good faith assignee exception, tied to border measures to prevent a negative impact on free movement of goods.

19.2 intangible objects and tangible goods in intellectual property and trade secret law 19.2.1 Intangible Objects Patents and copyrights protect innovative and creative endeavours as such. Trademarks protect investment in the creation of some intangible meanings represented by signs.10 These intangible objects are qualitatively defined in law, and the public is given notice of the title through various legal conventions such as disclosure and registration. Patents and trademarks are registered and disclosed so that the public is informed not only of the presence of the right, but also its substance (i.e. intangible object), represented in some form. Although copyright protection is not tied to any formal publication or notice, it is tied to the original works of expression. Communication of the copyrighted works to the public is so central to the creators aspect of copyright that it forms one of the core economic rights, and communication ultimately gives notice to the public of the substance of protected works.11 In sum, existence and substantive scope of intangible objects of intellectual property are communicated to the public, including potential infringers, ex ante. In contrast, trade secret protection seems to be a form of compensation for broken promises, or a punishment for wrongdoing or causing harm. Often trade secret 9

10

11

Guido Calabresi and Douglas A. Melamed, Property Rules, Liability Rules, and Inalienability: One View of the Cathedral, 85 Harv. L.R. 1089 (1972). See for a similar point that trademarks are a device for ‘market transparency’, Ansgar Ohly, Free-Riding on the Repute of Trade Marks – Does Protection Generate Innovation?, in Josef Drexl and Anselm Kamperman Sanders (eds.), The Innovation Society and Intellectual Property (Edward Elgar, 2019), at 146–159. On the international origin of the right, see Jane C. Ginsburg, The (New?) Right of Making Available to the Public, in David Vaver and Lionel Bently (eds.), Intellectual Property in the New Millennium: Essays in Honour of William R. Cornish (Cambridge University Press, 2004), at 234–247.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

Trading of Tangible Goods and Trade Secrets

245

protection is provided by unfair competition law, which targets commercially dishonest conduct of unfair competition. Before harmonization, some EU jurisdictions provided civil law remedies whereas others protected trade secrets by criminal law.12 Mostly, trade secrets seem to be protected by defensive rights against specific wrongdoers, who are given explicit or implicit notice of a duty of confidence or nondisclosure. Commercially dishonest conduct may be associated with an intangible object, but it is the wrongfulness of conduct that justifies the protection, not the copying of intangible objects. Moreover, not only do trade secrets not have categorically distinctive qualitative requirements, their communication is also discouraged, as non-disclosure is the key for their protection. In the TRIPs Agreement, a trade secret is defined as information that is ‘secret, has commercial value due to secrecy’ and has been subject to ‘reasonable steps’ of keeping it secret.13 The same wording is used in the text of Art. 2(1) of the EU Trade Secrets Directive, and these elements – secrecy, value and reasonable steps – are commonly found in national laws.14 Arguably, the reasonable steps of keeping the information secret at least give some notice about the existence of trade secrets to the relevant persons who deal with the information under secrecy measures.15 Nonetheless, neither the public nor even the relevant persons who may be subject to the liability of misappropriation are given specific notice of the scope, describing what exactly is protected, as non-disclosure is necessary to meet the secrecy requirement. Without notice of the substance, specific knowledge of the protected (intangible) subject matter is not communicated to the potential infringer. In other words, in trade secret law, the substantive communication of the protected intangible object can only be made ex post, at the time when there is misappropriation, and thus at the time of enforcement. In other words, there is no verification of existence nor delineation of its scope by law ex ante. To the contrary, the entity legitimately controlling the trade secret information may unilaterally

12

13 14

15

See for EU member state practice survey before the adoption of the EU Directive, Study on Trade Secrets and Confidential Business Information in the Internal Market (European Commission, 2013), http://ec.europa.eu/growth/content/study-trade-secrets-and-confidentialbusiness-information-internal-market-0_en (accessed 7 July 2020). TRIPS Agreement Art. 39.2. EC Trade Secret Study (2013), note 12 above. See also, for example, the USA’s DTSA in 18 USC § 1839(3). Japanese Unfair Competition Prevention Act (UCPA) Art. 2(6) requires secrecy (not known), kept secret and commercial utility. For example, Jeanne C. Fromer, Trade Secrecy in Willy Wonka's Chocolate Factory, in Rochelle C. Dreyfuss and Katherine J. Strandburg (eds.), The Law and Theory of Trade Secrecy: A Handbook of Contemporary Research (Edward Elgar, 2011), 3–17, at 14. Fromer noted that secrecy measures give notice by broadcasting ‘a message to both outsiders and insiders that the protected information ought not to be taken wrongfully’. See also M. Risch, Why Do We Have Trade Secrets, 11 Marq. Intell. Prop. L. Rev. 1 (2007). However, the requirement to keep the information secret seems to have become less stringent. Moreover, secrecy measures may not be considered a strong form of notice, as the notice given in this manner informs a potential infringer only of the existence of a secret and not of the substance.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

246

Nari Lee

claim its existence, by subjecting it to secrecy measures without verification and may impose a duty to respect it as such. This structure of trade secret protection highlights that trade secret protection is indeed a legal hybrid that falls between the property and the tort regimes. One consequence is that doctrines that are connected to the quality of an object, such as equivalences, similarity and identity, play little role in trade secret protection.16 In sum, the concept of ‘intangible object’ as subject matter of trade secret does not seem to be a part of the trade secret law. 19.2.2 Tangible Goods Intellectual property law protects an intangible object, disembodied from its tangible and concrete manifestation.17 However, tangible embodiments have a close qualitative connection both with respect to the requirement for protection and the confirmation of scope. Historically, intellectual property transitioned from actionfocused approach in pre-modern law to object-focused one in modern intellectual property law.18 The connection to the tangible may be explained by administrative and pragmatic necessity. As exclusive rights eventually allocate tangible resources, the administration of rights requires tools to connect rights to the resources and to monitor their uses.19 Moreover, pragmatically, the market value of the tangible goods may be considered in law, as the protected intangible good generates real-world value through tangible goods embodying the claimed subject matter, for which end users are willing to pay. Similarly, no matter how remote that connection may be in the digital age, creative works are appreciated by the public through communication and publication of the expressions which require some physical manifestation. Even the streaming of digital music requires some device to enable the reproduction of the digital signals and data that communicate the expression. In trademark law, the 16

17

18 19

See, however, Joseph P. Fishman and Deepa Varadarajan, Similar Secrets, 167 U. Pa. L. Rev. 1051 (2019), arguing that similarity should be a feature in trade secret law and proposing importation of similarity doctrine from copyright to US trade secret law. The invention of the legal fiction of ‘abstract’ intangible objects is explored in detail in, for example, Peter Drahos, A Philosophy of Intellectual Property (Ashgate, 1996), at 151–164. See also Brad Sherman and Lionel Bently, The Making of Modern Intellectual Property Law: The British Experience, 1760–1911 (Cambridge University Press, 1999), at 47–50; Alexander Peukert, Kritik der Ontologie des Immaterialgu¨terrechts (Geistiges Eigentum und Wettbewerbsrecht 134, Mohr Siebeck, 2018) (exploring different categories of abstract objects in intellectual property law in connection to communication practices). See generally, Sherman and Bently, ibid. On patented inventions, Pottage and Sherman observed that the legal construct of abstract objects originated from the need to represent them. They noted that disembodiment became possible ‘when patent lawyers and administrators had access to the specific material media in which inventions could be represented; namely, scale models, formalized texts, technical drawings, and archives’: Alain Pottage and Brad Sherman, Figures of Invention: A History of Modern Patent Law (Oxford University Press, 2010), at 2.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

Trading of Tangible Goods and Trade Secrets

247

connection is even closer as the sign has no meaning as a trademark unless it is connected to particular goods or services that the consumer can associate with the sign.20 Furthermore, when the production cost of the tangible medium decreases, the value of goods is often determined by the value of the intangible (i.e. the strength of exclusivity). Thus, the values of tangible embodiment and the intangible objects cannot easily be separated.21 Connection to the tangible manifests itself more clearly in the tracing of the liability for infringement. Liability for intellectual property infringement is often constructed such that certain economic conduct that may serve as a proxy for possession of an abstract object is prohibited in connection to the finding of a ‘reproduction’ of the protected intangible object in its tangible embodiment.22 Even in the cases of process or method inventions, proving that the protected method has been used may require some connection to the physical resources related to that method.23 As a result, enforcement often calls for balancing the property interests in the tangible goods against the property interests in the intangible object of intellectual property, as for example the doctrine of exhaustion shows. As the interests of those who have legally acquired tangible goods have to be balanced against the interests of intellectual property right holders, it is logical to have a doctrine that leads to the exhaustion of the specific intellectual property rights in the tangible embodiments at the first consented sale. The balance is also crucial in order to allow recouping the costs of invention, creation and qualitative goodwill through the consented sale of physical goods in the market based on which the right holders may collect licensing fees for the use of the intangible object. In contrast, trade secrets traditionally do not have strong ties to the tangible goods embodying them. Trade secret law is often treated as tort law or conduct regulation and thus its focus is on actions and conduct. Thus identical, similar or equivalent tangible objects do not have a strong significance for trade secret protection. 20 21

22 23

See e.g. Barton Beebe, The Semiotic Analysis of Trademark Law, 51 UCLA L. Rev. 621 (2003). Perversely, more often than not, the exclusive rights determine the real-life value of a product beyond the material production costs. See, for example, a similar argument made in Glynn Lunney, Copyright’s Excess: Money and Music in the US Recording Industry (Cambridge University Press, 2018), at 49. Material value may be the initial point connecting the values of the tangibles to the evaluation of the intangibles; they have become nearly inseparable, as the values of the intangibles are fed back into the production costs. Sherman and Bently, note 17 above, at 51–52. The patent eligibility procedure is often tied to utilization of hardware resources or materials, even when the claim is to a process or method – an activity rather than a product. Under US law, although not the only test employed, the so-called machine-or-transformation test has been routinely used in the patent eligibility procedure. See Bilski v. Kappos, 561 US 593 (2010). Indeed, physical connection may no longer be strictly applied, but to assert and thus to prove the use of a protected process, some level of connection to the protected process will become necessary. TRIPS Art. 34, which provides for a reversal of burden of proof and thus a presumption of identical product produced by the plaintiff’s process without contrary evidence produced by the defendant, in the case of process patents is one example where such connection is viewed necessary.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

248

Nari Lee

Producing tangible goods embodying trade secrets may be provided as one of many examples of the proscribed wrongful conduct, but it is not central in the tracing of liability, nor essential for finding a conduct unlawful. Scholars have even stressed that one of the benefits of trade secret protection is that its liability rules regime allows concurrent uses.24 As it is not constructed as a right in rem, concurrent use of the same information is allowed in trade secrets. Even in cases where the same information is used to create identical goods, if there is no unlawfulness in the acquiring, disclosing or using of the information, trade secret protection would not reach either the conducts or the products. For example, the definition of misappropriation in Art. 39(2) of the TRIPs Agreement clearly reflects this consideration, as it provides for a right to prevent unauthorized and commercially dishonest disclosure, acquisition and use, and not a right to exclude. The EU Trade Secrets Directive goes a step further as Recital 16 states that the Directive should not create any exclusive right to know-how or information protected as trade secrets. The Directive extends primary liability to the unlawful acquisition, i.e. ‘unauthorised access to, appropriation of, or copying of any documents, objects, materials, substances or electronic files, lawfully under the control of the trade secret holder, containing the trade secret or from which the trade secret can be deduced’, or other forms of commercially dishonest acquisition,25 use and disclosure of trade secrets, without the consent of the trade secret holder, by a person who either unlawfully acquired them or acquired them against a duty of confidence or other duties limiting their use.26 Unauthorised access and appropriation of tangible goods such as objects, materials or substances is mentioned as one example of ‘unlawful acquisition’. However, as there is no stand-alone concept of protected intangible object, tangible goods themselves may not be ‘infringing’ as such, unless their acquisition is unlawful.27 This reflects the fact that secret information may not even be found in the tangible objects that are resulting from the use of trade secrets – unless protected trade secrets are related to product design or technology (i.e. technical secrets, manufacturing secrets). In case of secret business information such as customer lists, tangible objects may have little or no trace of protected trade secrets. Trading of tangible goods which contain similar or identical information as protected by the trade secret alone may not even be within the scope of trade secrets’ primary liability of misappropriation. One consequence of the lack of connection is that there is no need to balance the ownership interest in the tangible goods against that of the trade secrets holder. 24 25 26 27

See Reichman, note 2 above. EU Trade Secrets Directive, Art. 4.2(a) and (b). Ibid., Art. 4.3. This aspect of trade secret law has led some commentators to claim that it is a parasitic right. See Robert G. Bone, A New Look at Trade Secret Law: Doctrine in Search of Justification, 86 Cal. L. Rev. 241 (1998), at 245.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

Trading of Tangible Goods and Trade Secrets

249

Indeed, trade secret law does not have a doctrine of exhaustion, which categorically excuses the trader of genuine goods placed in the market with the consent of the right holder. Traditionally, the doctrine of exhaustion may have little justification in trade secret law, as the protection does not extend to all tangible goods that embody similar or even identical information as the trade secrets. As concurrent use is allowed under trade secret law unless there is a clear case of misappropriation, the presence of identical information in the tangible goods does not mean that there is trade secret misappropriation. However, in the exceptional case where there is a clear connection between the tangible objects and the trade secrets, as in the case where the objects have been derived from the trade secrets, there may be a necessity or even possibility to trace the liability of misappropriation via tangible objects. The need to balance the interests of conflicting rights to tangible goods and to trade secrets may arise when the tangible item is used as a way to trace the liability, which may be the case in the liability under Art. 4(5) of the Directive.

19.3 trading of ‘infringing goods’ Intellectual property uses a concept of intangible objects connected to tangible objects to trace primary liability for infringement. Indeed in patent law, primary liability is often explained as strict liability, which ‘requires no knowledge or intention on the part of the alleged infringer, whose state of mind is wholly irrelevant’ to infringement.28 Knowledge and intent are often considered subjective requirements in intellectual property infringement liability, often associated with secondary or third party liability. However, unless there is an identical or similar intangible object found at some point in the value chain (buyer, supplier or end user) implying actual or constructive direct infringement, subjective elements of knowledge and intent alone may not mean infringement. In contrast, misappropriation of trade secrets, as seen above, is often constructed as a subjective wrong. The production or marketing of identical or similar objects alone does not lead to primary liability for misappropriation. The mental state of the wrongdoer – in particular after receiving the notice of trade secrets which have been kept secret by the trade secret holder – is important even for the primary liability of trade secrets misappropriation. In the TRIPs Agreement, footnote 10 to Art. 39(2) requires the member states to extend liability to third parties, as at minimum, liability should extend to ‘the acquisition of undisclosed information by third parties who knew, or were grossly negligent in failing to know, that such practices were

28

Citation is to the UK’s Supreme Court, Lord Neuberger, Vestergaard Frandsen A/S and others v. Bestnet Europe Limited and others [2013] UKSC 31, at para 37. This type of statement is found in the majority of patent textbooks.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

250

Nari Lee

involved in the acquisition’. However, it does not extend such liability to those who are importing or exporting the tangible goods. The EU Trade Secrets Directive goes further. In Art. 4(4), the liability of third parties extends not only to the acquisition, but also to use and disclosure of the trade secrets, subject to an actual or constructed knowledge requirement.29 While this is clearly an expansion of the TRIPs obligation under Art. 39(2), these subjective elements do not deviate much from the conduct-oriented focus of traditional trade secrets regimes. More importantly, a new type of liability is now imposed. Art. 4(5) reads: ‘The production, offering or placing on the market of infringing goods, or the importation, export or storage of infringing goods for those purposes, shall also be considered an unlawful use of a trade secret where the person carrying out such activities knew, or ought, under the circumstances, to have known that the trade secret was used unlawfully within the meaning of paragraph 3.’ It imposes liability on the knowing traders of ‘infringing goods’, defined as ‘goods, the design, characteristics, functioning, production process or marketing of which significantly benefits from trade secrets unlawfully acquired, used or disclosed’.30 This definition of infringing goods makes it necessary to determine what such a significant benefit would be in the case of directly or indirectly obtained trade secrets. Moreover, as ‘marketing’ is included in the definition of the benefit, the notion seems to go beyond technical features that are directly derived from the trade secret, and includes business secrets. This is problematic, as it requires a judgement of significance in areas which may seem directly unrelated to the goods themselves and introduces more uncertainty. For complex-system goods, such as consumer electronics, mixed goods may be traded embodying a technical secret in some minor part. There may be cases where marketing efforts are made using business secrets only in part, or to market perfectly legal products. These cases of mixed goods require a kind of proportionality analysis for technical or business significance of trade secrets in comparison to other factors.31 The combined reading of Art. 2(4) and Art. 4(5) suggests that the liability for traders of infringing goods could be quite broad. As the application of Art. 4(5) is not limited to technical or manufacturing secrets, it can potentially be applied to all types of business data and information. Moreover, all tangentially related products, such as goods using technical or business secrets in parts, could be included. While

29 30

31

Art. 4(4) of the EU Trade Secrets Directive. Art. 2(4) of the EU Trade Secrets Directive, emphasis added. See for a discussion of various versions of the Directive, Tanya F. Aplin, A Critical Evaluation of the Proposed EU Trade Secrets Directive (18 July 2014), King's College London Law School Research Paper No. 201425, available SSRN: https://ssrn.com/abstract=2467946 or http://dx.doi.org/10.2139/ssrn.2467946. For a concerned comment on the infringing goods misappropriation, see Richard Arnold, Lionel A. F. Bently, Estelle Derclaye and Graeme B. Dinwoodie, The Legal Consequences of Brexit through the Lens of IP Law, 101 Judicature 65 (2017).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

Trading of Tangible Goods and Trade Secrets

251

indirect patent infringement liability similarly extends liability to partially knowing users, it is limited to third parties who provide means relating to an essential element for putting the protected invention into effect.32 The test for determining when a product is infringing for patents thus always requires objective analysis, whilst the test for trade secret misappropriation is mostly subjective. The liability for infringing goods under Art. 4(5), however, would introduce objective elements, such as how similar or dissimilar tangible goods are, as in other areas of intellectual property, and to what extent differences would matter in the understanding of ‘significantly benefits’. As trade secret law does not have conceptual tools to delimit the scope of liability for the marketing of infringing goods, either limitation should be set clearly on the infringing conduct or the category of relevant type of trade secrets should be limited. In the case of technical or manufacturing secrets tracing is possible through derivation, hence the liability may be limited by facts. But in the case of business secrets, liability may extend to the marketing of goods which lack an inherent connection to the trade secret. This highlights that there is a clear need to limit the scope of ‘infringing goods’ in the implementation of Art. 4(5). In the current text, the only limitation is the knowledge requirement. As the Directive is unclear about the level of ‘knowledge’ required of the infringer, and since there is no case law so far, we cannot be sure yet what the extent of knowledge and the notice required for the right holder to prevail against the trader of goods potentially embodying trade secrets would be. Whilst Art. 3(1)(a) and (b) provide lawfulness of acquisition by independent discovery or creation, and reverse engineering, using such lawfully acquired information33 is not explicitly exempted from the liability. Furthermore, with Art. 4(4), without an explicit exception, trading of goods as such may constitute unlawful ‘use or disclosure’, if a member state would define commercial use of reversely engineered trade secrets information as unlawful34 or if the right holder restricts it.35 Comparatively, in countries where similar wrongs exist, the wrongs are limited to strictly technical secrets, and there are strong good faith defences. Indeed, exceptions or defences for the good faith purchasers of tangible goods may in effect 32

33 34

35

Compare, for example, Agreement on a Unified Patent Court (UPC), Official Journal of the European Union, C 175, 20 June 2013, which provides in Art. 26: ‘A patent shall confer on its proprietor the right to prevent any third party not having the proprietor’s consent from supplying or offering to supply . . . with means, relating to an essential element of that invention, for putting it into effect therein, when the third party knows, or should have known, that those means are suitable and intended for putting that invention into effect.’ EU Trade Secrets Directive, Art. 3. Compare EU Software Copyright Directive, which takes this position on the reversed engineered SW codes. See Art. 6.2 of Directive 2009/24/EC of the European Parliament and of the Council of 23 April 2009 on the legal protection of computer programs (Codified version) (Text with EEA relevance), OJ L 111, 5.5.2009, pp. 16–22. The Directive, however, provides for a first sale exhaustion doctrine for distribution right, which may function as a general good faith purchaser’s exception. EU Trade Secrets Directive, Art. 4.3(c).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

252

Nari Lee

function as a trade secrets exhaustion doctrine for both technical and business secrets. For example in Japan, third party liability for the importers and exporters of products that result from trade secrets is limited to technical information, and the legislative history shows that the liability originally meant to cover object code of digital products produced by using secret source code.36 The Japanese statute also limits its scope by the expression ‘produced by’ which implies that liability is limited to direct results of the use of the trade secret, and good faith purchasers of the goods are explicitly excused. The trading of ‘infringing’ goods as well as of derivative goods that may embody trade secrets presents challenges to international trade and to the free movement of goods in the internal market. The paradoxical situation may arise that all intellectual property rights with respect to a traded product are exhausted, but that it still contains a trade secret which is not. A more flexible interpretation of the Directive may be called for. For example, it is possible to read into Art. 3(1) (d) of the Directive an explicit and general good faith purchaser exception, as it allows practices which are in conformity with ‘honest commercial practices’. It is also possible to read from Art. 3(2), which allows EU law based exception, exploration of a limitation such as doctrine of exhaustion in trade secrets embodying goods. Furthermore, as infringing goods necessarily influence the traders of goods in the internal market, the enforcement of this particular liability may have an influence on border measures, which the Directive does not deal with.

19.4 concluding remarks As seen above, the EU Trade Secrets Directive may now impose liability on openly traded tangible goods. Tracing the liability for a trade secret by tracing the tangible goods embodying or related to the trade secret may have unintended consequences. As the rationale for protection appears to be based on consideration for fairness in competition (i.e. commercial honesty) rather than objective qualitative values, trade secret protection seems to be versatile enough to provide a direct investment protection without the cost and burden of registering the right. As seen above, in an intangible and tangible goods based property rules regime such costs and administration may be necessary. Tracing of liability based on infringing goods may create such need for the administration of trade secrets, which may remove one of the benefits of using trade secrets. There are further policy concerns for using tangible objects as the basis for tracing liability. It has been noted by scholars that by keeping the know-how related to a particular embodiment of patents secret it is possible to extend the exclusivity of 36

Japanese Unfair Competition Prevention Act, Art. 2(1)10.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

Trading of Tangible Goods and Trade Secrets

253

patents even after the expiry of patent or in parallel to patents.37 Moreover, as trade secrets liability often includes criminal remedies, any expansion of trade secrets liability should be approached with caution. Harmonization done without heeding such essential differences may create more uncertainty. As the CJEU would be asked to interpret many of the open definitions and concepts that the Directive uses, there is also a concern that the Directive may end up creating a sui generis protection of trade secrets tied to infringing goods, as a separate species of quasiexclusive autonomous right protected in the EU, despite the contrary declaration in Recital 16 of the EU Trade Secrets Directive. One of the main benefits of the trade secrets regime as opposed to the intellectual property regime is that it is a ‘loosely constructed liability regime’.38 The benefit of having a loosely constructed liability regime is the possibility of mediating between exclusive rights and contractual rights. As the substance of trade secrets is not examined or communicated ex ante, when trade secrets are not found in openly traded goods – as is the case with business secrets – trade secret protection may be used to create a property rules regime ex post. It then no longer mediates between exclusive rights and contractual rights.

37

38

Lionel Bently, Patents and Trade Secrets, in S. Basheer and N. Wilkof (eds.), Overlapping Intellectual Property Rights (Oxford University Press, 2012), 57–82. Reichman, note 2 above, at 2439.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.028

20 From Smörgåsbord to New Nordic Cuisine: EUHarmonization of Trade Secrets Protection in the Nordic Countries Jens Schovsbo1

20.1 menu Prior to the EU Directive on the protection of trade secrets (TSD),2 trade secrets protection in the EU was like the result of a visit to a traditional Nordic smörgåsbord: Even though all had picked their dishes from the same buffet, everyone ended up with different meals. By way of example, Denmark, Finland and Norway have traditionally protected trade secrets via the provisions in their (rather similar) national Marketing Practices Acts (i.e. “unfair competition law”) and Penal Codes, whereas Sweden has had for a long time (and as the only EU country) an Act dedicated to the protection of trade secrets.3 With the adoption of the TSD this is going to change. The number of available dishes has been reduced and countries have to pick and choose within the selections found in the Directive. This, however, does not mean that all national fares are 1

2

3

Professor, LL.D., Ph.D., Center for Information and Innovation Law (CIIR), University of Copenhagen. Thanks to Vibe Ulfbeck and Thomas Riis, both of the University of Copenhagen, for discussions and comments; and to Ansgar Ohly for comments on an earlier draft. Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the protection of undisclosed know-how and business information (trade secrets) against their unlawful acquisition, use and disclosure, OJ L 157, 15.6.2016, 1. On the implementation of the TSD in the Nordic countries, see Nordiskt Immateriellt Ra¨tsskydd (2018), 463 (Denmark by Thomas Riis, Jens Schovsbo and Henrik Udsen), 486 (Finland by Niklas Bruun and Marcus Norrgård), 515 (Norway by Harald Irgens-Jensen) and 538 (Sweden by Bengt Domeij). See also, for an analysis in English, Jens Schovsbo and Niklas Bruun, The Implementation of the Trade Secrets Directive in the Nordic Countries, in J. Schovsbo, T. Minssen and T. Riis (eds.), The Harmonization and Protection of Trade Secrets in the EU: An Appraisal of the EU Directive (Edward Elgar, 2020). See generally about the legal situation in the EU prior to the TSD, the Study on Trade Secrets and Confidential Business Information in the Internal Market (Final Study, April 2013. Prepared for the European Commission, Contract No. MARKT/2011/128/D. Available at file:///C:/Users/ Tmc889/Downloads/130711_final-study_en.pdf).

Downloaded from https://www.cambridge.org/core. 254 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

From Smörgåsbord to New Nordic Cuisine

255

going to be identical. Rather than a smörgåsbord, the differences can be likened to the New Nordic Cuisine.4 According to this, the territoire is of central importance and chefs should do their utmost to use local ingredients to prepare their “traditional” dishes. With the passing of the new Acts, a Nordic judge is going to find herself in a position where she too has to “combine local self-sufficiency with regional exchange of high-quality goods”.5 Like the chef, a judge will have to go into the local environment surrounding her court and find local ingredients; not live lemon ants,6 of course, but other equally exotic legal principles can no doubt be found outside of most court rooms. As good as this may be in the world of fine dining, too much local selfsufficiency is hard to digest when it comes to the harmonization of the laws of EU Member States. With a New Nordic Cuisine approach, companies and citizens are going to have a different meal in all countries even though they are looking for the same dish. These difficulties, which are new to the area of trade secrets law, are of course very well known to EU law. In particular, they have long been of special interest to Intellectual Property Rights (IPRs), where internationalization came early and EU harmonization has come far. Here it is well known that the transition of the law as it appears in the EU statutes to the law in action is always hard and unpredictable, not least because of the interaction with local rules or the application of national practices. Important nuances may be lost or solutions provided for in the common system overridden by overlapping local norms. At the same time, there is a risk that over-eagerness to harmonize may lead to the discarding of national solutions which are worth keeping. After all, national tastes vary. Few scholars have written more intensely on legal harmonization than Annette Kur. In this essay honoring Annette, I will focus on the newly adopted TSD, as it provides good examples of the questions which may arise in the translation of EU norms to the national level in the borderline area of unfair market law which seems to be particularly problematic. To further draw on an area of Annette’s expertise, I use examples from some of the Nordic countries. Finally, the food allegory hopefully reminds Annette of the many nice meals we have shared over the years; in the Nordic countries and beyond.

4

5 6

The New Nordic Cuisine was introduced in the early 2000s and led to a number of successful restaurants, including “Noma” in Copenhagen, which was ranked as the best restaurant in the world in 2010, 2011, 2012 and 2014 (see Wikipedia). The movement is based (at least in part) on a Manifesto consisting of ten principles – see e.g. https://www.meyersmad.dk/english/about/thenew-nordic-cuisine-movement/ and http://www.newnordicfood.org/. Point 9 of the Manifesto for the New Nordic Cuisine, ibid. https://www.telegraph.co.uk/foodanddrink/foodanddrinknews/11379809/Worlds-best-restaurantNoma-serves-live-ants.html.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

256

Jens Schovsbo

20.2 starters: floors and ceilings in the tsd According to TSD Art. 1, the Directive “lays down rules on the protection against the unlawful acquisition, use and disclosure of trade secrets”. Art. 2 defines a trade secret as information which is secret, has commercial value because it is secret and has been the subject of reasonable steps to keep it secret. All three prongs have to be satisfied. In order to become binding on individuals, directives normally have to be implemented into national law (TFEU Art. 288). In this respect, TSD Art. 1 states that Member States may provide for more far-reaching protection against the unlawful acquisition, use or disclosure of trade secrets than that required by the Directive. This makes it clear that the TSD provides for minimum protection. Minimum obligations establish the floor of the protection.7 States may deviate from such provisions in the Directive but only in order to protect more. The freedom to protect more, however, is limited by the obligation to “secure compliance” with a number of safeguards, including Art. 3 (on lawful use, etc.) and Art. 5 (exceptions), and on procedural matters. These rules impose mandatory limitations on the protection of trade secrets and in this way constitute ceilings.8 Other important steers on the EU countries’ implementation flow from general principles of EU law, including the principle of the effectiveness of the EU legal order (effet utile).9 Following this principle, the Court of Justice of the European Union (CJEU) has described EU law as “a structured network of principles, rules and mutually interdependent legal relations”,10 and has stated that terms found in EU norms which make no express reference to the law of the Member States should normally be given an independent and uniform interpretation throughout the European Union.11 Therefore, the central concepts of the TSD – including “trade secret” – should be regarded as autonomous concepts of EU law. For such concepts, the freedom of legislators to deviate from the model found in the Directive or for national courts to continue to rely on pre-existing national conventions or principles is very limited. These concepts represent elements of complete (or, full) 7

8 9

10 11

See on the terminology Annette Kur and Henning Grosse Ruse-Khan, Enough is Enough – The Notion of Binding Ceilings in International Intellectual Property Protection, 359 ff, in Annette Kur and Marianne Levin (eds.), Intellectual property Rights in a Fair World Trade System (Edward Elgar, 2011). Kur and Grosse Ruse-Khan, ibid., 359. See e.g. V. Cassiers and A. Strowel, Intellectual Property Law Made by the Court of Justice of the European Union, in C. Geiger et al. (eds.), Intellectual Property and the Judiciary (Edward Elgar, 2018), 175 –206, at 177 et seq. The principle of proportionality also restricts the wriggle room of national states beyond the exact wording of EU legislation, e.g. Case C-467/17, Funke Medien v. Bundesrepublik Deutschland, ECLI:EU:C:2019:623, para. 47. Case-621/18, BREXIT, ECLI:EU:C:2018:999, para. 45. BREXIT case, ibid., para. 47 and Case C-467/08, Padawan SL v. Sociedad General de Autores y Editores de España (SGAE), ECLI:EU:C:2010:620, para. 32; and for more examples, Cassiers and Strovel, supra note 9.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

From Smörgåsbord to New Nordic Cuisine

257

harmonization (or to put it another way, they constitute a floor and a ceiling at the same time). Here no freedom is left for national deviations, not even to “protect more”, for example by lowering the threshold for what constitutes a trade secret.12 These principles are enshrined in the Preamble to the TSD, which stresses the importance of establishing a homogenous definition of a trade secret (Recital 14) and makes it clear under which circumstances legal protection of trade secrets is justified (Recital 15). Outside of the scope of the TSD, Member States and courts are free to provide for or apply national solutions. However, defining the boundaries of an EU obligation is often hard, not least because the CJEU has stated that limitations on national freedom are not restricted to the harmonized areas strictu sensu but apply to the “national legal system” in a broader sense. Thus, within the area of trademark law, the CJEU ruled in Martin y Paz (and in the words of Annette Kur) “that national unfair competition law may not be employed to counteract a solution found on the basis of harmonized trademark law”.13 Here the harmonization obligations which stem from the Trademark Directive had a spillover effect into the realm of (otherwise mostly) unharmonised unfair competition law. In a similar vein, the Court has explained that within the scope of the total harmonization obligations of the Product Liability Directive,14 States could not maintain “a general system of product liability” which is different from that of that Directive.15 Again, there is a spillover (or ripple) effect from the harmonization obligation not to disrupt or sidestep the general system represented by the Directive.

20.3 main course: the implementation of the tsd in the nordic countries As long as Member States and national courts stay within the “house of the TSD” as it is defined by the floor and ceiling of the TSD and the general principles of EU law described above, they are free to devise and apply their own legal principles. However, within the areas of complete harmonization or those covered by the mandatory limitations, there is very little room to maneuver. Member States cannot decide to, for example, disregard one of the prongs in the definition of a 12

13

14

15

See from the realm of copyright law Case C-310/17, Levola Hengelo BV v. Smilde Foods BV, ECLI:EU:C:2018:899, para. 37: “only something which is the expression of the author’s own intellectual creation may be classified as a ‘work’ within the meaning of Directive 2001/29”. See Annette Kur, Trade Marks Function, Don’t They? CJEU Jurisprudence and Unfair Competition Principles (2014) IIC 434, especially at 452 writing on the judgment of the CJEU in Case C-661/11, Martin y Paz Diffusion SA v. David Depuydt and Fabriek van Maroquinerie Gauquie NV, ECLI:EU:C:2013:577. Council Directive 85/374/EEC of 25 July 1985 on the approximation of the laws, regulations and administrative provisions of the Member States concerning liability for defective products, OJ L 210, 7.8.1985, 29. Case C-402/03, Skov Æg v. Bilka Lavprisvarehus A/S, ECLI:EU:C:2006:6, para. 39.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

258

Jens Schovsbo

“trade secret” in Art. 2 or outlaw reverse engineering required by Art. 3, not even if the result is to strengthen the protection offered to the trade secrets holder. The freedom to “protect more” is only open as long as it exceeds on the system established by the TSD without changing its basic parameters.16 Complications may arise when Member States decide to grant protection in areas which are not clearly covered by the scope of the TSD, for example by (i) including information which does not qualify as a trade secret under the protection of the national trade secrets Act, or (ii) offering additional protection, via unfair competition law for instance.

20.3.1 Local Specialties: Entrusted Technical Drawings, etc. Denmark and Finland have traditionally protected trade secrets primarily via specifically dedicated provisions and the general clause in their Marketing Practices Acts.17 Despite the lack of statutory definitions, all countries defined “trade secrets” more or less in accordance with the definition contained in what is now TSD Art. 2. However, in addition to the protection of trade secrets, these countries also protected technical drawings, etc. which had been “entrusted” to someone against unfair uses even if those drawings did not contain “trade secrets”.18 For this category of information, protection was not triggered by the nature of the information but rather by the nature of the conditions under which the information had been shared and the relationship between the parties. Finnish courts have relied on the provisions on technical drawings in several cases (unlike in Denmark).19 The new Danish Trade Secrets Act has arguably20 not continued the additional protection for technical drawings. However, the Finnish Act has kept the rule but made it clear that it does not apply to “trade secrets” (which are defined as by TSD Art. 2).21 This raises the question whether the TSD allows for the Finnish special rule which continued the traditional protection. On the one side, one could argue that such a deviation falls within the wriggle room left by the principle of minimum harmonisation which allows Member States to “protect more”.22 This argument, however, is hardly convincing. As explained above, the freedom to increase protection should thus be seen within the system 16

17 18

19 20

21 22

By way of example, Denmark was no doubt within the wriggle room when it decided to exceed protection by making grave infringements of trade secrets (as defined by the TSD) punishable under the Penal Act with imprisonment of up to six years; see supra note 3. See supra note 3. Supra note 3. The model was inspired by the German Unfair Competition Act, which had a provision to this effect, dating back to 1896 (“Vorlagenfreibeuterei”). Schovsbo and Bruun, supra note 3. But see differently Palle Bo Madsen, Markedsret – Del 2 Markedsføringsret og Konkurrenceværn (DJØF Publishing, 2019), 310. Bruun and Norrgård, supra note 3. This is the justification given in Finland; see Bruun and Norrgård, ibid.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

From Smörgåsbord to New Nordic Cuisine

259

established by the TSD for the protection of “trade secrets”, and arguably just opens for an increase of the level of protection within the space defined by Art. 2. Since it seems to be clear that the Finnish provision is limited to information which does not constitute trade secrets, this justification does not apply. This leads to the next and more complicated question, viz. whether the Finnish special protection falls outside of the scope of the TSD. If so, the Finnish government is at liberty to uphold the rule. As it has been pointed out above, the “scope” should not be defined in a narrow way and should in any event only be determined after a careful examination of the “general system” which has been established by the total harmonisation obligation. Therefore, an argument along the lines that since the Finnish rule does not pertain to “trade secrets” as such it is outside of the scope of the TSD altogether falls short. Instead, the analyses should focus on identifying the “general system” which has been established by the TSD. Recital 14 reads: It is important to establish a homogenous definition of a trade secret without restricting the subject matter to be protected against misappropriation. Such definition should therefore be constructed so as to cover know-how, business information and technological information where there is both a legitimate interest in keeping them confidential and a legitimate expectation that such confidentiality will be preserved. Furthermore, such know-how or information should have a commercial value, whether actual or potential. Such know-how or information should be considered to have a commercial value, for example, where its unlawful acquisition, use or disclosure is likely to harm the interests of the person lawfully controlling it, in that it undermines that person’s scientific and technical potential, business or financial interests, strategic positions or ability to compete. The definition of trade secret excludes trivial information and the experience and skills gained by employees in the normal course of their employment, and also excludes information which is generally known among, or is readily accessible to, persons within the circles that normally deal with the kind of information in question.

The Recital stresses the importance of a “homogenous” definition of a trade secret. This clearly makes it very problematic if states apply definitions which differ from the one found in Art. 2. Therefore, had the Finnish Trade Secrets Act included the protection of technical drawings as part of the definition of “trade secrets”, this would most likely have conflicted with the obligation under the TSD. Importantly, however, the Recital not just makes the case for total harmonization of the concept of a trade secret. It also singles out “trade secrets” as the relevant connecting point for the system as such. It is thus exactly for the category of information which qualifies as “trade secrets” that there exists both a legitimate interest in keeping them confidential and a legitimate expectation that such confidentiality will be preserved. It is for this reason that information which is, for example, trivial, generally known or readily accessible is excluded from protection. In this way, it is clear that the purpose of the definition of a trade secret is not just to indicate what is protected. It is also to indicate what is not. Therefore, the Finnish decision to offer Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

260

Jens Schovsbo

protection to technical drawings, which may well extend protection to cover information which has been deliberately left out of protection by the definition in Art. 2, appears to be problematic vis-a-vis the system established by the TSD. For similar reasons, the Finnish rule is seemingly also at odds with Recitals 15 and 16 (mentioned above), according to which it is just for information which qualifies as trade secrets that the protection scheme established by the TSD is “justified”. These remarks cut both ways; they make the case both for restricting access and use of certain types of information (i.e. trade secrets) and for not restricting access and use of other types of information. The importance of open innovation as a “catalyst for new ideas” is even stressed explicitly in Recital 3. By opening the door to protection of information which does not qualify as a trade secret, the Finnish provision surely cannot rely on the legitimacy provided by the TSD. Furthermore, the expansion of the protection to cover information which does not qualify as trade secrets may even be said to conflict with the general system established by the TSD, as it not only opens the way for the protection of sub-secret information, but also introduces a justification for the protection which is not based on the inherent qualities of the information but instead on the nature of the relationship between the parties when the information was shared. It will most likely eventually fall upon the CJEU to decide whether the TSD allows Member States to diversify protection in this way. It is hard to predict the outcome of such a case, but for the reasons indicated above, it seems most likely that the Court would find that the rule goes beyond the wriggle room left by the TSD. 20.3.2 Surf and Turf: Supplementary Protection via the Marketing Practices Acts? In addition to the special rules in the Marketing Practices Acts, the Nordic countries have traditionally offered supplementary protection via the general clause in these Acts, which prohibits traders from engaging in “unfair trading practices”.23 This raises the question whether and to what extent the TSD has pre-empted the use of the general clause regarding trade secrets. TSD Recital 17 addresses the relationship to “the law on unfair competition” (for the Nordic countries this means their Marketing Practices Acts): In some industry sectors, where creators and innovators cannot benefit from exclusive rights and where innovation has traditionally relied upon trade secrets, products can nowadays be easily reverse-engineered once in the market. In such cases, those creators and innovators can be victims of practices such as parasitic copying or slavish imitations that free-ride on their reputation and innovation efforts. Some national laws dealing with unfair competition address those practices. While this Directive does not aim to reform or harmonise the law on unfair competition in 23

Supra note 3.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

From Smörgåsbord to New Nordic Cuisine

261

general, it would be appropriate that the Commission carefully examine the need for Union action in that area.

The Recital is not overly precise. Assuming, however, that it does give input to the interpretation of the norms of the TSD (and is not just an instruction to the Commission), it no doubt recognizes that unfair competition law may be used to supplement trade secrets protection. Recital 17 also makes it clear, however, that this area is not unlimited. The reference to protection against “parasitic copying or slavish imitations” which free-ride on “easily reverse-engineered” products indicates the area for unfair competition law to be related to restrictions on the use of already obtained information. Restrictions on the obtaining of such information are not foreseen by the Recital. Such a reading is also supported by the CJEU’s decisions in Martin y Paz (see 20.2 above). Here the Court ruled that national unfair competition law may not be employed to counteract solutions found on the basis of harmonized law. Therefore, ways of obtaining information which are explicitly allowed for by mandatory provisions in the TSD (in particular Art. 3, including reverse engineering) cannot by themselves be banned via unfair competition law as this would “upset” the balance of the TSD. This puts a steer on the national application of the general clause. It does not, however, exclude its application, but it stresses that assessment in unfair competition law should be subscribed under the solutions prescribed by the TSD. In the words of Matthias Leistner (writing on the interface between unfair competition law and the Database Directive), the general clause may be applied but only to the extent that the unfair competition application is based on genuinely additional aspects of the case (such as risk of confusion) which have not been comprised by the balancing of interests on which the EU legislator based the protection and its limitations.24 To some extent this is also reflected in the TSD, which would seem to open for contractual limitations on reverse engineering, but only for products which have not been made available to the public (TSD Art. 3(1)(b)). For mass-market products contractual limitations on reverse engineering apparently are not possible.25 Nor may such activities be limited via unfair competition law. What then about unprotected information? The TSD does not imply any direct obligations on to what extent national courts may rely on their Market Practices Act regarding information which does not constitute a trade secret. Indirectly, however, the TSD may have a ripple effect. It would, for instance, be hard to argue that reverse engineering or whistleblowing, which are lawful regarding trade secrets, should be unfair when it comes to information which does not even qualify as a 24

25

Matthias Leistner, Big Data and the EU Database Directive 96/9/EC: Current Law and Potential for Reform (September 7, 2018), available at SSRN: https://ssrn.com/abstract= 3245937 or http://dx.doi.org/10.2139/ssrn.3245937, 17 with references. But see differently Wilhelm Schröder, Transformation from a Patchwork Quilt to a Unified Fabric: Discussion on a Few Particular Aspects of the New Finnish Trade Secrets Act and the EU Trade Secrets Directive, Nordiskt Immateriellt Ra¨ttskydd (2018), 505 f.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

262

Jens Schovsbo

trade secret. Moreover, even the overall harmonisation objective of the TSD may restrict the freedom of Member States to rely on their Marketing Practices Acts. To the extent that unfair competition law is relied upon to create a second tier for the protection of sub-trade secrets information, this may at some point amount to a “counteraction” of the general system of trade secrets protection found in the TSD, which aims not just at protecting certain information, but – and equally important – at leaving non-protected information free from restrictions (see 20.3.1 above).

20.4 dopo The New Nordic Cuisine where well-known dishes are given a local flavor but are still (somewhat) recognizable has served Nordic restaurants and foodies from around the world well. In the legal field the approach is to some extent supported by the general principle of subsidiarity found in TFEU Art. 5(3), which aims to strike a balance between the national and the regional interests and the need for uniformity versus the respect for national traditions. Defining the room to maneuver for national legislators and courts is very complicated. As has been pointed out above, to do so in the area of the TSD involves a number of difficult principles which are hard to align; some constitute “floors”, others “ceilings”, and some even both at the same time. It adds considerably to the difficulties that the obligation to secure the appropriate level of harmonization is not limited to specific areas of the law but and affects broader parts of the national legal systems. It falls outside the scope of the current contribution to try to provide more general answers to these general problems. To some extent the difficulties are the unavoidable consequences of the whole process of EU legal harmonization. It is worth pointing out, however, that as far as the relationship between the TSD and the national rules on unfair competition law is concerned, the system established by the TSD does actually seem to be rather well suited for finding the boundaries and avoiding under- or over-protection or erosion of the safeguards found in the TSD. The non-proprietarian base line of the protection arguably invites a natural balancing of interests and this is, furthermore, underscored by the combination of Art. 3 on lawful acts and Art. 4 on unlawful acts. Add to this a not overly broad, yet precise and even relatively open inclusion of the EU Charter, and the TSD – at least on paper – seems to contain the ingredients which skillful chefs/judges need to make dishes which are both digestible and acceptably recognizable, yet distinct.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.029

21 Trade Mark Rights and Parallel Imports vis-à-vis the Never-Ending Evolution of the Behavior of Firms: Transition and Coherence Put to a Test Vincenzo Di Cataldo1

21.1 different markets, different trade marks, different prices: parallel import and exhaustion principle to fight off the price differential It has been, and is, quite frequent practice for a firm, producing goods directed at different national markets, to label them with different trade marks. Sometimes there is no alternative to this practice, as when the trade mark registered in one country cannot be used in other countries. This might arise, for example, when a trade mark consisting of one word has, in the language of other countries, a commercially unacceptable meaning, or when a trade mark used in one country is already in the hands of somebody else in other countries or is susceptible of confusion with a preexisting trade mark in other countries. Sometimes this practice is only the fruit of specific events, like mergers of previously independent structures. Sometimes the firm intentionally seeks to keep separate markets, so as to exploit the possibility of different prices and thus achieve globally higher profits. Even when the firm labels all its products, directed at different countries, with the same trade mark, it frequently decides to keep the various markets separate. This corresponds to the general interest of the firm in monitoring, understanding and interpreting separately the various markets – their tendencies, their needs, their wishes. In such a way, the firm can better react to the possibly different evolutions of the markets. Again, this leads to the possibility of different prices. Different prices for the same product in different countries is not a good thing from the point of view of the interests of the market – at least, if these countries try to establish a “common market”, as in the European Union. But this is maybe one of the cases to which a “healthy” market can adequately react without turning to authoritative interventions. If there is a substantial price differential for the same 1

Professor of Commercial Law, University of Catania.

Downloaded from https://www.cambridge.org/core. 263 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

264

Vincenzo Di Cataldo

product in two markets, somebody will try to buy products in the market where the price is lower and transfer the goods into the market where the price is higher. This is what has been called a “parallel import”: the market where the price is higher is the point of arrival of – at least – two parallel flows of the same products, the first one originated by the producer of the goods and the second one originated by somebody importing the goods from the country where the price is lower. If this really happens in significant respects, the prices of the two markets will tend to get closer, and the original differential between them will tend to disappear – or, at least, to decrease substantially. Needless to say, a “perfect” result will be attained only in a “perfect” market. In a market with more pronounced imperfections, there will be parallel imports involving a more limited amount of goods; hence, the price differential will not be entirely eliminated, and some partitioning between the two markets will remain, though perhaps in acceptable – or in not unacceptable – terms. A parallel import occurs “naturally” if there is a significant price differential. As a matter of principle, it is correct to favor the parallel import, if only because it induces markets having different prices for the same product towards an identical price, and thus it cooperates in establishing a common market. In this perspective, the private profit of the parallel importer – the profit he extracts from his activity, obtaining the differential between the two prices – is justified by the contribution his activity makes to the public interest. Moreover, this private profit of the parallel importer is limited in temporal terms to the first period of his intervention. It will cease once the prices in the two markets become identical. Obviously, the producer – if his intent is to enjoy different prices in two markets – will try to impede passage of any products from one market to the other. He can act against the parallel import using different tools. He can seek to prevent the wholesale trade of his products into the country where the price is lower; he can insert in the distribution contracts with his distributors clauses prohibiting, directly or indirectly, the diversion of the products into a different country; or, he can oppose the importation of the goods by using an intellectual property right, for example a trade mark right registered in the country where the price is higher. In the 1960s, the European Economic Community (as it then was) perceived the risk these reactions to parallel imports posed for the “European common market” then under construction. The European Community has established in the course of time various rules preventing these reactions: among them rules declaring void clauses in distribution contracts prohibiting the re-export of products out of a certain territory – be it assigned in exclusivity or not to a distributor, and rules preventing the trade mark holder from resorting to his trade mark right to oppose the circulation of his products inside the territory of the European Community once he has put those products on the market in one of the countries of the Community. This second set of rules, justified in conceptual terms by the exhaustion principle found in intellectual property rights, was created by the Court of Justice case law Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

New trade mark problems in parallel import of drugs

265

and was later written into trade mark legislation.2 It has been given further detail by clarification of fragments of the rule, such as specifying what constitutes the act of putting a product on the market, when there is consent of the right holder, and so on. This rule is generally considered a positive one, for the reasons mentioned briefly above. Indeed, it could perhaps be discussed why exhaustion applies only at the European Union level, and not at international level.3

21.2 different markets and different prices for products – like drugs – whose price is fixed by a public authority This set of rules has been in force for some decades now. Time goes by. Commerce in Europe – and elsewhere – evolves. We too evolve, we change our ideas and our projects. From then on, we have had the opportunity to apply these rules to the commercialization of different products and in different markets. We have also applied them to the development of new forms of trade, new ways of organization of production and advancements in transportation of goods. The evolution in our understanding, the evolution and transition of “things” and “facts”, gives way to a continuous evolution of the rules. European Union law has evolved to keep pace with the evolution of trade. It is interesting now to consider some developments in these rules and further possible developments that might occur. The evolution of rules must be guided by coherence; that is not only a formal, dogmatic requirement, but something important to ensure that the rules positively solve the problems for which they were created. It is ever more frequent, due to the complexity of our world, to our inability to look into the future, and to the intricacies of our legal systems, that the effects of a rule are significantly different from those that were predicted. Hence the interest, the need, to engage in constant control and revision of our legal system. In the following pages I will suggest that European Union rules favoring the parallel import of goods cannot – or risk not being able to – reach their specific goals in some markets, under some circumstances. Most notably, this is true in markets where the prices are not free, being instead fixed by a public authority. Hence, it would be wise not to extend these rules to certain new cases brought to our attention by some recent decisions (see 21.5 below).

2

3

For the position of current EU law in point of exhaustion of trade mark rights, see A. Kur and M. Senftleben, European Trade Mark Law (Oxford University Press, 2017), 435. See the report written for the EU Commission by National Economic Research Associates (NERA), 8 February 1999, synthetized by a Communiqué from Commissioner Bolkenstein on the issue of exhaustion of trade mark rights, 7 June 2000.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

266

Vincenzo Di Cataldo

21.3 when the price is fixed by a public authority, the parallel import has no effect on the price differential Parallel import rules are currently applied in EU Member States even to parallel imports of products whose price is fixed by some public authority, as is the case with drugs, for example. A parallel importer may also re-import the products bearing the trade mark of the producer into the country of origin without the consent of the producer. All European countries provide for a complicated system of fixing some or all drug prices. The criteria for fixing the prices of drugs in the European countries are in theory quite similar, but the result of their application by authorities in different states frequently produces different prices, and frequently with important differences from one state to another. In these cases, there is a strong incentive to import the drug from the country where the price is lower into the country where the price is higher. In fact, we assist in these crossed flows of drugs leaving one country for another. Regardless of the volume of products that are the object of parallel import, the market cannot react as it would “naturally” to parallel imports, which would involve modifying the price of the product. The price of the producer will remain as it was, unless the producer himself applies to the regulatory national authority to change it – presumably, to reduce it. This happens rarely, and as late as possible. It is easy to understand why. The price of the parallel importer will be fixed by the same authority that has fixed the price of the original producer. This authority could fix a price identical to or lower than the price of the producer. I will consider briefly both cases. Since it is unlikely that the price fixed for the parallel importer will be higher, this third hypothesis does not need to be considered here. We will see that in the two cases to be examined, the parallel import does not have the positive effects it has in a market with unregulated prices. If the price fixed for the parallel importer is lower than the price of the producer, the two prices will remain different – and the availability of parallel imports will not have its natural positive effect in the markets at stake: the prices of the two products will remain different. The parallel import will increase the presence of the product in the country where the price is higher.4 This is not particularly interesting for that market, as generally the drug is already present there in an adequate measure, though at a higher price. In this market, the public will find a certain quantity of products at a lower price, and the original producer will sell less, because his prices are higher. On the contrary, frequently the parallel import generates a negative effect in the country where the price is lower, reducing the quantity of the drug to a level which could be inadequate vis-à-vis the need of that country. This has happened frequently, and can be particularly dangerous. 4

Unless the original producer gives in, and asks the authority to set a lower price for his drug.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

New trade mark problems in parallel import of drugs

267

There will also be no positive effects if the price fixed for the parallel importer is identical to the price of the producer. The country receiving the parallel import sees no benefit; the price of the two products will be fixed in the higher amount. The country from which the parallel import originated will also risk being deprived of that drug. The only benefit will go to the parallel importer. He will gain the difference in prices in the two countries, without time limits. It is difficult to think that this is the real goal of European Union law. It has to be recognized that the different prices for the drug in two countries is the result of questionable behavior of at least one – and sometimes both – of the national authorities that have fixed different prices for the same drug in spite of the similar criteria used by them (and this is a rather frequent scenario in the EU). Everything is in the hands of the national authorities in charge of fixing the prices. The European Union has no power on this point. To be sure, rules favoring parallel import – such as rules on the exhaustion of trade mark rights – have a positive effect in a system of free prices, where prices can react to competition and to differences in the prices and quantity of the products put on the market. On the contrary, rules favoring the parallel import do not have such a positive effect in a system of prices fixed by public authorities. If this is true, it is worth asking if there is an interest in favoring the parallel import of the drugs and of all other products whose price is fixed by a public authority. This question is worth a clear answer, at least if we share the idea that a rule deserves to exist only if it gives a benefit to the public – if the balance of its costs and its benefits is positive. This suggests that – instead of focusing on parallel imports in the drug market – the European Union should enact a common system of fixing the prices of drugs for all Member States. This would further coherence. Rules favoring the parallel import of products, like drugs, subject to national price regulation by different national authorities, are ineffective; a system establishing the prices of drugs at the European Union level would be more desirable – and could be done, despite the differences between the health systems of the Member States. This would be, at the end of the day, a real benefit for all – for the economic system, for society at large, for all of us. No such thing is on the agenda of the European Union today, and it seems improbable that this will happen in the near future given the current weakness of the Union. But we should know that only a common system of prices for drugs in the European Union can solve today’s problems. It is not seriously possible to solve them by rules favoring the parallel imports of drugs alone.

21.4 parallel import of drugs, repackaging and rebranding The pharmaceuticals sector has created, in respect of parallel imports, a series of not easily resolvable sub-problems due to the existence in Member States of Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

268

Vincenzo Di Cataldo

different national regulations imposing regulatory obligations for each drug, such as certain packaging, certain dosages, and the provision of certain information to the public. These rules are all, as a principle, obstacles to the free circulation of the products. In principle, parallel imports are only possible if the parallel importer is free to repackage the product in such a way as to meet the conditions posed by the country of import. The European Union has not objected to different national rules on the packaging of drugs, though – here too – a move in this direction would be useful. For example, it would be useful to ban unjustified and purely “formal” differences in packaging among the Member States.5 Instead, the European Union has tried to liberalize the repackaging of drugs that is a precondition for the import of drugs into a country different from the one where they have been put on the market by their producer. Hence, European Union law prevents drug producers from opposing the repackaging of the drugs if the author of the repackaging satisfies a complicated set of five conditions. These rules have been written by the Court of Justice, in the Bristol-Myers Squibb case,6 and, later, in the Upjohn case.7 I do not want to comment on these rules in all their details. Their starting point is the idea that the opposition of the trade mark holder to the import of repackaged drugs would “contribute to the artificial partitioning of the markets between member States” (paras. 49 and 17 respectively). The Court does not consider that here, in the case of drugs, the partitioning of the markets is directly created not by the behavior of the pharmaceutical firms who own trade marks, but by the regulatory system fixing drug prices. The behavior of the trade mark holder is scarcely relevant, if not absolutely irrelevant, both in terms of prices and in terms of markets. This doctrine has been interpreted by some national courts as enabling – when the producer of a drug sells it in different Member States with different trade marks – the importer of the drug from one state into another to rebrand it with the trade mark the producer uses in that country.8

5

6

7

8

Obviously, the differences between the languages and their consequences for labelling would remain. CJEU, 11 July 1996, Bristol-Myers Squibb and Others v. Paranova, Cases C-427/93, C-429/93 and C-436/93. CJEU, 12 October 1999, Pharmacia & Upjohn SA, formerly Upjohn SA v. Paranova, Case C379/97. England and Wales Court of Appeal, 6 February 2015, Speciality European Pharma Ltd v. Doncaster Pharmaceuticals Group Ltd and Madaus GmbH (2015 EWCA Civ 54). In this case, the producer of the drug (Madaus) sold the drug in France under the trade mark CÉRIS, in Germany under the trade mark URIVESC, and in the UK under the trade mark REGURIN. Doncaster bought the drug in France and in Germany, rebranded it with the UK trade mark REGURIN, and imported it into the UK.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

New trade mark problems in parallel import of drugs

269

21.5 new and different cases: export of loose drugs, export of active ingredients The last years have seen the blossoming of new and different cases, similar but not identical to those mentioned above. I will briefly discuss them here. Their disposition is uncertain. Indeed, it is not clear if these can be described as cases of parallel imports, that is, if these cases ought to be governed by the same rules established for parallel imports. Given different national regulation of packages, dosages, labels and many other formal details of the drugs, the producer of a drug must package different stocks of his production according to the regulations of the different countries where he aims at distributing the drug. He can instead – and this happens in many cases, being frequently more reasonable and more profitable – sell the drug directly, duly packaged, in his country only, and supply the loose product to a firm in a different country that will package it according to the regulations of that country (and sell it there). A possible – and frequent – variant occurs when the producer of a drug supplies only the active ingredient of the drug to a buyer that, with it, will produce the drug, will package and label it, and will sell it in its own country. In both these cases, the buyer of the loose drug, or the buyer of the active ingredient of the drug, might be interested in exporting his product from his home country, so as to put it on the market in the country of the producer of the drug, or in the country of the producer of its active ingredient. This interest depends, quite obviously, from the existence of a price differential for that drug in the two countries. Naturally, the buyer must respect the rules on packaging of the drug in the country where he introduces the product. But could the buyer, in the country of the originator, use – without the consent of the right holder – the trade mark used there by the originator?9 And could the originator oppose this use? In a recent Italian case,10 an Italian pharmaceutical firm produced and sold in Italy a drug under the trade mark DIBASE. This firm supplied the loose drug to a Spanish firm that packaged it and sold it in Spain under its own Spanish trade mark, DELTIUS. The Spanish firm decided to export the drug into Italy, and applied for authorization of the sale of the drug from the Italian regulatory agency, Agenzia Italiana per il Farmaco (AIFA). The Spanish firm could not sell the drug 9

10

A positive answer can seem implicit in the parallel import rules. But the cases at hand are not, in my view, real cases of parallel import, never are cases of repackaging. In the cases considered, the buyer is not the buyer of a drug ready for commercialization. The buyer has bought only the loose drug, and has packaged and labeled the commercial product; or has bought only the active ingredient of the drug, and has produced the drug with it. In both cases, the “product” bought by the buyer, and put on the market by him, has never been packaged and labeled by the producer with his own trade mark. The case was decided at first instance by TAR Lazio, 2 July 2015, Giurisprudenza annotata di diritto industriale, 2015, n. 6330, p. 1565, and at second instance by Consiglio di Stato, 28 May 2018, Giurisprudenza annotata di diritto industriale, 2018, n. 6702, p. 1223.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

270

Vincenzo Di Cataldo

in Italy under its own Spanish trade mark, because there was already on the Italian market a different drug labeled by a third party with a trade mark that could be confused with the trade mark DELTIUS. The Spanish importer applied to AIFA for authorization to use the Italian trade mark of the original producer of the drug (DIBASE). AIFA gave the authorization. The Italian producer of the drug filed a complaint against this authorization before the Tribunale Amministrativo Regionale del Lazio (TAR Lazio), which invalidated the AIFA authorization. On appeal by the Spanish firm to the Consiglio di Stato, the decision was reversed. Interestingly, the TAR Lazio has confirmed its decision in a subsequent identical case.11 Setting aside the problem of the competence of AIFA to authorize the use of a trade mark against the will of the right holder – in my view, AIFA simply lacks any power for this: no rule, in Italian Law, gives AIFA the power of authorizing a firm to use a trade mark of a third party12 – the crucial question can be put in these terms: is the buyer of the loose drug, that has bought the product with no trade mark on it, in case of import of the drug into the country of the originator, entitled to use in that country, without the consent of the originator, the trade mark under which the originator commercializes the drug? It is hard to see persuasive arguments for a positive answer. The Italian Consiglio di Stato has said that the market could be “confused” by the presence of different stocks of the same drug labeled with different trade marks. This argument is not really persuasive. The presence of identical or similar products with different trade marks is normal in any market economy, and is particularly normal in the market for drugs. This is the case of the generic drugs, absolutely bioequivalent, namely identical, to the drug of the originator, but labeled with different trade marks. Moreover, and more generally, in any market, and for any product, any firm is free to use different trade marks for different stocks of the same product, and is free to sell some products under a trade mark and some other, identical, products without a trade mark. Any favored parallel import should be per se a benefit for all, and it is correct to widen as much as possible the rules allowing the parallel import of goods, applying them to any case that can seem somehow similar, even by implying an analogous set of interests. But this is not always true. Parallel import can be useful in a context of free prices, but may be less useful – or not useful at all – in a market of regulated prices. 11

12

TAR Lazio, 29 August 2018, Giurisprudenza annotata di diritto industriale, 2018, n. 6703, p. 1230. The only power expressly given by law to AIFA is the power of authorizing the firm to use its own trade mark. The authorization can be denied if the trade mark is misleading or can be confused with other trade marks used by other firms for other drugs. In those cases, the firm is simply invited to choose a different mark. It should be obvious that this must be a trade mark not already in the hands of another firm.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

New trade mark problems in parallel import of drugs

271

21.6 sales of loose drug, sales of active ingredient: is the buyer entitled to use the producer’s trade mark? The buyer of the loose drug should not have the right to use the trade mark of the producer without the consent of the producer itself, for several reasons. The same applies, a fortiori, to the buyer of the active ingredient of the drug. First, this is a fact pattern that is clearly different from the case of a true parallel import. The exhaustion principle was created for – and applies to – cases where the product has been supplied by the originator to the distributor already labeled with the trade mark of the producer. It gives the parallel importer the right to maintain on the product the trade mark that had been put on it by the producer. The right of the distributor comprises the right of relabeling the product, in case of repackaging, with the trade mark of the producer – but it is still a trade mark that had been put on the product by its right holder. In the case at hand, the distributor has been given a product not labeled with the trade mark of the originator. It is not easy to see how the exhaustion principle could give the parallel importer the right to label the product with a trade mark that had never been put on it by the producer. Moreover, if the case at hand is considered in the context of a single country, the solution is absolutely clear. Suppose that the producer of a drug – or of any other kind of product – puts on the market part of his production under a trade mark, and part without trade mark. Is the buyer of the second part entitled to put the trade mark of the producer on the products? The answer is surely no.13 The buyer of the product without the trade mark cannot pretend to put the trade mark of the producer on it. This flows from the principle that the trade mark holder is free to decide the quantity of his production to be labeled with his trade mark, and to sell part of his production without his trade mark. And there is no reason for deciding differently if the sale of the products without trade mark involves the export of them into another country. To sum up, these new cases of parallel import cannot coherently be regarded as cases of parallel import. It would be irrational to subject them to the rules of parallel import. Their resemblance with the true cases of parallel import is superficial and deceptive. If we look for a good level of coherence in the transition of existing rules to new forms of international trade with the rigor suggested by scholars like Professor Annette Kur, to whom this short essay is dedicated, there should be no doubt as to the solution we should find.

13

See, on this point, V. Di Cataldo, Prerogative del titolare del marchio, quantita` dei prodotti e segmentazione del mercato: Studi in onore di Gerhard Schricker (Giuffré, 2005), 70.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.030

22 Legal Concept of “Exhaustion”: Exhausted? Reto M. Hilty1

22.1 incoherent legal concept: a transitory phenomenon? The legal concept of exhaustion is a unique attribute of intellectual property law. In essence it means that although an exclusive right in fact is granted, it can no longer be claimed if certain conditions are met – precisely because this right is “exhausted”. First and foremost,2 this is about the exclusive right of an IP right holder to disseminate an object embodying an IP right.3 In principle, this right is exhausted if the object in question has been put on the market4 lawfully, namely by the right holder himself or with his consent.5 Subsequent to this transaction, the exclusive right with respect to the object in question can no longer be asserted, in other words the right holder cannot prevent its resale, including resale for commercial purposes. Today, the existence of this legal concept is simply taken for granted, especially since it is now explicitly stipulated in most jurisdictions.6 However, this was not always the case, especially not in copyright law. In many legal systems, in fact, the entitlement of the right holder to prohibit third parties from distributing a copy of

1

2 3

4

5 6

Prof. Dr. Dr. h.c., Director of the Max Planck Institute for Innovation and Competition, Munich; Professor at the Universities of Zurich and Munich. The author wishes to thank Ansgar Kaiser, Ph.D. student and Junior Researcher at the Max Planck Institute, for his valuable support, especially in document enquiry and analysis. All quoted websites were last accessed on 13 July 2020. But see below under section 22.3. E.g. a copyright-protected work on a carrier (printed book, recorded DVD), a patented object, a branded product, etc. Depending on the respective legal system exhaustion requires marketing on national, regional (like the EU) or any territory. Kur/Dreier, European Intellectual Property Law, 2013, 244. E.g. Art. 4 (2) InfoSoc Directive [EU], § 17 (2) UrhG [Germany], § 18 (3) CDPA [UK], Art. L 122-3-1 CPI [France], 17 U.S.C. § 109 (a) [USA].

Downloaded from https://www.cambridge.org/core. 272 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

Legal Concept of “Exhaustion”: Exhausted?

273

the work was established only later, namely not at the same time as the reproduction right.7 In most cases the exhaustion of the distribution right as a consequence of putting the copy on the market was stipulated even later.8 Before that, legal practice had to develop the principle of exhaustion on its own, out of necessity to ensure the tradability of objects already put on the market, but embodying an intellectual property right despite the distribution right.9 From the perspective of general civil law, this “revocation” of a right granted by law on the basis of a legal concept developed by practice – but without a legal foundation – is quite remarkable. But also from the point of view of IP law, the recourse to what at least in the early days was an extra-legal concept attracts attention. It is true that property rights are often not granted without boundaries; it is precisely copyright law that provides for limitations authorising third parties to carry out certain actions which in themselves affect the property right. However, such limitations are then – at least in jurisdictions shaped by the continental European system – explicitly stated; even in common law, the case law that establishes permissions based on fair use derives from a corresponding legal norm.10 Against this background, it would have been coherent to subject the distribution right – like any other exclusive right – to a limitation11 if it was felt to be too farreaching. At the latest at the time when the legal concept developed by practice was stipulated, an ordinary limitation could have been established.12 Admittedly, the intended limitation of protection only applies to individual copies of a work, and this also only takes effect after the right holder has undertaken something (in the form of a sale of this copy). The former however also applies to the freedom of panorama – where the position of the single copy of a work matters – and the latter also applies to the right of quotation, for example, which likewise depends on an act of the copyright holder (his publication of the work). No one would declare such uses permissions by means of exhaustion; they are clearly based on limitations. The fact that, in contrast, the right to redistribute was derived via a special route is probably due not least to the fact that the existence of “unwritten limitations” has been rejected to this day. Significantly, recital 32, sentence 1, of the InfoSoc 7

8 9

10 11

12

In Germany, for example, the distribution right in terms of an own property right was stipulated only in 1901 (§ 11 LUG); prior to that, a rather ambiguous concept based on criminal law and rudimentary civil law protection existed. In the UK the distribution right itself – the “right to issue” – as an explicit prohibition standard was not established before 1988; Garnett/Davies/ Harbottle, Copinger and Skone James on Copyright, vol. i, 16th ed. 2011, para. 7-87 et seqq. In Germany: 1965; in France: 2006; in USA: 1909. RG, 16.6.1906, RGZ 63, 394, 398 – Koenigs Kursbuch [Germany]; Supreme Court, 1.6.1908, 210 U.S. 339, 350 – Bobbs-Merrill Co. v. Straus [USA]. E.g. 17 U.S.C. § 107 [USA]. “Limitations” in the present context are understood as restrictions of exclusive rights as listed e.g. in Art. 5 InfoSoc Directive [EU]. In fact, in the USA the exhaustion rule in 17 U.S.C. § 109 (a) is systematically listed under the term “limitations”.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

274

Reto M. Hilty

Directive, for example, declares the list of limitations in Art. 5 (with regard to the right of reproduction and the right of communication to the public) to be an “exhaustive enumeration”.13 The call from academics for a kind of general clause – for example in the sense of a positively applied three-step test – has so far gone unheard.14 At the time when the principle of exhaustion was being developed, this was no different, which is why it may be obvious that the doctrine developed an independent – quasi extra-legal – concept. And once this approach found acceptance, there was no longer any sense in switching to a traditional legal limitation, which might only have provoked new questions. The result as such is undisputed, and the consensus that it is based on exhaustion still seems unshaken today, despite the fact that this legal concept is doctrinally hardly conceivable.

22.2 unanswerable questions At the same time, it cannot be overlooked that an argument based on the exhaustion of intellectual property rights reaches its limits or fails completely as soon as the marketing of physical goods ceases. In copyright law in particular, however, business models based on data transfer via the internet are coming to the fore.15 Initially, data was only transported in this way to be recorded on the user’s own medium; but even here, the logic behind the exhaustion is challenged because the right holder no longer sells anything. Of course, the user still acquires an empty carrier, but not from the right holder, thus the specific subject matter which is linked to the exhaustion with respect to the property right is missing.16 Now more than ever, this line of reasoning has had its day; in pure streaming, where the carrier is completely eliminated, the only issue is to have temporary access to data stored elsewhere.17 But even if there is still a carrier in the game on the user’s side, the arguments are no longer convincing as to why intellectual property rights in what is downloaded

13

14

15

16

17

Also recognized by the German Federal Court of Justice: BGH, 28.11.2013, ZUM 2014, 524, 530 – Meilensteine der Psychologie. Fo¨rster, Fair Use, 2008, 211 et seqq.; Geiger/Griffiths/Senftleben/Bently/Xalabarder, IIC 2015, 93, 95; Hugenholtz/Okediji, Conceiving an International Instrument on Limitations and Exceptions to Copyright, 2008, 25 et seq., accessible at https://www.ivir.nl/publicaties/down load/limitations_exceptions_copyright.pdf; Senftleben in: Dinwoodie, Methods and Perspectives in Intellectual Property, 2014, 5 et seqq., accessible at https://ssrn.com/ abstract=2241284; regarding the three-step test in the TRIPS Agreement: Kur in: Kur/Levin, Intellectual Property Rights in a Fair World Trade System, 2011, 208, 237 et seqq.; ibid., Richmond Journal of Global Law & Business 2009, 287, 328 et seqq.; Geiger/Gervais/ Senftleben, American University International Law Review 2014, 581, 597 et seqq. See for example the increasing number of Netflix subscriptions since 2001, https://de.statista .com/statistik/daten/studie/183340/umfrage/abonnenten-von-netflix-seit-2003/. Becker, UFITA 2015, 687, 694; Bergmann in: Ahrens/Bornkamm/Gloy/Starck/v. UngernSternberg, FS Erdmann, 2002, 17, 17 with additional references. Hilty, GRUR 2018, 865, 875; Becker, ZGE 2016, 239, 264.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

Legal Concept of “Exhaustion”: Exhausted?

275

should no longer be enforceable if this carrier is resold.18 This is because this situation differs fundamentally in that there is no need to purchase every single item to which the exhaustion of the property right relates. If legal access exists, downloads can in principle be made to an indefinite number of media without having to pay an additional fee. For sure, a certain parallelism could be achieved by permitting only the resale of the carrier used for the first download; thus, other carriers would no longer be covered by the principle of exhaustion. With such differentiations, however, this legal concept would be overstrained, since it is directed at a simple “either-or”.

22.3 further applications of the legal concept However, the legal concept of exhaustion is broader than the right of resale. Although it is this aspect that has shaped the discussion in copyright law from the beginning, it is also copyright law where this aspect loses importance because of new business models. Such models increasingly only depend on access to the subject matter of protection, but do not require the acquisition of an item in which the property right is embodied. In patent law, on the other hand, the resale perspective will continue to be relevant because objects embodying the patent are naturally acquired physically. However, there is another aspect at stake in patent law: that of the right of use. Generally, the right of use is a core element of an exclusive right. In the case of a product patent this means that a (physical) object in which the IP right – here the patent – is embodied may not be reproduced without the consent of the right holder. It is less clear, however, what that means in relation to the mere use of the product as such. An IP right in principle does not cover the use of a physical good by which the protected (intangible) subject matter is not executed – although Art. 28 (1) of the TRIPS Agreement appears to say the opposite in lit. a (“to prevent third parties . . . from the acts of: making, using . . . that product”). In any case, there seems to be a consensus that the acquirer of a product may use it (irrespective of the private or commercial purpose of that use) without the consent of the right holder – whichever legal construction this is based on. This changes, however, if the patent does not relate to a product but to a process. In this case, “use” no longer means the manufacture of something (in which the IP right is embedded). Rather, the use of a suitable device – which itself does not have to be protected (e.g. a machine that permits the application of the patented process) – constitutes the very act of use of the IP right. This use of the device 18

The exhaustion doctrine can principally be based on three arguments: The possibility of remuneration at the occasion of the first sale, the autonomy of the owner of personal property, and the facilitation of trade: Becker, UFITA 2015, 687, 688 et seqq.; Mysoor, IIC 2018, 656, 660 et seqq.; Reimer, GRUR Int 1972, 221, 225 et seqq.; de la Durantaye/Kuschel, ZGE 2016, 195, 210 et seqq.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

276

Reto M. Hilty

therefore would constitute a patent infringement if it were not possible to explain why it is lawful. The solution found was similar to that for the resale right; here, too, the lawful acquisition of such a device should have the effect that the process patent can no longer be asserted against the acquirer and user of the device. In practice, this result is obvious because the acquisition of the device without a right of use would not make sense; however, this insight does not yet constitute a doctrinal reason. The well-known German jurist Josef Kohler explained this solution with the legal concept of the “implied license”: if the right holder (or with his consent a third party) sells the device, he tacitly consents to its use, so that he can no longer assert his property right against the acquirer.19 In the beginning, Kohler also tried to apply this “implied license” to the situation of resale – be it of a patented subject matter, be it of a work – but this met with criticism. He himself then turned to the explanatory model of exhaustion.20 Conversely, the legal view according to which a lawfully sold device may be used, even if thereby a patented process is carried out, occasionally was also explained by the legal concept of the exhaustion of the patent.21 This did not help to clarify the doctrinal background of this legal concept. However, the uniform reference to the lawful sale of the device made it easier to explain why the subsequent purchaser of the same device must also be entitled to the right of use. The fact that this aspect of the right of use has not been discussed much in copyright law22 so far may have something to do with the fact that the pure consumption of a work – namely the mere reading, listening to or seeing of a work – is not considered to be covered by copyright, and is therefore generally seen as copyright free.23 This does not apply to other acts, for example the public lecture or 19

20

21

22 23

Kohler, Deutsches Patentrecht, 1878, 160 et seq.; ibid., Patent- und Industrierecht, 1889, 70 et seq. Kohler, Handbuch des deutschen Patentrechts in rechtsvergleichender Darstellung, 1900, 454 et seqq.; ibid., Lehrbuch des Patentrechts, 1908, 131 et seqq.; ibid., Musterrecht, 1909, 121 et seq.; ibid., GRUR 1906, 269, 272; ibid., Kunstwerkrecht, 1908, 103. See Maila¨nder in: Wild/Schulte-Franzheim/Lorenz-Wolf, FS Gaedertz, 1992, 369, 372; Graf, GRUR 1973, 55, 55; Blum, GRUR 1932, 1, 9; Brandi-Dohrn, GRUR 1980, 757, 758 et seq.; Hubmann, Gewerblicher Rechtsschutz, 3rd ed. 1974, 130. In 2008, the US Supreme Court followed this line of reasoning: 9.6.2008, 553 U.S. 617 – Quanta Computer, Inc., accessible at https://supreme.justia.com/cases/federal/us/553/617/. Nowadays, however, at least the German doctrine explains the right to use based on an “implied license”. Kraßer/Ann, Patentrecht, 7th ed. 2016, § 33 para. 318; Benkard, PatG, 11th ed. 2015, § 9 para. 25; Busse/ Keukenschrijver, PatG, 8th ed. 2016, § 9 para. 155; Mes, PatG, 4th ed. 2015, § 9 para. 80; Schulte, PatG, 10th ed. 2017, § 9 para. 24; BGH, 24.9.1979, GRUR 1980, 38, 39 – Fullplastverfahren; BGH, 27.2.2007, GRUR 2007, 773, 776 – Rohrschweißverfahren. As to the Swiss approach see also fn. 48. Hilty, GRUR 2018, 865, 869 et seq. Entwurf eines Gesetzes über Urheberrecht und verwandte Schutzrechte (Urheberrechtsgesetz), 1962, Deutscher Bundestag Drucksache IV/270, 28; Sucker, Der digitale Werkgenuss im Urheberrecht, 2014, 53 et seqq.; Stieper, Rechtfertigung, Rechtsnatur und Disponibilita¨t der Schranken des Urheberrechts, 2009, 8, 111 et seq., 147, 411;

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

Legal Concept of “Exhaustion”: Exhausted?

277

presentation of a work, but for such acts an authorisation of the right holder is required anyway. Whoever buys a book may therefore read it, but not read it publicly. However, the freedom to consume a work ends with the computer program. Its use is always relevant in terms of copyright, even if it is for private purposes.24 For this reason, the use of a program must be permitted, which is the case according to Art. 5 (1) Software Directive if a program copy is lawfully acquired. This, however, also opens up the question of the right to use in copyright law, which requires specific considerations to be made as soon as the right holder fails to sell a physical carrier. In this case, the statutory condition from which the right to use a computer program (in the sense of application) after sale is derived is not fulfilled. Even if the user still downloads a computer program, the link to that particular carrier doesn’t lead any further, because in terms of the right of use it is not really about that carrier – in contrast to the right of resale, which naturally involves something physical. The explanatory model of the right of use based on the program copy collapses even more if a computer program is only used by accessing an external server, which means that it is only temporarily available to the user (similar to streaming in traditional work categories). From a practical point of view, the right of use in such a situation will of course derive from a contract anyway. If the right holder (or a third party with his authorization) grants access to the computer program,25 he of course does so for the purpose that the person granted access may also use it – which is to be assumed even if there is no explicit agreement to this effect. Similar to the sale of a device enabling a patented process to be used, granting access would make no sense if the use of the program were not also permitted. In contrast to a device where the right of use can be transferred to a third party by reselling the device, in this case it is not easy to argue whether and to what extent the right to use the computer program can be transferred to a third party. Should this be the case, the change of party would in any case have to take place by the fact that the access facilities would be used by another person from a certain point in time. The question then arises, however, as to whether the right holder will have to watch this just as passively, as he cannot – legally – prevent a device from being resold and subsequently used by a third party.

24

25

Dustmann in: Fromm/Nordemann, Urheberrecht, 12th ed. 2018, § 15 para. 13; Kroitzsch/ Go¨tting in: Möhring/Nicolini, UrhG, 4th ed. 2018, § 15 para. 11; v. Ungern-Sternberg in: Schricker/Loewenheim, UrhG, 5th ed. 2017, § 15 para. 188. Because temporary reproductions necessarily come with the use of a software; see also Proposal for a Council Directive on the legal protection of computer programs, COM(88) 816 final in: Official Journal of the European Communities, C 91/05; Hilty, CR 2012, 625, 630; ibid. in: Calboli/Lee, Research Handbook on Intellectual Property Exhaustion and Parallel Imports, 2016, 64, 67; Schack, Urheber- und Urhebervertragsrecht, 8th ed. 2017, 213; Gru¨tzmacher in: Wandtke/Bullinger, Praxiskommentar zum Urheberrecht, 4th ed. 2014, § 69c para. 5. See section 22.2.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

278

Reto M. Hilty

22.4 transitional and sustainable business models The last question makes it clear that the legal concept of exhaustion in the online world is deprived of its significance. Not only is it inappropriately tied to the sale of a physical medium due to its history.26 Furthermore, it does not hinder the right holder from taking measures to prevent the occurrence of the effect intended by exhaustion. In the online world, the focus is no longer on a device whose destination and use the rights owner actually loses control of as soon as he sells it; what is decisive is rather access to the subject matter of protection, the control of which the rights owner can permanently exercise.27 This applies even if a physical subject matter still comes into use on the user’s side by downloading the work or the computer program.28 Here, too, the decisive precondition is that there is access at all. Even if this situation is still reminiscent of the one on which the principle of exhaustion is based, because this subject matter is outside the actual control of the right holder, the right holder has it in his hands to apply technical measures on his side when granting access; for example, it can be identified whether the first download took place and any further download can be made impossible. However, it appears that carriers recorded by the user himself are also likely to represent a transitional phenomenon.29 If rights owners obtain maximum control by no longer enabling downloads at all, but only allowing purely online use (via streaming or similar technologies), they have good reasons to push this technology.30 With the full possibility of control thus created, it even becomes irrelevant whether an act of use is at all of relevance under copyright law. If reading, hearing or seeing is only made possible if the user submits to certain contractual conditions, the limits of intellectual property rights no longer play any role.

26 27

28

29

30

See fn. 16. Hilty, GRUR 2018, 865, 875; Perkins, Northern Illinois University Law Review 2014, 1, 15; Riehl/Kassim, William Mitchell Law Review 2014, 783, 809 et seq.; Becker, ZGE 2016, 239, 264. The question whether the principle of exhaustion applies in the case of e-books downloaded by the user is the subject of a preliminary ruling requested by the Rechtbank (District Court) Den Haag, Netherlands (C-263/18 – Tom Kabinet), for which the Advocate General delivered his opinion on 10 September 2019, accessible at http:// curia.europa.eu/juris/document/document.jsf?text=&docid=217552&pageIndex=0&doc lang=EN&mode=req&dir=&occ=first&part=1&cid=14315583. See also the opinion of the Advocate General in C-263/18 – Tom Kabinet (fn. 28), No. 96: “By recognising the rule of exhaustion of the right of distribution in the internet environment, the Court would thus resolve a problem that does not really need to be resolved and that to a large extent belongs to the past.” See fn. 15.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

Legal Concept of “Exhaustion”: Exhausted?

279

22.5 rescue attempts of doctrine and jurisprudence In the doctrine, several attempts can be discerned that do not need to be described in detail here to keep alive the idea behind the legal concept of exhaustion in the online world.31 At first glance, this may seem sensible if one assumes that the primary focus is on user interests and thus limiting excessive rights of the holders of intellectual property rights. However, a kind of analogous application of the exhaustion principle can only be convincing if the actual framework conditions have remained comparable to the extent that similar interests of the parties involved have to be reconciled, as is the case when an item is sold which embodies an intellectual property right.32 For a physical carrier of the work, for example, the decisive factor is that it may only be needed for a certain period of time – a textbook, for example – but that it still has a remaining value (albeit slowly decreasing) which the purchaser should be able to realise by reselling the object in question. It is rightly claimed here that the right holder has already been fully compensated with the first sale.33 On the other hand, in the interest of the acquirer, the legal consequence of the sale act must be that the subject matter sold may be used, even if this use is relevant from the point of view of the intellectual property right and if the right holder does not explicitly consent; otherwise the acquisition of the subject matter would be pointless. In the online world, however, the circumstances are quite different. Typically, the payment of a full price at the very beginning which compensates the right owner once and for all fails to happen. Rather, streaming models are based, for example, on monthly payments, which allow the user continuous, but also varying uses (e.g. constantly newly available music titles). The user therefore does not invest a larger sum once, and there is therefore no remaining value that he would have to be able to realise if he wanted to waive the use of the content at any point in time.34 Instead, his obligation to make monthly payments simply ends. Even an explicit contractual user authorisation – for example relating to a computer program – typically is based on a continuing obligation with periodic payments. Despite this, some courts occasionally also explore the principle of exhaustion in the online world.35 In some cases hybrid models are concerned, meaning that 31

32 33 34

35

E.g. Spedicato in: Caso/Giovanella, Balancing Copyright Law in the Digital Age, 2015, 27, 45 et seqq.; Sganga, JIPITEC 2018, 211, 228 et seqq.; Ganzhorn, CR 2014, 492, 496 et seq.; Hoeren/Jakopp, MMR 2014, 646, 647 et seqq.; Hoeren/Försterling, MMR 2012, 642, 647; Redeker, CR 2014, 73, 77; Hartmann, GRUR Int 2012, 980, 984; Dreier in: Dreier/Schulze, UrhG, 6th ed. 2018, § 69c para. 24b; Terhaag/Telle, K&R 2013, 549, 553; Ohly, JZ 2013, 42, 43. Similar the opinion of the Advocate General in C-263/18 – Tom Kabinet (fn. 28), No. 79. With this argument, the so-called “reward theory” attempts to justify the exhaustion; see fn. 18. Hilty, GRUR 2018, 865, 879; ibid. in: Calboli/Lee, Research Handbook on Intellectual Property Exhaustion and Parallel Imports, 2016, 64, 81; Kerber, ZGE 2016, 149, 167. Germany had one of the first and therefore internationally much noted cases: LG Bielefeld, 5.3.2013, GRUR-RR 2013, 281; for more cases see Hilty, GRUR 2018, 865, 868. A pending

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

280

Reto M. Hilty

downloads still take place on the user side but are only available for a limited time. The undoubtedly most discussed decision, UsedSoft, comes from the Court of Justice of the European Union (CJEU) and refers to a computer program.36 Here, too, it was a hybrid technical solution in that it was essentially a question of whether it was permissible for the primary user, who did not utilize the program in part, to allow a third party to use the program to the respective extent by downloading it again himself (i.e. no physical carrier was ever handed over to any party).37 The decision has been discussed in detail elsewhere.38 Reduced to the result it is noticeable that the CJEU seemingly argues based on the principle of exhaustion, but properly read it answers the question whether on the side of the user – without the consent of the right holder – a (partial) change of party is permissible.39 This question can be generalized. What is special about this case is that, as mentioned above, the use of a computer program is relevant under copyright law in general, which means that the user contractually authorised by the right holder is to be regarded as the licensee.40 At first glance this seems to be different with traditional categories of works, for example in the case of a pure enjoyment of a work by a private individual. However, although such uses may be irrelevant from a copyright perspective, the person in question also has a contractual relationship with the right holder – at least in a certain phase – in order to be granted the necessary access at all. Therefore, the question also arises here as to whether the relevant party status can be transferred to a third party – without the consent of the right holder.

36

37

38

39 40

preliminary ruling by the CJEU, however, will have fundamental importance for future decisions (Case C-263/18 – Tom Kabinet, see fn. 28). CJEU, 3.7.2012, Case C-128/11 – UsedSoft, accessible at http://curia.europa.eu/juris/document/ document.jsf?text=&docid=124564&pageIndex=0&doclang=EN&mode=lst&dir=&occ=first& part=1&cid=14219885. UsedSoft, see fn. 36, para. 26; Hilty, CR 2012, 625, 627; ibid., GRUR 2018, 865, 869; Stieper, ZUM 2012, 668; ibid. in: Staudinger, BGB, 2017, § 90 para. 13; Specht in: Bräutigam/Rücker, E-Commerce, 1st ed. 2017, 574 et seq. para. 22 et seqq. Kur/Dreier, European Intellectual Property Law, 2013, 309; Hilty/Köklü/Hafenbrädl, IIC 2013, 263, 272 et seqq.; Hilty, CR 2012, 625, 627 et seqq.; ibid. in: Calboli/Lee, Research Handbook on Intellectual Property Exhaustion and Parallel Imports, 2016, 64, 72 et seqq.; ibid., GRUR 2018, 865, 868 et seq., 872 et seq.; Spedicato in: Caso/Giovanella, Balancing Copyright Law in the Digital Age, 2015, 27, 40 et seqq.; Sganga, JIPITEC 2018, 211, 224 et seq.; Determann, Berkeley Technology Law Journal 2018, 185, 202 et seq.; Mysoor, IIC 2018, 656, 678 et seqq.; de la Durantaye/Kuschel, ZGE 2016, 195, 200 et seqq.; Stieper, ZUM 2012, 668, 668 et seqq.; Specht in: Bräutigam/Rücker, E-Commerce, 1st ed. 2017, 573 et seq. para. 20 et seqq. Hilty, GRUR 2018, 865, 873; Hilty/Köklü/Hafenbrädl, IIC 2013, 263, 279. It should be noted that a software contract is actually a license agreement; see Hilty, MMR 2003, 3, 13 et seqq.; ibid., CR 2012, 625, 637; Hilty/Köklü/Hafenbrädl, IIC 2013, 263, 274 et seqq.; contrary the German jurisdiction since BGH, 18.10.1989, NJW 1990, 320, 321; also Redeker, CR 2014, 73, 74.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

Legal Concept of “Exhaustion”: Exhausted?

281

22.6 consequences for the concept of exhaustion and beyond It does not require in-depth technical knowledge to understand that this legal question is of a purely theoretical nature. In the case of UsedSoft, the right holder Oracle failed to take the necessary technical measures to prevent a third party from downloading the respective program itself – without its own license. However, it is easy for any online service based on pure streaming or similar technology that a priori prevents a download on the user’s side to be secured in such a way that only a certain party obtains access and the corresponding use options. This implies that any legal justification aimed at explaining the transfer of a right of use – which in the online world is the only issue for computer programs – is, at least from a practical point of view, futile.41 Seen in this light, the legal concept of exhaustion in the online world is obsolete in two respects. On the one hand, the primary situation – the sale of an item in which an intellectual property right is embodied – no longer occurs at all. If, instead, it is only a matter of granting access, it is from the outset also only the second aspect that the legal concept of exhaustion should explain: the right of use.42 On the other hand, even with regard to this right of use, it is no longer possible to fall back on the original explanatory model which was based on the implied license. Even if certain rights of use were supposed to be granted based on particular activities of the right holder (at that time it was the sale, now it would be the granting of access), such a legal construction would remain ineffective if the use so permitted could be prevented by technical means, possibly coupled with contractual prohibitions.43 Of course, life will not only take place online in the future. With regard to patent law or trademark law in particular, the focus will continue to be on objects which are sold and which must also be able to be resold; in patent law, the explanation remains necessary as to why a device may be used if it allows a patented process to be used. Also, with regard to copyright law – in addition to the analogue area – business models based on downloads on the user’s side may continue to exist for the time being, hence it is obvious that such carriers might also be passed on at some point. But already in this case, as has been shown,44 doubts arise as to whether the principle of exhaustion can be used to achieve a robust weighing of interests that reflects these specific circumstances. If such a weighing of interests generally does not take place via the legal concept of exhaustion, but by means of more or less differentiated limitations oriented to specific circumstances, their advantage lies above all in the fact that fundamental decisions are shifted to the legislature. This is particularly important when a new technical situation 41

42 43 44

Hilty, GRUR 2018, 865, 879 et seq.; Oprysk/Matuleviˇcius/Kelli, JIPITEC 2017, 128, 131; Moritz, K&R 2012, 456, 458; Hartmann, GRUR Int 2012, 980, 985. See section 22.3. See section 22.4. See section 22.5.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

282

Reto M. Hilty

has to be dealt with and valuations of interests are necessary. If, instead of establishing a specific limitation, the attempt is made to apply a legal concept beyond its traditional scope, it is up to the courts to weigh up general interests. Admittedly, case law makes sense in a first phase,45 but when the fundamental questions are on the table, it is at least in line with continental European legal tradition to establish legal certainty via statutory law within a reasonable time frame. Apart from the fact that the legal concept of exhaustion, as it is stated in positive law today, does not fit contemporary circumstances, one has to wonder why a singular, doctrinally hardly explainable construction should be adhered to at all, if in the end nothing else is achieved than the restriction of an exclusive right, which typically represents the task of a limitation. Instead, reasons of coherence suggest that existing legal norms stating the principle of exhaustion – irrespective of their mostly atypical systematic placement at least in copyright legislations – should be regarded in doctrinal terms as limitations.46 If, however, such provisions do not even tie in with contemporary practices – such as the right of use in relation to computer programs, where the sale of a copy of a program is still the decisive criterion – then a new limitation (if a provision is even necessary) should be established anyway.47 Art. 9a para. 2 of the Swiss Patent Act48 shows, for example, how easily such a systematic positioning of a limitation can be carried out with regard to the right of use of a device for the application of a patented process. However, from a practical point of view, even the doctrinally most conclusive limitation remains ineffective if the right holder can prevent a legally permitted act of use through de facto measures. Copyright lawyers are well aware of this problem, as technical protection measures have considerable potential to undermine the balance of interests provided for by legislation. The way EU law imposes here in order to counteract49 is unconvincing to a large extent,50 but at least it shows that it is necessary to design certain limitations in an “enforceable” way, namely to provide 45 46

47

48

49 50

See fn. 35, but also Hilty, GRUR 2018, 865, 868 et seq. Kur in: Kur/Levin, Intellectual Property Rights in a Fair World Trade System, 2011, 208, 211 et seq.; Mezei, JIPITEC 2015, 23, 23; Fo¨rster, Fair Use, 2008, 5 et seq.; Stieper, Rechtfertigung, Rechtsnatur und Disponibilita¨t der Schranken des Urheberrechts, 2009, 7. Generally in favor of using limitations to fine tune the weighing of interests in copyright law: Hofmann, GRUR 2018, 21, 21; Dreier, GRUR Int 2015, 648, 650 et seq. Under the title “Exceptions to the effects of the patent” (in the context of further limitations), it is essentially stated that the first and any subsequent acquirer of a device allowing the use of a process patent is entitled to do so if the device has been lawfully put on the market. Para. 1 of the same provision also provides for the right of resale in terms of an ordinary limitation rule. Art. 6 (4) InfoSoc Directive. Hilty in: Dreyfuss/Pila, The Oxford Handbook of Intellectual Property Law, 898, 918; Finck/Moscon, IIC 2019, 77, 87 et seq.; Dusollier, IIC 2003, 62, 73 et seq.; Reichman/ Dinwoodie/Samuelson, UC Berkeley Public Law Research Paper No. 1007817, 2007, 47 et seqq., available at http://ssrn.com/abstract=1007817; Specht in: Dreier/Schulze, UrhG, 6th ed. 2018, § 95b para. 17; more reserved: Peukert/Rehbinder, Urheberrecht, 18th ed. 2018, para. 1007.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

Legal Concept of “Exhaustion”: Exhausted?

283

the authorised user with tools to actually make use of limitations. Where and how such safeguards are necessary – or, if need be, whether the use of technical means should even be prohibited altogether – must ultimately be decided according to legal policy. The principle of exhaustion no longer provides a sound basis for this. At any rate outside the world of physical goods, its scope of application is exhausted – and within that world this legal concept is no longer necessary if all acts which are to be permitted to third parties although affecting intellectual property rights are covered by doctrinally coherent limitations.

addendum On 19 December 2019, after completion of this contribution, the ECJ issued the decision announced in footnote 28. According to this decision, the “Distribution Right” under Art. 4(1) of Directive 2001/29 comprises only copies “that can be put into circulation as tangible objects” (C 263/18 – Tom Kabinet, No. 40, referring to Art. 6 of the WCT, in particular footnote 5). Any form of “interactive activities” “other than the distribution of physical copies”, in contrast, falls under the “Right of communication to the public” under Art. 3 of the Directive (No. 41-44). Thus, the exhaustion effect under Article 4(2) of the Directive does not apply. In the case to be assessed, no physical copies were distributed at all. Instead, an intermediary enabled third parties to download “used” e-books again. However, the ECJ made it generally clear that there is no room for an interpretation of Art. 4(2) and to apply the principle of exhaustion beyond the legitimate distribution of physical copies.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.031

23 Building Coherence in Technological Transitions: Putting Exploitation at the Core of IP Séverine Dusollier1

A common criticism that has been expressed across the whole range of intellectual property rights in recent decades, in the wake of technological developments (whether digital or biotechnological changes), is what is called the ‘propertization’ of IP by some2 or an ‘over-expansionism’ of IP by others.3 Annette Kur herself has participated in projects attempting to provide for a more consistent regime of intellectual property to counter excessively far-reaching protection.4 The consequence of such expansion in (mostly) all IP rights is that potentially any use of the protected intangible could be considered an infringement and be subject to the control of the IP owner. Whether in trademark or copyright law, though less often in design or patent law, some borderline cases have puzzled the Court of Justice of the European Union (CJEU) or national courts. In copyright, mere technical copies, hyperlinks or ancillary activities have been considered as falling under the scope of exclusive rights. In trademark law, the attention of scholars, including Annette Kur, has focused on double identity infringement cases where the scope of protection aligned with the harm to any function of trademark. Patents in biotechnological inventions have triggered issues of proper delimitation of the scope of protection. 1 2

3

4

Professor, Sciences Po Law School. See for instance for recent scholarship, O.-A. Rognstad, Property Aspects of Intellectual Property, Cambridge University Press, 2018; C. Sganga, Propertizing European Copyright, Edward Elgar, 2018. See, using that exact expression or not, W. Cornish, Intellectual Property – Omnipresent, Distracting, Irrelevant?, Oxford University Press, 2004; G. Ghidini, Rethinking Intellectual Property, Edward Elgar, 2018; H. Ullrich, Expansionist Intellectual Property Protection and Reductionist Competition Rules: A TRIPS Perspective, J. Int’l. Econ. Law, 2004, p. 401. See for instance her implication in the research project IP in Transition (with Marianne Levin and Niklas Bruun) that resulted in the following publication: A. Kur and M. Levin (eds.), Intellectual Property Rights in a Fair World Trade System – Proposals for Reform of TRIPS, Edward Elgar, 2011.

Downloaded from https://www.cambridge.org/core. 284 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

Putting Exploitation at the Core of IP

285

In order to withstand such problematic expansion of IP and to reintroduce some coherence, several routes could be explored, and probably a mix thereof might be necessary to tame an uncontrolled IP regime. First, one could analyse what is property in IP and how alternative conceptualization of property could lead to a different legal organization of the regime.5 A second strategy is to strengthen the instances and normative force of IP limitations.6 Finally, one could limit the scope of control granted to IP owners. This chapter is situated in the third avenue. It seeks to explore what notion of exploitation could be commonly applied to all IP rights in order to make sense of the specific property of which IP consists. It does not go so far as to imagine a unique IP Right, as Annette Kur suggested envisioning in a fascinating conference she organized in Helsinki in October 2008.7 My modest ambition is rather to propose a methodology for defining the parameters of the scope of protection, around a notion of ‘exploitation’ of the IP subject matter, supported by the function of the right and the specific value of the diverse protected subject matters.8 I have already explored at length the concept of exploitation in copyright as an operative tool to better assess the scope of exclusive rights.9 Expanding that thinking to the other fields of IP is not easy for many reasons. So, I concede at the outset that I still have a great deal of work to do to transform my proposal into a comprehensive reflection about all IP rights. And assuming a systemic coherence of IP regimes probably ought not be my objective, as it would be far-fetched and illusory. Accordingly, my suggestion does not aim to rigidify the necessary diversity of the many IP rights we know into one unitary approach that would soon show its defaults, leaks and fissures. My only purpose is to offer a compass by which to navigate the scope of protection in borderline cases.

5

6

7

8

9

S. Dusollier, Intellectual Property and the Bundle-of-Rights Metaphor, in P. Drahos, G. Ghidini and H. Ullrich (eds.), Kritika: Essays on Intellectual Property, vol. iv, Edward Elgar, 2020., p. 146–179 See for instance A. Kur and H. Grosse Ruse-Khan, Enough is Enough – The Notion of Binding Ceilings in International Intellectual Property Protection, MPI Research Paper No. 09-01, available at http://ssrn.com/abstract=1326429. One Right System for IP – Vision Impossible?, IPR University Center International Conference, Helsinki, 3 October 2008. If Annette had doubts about the feasibility of such a system in her talk entitled ‘Differentiated Protection Based on Unitary Ground – A Feasible Approach?’, Andrew Christie was more direct in proposing ‘A Possible Vision of a One Right System for IP’. Also looking to align IP rights along the function they pursue, see M. Vivant, Intellectual Property Rights and their Functions: Determining their Legitimate ‘Enclosure’, in P. Drahos, G. Ghidini and H. Ullrich (eds.), Kritika: Essays on Intellectual Property, vol. ii, Edward Elgar, 2017, p. 47. S. Dusollier, Realigning Economic Rights with Exploitation of Works: The Control of Authors on the Circulation of Works in the Public Sphere, in B. Hugenholtz (ed.), Copyright Reconstructed: Rethinking Copyright’s Economic Rights in a Time of Highly Dynamic Technological and Economic Change, Kluwer Law International, 2018, at 163–201.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

286

Séverine Dusollier

In this short chapter, I can offer only a rough sketch, where subtleties and nuances might be missing. I hope the reader, and mostly Annette, who is never shy when it comes to delving into the many dimensions and nuances of the questions she researches, would forgive me for being somewhat unsophisticated or reductive. Thus, only the key lines of this approach will be outlined here. I will start by justifying why I opted for ‘exploitation’ as the central notion that could determine the proper scope of IP protection10 (23.1). Then I will explain the two prongs of an approach based on exploitation, namely the specific subject matter the right aims at protecting and the function it seeks to achieve (23.2).

23.1 the need for an ‘exploitation’ standard in ip I have elaborated elsewhere at length on the need to reconnect the scope of economic rights in copyright to the notion of exploitation.11 The evolution of copyright in the EU has severed the link between the exclusive rights granted to right holders and their very purpose. Many decisions of the CJEU have built a system where technical acts of reproduction or technical steps in a chain of communication or distribution12 have been held sufficient to trigger protection, irrespective of the reality of the public circulation of the work. Part of the reasons for such an extension of the exclusive rights lie in technological development. Particularly for the reproduction right, the mere fact that any use in a digital format requires multiple copies reduces the exclusive right of reproduction to its basic element, with no need to verify whether the copy matters in terms of exploitation of the work. Such a technical approach has a supplementary effect, for it splits the exploitation into many fragments corresponding to separate acts of reproduction or communication. Whereas, in the early age of copyright, the acts of reproduction and communication represented the key activities by which the work was exploited and transmitted to the public (i.e. by making and selling copies or by performing the work publicly), the exclusive rights appear today as being disconnected from the ultimate purpose of giving to authors the control over the dissemination of their works. Instead of being means to such ends, the economic rights have become the ends themselves. This contrasts with other IP rights, where the activities that could be deemed infringing are not technically defined, but correspond to any use of the protected subject matter. In trademark law, the standard of infringement is the use in the course of trade. Some examples of the types of use that are relevant are then given, such as affixing the sign to the goods, importing the goods under that sign, etc., but they are only illustrative. In patent law, any making and any use of the invention amounts to an infringement. 10 11 12

I will not touch upon design law, due to my limited knowledge in the field. Ibid. CJEU, 21 June 2013, Donner, C-5/11; CJEU, 13 May 2015, Dimensione Direct Sales, C-516/13.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

Putting Exploitation at the Core of IP

287

‘Use’, which is common to all such rights, but not to copyright and related rights, is not synonymous with exploitation. If one construes the ‘use’ as broadly as in property in tangibles, any act carried out with the protected sign, invention or design would be infringing. One knows perfectly that the protection is not that broad and the specific acts given as examples of what each right means by ‘use’, combined with the usual exclusion of acts carried out privately and for non-commercial purposes, lead to an attenuated picture. In order to avoid an excessive understanding of use as encompassing any activity dealing with a protected sign or invention, some reconnection with the ultimate purpose of the protection could instead be reinstated, through a notion of ‘exploitation’. One generally understands ‘exploitation’ as referring to the extraction of economic value from something, as using it for one’s advantage, as yielding some benefit out of its use. The Oxford Dictionary defines exploitation as ‘the action of making use of and benefiting from resources’.13 Beyond such economic interpretation, ‘exploitation’ should, in my view, connect a specific protected intangible with the objective of its protection.14 It has thus a twofold dimension: it limits the scope of economic rights in an IP subject matter to some activities, but also determines what in a specific subject matter should be the object of those activities. Therefore, a use of a protected subject matter, be it a work, an invention or a sign, would be considered as being an infringing exploitation when two conditions are met: (1) the use pertains to the element of the subject matter that the IP right aims at protecting; and (2) the use harms the function of the right. This two-prong infringement test resembles what has been gradually developed in trademark law,15 but could also be applied in the other IP fields to properly assess the scope for protection.

23.2 the two dimensions of exploitation 23.2.1 Exploitation of the Protected Dimension of the Subject Matter IP owners should only be entitled to prohibit the use of their protected work, invention or sign, when such use exploits the very value that the IP protection aims 13 14

15

See https://www.lexico.com/definition/exploitation. There is no consensus on the overall objective of copyright in the EU acquis, compared to the objective of trademark or patent rights. Different justifications of copyright can be found in the recitals of the relevant Directives, from ensuring ‘the maintenance and development of creativity in the interests of authors, cultural industries, consumers and society as a whole’ (recital 11 of the Infosoc Directive) to stimulating ‘innovation, creativity, investment and production of new content . . . in order to avoid the fragmentation of the internal market’ (recital 2 of the DSM Directive). For a comparison of the different justifications of IP rights in the EU Directives, see S. Dusollier, Pruning the European Intellectual Property Tree: In Search of Common Principles and Roots, in C. Geiger (ed.), Constructing European Intellectual Property, Edward Elgar, 2013, pp. 30–33. Not without critique though, that I will address later on.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

288

Séverine Dusollier

at controlling. If the subject matter is not used ‘as such’, but only incidentally or not for the value that the right protects, it should remain outside of the IP ambit. In copyright, this condition has as a consequence that only reproducing or communicating the expressive content of a work should be considered as infringing, and not use of the work only as information or reference. To put it differently, copyright is about exploiting the work as a work, the creative and expressive content being the object of its protection.16 Exploiting the work is a communicative act that has the work as object, and not the tangible item incorporating the work, nor the information, ideas and mere data it conveys. This necessary limitation could easily exclude from copyright scope most activities of text and data mining, of hyperlinking, as well as of search engines operations, where protected text or images are only used for their informational content. Similarly, the primary value of the sign protected by trademark is its distinctive capacity, and a general prerequisite for protection should be that the sign is ‘used as a trademark’. As explained by Senftleben, ‘the entrance requirement of trademark use would already serve as a filter to exclude claims that are unrelated to the identification and distinction of goods and services’,17 except for instances of dilution upon some strict conditions. When the protected sign is used as a necessary reference, for instance in fair competition or to denounce or parody the trademark owner, it should not belong to the realm of exclusive control. That would not exclude, however, further application of any relevant exception or external limitations that could excuse such use in a more nuanced way. A patent protects the invention that is defined as a technical teaching,18 in other words as the technical process or product that provides a solution to a technical problem. There is no protection if the invention itself does not operate or is not ‘active’. That was illustrated by the CJEU, albeit indirectly, in a singular case dealing with a biotechnological invention consisting in a genetically engineered form of soy.19 Flour produced with the soy was seized by EU customs authorities and suspected of infringing the patent in the soy. The CJEU refused to extend the patent to non-organic products derived from the biological invention, by insisting that the patent only covered the invention ‘in action’, meaning the invention performing its technical function. The Court held that the patent cannot extend 16

17

18 19

I developed that point further in Dusollier, note 9 above, pp. 187–188. See also A. Strowel, Reconstructing the Reproduction and Communication to the Public Rights: How to Align Copyright with its Fundamentals, in B. Hugenholtz (ed.), Copyright Reconstructed: Rethinking Copyright’s Economic Rights in a Time of Highly Dynamic Technological and Economic Change, Kluwer Law International, 2018, p. 202; Vivant, note 8 above, pp. 66–67. M. Senftleben, Adapting EU Trade Mark Law to New Technologies – Back to Basics?, in C. Geiger (ed.), Constructing European Intellectual Property: Achievements and New Perspectives, Edward Elgar, 2013, p. 146. See Ghidini, note 3 above, p. 74. CJEU, 6 July 2010, Monsanto, C-428/08.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

Putting Exploitation at the Core of IP

289

to a DNA sequence that ‘is not able to perform its function when it is incorporated in a dead material such as soy meal’.20 This decision deals more with the subject matter protected than with the proper condition for infringement, and it would be far-fetched to interpret this very peculiar case as requiring as a general principle that the infringing activity uses the patent as a technical invention. Having said that, one can read in the lines of that decision that patents cannot cover inventions when they are not operating as technical mechanisms,21 which aligns with our principle of pairing the scope of IP to the protected dimension of each specific subject matter.

23.2.2 Exploitation Delimited by the Function of the Right The value of the protected subject matter should then be associated with the function of the right, and of exclusive rights, for the use to count as an exploitation. Here also, diversity reigns. Though all IP rights aim at incentivizing the production of creation, innovation or at organizing competition, each could be said to pursue a specific function. Yet, lawmakers, if they sometimes agree on the particular function of intellectual property, rarely organize the scope of protection around this function. Insisting on the function of IP rights as a tool to better delineate what should be controlled by owners does not compare with the social function of property that is sometimes advocated in order to give more prevalence to limitations of property.22 The social function of property might curb the ‘absolute’ power of the property owner to exclude all others from the benefit of her asset and justify some encroachments or limitations over her exclusivity. Under that doctrine, property obliges its owner to conform with the social welfare society and the fundamental rights of others. Whereas the doctrine of social function of property justifies limitations to the right of the owner or scrutinizes its exercise under the general interest test,23 the recourse to the function of IP would consist in drawing the scope of the right itself. 20 21

22

23

ibid, §49. I should say that the Court does not go as far as drawing conclusions for the scope of rights under patent. The dearth of patent cases before the CJEU, due to the rare intervention of the EU lawmaker in patents through the Biotechnological Inventions Directive of 1998, also explains the inability to have a comprehensive preview of what an EU regime of patents could look like, notably as to the scope of exclusive rights. The social function of property was advocated by Duguit, at the beginning of the twentieth century (L. Duguit, Les transformations ge´ne´rales du droit prive´ depuis le Code Napole´on, Librairie Félix Alcan, 1912), with some scholars constantly coming back to his doctrine, but with not much success before courts. The Italian Constitution explicitly embraces such a social function by mandating law to put the necessary limits to property to guarantee its social function. The German Grundgesetz links the property with the obligation to contribute to the collective welfare, but this is construed in Germany as allowing for limitations to property for general interest. For a development of the doctrine of social function of property in those countries, and for the differences in this development, see Sganga, note 2 above, p. 194 et seq. Sganga, ibid., p. 224.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

290

Séverine Dusollier

The functions of trademarks are very well established by law, case law and doctrine. Trademarks’ principal function is to distinguish products and services, as recalled by recital 11 of the EU Trade Mark Directive (‘the protection afforded by the registered trademark the function of which is in particular to guarantee the trade mark as an indication of origin’). But other functions – such as those of communication, investment, advertising or guarantee of quality of the products bearing the sign – have been added, not without some criticism,24 by several CJEU decisions.25 The EU jurisdiction is now systematically insisting on the function(s) of trademark protection as a compass for any rule applicable in trademark law, from the conditions of protection to the scope of infringement.26 A few decisions started to require, as a condition for trademark infringement, verification that the use of the protected sign has harmed one of the functions of the trademark, thereby establishing a link between the scope of protection and the very objective of the trademark grant.27 The control of the trademark owner is limited to cases where the trademark has been exploited without any authorization, and not merely used, exploitation referring to the capacity of the trademark to accomplish one of its functions through the use made by a third party. Contrary to the protection of trademarks as a mere property right, which could lead to any use or reference to a protected trademark being infringing, such a functional approach, if reasonably applied, limits the intellectual property right to a purpose-bound property right that only confers on its owner a right to control the exploitation of the right, defined as what the grant of the right aims at.28 Annette Kur and Martin Senftleben similarly explain that ‘trademark protection, unlike patent, copyright or design law is not oriented towards objects, but towards functions, concerning first and foremost the function of identifying and guaranteeing the genuineness of commercial origin’.29

24 25 26

27

28

29

See A. Kur, Trademarks Function, Don’t They?, IIC, 2014, p. 434. See mostly CJEU, 18 June 2009, C-487/07, L’Oréal v. Bellure, §58. From the vast array of CJEU decisions using the function of a trademark as a key factor, see in particular ECJ, 12 November 2002, Arsenal v. Reed, C-206/01; ECJ, 25 January 2007, Adam Opel, C-48/05; ECJ, 18 June 2009, L’Oréal v. Bellure, C-487/07; CJEU, 23 March 2010, Google, C-236/08, C-237/08, C-238/08; CJEU, 22 September 2011, Interflora, C‑323/09; CJEU, 19 July 2012, Pie Optiek, C‑376/11; CJEU, 19 December 2012, Leno Merken, C‑149/11. See the Opinion of Advocate General Kokott in the Viking Gas case, 7 April 2011, C-46/10, §55: ‘The immediate purpose of trade mark rights is to ensure exclusive use of the mark such that the proprietor may strengthen the connection between that sign and its goods and services’; and § 59: ‘However, not every adverse [e]ffect on those functions justifies the application of Article 5 (1) of Directive 89/104. The protection of those functions on the basis of that provision, first, must not undermine the requirements of specific protective rules and, second, must respect overriding other interests.’ (footnotes omitted). Those functions can be discussed, as not all seem relevant for trademarks, despite the case law of the CJEU, but here is not the place for such a discussion. A. Kur and M. Senftleben, European Trademark Law – A Commentary, Oxford University Press, 2016, §1.03.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

Putting Exploitation at the Core of IP

291

By contrast, copyright does not generally follow a functional approach,30 except in cases of abuse under competition law.31 In one such decision, the CJEU has ruled that ‘the specific subject-matter of the intellectual property is intended in particular to ensure for the right holders concerned protection of the right to exploit commercially the marketing or the making available of the protected subject-matter, by the grant of licences in return for payment of remuneration’.32 The Court has hence anchored the function of copyright to the commercial marketing or making available of works and the possibility of obtaining adequate remuneration therefor. In two other decisions, the EU judges associated acts of public performance or acts of reproduction and marketing of the copies of works to the exploitation.33 The specific rights of communication or reproduction were then considered as means to achieve exploitation, not ends in themselves. Advocate General Kokott, in the FAPL case, has perhaps expressed the relationship with exploitation in the most convincing way. Indeed, she considered that the exemption of temporary copying allowed under Article 5(1) of the InfoSoc Directive is precisely ‘to permit acts of reproduction which are the condition for actual exploitation’,34 thereby anchoring the right of reproduction not to a technical act but to exploitation, which, in the case of broadcasts, she defines as the possibility of reception of the work by a public. Following that insistence on exploitation, I would however define it a bit differently, in less strictly economic terms. In my opinion, copyright serves to enable the circulation of the work in the public sphere (and to associate the author to the benefits it generates).35 As a consequence, not all acts of use or all copies of a work should be covered by copyright control, but only those that aim at a public circulation. Economic rights should then refer to any activity enabling the work to be made available to a public. The perimeter of exploitation should be drawn by reference to the public dimension of the use of works. The bundle of rights gathered under the notion of exploitation should enable the author to enjoy an economic reservation of her work and associate her with the constitution of a public space where her works circulate. Practically it could mean that temporary incidental copies or text and data mining would for instance remain outside the copyright realm, as they are not instrumental in making the work circulate in the public sphere.

30

31

32

33

34

35

The insistence of the CJEU, in its copyright decisions, on the need for ‘a high level of protection’, does not amount to the affirmation of a function of any kind. See the early case law relating to competition law in copyright, such as ECJ, 8 June 1971, Deutsche Grammophon, C-78/70. CJEU, 4 October 2011, Football Association Premiere League, joined cases C‑403/08 and C‑429/ 08, §107 (my emphasis). CJEU, 22 September 1998, Foreningen af danske Videogramdistributører, C-61/97, §14; CJEU, 22 January 2015, Art & Allposters, C-419/13, §36. Opinion of the Advocate General, 3 February 2011, Football Association Premier League Ltd and Others v. QC Leisure and Others, C‑403/08 and C‑429/08, §92. I have developed this argument in Dusollier, note 9 above, esp. pp. 177–183.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

292

Séverine Dusollier

Protection by patents aims at reserving to the patent owner the use of the technical solution she invented, either in applying the solution in technical processes of any kind, or in manufacturing and selling products. The function of patents is then to reserve to the patent owner the application of the technical invention. There have not been many borderline cases in recent years that the function of patent so defined could help solve, perhaps because the subject matter, including the intended function of the invention and its fields of use, are well circumscribed by the claims of the patent itself, which leaves less room for uncertainty. The exception for research might be one case where an exploitation-based regime of IP would necessarily lead to allow for acts of experimentation with the patented invention as they are only operating the technical knowledge to understand it, not to exploit it.36 Exceptions for regulatory review or even stockpiling, as considered in the Canada – Pharmaceutical Products37 case, could also be admitted, as the invention is not manufactured or used for its technical teaching, but only to prepare for marketing purposes after the expiration of the exclusive right. As a conclusion and by joining the two dimensions of the exploitation, one could propose that the scope of each intellectual property right be formulated as follows: – economic rights in copyright should only cover acts of circulation in the public sphere, that are communicative of the expressive dimension of the work, and not only using its informational or data contents; – exclusive rights in patent should encompass the use of the invention as technical teaching, applying the protected process or manufacturing or selling protected products; and – exclusive rights in trademark should only cover the acts of use of the trademark, in relation to products or services, as a distinctive sign, to the extent they harm the essential function of distinguishing products and services. Other functions of the trademark could be relevant in some cases.

23.3 conclusion Designing the scope of exclusive rights strictly around their purpose and function runs counter to how the EU makes law today. The recent Directive on copyright in the Digital Single Market38 is a perfect illustration of a policy-inspired legislation, relying on a balance corresponding to what all negotiating forces have managed to 36

37

38

Vivant, note 8 above, at p. 57. See also the distinction made by Ghidini (note 3 above, p. 21) between ‘use or access to know’ and ‘use or access to exploit’, only the latter one being in his opinion within the scope of IP. WTO Panel, Canada – Patent Protection of Pharmaceutical Products, WT/DS114/13, 18 August 2000. Directive 2019/790 of the European Parliament and of the Council of 17 April 2019 on copyright and related rights in the Digital Single Market, OJ L 130, 17.5.2019, p. 92.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

Putting Exploitation at the Core of IP

293

get one after the other, detached from any legal logic or conceptual frames, like its Article 17 declaring that some operators carry out an act of communication to the public without further justification or discussion.39 The exploitation-bound methodology exposed here follows instead a logic of proportionality and consistency between the means and the ends of protection. The scope of protection is delineated by the specific function pursued by each right and by the particular value of the protected creation, innovation or sign. A conceptual return to ‘exploitation’ confirms that intellectual property is not an absolute property but a functional one. Also, it could serve as a general mandate, particularly when new rights are envisioned, to identify what the value the right aims at capturing is and to strictly limit the scope of the right to what is needed to exploit such value. The holistic effort presented here is largely incomplete and would merit further exploration and refinement. And it does not say anything about the outer limits and balancing interests that should be preserved across all IP regimes, such as of freedom of expression and freedom of commerce, which are of course important parts of the design of intellectual property regimes.40 If done properly, an exercise in conjugating the scope of protection with the notion of exploitation could maybe help the diverse intellectual property rights to endure in a reasonable manner through the technological changes and maintain their legitimacy. Isn’t this what Annette Kur has always tried to achieve?

39

40

I do consider that video sharing platforms communicate works to the public and exploit works for their expressive value (see on that point Dusollier, note 9 above, pp. 197–198); I regret, though, that instead of justifying this point on such rationale, the Directive only seems to be dictated by policy and economic arguments. Kur and Senftleben, note 29 above, §1.45 – 1.46.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.032

24 “Accessory Exhaustion” – and Use of a Work as a Work Ole-Andreas Rognstad1

24.1 introduction The question of how to handle overlapping IP rights is central to Annette Kur’s scholarship. In addition to her unique overview of and deep insights into all disciplines of IP, she has always been concerned with fairness and balance in IP law. This is a concern that I share, and in the late 1990s, when I was working on finishing my doctoral dissertation on copyright exhaustion, I came to reflect on the interrelationship between the Dior/Evora decision of the Court of Justice of the European Union (CJEU) and the Astra decision of the European Free Trade Association (EFTA) Court.2 Although both the facts of and the reasoning in these two cases were quite different, the common feature of the decisions was the conclusion that the trade mark holder3 was unable to invoke copyright to ancillary material in order to prevent parallel imports of the goods. While reflecting on this, it hardly came as a surprise when Annette simultaneously published an article where she analyzed the trade mark/copyright interface under the viewpoint that the Dior/Evora decision in effect provided for the exhaustion of the copyright holder’s reproduction right,4 also making the link to the Astra decision of the EFTA Court.5 Her article inspired me to introduce the term “accessory

1 2

3 4

5

Professor dr. juris, Department of Private Law/Centre for European Law, University of Oslo. Case C-337/95, Parfums Christian Dior SA and Parfums Christian Dior BV v. Evora BV, ECLI: EU:C:1997:517; Case E-01/98, The Norwegian Government, represented by the Royal Ministry of Social Affairs and Health and Astra Norge AS, 24.11.98. Noting, however, that trade mark rights were not an issue in the Astra case. Annette Kur, Händlerwerbung für Markenartikel aus Urheberrechtlicher Sicht – Präsentationsrecht als neue Schutzschranke. Bemerkungen zu i.S. Dior ./. Evora, 48 GRUR Int. (1999), 24–30. See also Annette Kur, Advertising for Parallel Imports in the EEA, in Lars Gorton et al. (eds.), Festskrift till Gunnar Karnell, Jure AB 1999, 379–395, 394. Kur, Händlerwerbung für Markenartikel aus Urheberrechtlicher Sicht (supra fn. 4), 26 (fn. 21) and 27 (fn. 32).

Downloaded from https://www.cambridge.org/core. 294 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

“Accessory Exhaustion” – and Use of a Work as a Work

295

exhaustion” to describe the non-application of copyright as an accessory to trade mark exhaustion.6 When showed the honour to contribute to this book, readdressing the issue of “accessory exhaustion” came to my mind as the appropriate way of honouring Annette’s eminent scholarship. Like much of her writings, the topic touches upon both coherence and transition in IP law. Coherence, because the fact that application of various IP rights leads to different results may be – although not necessarily – a coherence problem. Transition in the sense that the dynamic field of IP law is in constant need of solutions to problems that occur as a result of commercial, economic and technological development. Readdressing the topic, in light of an article Annette wrote over twenty years ago, I find it appropriate to link it to current discussions about the proper scope of copyright law. Also in this context, an observation made by Annette in her 1999 article is right on target. Discussing the “copyright aspect” of the Dior/Evora decision, she drew a parallel to the Magill decision decided two years earlier, and remarked that “the reason behind the problems [occurring in these cases] is ultimately that copyright more and more often applies to objects of protection that earlier were not considered as worthy of protection”.7 This is a symptom which to a large extent explains the two decisions that form the basis of a principle of “accessory exhaustion” in EU/EEA law and similar problems in other jurisdictions, but which also have repercussions beyond the problem of parallel imports. Recently, scholars have pointed out that one problem with current copyright law is the lack of mechanisms similar to those in trade mark law where exclusive rights are limited to the use of the trade mark as a trade mark.8 Annette has on her part pointed to the “asymmetric convergence” implicit in the fact that the lowering of thresholds for copyright protection has not yet been offset by exceptions or limitations aimed at meeting challenges as a result of

6

7

8

Ole-Andreas Rognstad, Spredning av verkseksemplar: Om konsumpsjon av rettigheter i eksemplar av vernede a˚ndsverk, Cappelen Akademisk 1999, 750–751. See also OleAndreas Rognstad, Intellectual Property Law, in Carl Baudenbacher (ed.), The Handbook of EEA Law, Springer 2015, 703–720, 719 and Ole-Andreas Rognstad, Article 65(2), in Finn Arnesen et al. (eds.), Agreement on the European Economic Area: A Commentary, Nomos 2018, 641–659, 653–654. Kur, Händlerwerbung für Markenartikel aus Urheberrechtlicher Sicht (supra fn. 4), 26. Translated from German: “Bemerkenswert erscheint dabei, daß – wiederum ähnlich wie im Magill-Fall – die Ursache des Problems letztlich darin zu sehen ist, daß sich das Urheberrecht immer häufiger auf Schutzobjekte erstreckt, die früher nicht als urheberrechtlich schutzwürdig galten.” See, in particular, Alain Strowel, Reconstructing the Reproduction and Communication to the Public Rights: How to Align Copyright with its Fundamentals, in P. Bernt Hugenholtz (ed.), Copyright Reconstructed: Rethinking Copyright’s Economic Rights in a Time of Highly Dynamic Technological and Economic Change, Kluwer 2018, 203–240, 223 ff. See also, in the same volume, Se´verine Dusollier, Realigning Economic Rights with Exploitation of Works: The Control of Authors over the Circulation of Works in the Public Sphere, 163–201, 172–176.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

296

Ole-Andreas Rognstad

this lowering.9 The link between the cases on accessory exhaustion and recent developments in the CJEU’s interpretation of existing exceptions and limitations in EU copyright law may, however, point to certain remedies in this respect, on which I will elaborate in the following section.

24.2 “accessory exhaustion” – and its symptoms When speaking about “accessory exhaustion” in European law, the point of departure is the reasoning of the CJEU in the Dior/Evora case of 1997 concerning parallel imports of perfume products from France to the Netherlands. Here, the CJEU held that the trade mark holder Chr. Dior was not entitled to invoke its trade marks to prevent a parallel importer’s use of a leaflet containing photographs of Dior products under Article 7(2) (now Article 15(2)) of the Trade Mark Directive,10 unless this use of the trade mark seriously damaged the reputation of the trade mark.11 In other words, the trade mark rights would be exhausted without such damage occurring. To the following question whether invoking copyright to the bottles and the packaging would be an illegitimate restriction under the rules on free movement of goods (corresponding to Article 30 and 36, TFEU), the Court stated: Having regard to [previous] . . . case law – there being no need to consider the question whether copyright and trade mark rights may be relied on simultaneously in respect of the same product –, it is sufficient to hold that, in circumstances such as those in point in the main proceedings, the protection conferred by copyright as regards the reproduction of protected works in a reseller’s advertising may not, in any event, be broader than that which is conferred on a trade mark owner in the same circumstances.12

A key phrase in this respect is obviously “in circumstances such as those in point in the main proceedings”. Thus, the Court’s statement cannot be interpreted as giving trade mark rights precedence over copyright from a general point of view, or as a general prescription of how copyright and trade mark overlaps are to be solved.13 Both the subject matter and the purpose of trade mark and copyright law are substantially different, so that letting copyright protection depend on trade mark protection would as a general rule make no sense. Therefore, the Court was careful enough to underscore that the solution in question applied to the circumstances of 9

10 11 12 13

See Annette Kur, Funktionswandel von Schutzrechten: Ursachen und Konsequenzen der inhaltlichen Annäherung und Überlagerung von Schutsrechtstypen, in Gerhard Schricker, Thomas Dreier and Annette Kur, Geistiges Eigentum im Dienst der Innovation, Nomos 2001, 25 ff. Directive 104/89/EEC, OJ L 40 (1989) 1–7; Directive (EU) 2015/2436, OJ L 336 (2015) 1–26. Case C-337/95 (Dior/Evora), paras. 42–48. Case C-337/95 (Dior/Evora), para. 58. See also Annette Kur, Ersatzteilfreiheit zwischen Marken- und Designrecht, 118 GRUR (2016), 20–30, 25.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

“Accessory Exhaustion” – and Use of a Work as a Work

297

the case. The special circumstances of this case were that copyright was invoked in a setting that is remote to the purpose of copyright protection.14 The point of prohibiting the use of the leaflet was not to protect the artistic expression of the bottles and packaging, but to deprive the parallel importers of the benefit of marketing the parallel imported products. In such circumstances, invoking copyright would disturb the balance inherent in the set of rules that is meant to regulate the marketing of the products, in this case trade mark law and the “legitimate reasons” rule, which marks the outer limit of exhaustion of trade mark rights. Pursuant to the Court’s reasoning, copyright cannot extend beyond trade mark rights under such circumstances, and the exhaustion of the reproduction right occurs as an accessory to the exhaustion of trade mark rights. The less known EFTA Court decision in the Astra case, delivered the year after the Dior/Evora decision, expresses something similar, although the EFTA Court itself did not draw the parallel. The reason is obviously that the case was not about trade mark rights, but rather rules on marketing authorization for pharmaceuticals. According to the Norwegian Medicinal Act, an applicant for marketing authorization is obliged to produce a short description of the products, named Summary of Product Characteristics (SPC).15 At the same time, the Norwegian authorities decided that the same SPC also applies to parallel imported products, for security reasons. The pharmaceutical company Astra, by their Norwegian subsidiary, brought proceedings before Norwegian courts, demanding that the Norwegian medicinal authorities be prohibited from giving out SPCs for parallel imported products on the ground of copyright law. The question before the EFTA Court, on a request for an advisory opinion, was whether the application of copyright law in order to prohibit the use of the SPC, amounted to an unjustified restriction of the free movement of goods. The EFTA Court answered the question in the affirmative, with the following reasoning: The case at hand is not about trade in SPCs but trade in the medicinal products themselves. If the competent authority should be prevented from continuing its practice described above, trade in medicinal products between the Contracting Parties of the EEA Agreement would be restricted. In other words, the enforcement of a national copyright in an SPC, with the consequence that the competent national authority would be prevented from giving out/approving/laying down the same SPC for a product imported by way of parallel import as for a directly imported medicinal product, would lead to an artificial partitioning of the market in the European Economic Area. This would be disproportionate to the aim of protecting the copyright in the SPC. In addition, the Court finds that this would amount to a disguised restriction on trade between the Contracting Parties. A justification under Article 13 EEA is, therefore, not possible.16 14 15 16

Ibid. As required by EU law – see Directive 2001/81/EC, OJ L 311 (2001) 67–128, Article 6(1). Case E-1/98 (Astra), para. 26.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

298

Ole-Andreas Rognstad

The reasoning captures an important point that also is present in Dior/Evora, namely that the application of copyright in the SPC is not aimed at protecting the interests in the SPC as a work, but at preventing the distribution of the parallel imported goods. Hence the disproportionality observation. Although references to trade mark law or other IPRs than copyright are absent in the reasoning of the EFTA Court, the rationale behind it and its effect is compatible with a principle of accessory exhaustion.17 In the situation in question all relevant IPRs related to the parallel imported good, notably trade mark and patent rights, would be exhausted since the products were put on the market in the EEA with the consent of the right holder. Invoking copyright to the SPCs in this situation would make parallel imports of pharmaceuticals impossible. Linking the disproportionality assessment of the EFTA Court to the exhaustion of the IPRs in the parallel imported goods means that the reproduction right in the SPC as a copyrighted work,18 would be exhausted as an accessory to the exhaustion of trade mark and patent rights. Like in the Dior/Evora case, such accessory exhaustion would apply to the circumstances of the case, where the invoking of copyright would serve a different purpose than safeguarding the interest in protecting the work as a work. Read with this construed ratio decidendi, the EFTA Court’s decision forms, together with Dior/Evora, the basis for an EU/EEA principle of accessory exhaustion. The problems dealt with in these two cases are not unique to EU/EEA law, although the legal foundation for the development of a principle of accessory exhaustion – the rules on free movement of goods – is confined to the EU/EEA setting. In Australian law, for example, since 2000 there has been a statutory exception in the Copyright Act (Section 44C) for importation of so-called “noninfringing accessories”, i.e. works embodied in accessories to an article, including labels, packaging, written instructions and other information provided in the article.19 In the United States, a similar problem to what occurred in Dior/Astra arose in the L’anza case, concerning invoked copyright to labels of shampoo bottles, handled by applying the first sale (exhaustion) rule in the US Copyright Act to the re-importation of the labels.20 However, the current problem is, after the Kirtsaeng decision of the US Supreme Court,21 rather the opposite than what follows from the European accessory exhaustion principle, namely whether, or to what extent, trade 17

18

19

20

21

See Rognstad, Intellectual Property Law, 718–719, and Rognstad, Article 65(2), 653–654 (supra fn. 6). Although it is possible to claim that the Norwegian first instance court erred in its assessment of the SPC as a copyrighted work – see Rognstad, Intellectual Property Law (supra fn. 6), 717–718. Australian Copyright Act (1968, as amended) Section 10(1). See Mark Davison, Ann Monotti and Leanne Wiseman, Australian Intellectual Property Law, Cambridge University Press (2015) 303. Quality King Distributors v. L’anza Research International Inc., 523 (US) 135 (1998). For a comment to the decision, see for example Christopher Morris, Quality King Distributors, Inc. v. L’Anza Research International, 14 Berkeley Tech. L.J. (1999) 65. Kirtsaeng v. John Wiley & Sons, Inc., 568 US 519 (2013).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

“Accessory Exhaustion” – and Use of a Work as a Work

299

mark rights can be used to block imports of copies of copyright-protected works.22 In Canada, a majority of the Supreme Court (five of nine judges) in the Kraft Food case (2007) declined to reject copyright protection for the artworks in chocolate wrapping as a ground for permitting parallel imports of chocolate bars where the producer was unable to rely on trade mark rights.23 The majority (seven judges) still voted in favour of the parallel importer, but on different grounds. Although a principle of accessory exhaustion derived from the Dior/Evora and Astra cases applies to the situation where copyright is invoked to override results that follow from other IP rights, the application of a similar principle in situations where IP rights other than copyright are invoked to prevent parallel imports should not be excluded. However, it is questionable to what extent there is a need for such extension of the accessory exhaustion principle. Compared with, for example, the United States, where the problem of invoking trade mark law in order to override copyright exhaustion has occurred as a result of the Supreme Court’s Kirtsaeng ruling,24 the legal situation in Europe is different. Here, the exhaustion rules are harmonized across the fields, in the sense that EEA-regional exhaustion applies to copyright, trade mark and design law. Consequently, a “Kirtsaeng situation” would not occur in EU/EEA law regarding importation from countries outside the EEA. Moreover, the provision in Article 15(2) of the Trade Mark Directive and Regulation respectively provides for an assessment of whether there exist reasonable reasons to oppose further commercialization of the goods. This rule, combined with the “function test” in trade mark law, implies that there is no equivalent need of a doctrine of accessory exhaustion to deal with situations where trade marks are invoked to parallel importation of copyright goods. Also, with design/copyright overlaps, the problem is rather the extension of copyright into the design field than the other way around. A more relevant problem in the context of European law is whether the rationale behind the application of accessory exhaustion of the Dior/Evora and Astra cases 22

23

24

See e.g. Mary La France, Using Trade Mark Law to Override Copyright’s First Sale Rule for Imported Copies in the United States, in Irene Calboli and Edward Lee (eds.), Research Handbook on Intellectual Property Exhaustion and Parallel Imports, Edward Elgar Publishing (2016) 390–407 (Ch. 20). Euro-Excellence Inc. v. Kraft Canada Inc., 2007 SCC 37, [2007] 3 SCR 20. See further Robert Tomkowitz, Intellectual Property Overlaps, Routledge (2012) 150–152. See also Robert Tomkowitz, Copyrighting Chocolate: Kraft Canada v. Euro Excellence (2007), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1196542, and Abraham Drassinower, The Art of Selling Chocolate: Remarks on Copyright’s Domain, in Michael Geist (ed.), From “Radical Extremism” to “Balanced Copyright”: Canadian Copyright and the Digital Agenda, Irwin Law Inc. (2010). See Mary LaFrance, A Material World: Using Trade Mark Law to Override Copyright’s First Sale Rule for Imported Copies, 21 Mich. Telecomm. & Tech. L. Rev. (2014) 43–77. See also, from a broader perspective, Mary LaFrance, Wag the Dog: Using Incidental Intellectual Property Rights to Block Parallel Imports, 20 Mich. Telecomm. & Tech. L. Rev. (2013) 120–151.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

300

Ole-Andreas Rognstad

may extend to contexts other than parallel imports, exhaustion and free movement of goods. One lesson to learn from the two cases is that copyright protection may be curtailed in situations where the purpose of invoking copyright is not to protect the creative effort inherent in the work, but to exploit low thresholds for copyright in order to achieve a different purpose. The general relevance of this observation will be discussed in the following section.

24.3 use of the work as a work Arguably, a common trait of the Dior/Evora and Astra cases is that the copyright holder was unable to invoke his copyright in situations where the work was not used as a work,25 but as a marketing tool. If taken for granted that the photographed perfume flacons and the SPC were to be considered as artistic and literary works respectively, the very point of using them was to promote the parallel imported products and not to exploit the works as such. Conversely, invoking copyright in these situations was the pretext for preventing parallel imports. To be sure, the Dior/ Evora and Astra cases were about IP coherence in the sense that copyright in ancillary material could not be relied on in order to undermine the limits of IP rights inherent in the goods as such. Still, what the courts nevertheless did, in order to achieve such coherence, was to curtail the scope of national copyright law on the basis of the rules on free movement of goods. The question one could ask is whether such an approach is legitimate also beyond the problem of accessory exhaustion and free movement of goods in situations where courts find that copyright is invoked as a pretext for other purposes – in other words, in cases where the purpose of invoking copyright is alien to what copyright is aimed at protecting, namely the creative efforts of the author. In this respect, Advocate General Szpunar’s Opinion in the recent Funke Medien case may be held to represent a similar approach, although in a quite different context and based on different legal grounds. In this case, the German government invoked copyright in order to prevent the publication of certain military documents in a situation where it had considered that protecting the documents as confidential information did not justify the restriction on the freedom of expression enshrined in Article 11 of the EU Charter of Fundamental Rights. Having serious doubts about whether the military report qualified as an original work in the first place,26 the Advocate General went on to consider the possible scope of protection of the documents. In addition to raising doubts as to whether the German state could rely on the fundamental rights to property under Article 17 of the EU Charter as a means 25

26

Paraphrasing Alain Strowel, Reconstructing the Reproduction and Communication Rights (supra fn. 8), 220 et seq. Case C-479/17, Funke Medien NRW GmbH v. Federal Republic of Germany, AG Opinion, ECLI:EU:C:2018:870, paras. 17–22.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

“Accessory Exhaustion” – and Use of a Work as a Work

301

of restricting another fundamental right (the right to information),27 the Advocate General held that the restriction on freedom of expression was not necessary for the protection of property. In this respect the Advocate General emphasized that “the sole objective of the action taken by the Federal Republic of Germany in the main proceedings was to protect the confidential nature . . . of certain information deemed to be sensitive”, and that this objective “falls completely outside the scope of copyright”. Thus, “[c]opyright is . . . used here to pursue objectives that are entirely unrelated to it”.28 Consequently, the German government could not invoke its copyright in order to restrict the freedom of expression. Here, the fundamental right of freedom of expression backed up the view that copyright protection could not be invoked for purposes alien to copyright and applied in a situation where the purpose was not to protect the creative effort of the author, but to pursue a different goal – in this case the prevention of distribution of sensitive information. It shall here only be added that the Advocate General noted that the German state was not the “real” author, as the report was prepared by anonymous drafters pursuing their obligations as servants or officers, and that in this case there was “no question of protecting the author’s link with the work”.29 Furthermore, the fact that copyright was invoked in a situation where the real ground for opposing the publication was the protection of confidential information – an aim completely unrelated to copyright –was not acceptable in the Advocate General’s view.30 Hence, the result of the Advocate General’s approach was that copyright could not be invoked as a pretext for purposes not corresponding to its rationale.31 Here, the result was grounded in the fundamental rights of freedom of expression and information in a similar way as the free movement of goods forms the basis for accessory exhaustion. The CJEU did not follow the Advocate General’s approach, as the Court held that Article 11 of the EU Charter of Fundamental Rights was not capable of justifying exceptions to the exclusive rights laid down in the EU Copyright Directive (“Infosoc Directive”) beyond the exceptions or limitations provided for in Article 5(2) and (3) of that Directive.32 Hence, the position of the Court is that the provision in the EU Charter cannot play the same role as the rules on free movement of goods did in the Dior/Evora and Astra cases. Still, the Court left much room for interpretation under the exceptions and limitations, notably in this case the press-publishing and quotation exceptions enshrined in Article 5(3)c) and d) of the Directive. The Court emphasized that these provisions do not “constitute full harmonization of 27 28 29 30 31

32

Ibid., paras. 53–55. Ibid., para. 61. Case C-479/17 (Funke Medien), AG Opinion, para. 59. Ibid., para. 56. Compare Christophe Geiger and Elena Izyumenko, Freedom of Expression as an External Limitation to Copyright Law in the EU: The Advocate General of the CJEU Shows the Way, 43 EIPR (2019) 131–137, 136, fn. 56, with further references. Directive 2001/29/EC, OJ L 167 (2001) 10–19.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

302

Ole-Andreas Rognstad

the scope of the exceptions and limitations” that they contain, referring to phrases like “to the extent justified by the informatory purpose” and “in accordance with fair practice, and to the extent required by the specific purpose”, which leave the Member States with “significant discretion allowing them to strike a balance between the relevant interests”.33 This balance includes, inter alia, the compliance with the general EU principle of proportionality.34 Furthermore, the principles enshrined in the EU Charter on Fundamental Rights apply to the Member States when implementing EU law, and Member States must “ensure that they rely on an interpretation of the directive which allows a fair balance to be struck between the various fundamental rights protected by the European legal order”.35 It is also to be noted that the Court emphasized that the provisions in Article 5(3)c) and d) of the Infosoc Directive “are specifically aimed at favouring the exercise of the right to freedom of expression by the users of protected subject matters and to freedom of the press, which is of particular importance when protected as a fundamental right, over the interest of the author in being able to prevent the use of his or her work”.36 Hence, one could argue that also the Court’s decision leaves room for the same “use of the work as a work” approach as what followed from the Advocate General’s Opinion, although within the framework of Article 5(2) and (3) of the Infosoc Directive and not on the basis of Article 11 of the Charter alone. The argument could in that case be that invoking copyright in order to oppose the publication in the press of the military report as a pretext for the protection of the confidential information inherent in the report represents a disproportionate restriction of the freedom of expression. Thus, the publication of the report in this situation would be justified by the informatory purpose of the reporting of current events. Conversely, the balance between the relevant interests could imply a different solution if the author sought to protect the creative effort inherent in the report instead of the information, which does not enjoy copyright protection as such. Considered in this way, there is a line between the cases on “accessory exhaustion” and the Funke Medien case, although the legal context and grounds for delimiting copyright protection are different. A common feature is the view that copyright, like any other right, should not expand beyond its justifications. Both the low originality threshold and the sweeping definitions of exclusive rights represent challenges in this respect. This hardly means that there is sufficient legal basis for a general principle that copyright protection should be limited to the use of a work as a work. However, both the cases on accessory exhaustion and the Funke Medien case show the relevance of the argument under EU law.

33 34 35 36

Case C-479/17 (Funke Medien), CJEU, ECLI:EU:C:2019:623, para. 43. Ibid., para. 49. Ibid., para. 53. Ibid., para. 60. See also para. 68.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

“Accessory Exhaustion” – and Use of a Work as a Work

303

This leads us finally to the overarching topics of this book – transition and coherence; the potential to delimit the scope of copyright law in accordance with its purpose, and the rationale that the rights should not expand beyond their justifications. In this respect courts may be better served to initiate transitions in IP law than the legislature with its narrowly construed, or asymmetric, exceptions and limitations. At the same time the principle of accessory exhaustion, along with other efforts to bring the scope of copyright law in line with its purpose, in similar ways as in other IP rights, aims at coherence of the IP system. Thus, principles like accessory exhaustion do not target the overlapping of IP rights as such, since overlaps are accepted, and justified, as long as they are grounded in the purpose of the various rights. Rather, they are tools to secure that IP rights are tailored and exercised in accordance with their purpose. Hence, the better view is to look at the decisions of the CJEU and the EFTA Court in the Dior and Astra cases as early, but timely, examples of refusing copyright protection in situations where the work is not used as a work.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.033

a

Intellectual “Property” and its Limits

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms 212 of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

25 Intellectual Property in Transition: The Several Sides of Overlapping Copyright and Trademark Protection Jane C. Ginsburg1 and Irene Calboli2

25.1 introduction: the reason for overlapping copyright and trademark protection In this chapter, we address the phenomenon of overlapping trademark and copyright protection under United States (U.S.) law, which represents, in our view, an important example of intellectual property (IP) rights in transition. The Lalique bottles created for Nina Ricci perfumes, for example, may enjoy both copyright and trademark protection.3 Similarly, cartoon characters are components of copyrightable works (and in some jurisdictions, may be copyrightable works in their own right),4 but many have also long been registered as trademarks for entertainment services or merchandise.5 Copyright and trademark protections will overlap when 1

2

3

4

5

Morton L. Janklow Professor of Literary and Artistic Property Law, Columbia Law School. This contribution is based in part on Intellectual Property as Seen by Barbie and Mickey: The Reciprocal Relationship of Copyright and Trademark Law, 65 J. Copyr. Soc. USA 245 (2018). Professor of Law, Texas A&M University School of Law; Academic Fellow, School of Law, University of Geneva. This contribution is based in part on Overlapping Copyright and Trademark Protection: A Call for Concern and Action, University of Illinois Law Review Slip Opinions 25 (2014) and Contributory Trademark Infringement on the Internet: Do We Finally Know What Intermediaries Need To Know?, in Research Handbook on E-Commerce 211 (John A. Rothchild ed. 2016). We would like to thank the students of the International Intellectual Property Colloquium, spring 2020, Columbia Law School, and Barton Beebe for useful comments. The bottle designs may be protectable under copyright as “works of artistic craftsmanship,” see Copyright Act, 17 U.S.C. § 102(a)(5) (2019) (enumerating category of pictorial, graphic and sculptural works, of which works of artistic craftsmanship are a subset, see §101 (definition of “pictorial, graphic or sculptural work”). See, e.g., DC Comics v. Towle, 802 F.3d 1012 (9th Cir. 2015) (protecting Batmobile as a copyrighted character). For example, MICKEY MOUSE (design mark of the Mickey Mouse cartoon character), Registration No. 3598848 (registered, inter alia, for Audio and visual recordings, namely, animated and live action motion picture films, and pre-recorded compact discs all featuring music, stories and family entertainment; computer game discs and software featuring music

Downloaded from https://www.cambridge.org/core. 307 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

308

Jane C. Ginsburg and Irene Calboli

the same object is both an original work of authorship but also serves to identify and distinguish one producer’s goods or services. In theory, the regimes promote different objectives: creativity in the case of copyright, and source identification (or more abstractly, honest market information) in the case of trademarks. The contours of protection differ as well. U.S. copyright law covers original works of authorship; these include a large array of creations, including pictures, architectural works, computer software, and even some works of applied art, so long as they are minimally creative. By contrast, under U.S. law, a trademark need not be creative (though many in fact might be), but it must be “distinctive,” because a sign that does not identify and distinguish goods or services does not fulfil its function of identifying products in the market place as originating with a single source.6 Accordingly, even though creativity is different from distinctiveness, some trademarks can be original works of authorship and some creative works can be distinctive signs. Today, the convergence of copyright and trademark rights in the same objects may be even more prevalent, as artistic elements may enhance the attractiveness to consumers of a variety of goods and services.7 This trend toward subject matter convergence has been further accentuated by the growth of copyright and trademark merchandising – the use of trademarks and copyrighted works on promotional and collateral products – as a common practice not only to increase revenues in creative industries but also in building a brand. IP lawyers routinely advise their clients that, since copyrights and trademark rights “aim to protect different subject matter and

6

7

videos, trivia games, interactive games, activities and stories for children[; consumer electronics, namely, MP3 players, computer mouse, audio speakers, microphones, headphones, and digital cameras]; sunglasses); BATMAN (design mark of the Batman cartoon character), Registration No. 4871024 (registered, inter alia, for Entertainment services, namely, providing a website featuring non-downloadable film clips, photographs and other multimedia materials featuring entertainment; conducting contests online); RONALD MCDONALD (design mark of the Ronald McDonald character), Registration No. 0874861 (registered, inter alia, for Restaurant services, drive-in restaurant services). See 15 U.S.C. § 1051 (2006); see also Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4, 9 (2d Cir. 1976) (distinguishing fanciful, arbitrary, and suggestive marks as inherently distinctive and descriptive marks as marks that required secondary meaning in order to be protected). U.S. courts addressed the possibility of cumulating the two forms of protection on the same subject matter over a half century ago. In 1962, in Planters Nut & Chocolate Co. v. Crown Nut Co., 305 F.2d 916 (C.C.P.A. 1962), the Court of Customs and Patent Appeals stated that copyright and trademark protection could co-exist in “a fanciful picture of a humanized peanut . . . that . . . is also used as a design mark to identify and distinguish the source of peanuts.” 1 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 6:18 (5th ed. 2016) [hereinafter McCarthy]. In 1979, the District Court for the Southern District of New York repeated in Frederick Warne & Co. v. Book Sales, Inc., 481 F. Supp. 1191, 1196 (D.C.N.Y. 1979), that “dual protection under copyright and trademark law is particularly appropriate for graphic representations of characters” with respect to the famous character of Peter Rabbit. See infra 25.2.1.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

Overlapping Copyright and Trademark Protection

309

restrict different activity,”8 it is to the advantage of IP owners “to have access to both sets of protection, whenever possible.”9 However, when trademark and copyright subject matter protection overlap, courts should take care to ensure that each right remain within its proper domain: temporally limited protection for works of authorship, and protection against only such copying as leads to confusion of source identification for trademarks. Otherwise rightowners might succeed in conferring perpetual protection on works of authorship, and in-gross protection on distinctive signs. But if courts should guard against the kind of boundary-blurring of IP regimes that risks “mutant”10 copyrights and trademarks, the convergence of copyright and trademark subject matter might also usefully facilitate the importation of limiting doctrines from copyright to trademark. Ultimately, exploitation of copyright/trademark overlaps may herald a transition towards an integrated IP framework that reflects changes in marketing and product development and underscores the increasingly indistinct lines between creative works and distinctive signs. But this framework need not imply a one-way ratchet toward ever-increasing protection. Rather, as we elaborate in this chapter, the reciprocal influence of copyright and trademark protection can lead to a new balancing of trademark rights through importation of copyright limitations based on fair use and free expression, as well as through incorporation of U.S. online service provider liability limitations from copyright into trademark.

25.2 one side of the transition: when overlapping rights might lead to more protection 25.2.1 Creative Works Meet Trademark Protection: The Road Toward Eternal Mickey Mouse? In this chapter, we focus primarily on overlapping protection of visual characters. Visual characters are subject matter for which copyright and trademark protection are most commonly cumulated. Moreover, the two rights tend to be licensed together.11 Still, even if the subject matter may merge, the nature of the rights conferred, their duration, and their rationales all diverge. Trademarks symbolize the goodwill of the producer, and are largely protected only insofar as copying that 8

9 10 11

Craig S. Mende and Belinda Isaac, When Copyright and Trademark Rights Overlap, in Overlapping Intellectual Property Rights 137, 158 (Neil Wilkof & Shamnad Basheer eds., 2012). Id. Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23, 34 (2003). See, e.g., Mende and Isaac, supra note 8, 158. See also DC Comics v. Towle, 802 F.3d 1012 (9th Cir. 2015), cert. denied, 136 S. Ct. 1390 (2016) (holding the “Batmobile” both a copyrightable character and a protected trademark).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

310

Jane C. Ginsburg and Irene Calboli

symbol is likely to confuse consumers as to the source or approval of the goods or services in connection with which the mark is used.12 The object of protection is the producer’s goodwill (in theory, as a proxy for consumer source identification), not (again, in theory) the mark per se. Copyright, by contrast, is a right “in gross” allowing its owner to prohibit the copying of the work without regard to source confusion.13 Copyright protects the work of authorship itself, not the identification of that work with a single, if anonymous, source of origin.14 Pursuant to the U.S. Constitutional grant to Congress of power to secure authors’ exclusive rights “for limited times,” copyright lasts for a term of years;15 trademarks are protected for so long as they are used and continue to represent a single producer’s goodwill. Subject to those prerequisites, registered trademarks may be renewed indefinitely.16 This durational disparity prompts the question whether a trademark owner may effectively perpetuate the life of the copyright in a visual character by controlling the use of the trademark in the same image(s). For example, once the copyright in Mickey Mouse – or more accurately, in the motion picture Steamboat Willie, in which Mickey’s forebear appears – expires, can Disney successfully invoke its trademark rights in various depictions of The Mouse to prevent third parties from exploiting the now public domain Steamboat Willie (including its depiction of Mickey’s forebear)?17 Thus posed, the question suggests a pernicious outcome, which has been criticized by several in the academy.18 The Supreme Court itself expressed concerns over this outcome in Dastar v. Twentieth-Century Fox.19 Ultimately, the Court rejected the attempt by the producer of a copyright-expired motion picture to effectively end-run the work’s public domain status by seeking injunctive relief under trademark law to prevent “reverse passing off.” Yet, the Court did not forbid the former owner of a copyright-expired original work of authorship from claiming 12

13 14

15 16

17 18

19

15 U.S.C. § 1114; § 1125(a) (2013). Famous marks also enjoy dilution protection under § 1125(c), which affords relief for blurring or tarnishing the distinctive character of the famous mark, even in the absence of confusion. See 17 U.S.C. § 501 (2013) (listing the elements of an infringement claim). Aside from the narrow category of Works of Visual Art, the U.S. Copyright Act imposes no obligation to credit the author as the creator of the work. U.S. Const. art. I, § 8, cl. 8. 15 U.S.C. § 1059(a) (2018). The statute imposes no limit as to the number of times the registrant may renew. See generally Joseph P. Liu, The New Public Domain, 2013 U. Ill. L. Rev. 1395, 1427–40. See, e.g., Irene Calboli, Overlapping Rights: The Negative Effects of Trademarking Creative Works, in The Evolution and Equilibrium of Copyright in the Digital Age 52 (Susy Frankel and Daniel Gervais eds., 2014). Viva R. Moffat, Mutant Copyrights and Backdoor Patents: The Problem of Overlapping Intellectual Property Protection, 19 Berkeley Tech. L. J. 1473 (2004); Estelle Derclaye and Matthias Leistner, Intellectual Property Overlaps: A European Perspective 12–15, 47–60, 130–338, 200–05, 237–54, 290–92, 295–303, 306–33, and authorities cited therein (2011) (surveying copyright/trademark overlaps in the EU and member states). Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23 (2003).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

Overlapping Copyright and Trademark Protection

311

the work as a trademark to designate the source or sponsorship of its goods or services. Nor does Dastar prevent third parties from appropriating copyright-expired characters or other works as trademarks. A pre-Dastar decision, Frederick Warne & Co. v. Book Sales, Inc., a case involving a famous visual character, illustrates the point. In Warne, the court made it clear that the public domain status of the character of Peter Rabbit “should not preclude protection under the trademark laws so long as it is shown to have acquired independent trademark significance, identifying in some way the source or sponsorship of the goods.”20 Warne’s use of the Sitting Rabbit on book covers and as a colophon on pages of stories that did not feature that particular image of the Sitting Rabbit exploited the image in ways distinct from that image’s role in the story from which Warne extracted it. Thus, this use signaled that this incarnation of the character was source-identifying rather than narrative. The court cautioned, however, that the first publisher could not bootstrap all Peter Rabbit images to the single image used in the colophon and book covers. If Warne implies that trademark rights might attach to a particular iteration of the character, leaving other visual representations in the public domain, might a trademark owner, consistently with Dastar, nonetheless claim trademark rights not just in one representation of the character, but in all iterations of that character? For example, if, through Disney’s efforts to build brand identification in Steamboat Willie, by registering the character as a logo, and by disseminating a short video depicting Steamboat Willie morphing into Mickey Mouse, Steamboat Willie were to become a brand in its own right, could Disney successfully claim that its unlicensed appearance would confuse the public as to the source of the copies or transmissions of the work containing any incarnation of the character? But in that event, the trademark claim would clash with the public domain principle of copyright, and Dastar establishes that the latter prevails.21 The public domain makes the character free for competition. Assuming the character has also served as a trademark, trademark law should enable the former copyright owner to ensure that free competition be fair competition. But remedies short of full injunctions may achieve “fairness”: A court should decline to grant the trademark owner any relief other than the clear labelling of disclaimers as to the source of the competing goods.22 However, this solution may not be so straightforward. Creative industries routinely turn to trademark rights as a complement to copyright, particularly for characters, titles, songs, and video clips. Moreover, even though the current Trademark Manual of Examining Procedures (TMEP) provides for the refusal to register a mark 20 21 22

Frederick Warne & Co. v. Book Sales, Inc., 481 F. Supp. 1191, 1196 (D.C.N.Y. 1979). Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23 (2003). See Liu, supra note 17, at 1433 (“[A]t most the appropriate remedy would be a disclaimer, which would address the confusion without limiting the ability of a third party to use the underlying work.”).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

312

Jane C. Ginsburg and Irene Calboli

consisting of “the title, or a portion of a title, of a single creative work,” this exclusion does not apply to titles which have “been used on a series of creative works.” 23 The TMEP imposes a higher burden for applicants seeking to register series titles as marks.24 Still, a search in any trademark registry will reveal several marks consisting of series titles and/or visual characters that are, or were, protected under copyright.25 Rightholders thus seek the option to sue for both copyright and trademark infringement, or simply for the latter, even though the likelihood of success of such claims may vary.

25.2.2 Distinctive Signs Meet Copyright Protection: The Road to More Protection for Strawberry Shortcake and M&Ms? Trademark and copyright protection may also converge through transforming a trademark into a copyrightable work. This commonly involves the adaptation of trademarks into the object of copyright protection through endowing trademarkprotected images not only with brand significance, but also with personalities and other attributes that convert the source-identifier into a literary, artistic, or audiovisual character. This strategy is the logical endpoint of the recognition since at least the 1970s, that a trademark may be a thing of value independently of any particular goods or services with which it is associated.26 23

24 25

26

See TMEP § 1202.08 (8th ed. Oct. 2018) (directly addressing concerns about the copyright/ trademark protection overlap). TMEP § 1202.08(d)(iii). See, e.g., THE LORD OF THE RINGS (multiple registrations for standard character mark for a large number of goods and services); THE HOBBIT (multiple registrations for standard character mark for a large number of goods and services); HARRY POTTER (multiple registrations for standard character mark for a large number of goods and services); MICKEY MOUSE (design mark of the Mickey Mouse cartoon character), Registration No. 3598848; BATMAN (design mark of the Batman cartoon character), Registration No. 4871024; RONALD MCDONALD (design mark of the Ronald McDonald character), Registration No. 0874861; CINDERELLA (design mark of the Cinderella character), Registration No. 4874957; SNOW WHITE (design mark of the Snow White character), Registration No. 4552049; MINNIE MOUSE (design mark of the Minnie Mouse character), Registration No. 3915194; DONALD DUCK (design mark of the Donald Duck cartoon character), Registration No. 1119631; PRINCESS JASMINE (design mark of the Princess Jasmine character), Registration No. 4412568; ANGRY BIRDS (design mark of the Angry Birds video game characters), Serial No. 88215889; HELLO KITTY (design mark of the Hello Kitty character), Registration No. 1370105. See, e.g., Boston Professional Hockey Ass’n v. Dallas Cap & Emblem Mfg., 510 F.2d 1004 (5th Cir. 1975) (acknowledging that protecting the trademark in itself (on patches to be sewn on clothing) is a shift: “[a]lthough our decision here may slightly tilt the trademark laws from the purpose of protecting the public to the protection of the business interests of plaintiffs, we think that the two become so intermeshed when viewed against the backdrop of the common law of unfair competition that both the public and plaintiffs are better served by granting the relief sought by plaintiffs.”); Panavision Int’l v. Toeppen, 141 F.3d 1316 (9th Cir. 1998) (“Toeppen made a commercial use of Panavision’s trademarks. It does not matter that he did not attach the marks to a product. Toeppen’s commercial use was his attempt to sell the trademarks

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

Overlapping Copyright and Trademark Protection

313

For example, producers of breakfast cereals and fast food chains have long exploited characters in a combination of roles – branding consumer goods or retail services as well as featuring in associated commercials, toys, and even television shows. Characters such as Strawberry Shortcake were created to serve both as copyrighted works and as trademarks. More recent iterations of trademarks-as-characters differ from their predecessors in that these trademarks were not born as characters, but later underwent specific anthropomorphizing touches that turned uncopyrightable basic shapes into pictorial works. Consider a red circle. The Supreme Court has recognized that secondary meaning can give rise to trademark protection for such a sign.27 On the other hand, whether as an identifying indicium for shoe heels, or as the red disk of the 7-UP dot, the sign is too rudimentary to survive even a cursory Copyright Office examination for originality. But accessorize the dot with arms and legs and coif it with a pair of sunglasses, and a banal form blossoms into the “Spot” character.28 Or take the rudimentary, single-colored shapes of M&Ms candies, endow them with arms and legs and a range of highly expressive (if rather gender-stereotyped) facial features, and they become “spokescandies” with pre-packaged personalities and backstories.29 Given this new life, several trademarks-as-characters have starred not only in advertising for the trademarked goods, but also in independent works such as videogames and even feature films (for example, The Lego Movie and its sequels).30 Yet, this type of overlapping protection does not appear to pose the policy concerns articulated above regarding claims of trademark rights in works of authorship, probably because the flip side of the copyright/trademark coin does not risk compromising the public domain. In particular, the practice of converting trademarks into copyrightable works effectively protects the mark against copying that may not be source-confusing. Likewise, protection in gross for trademarks-ascharacters does not raise the same level of policy concerns so long as the mark manifests sufficient authorship to qualify as a “work.”31 As we have mentioned in

27 28

29

30

31

themselves.”). For an overview of the practice of trademark merchandising, see Irene Calboli, The Case for a Limited Protection of Trademark Merchandising, 2011 U. Ill. L. Rev. 865 (2011). Mishawaka Rubber & Woolen Mfg. Co. v. S. S. Kresge Co., 316 U.S. 203 (1942). 7Up, DR PEPPER SNAPPLE GROUP, https://www.drpeppersnapplegroup.com/brands/7up (last visited July 22, 2020). See generally Characters, M&M’s, https://www.mms.com/en-us/experience-mms/characters (last visited July 24, 2020) (presenting the characters). See The Lego Movie, http://www.thelegomovie.net/ (last visited July 24, 2020) (presenting the various Lego movies and the characters). Some concerns in this respect have been expressed in the context of parallel imports and the application of the principle of copyright exhaustion to consumer goods such as chocolates or liqueur bottles. The claim that the unauthorized importation of these products could represent copyright infringement (in the presence of a domestic rule of national copyright exhaustion versus a rule of international trademark exhaustion) has raised concerns in Australia and Canada. See R. A. Bailey & Co. Ltd. v. Boccaccio Pty. Ltd. (1986) 6 IPR 279 (Austl.); Polo/ Lauren Co. LP v. Ziliani Holdings Pty. Ltd. [2008] F.C.A.F.C. 195 } 5 (Austl.); Euro-Excellence

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

314

Jane C. Ginsburg and Irene Calboli

the Introduction (25.1), the influence of copyright protection over trademarks can lead to a new balancing of these rights by importing the limitations applicable in one domain to the other. Certainly, there are several benefits to a rightholder arising from this practice. First, there is little if any downside to recasting a trademark as a copyrightable work of authorship. Copyright does not diminish the protection the characters would enjoy as a matter of trademark law. In particular, because copyright is a right in gross, it strengthens the coverage available even to trademarks that are sufficiently famous to qualify for protection against dilution. Copyright is infringed by copying; trademark infringement requires a showing that the copying causes a likelihood of confusion or (for dilution) of association that is likely to “blur” or “tarnish” a famous mark.32 Even if it is true that U.S. trademark law has increasingly protected brand symbols in their own right, independent of the goods or services in connection with which they initially appeared, copyright protection directly protects the work per se without the need to prove established goodwill. The U.S. copyrights will expire ninety-five years from the first publication of the trademarks-as-works, if they are works made for hire, or seventy years after the death of their authors for other work. But during that very long stretch, the trademark owner of the work will have ample time to update the work and start another copyright clock rolling.33 Copyright lives of visual works in particular may last far longer than the copyright term of the works’ initial incarnation. Many characters evolve over time, and later manifestations may depart sufficiently from the original version to qualify as copyright-protected derivative works. Take, for instance, the character of the Michelin Man, “Bibendum.” First appearing in Michelin advertisements in the late nineteenth century, it has evolved from a bon vivant blob into a still-robust but more sportif assemblage of thicker forms showing greater

32 33

Inc. v. Kraft Canada Inc., [2007] 3 S.C.R. 20, 2007 SCC 37 (Canada). For a review of this aspect of overlapping rights, see Irene Calboli, The (Avoidable) Effects of Territorially Different Approaches to Trademark and Copyright Exhaustion, in Trademark Protection and Territoriality Challenges in a Global Economy 151 (Irene Calboli and Edward Lee eds., 2014). In the U.S., after the Supreme Court’s decision in Kirtsaeng v. Wiley, 568 U.S. 519, 133 S. Ct. 1351, 1358 (2013), overlapping trademark and copyright protection no longer seems to create concerns in this context, as the Court has clarified that the U.S. follows the principle of international copyright exhaustion; in other words, it is no longer possible for rightholders to end-run the trademarks grey market by asserting copyright protection against importing copies lawfully made abroad. 15 U.S.C. §§ 1114, 1125(a), 1125(c) (2012). 17 U.S.C. §§ 302, 304 (2019) (duration of copyright, including for works made for hire). Copyright in a derivative work extends only to the new matter, and does not affect the duration of the underlying material, § 103(b). In theory, therefore, the character as it originally appeared becomes free for unlicensed exploitation once the original copyright expires. But the original character may lack commercial value relative to its updates. So even if only the new matter added through the updates may claim copyright, the practical inextricability of the old and new content may enable copyright owners to prolong protection for the character.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

Overlapping Copyright and Trademark Protection

315

differentiation in torso, limbs, and especially facial characteristics. Both copyright and trademark law will protect the character in its revised instantiations.34 Finally, copyright offers significant remedial advantages over trademarks. The principal remedy for trademark infringement is injunctive relief.35 Damages require a showing of actual confusion; a trademark owner would be disinclined to wait to seek relief until consumers can be shown to have been confused in fact.36 By contrast, in copyright law, damages are a traditional remedy, and the prospect of an award of statutory damages of up to $150,000 per work willfully infringed, can provide significant monetary relief.37

25.3 the other side of the transition: when overlapping rights lead to less protection (and sometimes more laughs) We have seen how rightsowners may in different ways endeavor to strengthen their rights by supplementing one legal regime with another when the subject matter overlaps. But the reciprocal relationship of copyright and trademark law may also import some of copyright’s limiting doctrines into trademark law. We offer two examples of this phenomenon: first, the influence of copyright principles of fair use and free expression on expressive uses of trademarks; second, the recasting of trademark principles of secondary liability in light of amendments to the Copyright Act to immunize online service providers from direct or secondary liability for hosting infringing content. 25.3.1 Limitations Based in Freedom of Expression Since the late 1980s, when works of authorship are alleged to infringe trademarks, courts in the U.S. have progressively imposed a free-expression thumb on the scale of likelihood of confusion analysis.38 These decisions did not always expressly cite the copyright fair use defense, but the considerations underlying the copyright doctrine ultimately informed the trademark analysis as well. The spillover effect of copyright doctrine into the determination of trademark infringement may indeed have been inevitable, as several of the cases in which a copyright fair use defense prevailed involved copyright/trademark overlapping subject matter, and accordingly 34

35

36

37

38

In the U.S., Michelin may register revised Bibendums with the Patent and Trademark Office, but may also claim protection under section 43(a) of the Lanham Act for unregistered versions. 15 U.S.C. § 1116 (2008); 5 McCarthy, supra note 7, § 30:1 (“A permanent injunction is the usual and normal remedy once trademark infringement has been found in a final judgment.”). 15 U.S.C. § 1117(a) (2008) (relief for “any damages sustained” indicates that the trademark owner would in fact have had to have lost sales). 17 U.S.C. § 504(b)(c) (2012). A court may award statutory damages if the work was registered with the Copyright Office before the infringement occurred, id. § 412. See, e.g., Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

316

Jane C. Ginsburg and Irene Calboli

coupled copyright and trademark claims. Thus, it is not surprising that once a court has found the use to be “transformative” and to promote speech and/or learning, and thus most often to be “fair” in the copyright sense, that court is unlikely to find the same activity to violate the copyright holder’s trademark in the copied work. In particular, even though copyright is a property “right in gross,” while trademark is not (or is not supposed to be), the trademark and copyright fair use limitations have this in common: the users of the mark or of another author’s work are engaged in independent economic or creative activity. They are not simply redistributing another’s work of authorship or appending another’s mark to the same or confusingly similar goods. Rather, at least for traditional copyright fair use, they are making a new work, or, on the trademark side, proclaiming their own goods or communicative activities. By the same token, the uses do not unfairly usurp the copyright or trademark owner’s markets. A devastating review may dampen desire for the critiqued work, but that kind of harm is not cognizable in copyright. Similarly, a comparative advertisement may persuade consumers of the superior merits of the competitor’s goods or services, but if the advertisement is truthful no Lanham Act claim lies. Lawful unauthorized uses of trademarks enjoy a long pedigree. In 1924, in a case concerning the labelling of lawfully purchased and rebottled perfume, Justice Holmes famously declared: “When the mark is used in a way that does not deceive the public we see no such sanctity in the word as to prevent its being used to tell the truth. It is not taboo.”39 The Lanham Act codified part of this precept by permitting the use of descriptive terms in a registered trademark for the purpose of description, and not as a trademark.40 For example, if TRULY BEAUTIFUL is a trademark for a fragrance, the registrant cannot prohibit a competitor from informing the public that its fragrance, call it FRANKLY GORGEOUS, will make the wearer “feel truly beautiful.”41 But the Lanham Act does not explicitly exempt a competitor’s use of TRULY BEAUTIFUL to describe the competitor’s product, for example, to proclaim that FRANKLY GORGEOUS will make the wearer more alluring than TRULY BEAUTIFUL. Nor does the Lanham Act explicitly allow a non-competitor, such as the (fictitious) magazine Glamorous, to elicit its readers’ perfume preferences by asking them whether TRULY BEAUTIFUL is their most or least favorite fragrance. Former Ninth Circuit Judge Kozinski dubbed these latter denominations “nominative fair use,” and the name has remained.42 39

40 41

42

Prestonettes, Inc. v. Coty, 264 U.S. 359, 368 (1924) (citing Delaware & Hudson Canal Co. v. Clark, 80 U.S. 311, 327 (1871)). 15 U.S.C. § 1115(b)(4) (2002). See, e.g., United States Shoe Corp. v. Brown Grp., Inc., 740 F. Supp. 196 (S.D.N.Y. 1990) (plaintiff’s slogan “looks like a pump, feels like a sneaker” not infringed by advertising text stating that “when we say it feels like a sneaker, we’re not just stringing you along”). See New Kids on the Block v. News Am. Publ’g, 971 F.2d 302 (9th Cir. 1992); Playboy Enter, Inc. v. Welles, 279 F.3d 796 (9th Cir. 2002); Century 21 Real Estate Corp. v. Lendingtree, Inc., 425 F.3d 211 (3d Cir. 2005) (following Ninth Circuit in adopting the doctrine but applying a

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

Overlapping Copyright and Trademark Protection

317

The nominative fair use defense to trademark infringement proceeds through an analysis of factors. As we will see, these factors complement and echo the first, third, and fourth copyright fair use factors.43 Nonetheless, a notable early case finding nominative fair use in trademark law did not allege copyright infringement, though it did concern a parody. There, the song “Barbie Girl” by the Danish one-hit group Aqua in 1997 foisted on the airwaves lyrics like the following catchy refrain: “I’m a Barbie girl, in a Barbie world. Life in plastic, it’s fantastic. You can brush my hair, undress me everywhere. Imagination, life is your creation . . . . I’m a blond bimbo girl, in a fantasy world. Dress me up, make it tight, I’m your dolly.” To which the bass in the group would interject in a froggish croak (Aqua’s album was, after all, called Aquarium): “C’mon Barbie, let’s go party!”44 Holding that “the trademark owner does not have the right to control public discourse whenever the public imbues his mark with a meaning beyond its sourceidentifying function,” the Ninth Circuit rejected Mattel’s claim of likelihood of confusion.45 The court balanced the “public interest in free expression” against the “public interest in avoiding consumer confusion,” and accorded the former decisive weight unless the song title’s appropriation of Barbie “has no artistic relevance to the underlying work whatsoever, or, if it has some artistic relevance, unless the title explicitly misleads as to the source or the content of the work.”46 Observing that the Barbie doll was the target of the song, the court held the group was entitled to identify the butt of its joke, and had done nothing to mislead the public into thinking that Mattel authorized the song. (Ironically, Mattel subsequently licensed “Barbie Girl” for advertising Barbie dolls.)47 In its analysis, the Ninth Circuit

43

44

45 46 47

different test and assigning a different conceptual role to the nominative use doctrine in trademark law); Toyota Motor Sales, U.S.A., Inc. v. Tabari, 610 F.3d 1171 (9th Cir. 2009) (distinguishing from New Kids on the Block by holding that a defendant asserting a nominative fair use defense need only show that it used the mark to refer to the trademarked good, at which point the burden then reverts to the plaintiff to show a likelihood of confusion). Other circuits reach the same result through similar reasoning, albeit without applying the label “nominative fair use.” See, e.g., Universal Commun. Sys. v. Lycos, Inc., 478 F.3d 413, 424–25 (1st Cir. 2007); Swarovski Aktiengesellschaft v. Bldg. No. 19, Inc., 704 F.3d 44 (1st Cir. 2013) (refraining from endorsing any particular approach to the nominative fair use doctrine). See generally McCarthy, supra note 7, § 23:11 (5th ed.). See International Information Systems Sec. Certification Consortium, Inc. v. Security University, LLC, 823 F.3d 153 (2nd Cir. 2016); GlaxoSmithKline LLC v. Laclede, Inc., 2019 WL 293329 (S.D.N.Y. Jan. 23, 2019). Aqua, “Barbie Girl” (1997), available on Youtube at https://www.youtube.com/watch?v= ZyhrYis509A (last visited July 22, 2020). Mattel, Inc. v. Universal Music Int’l, 296 F.3d 894, 900 (9th Cir. 2002). Id. at 902 (citing Rogers v. Grimaldi, 875 F.2d 994, 999 (2d Cir. 1989)). See Stuart Elliott, Years Later, Mattel Embraces “Barbie Girl,” New York Times, August 26, 2009, https://mediadecoder.blogs.nytimes.com/2009/08/26/years-later-mattel-embraces-barbiegirl/ (last visited July 22, 2020).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

318

Jane C. Ginsburg and Irene Calboli

followed the precedent of the Second Circuit in the 1989 case Rogers v. Grimaldi,48 where the court found that the use of a mark in an artistic context is not infringement unless it “has no artistic relevance to the underlying work whatsoever” or unless it “explicitly misleads as to the source or content of the work.”49 Just a few months after the first Barbie case, a second Barbie case followed. This time, Mattel sought to enjoin the dissemination of the “Food Chain Barbie” series of photographs, which depicted the scantily clad dolls victimized by a variety of kitchen appliances. In Mattel, Inc. v. Walking Mountain Products,50 the claims implicated both copyright (the reproduction right in the image of the dolls) and trademarks (use of the Barbie name in the titles of the photographs). In granting summary judgment on the copyright claim, the court ruled on the first fair use factor that the Food Chain series “parod[ies] Barbie and everything Mattel’s doll has come to signify.” Undoubtedly, [the court continued] “one could make similar statements through other means about society, gender roles, sexuality, and perhaps even social class. But Barbie, and all the associations she has acquired through Mattel’s impressive marketing success, conveys these messages in a particular way that is ripe for social comment.”51 On the third factor, the court stressed that parodies need not restrict themselves “to take the absolute minimum amount of the copyrighted work possible,” and thus that, in context, reproducing the entire doll did not undermine the fair use defense.52 On the fourth factor, the court found it unlikely that Mattel would enter the market for “adult-oriented artistic photographs of Barbie.”53 The Ninth Circuit approvingly cited another Barbie copyright parody decision, Mattel, Inc. v. Pitt, in which the defendant’s “dungeon doll” website offered to customize Barbie dolls in a variety of dominatrix modes.54 The differently accessorized dolls arguably violated Mattel’s exclusive right to create derivative works based on Barbie, and the website’s photographs allegedly violated the reproduction right. Here, the Southern District of New York considered the statutory fair use factors, and ruled the defendant’s use and context of Barbie sufficiently transformative and unlikely to supplant one of Mattel’s markets for the work. As the district court quipped, “To the Court’s knowledge, there is no Mattel line of S&M Barbie.”55 Having held the defendant’s use to be “fair” as a matter of copyright law, notably because of the expressive values the parody advanced, the Ninth Circuit in “Food Chain Barbie” effectively sealed the fate of the trademark claim.56 With respect to

48 49 50 51 52 53 54 55 56

Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989). Id. at 999. Mattel, Inc. v. Walking Mountain Prods., 353 F.3d 792 (9th Cir. 2003). Id. at 802. Id. at 804. Id. at 806. Mattel, Inc. v. Pitt, 229 F. Supp. 2d 315 (S.D.N.Y. 2002). Id. at 322. Mattel, Inc. v. Walking Mountain Prods., 353 F.3d 792 (9th Cir. 2003).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

Overlapping Copyright and Trademark Protection

319

Food Chain’s incorporation of Barbie in the title of the series, the court relied on its prior decision in the “Barbie Girl” song case to reject a likelihood of confusion claim.57 Finally, the Ninth Circuit also rejected Mattel’s trade dress claim and stated: “Forsythe used Mattel’s Barbie figure and head in his works to conjure up associations of Mattel, while at the same time to identify his own work, which is a criticism and parody of Barbie. Where use of the trade dress or mark is grounded in the defendant’s desire to refer to the plaintiff’s product as a point of reference for defendant’s own work, a use is nominative.”58 “Conjure up,” notably, is a standard formulation of the minimum extent of copying that the copyright fair use caselaw permits in a parody.59 In other words, the court’s analysis of the second nominative fair use factor echoed its treatment of the third copyright fair use factor.60 By the same token, having found as a matter of copyright law that the defendant was not usurping a potential market for “adult photos” that Mattel was likely to enter, the court ruled on the third nominative fair use factor that “it is highly unlikely that any reasonable consumer would have believed that Mattel sponsored or was affiliated with [the defendant’s] work.”61

25.3.2 The Spillover Effect of Limitations of Copyright Liability for Online Service Providers on Trademark Liability for Contributory Infringement The subject matter convergence between copyright and trademark law is also relevant in the area of contributory infringement, in particular on the internet. Notably, a platform, such as eBay, Amazon, and the like, some of whose thirdparty vendors might offer infringing goods, could face different liability regimes for the same act of hosting if the goods are counterfeit copies of copyrighted works. For example, if the seller is proposing unlawfully copied items falsely labelled as branded jewelry, different standards of conduct might govern the host’s exposure to injunctive or monetary relief for the two different kinds of contributory infringement.62 57 58 59

60 61 62

Id. at 804 Id. at 810. Id. at 800 (citing Dr. Seuss Enters., L.P. v. Penguin Books USA, Inc., 109 F.3d 1394, 1400 (9th Cir. 1997)). Id. at 811. Id. at 812. The host would not incur this risk in the EU because the 2000 eCommerce Directive (Directive 2000/31/EC, OJ L 178) imposes a transversal liability limitation regime; in the U.S., these statutory liability limitations exist in the Copyright Act, but not in the trademark law. It is still unclear if the current liability regime will be affected by the new EU Copyright Directive, in particular Article 17 (Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on Copyright and Related Rights in the Digital Single Market and Amending Directives 96/9/EC and 2001/29/EC, OJ L 130), even though online marketplaces are explicitly excluded from the scope of the Directive.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

320

Jane C. Ginsburg and Irene Calboli

In trademark law, court decisions have traditionally referred to the test that the Supreme Court articulated in Inwood Laboratories v. Ives Laboratories to assess contributory infringement. A “manufacturer or distributor” could be found liable for contributory trademark infringement if “[it] intentionally induces another to infringe a trademark, or if it continues to supply its product to one whom it knows or has reasons to know is engaging in trademark infringement.”63 But the Court did not further detail the meaning of “know[ing] or ha[ving] reasons to know.” Thus, due to the uncertainty concerning the substance and proof of actual or constructive knowledge, a few courts have adopted another assessment, in effect looking to whether intermediaries are primarily non-infringing businesses or simply “bad guys.”64 This approach recalls that of the Supreme Court’s decision in Sony Corp. of America v. Universal City Studios, Inc.65 regarding copyright infringement. The Court there evaluated contributory infringement in copyright law by inquiring whether the accused technology used was capable of “substantial non-infringing uses.”66 The Second Circuit in Tiffany v. eBay67 explicitly referenced Sony in its analysis. eBay runs an internet marketplace, in which third parties sell items directly to internet users, with eBay making a profit from each transaction. eBay operates several buyer-protection programs in order to safeguard consumers and the legitimacy of the transactions that take place on its marketplace. It has also established a “Trust and Safety” department and a “fraud-engine” to detect illegal listings, as well as conducting manual reviews of listings.68 In addition, eBay offers a “notice and take down” system by which third parties can report instances of potential infringement and eBay will remove the potentially infringing listing within twelve hours of receiving a notice.69 Tiffany became aware that some eBay users were selling counterfeit Tiffany items and sued eBay for, inter alia, contributory trademark infringement.70 Both the Southern District of New York,71 and the Second Circuit, however, found that eBay was not liable for contributory trademark infringement.72 In 63 64

65 66

67 68 69 70 71

72

Inwood Labs., Inc. v. Ives Labs., Inc., 456 U.S. 844, 854 (1982). See Stacey L. Dogan, “We Know it When We See It”: Intermediary Trademark Liability and the Internet, 2011 Stan. Tech. L. Rev. 7 (2011) (comparing decisions in the area of secondary trademark infringement with similar cases in copyright law, including Sony’s progeny). Sony Corp. v. Universal City Studios, Inc., 464 U.S. 417 (1984). Id. at 442 (“The question is thus whether the Betamax is capable of commercially significant noninfringing uses.”). Tiffany, Inc. v. eBay, Inc., 600 F.3d 93 (2d Cir. 2010). Id. at 98–99. Id. Id. at 103. Tiffany, Inc. v. eBay, Inc., 576 F.Supp.2d 463 (S.D.N.Y. 2008). The district court focused much of its analysis on eBay’s efforts to suppress counterfeiting by implementing its VeRO program, fraud engine, and by conducting manual reviews of auction items. Id. Tiffany, 600 F.3d at 93.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

Overlapping Copyright and Trademark Protection

321

particular, the Second Circuit noted that eBay had removed many listings upon receiving take down notices and thus it did not “continue to supply” its service to the infringers with respect to those listings.73 The court also determined that eBay lacked sufficient knowledge to satisfy the test for contributory trademark liability with respect to the listings that it did not remove.74 Overall, eBay had only “generalized knowledge” that counterfeit items were being sold on its website; this did not satisfy the Inwood test, which requires “specific knowledge.”75 The Second Circuit also acknowledged that the Supreme Court in Inwood did not “establish the contours of the ‘knows or has reason to know’ prong.” The court thus evoked the Supreme Court’s substantial non-infringing use standard from Sony,76 concluding that “had the Inwood standard applied in Sony . . . [it] would have required knowledge by Sony of ‘identified individuals’ engaging in infringing conduct.”77 The court also rejected Tiffany’s argument that eBay had been willfully blind, and emphasized that eBay undertook significant efforts to combat the counterfeiting activities that were taking place on its website. Indeed, the measures eBay took to respond to notices regarding eBay’s sellers’ infringements closely resembled the “notice and take down” regime established under section 512(c) of the copyright act78 to immunize compliant online services from direct or contributory copyright infringement.

25.4 conclusion In conclusion, in the reciprocal relationship between copyright and trademarks, we have seen how these rights can both reinforce as well as limit each other. Ultimately, in an economy and society that is more visually oriented than ever before, producers rely on creative and aesthetic features to differentiate their commercial products. Naturally, they seek protection for these features under multiple IP rights. This has increased the instances of overlapping copyright and trademark protection. Yet, in this chapter we have described how, while on the one hand, overlapping rights can significantly enhance the protection that would otherwise be available under only one set of rights, on the other hand, the limiting doctrines of one right, notably copyright law, can offer additional means to cabin trademark claims.

73

74 75 76 77 78

Id. at 107 (“For contributory trademark infringement liability to lie, a service provider must have more than a general knowledge or reason to know that its service is being used to sell counterfeit goods. Some contemporary knowledge of which particular listings are infringing or will infringe in the future is necessary.”). Id. at 106. Id. at 107. Sony Corp. v. Universal City Studios, Inc., 464 U.S. 417 (1984). Tiffany v. eBay, 600 F.3d at 108 (discussing Sony, 464 U.S. at 439). 17 USC § 512(c) (2010).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.035

26 Cultural Heritage and the Public Domain: What the US’s Myriad and Mayo can Teach Oslo’s Angry Boy Rochelle C. Dreyfuss1

At a high level of generality, intellectual property (IP) regimes are much alike. They grant exclusivity to promote socially desirable activity, be it creating goodwill and maintaining quality, producing expressive works, or inventing technological products and processes. Of course details differ, for each system must take account of the demands of the sector it governs. Thus the gravamen of a copyright claim – copying – would never be suitable for trademark law because trademark’s focus on diminishing search costs is inconsistent with allowing similar trademarks (even if independently produced) to confuse consumers. Similarly, trademark’s term of protection – which continues as long as the mark is in use – would not be suitable for patent law, where one generation’s output is the next generation’s input and the goal is to promote rapid progress. Still, as a theme of this volume suggests, IP regimes have much to teach one another. Even though convergence may not always be normatively desirable, each regime can learn from the experience of others. A case in point is the recent move to safeguard the public’s interest in using its cultural heritage by denying trademark protection to symbols, such as Gustav Vigeland’s Angry Boy statue, on public policy and morality grounds,2 or as Martin Senftleben puts it, on the basis of interests grounded in aesthetic theory.3 US patent law similarly curtailed the scope of protection in order to leave certain works – the building blocks of science – “free to all men and reserved exclusively to none.”4 Accomplished over a series of four Supreme Court cases,5 the limits imposed on patent eligibility have not fared well.

1 2 3

4 5

Pauline Newman Professor of Law, New York University School of Law. EFTA Court, Case-E-5/16, Municipality of Oslo (April 6, 2017) (“Vigeland”). Martin Senftleben, Public Order in the Light of Aesthetic Theory – The Copyright/Trademark Interface After Vigeland (Chapter 27 in this volume). Funk Brothers Seed Co. v. Kalo Inoculant Co., 333 U.S. 127 (1948). Bilski v. Kappos, 561 U.S. 593 (2010); Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. 66 (2012); Ass’n for Molecular Pathology v. Myriad Genetics, Inc., 569 U.S. 576 (2013); Alice v. CLS Bank Int’l, 134 S. Ct. 2347 (2014).

Downloaded from https://www.cambridge.org/core. 322 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

Cultural Heritage, IP Overlaps & the Public Domain

323

Incentives to innovate have decreased, uncertainty has increased, and there are now concerted efforts to revise the law. What the US is learning is that permissive rules on coverage, when coupled with exceptions protective of the public domain and access interests, form a far more nuanced – and effective – approach to reconciling public and proprietary interests. The same, it can be argued, is true for trademark law. This chapter begins with a description of the Vigeland decision to deny trademark protection in order to safeguard the public domain status of Norway’s cultural heritage. It then outlines the US experience with the analogous move in patent law and argues that the problems encountered in the patent realm could easily arise in the trademark sphere. Drawing on my earlier work, the chapter suggests that defenses that focus on the significance of trademarks in the marketplace of ideas represent a superior way to protect the public interest in enjoyment of the “surplus value” of trademarks – that is, their non-source identifying dimensions.6

26.1 vigeland The controversy over monopolizing cultural symbols began when the City of Oslo became concerned with the lapse of copyright protection for works by important Norwegian artists, particularly a series of three-dimensional works by Gustav Vigeland. To acquire a legal basis to protect these materials, Oslo applied to register certain Vigeland works, including Angry Boy, as trademarks. The Norwegian Intellectual Property Office (NIPO) refused registration, claiming, among other things, that the works lacked distinctive character. On review, the Board of Appeal referred several questions to the European Free Trade Association (EFTA) Court, including questions as to whether registering works of great cultural value can be refused on public policy or morality grounds. In considering the question whether culturally prominent works can be trademarked, the EFTA Court began with the problem of overlapping intellectual property rights. Referencing trademark decisions involving Beethoven,7 Leonardo Da Vinci,8 and Picasso,9 it recognized the importance of allowing “anyone [to] draw from the ideas and creative content of others without limitation,” once their works had fallen out of copyright.10 Significantly, however, the Court did not render an opinion limited to the question whether the same work can be successively 6

7 8 9 10

Rochelle C. Dreyfuss, Expressive Genericity: Trademarks as Language in the Pepsi Generation, 65 Notre Dame L. Rev. 397 (1990) [“Genericity”]. See also Rochelle Cooper Dreyfuss, We Are Symbols and Inhabit Symbols, So Should We Be Paying Rent? Deconstructing the Lanham Act and Rights of Publicity, 20 Colum.-VLA J.L. & Arts 123 (1996). Shield Mark v. Kist, C-283/01, EU:C:2003:197 (2003). Case 24 W (pat) 188/96, GRUR 1998, p. 1021. Picasso v. EUIPO, C-361/04 P, EU:C:2005:531 (2006). Vigeland, } 65.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

324

Rochelle C. Dreyfuss

protected by copyright and then trademark law (important though that question may be11). Rather, it saw its task as safeguarding the storehouse of public knowledge: Once communicated, creative content belongs, as a matter of principle, to the public domain. In other words, the fact that works are part of the public domain is not a consequence of the lapse of copyright protection. Rather, protection is the exception to the rule that creative content becomes part of the public domain once communicated.12

Trademark protection, the Court said, is antithetical to the goal of freeing works for follow-on development, because it lasts indefinitely and during that time gives right holders considerable control over subsequent uses.13 Rejecting NIPO’s reliance on lack of distinctiveness, which can be cured through use,14 the Court considered it preferable to deny registration on the public policy and morality grounds found in the Trade Mark Directive because those objections cannot be overcome.15 As to morality, which the Court saw as enforcing subjective values of the public (and perhaps the artist), it stated: However, certain pieces of art may enjoy a particular status as prominent parts of a nation’s cultural heritage, an emblem of sovereignty or of the nation’s foundations and values. A trade mark registration may even be considered a misappropriation or a desecration of the artist’s work, in particular if it is granted for goods or services that contradict the values of the artist or the message communicated through the artwork in question. Therefore, the possibility cannot be ruled out that trade mark registration of an artwork may be perceived by the average consumer . . . as contrary to accepted principles of morality.16

It then called upon NIPO to engage in a case-by-case assessment to determine whether, given the nature of the goods for which trademark registration was sought, the use would misappropriate the work or desecrate it. As to public policy – an objective value – the Court also sent the case back to NIPO, this time to determine whether “exceptional circumstances” existed: An artwork may be refused registration, for example, under the circumstances that its registration is regarded as a genuine and serious threat to certain fundamental values or where the need to safeguard the public domain, itself, is considered a fundamental interest of society.17 11

12 13 14 15

16 17

Cf. Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23 (2003). See also Senftleben, supra note 3. Vigeland, } 66. Id., }} 68 & 71. Id., } 74. Directive (EU) 2015/2436 of the European Parliament and of the Council of 16 December 2015 to approximate the laws of the Member States relating to trade marks (Recast), art. 4(1)(f ). Vigeland, } 92. Id., } 96.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

Cultural Heritage, IP Overlaps & the Public Domain

325

Given that unrestricted access to cultural symbols could permit the public to use them in objectionable ways, there is a certain oddity in denying registration based on fears of desecration and threats to fundamental values.18 Nonetheless, Vigeland has met with considerable acclaim. In particular, Martin Senftleben considers the decision “a milestone” that “prevent[s] the pursuit of profit maximization from suppressing the evolution of free, autonomous art.”19 He argues that refusing registration is superior to offering defenses to those who wish to draw on cultural symbols in follow-on work. In his view, cease and desist letters and the prospect of litigation can exert a chill on many artists (particularly those who cannot afford legal advice) and interfere with their inclination to remix, reuse, and engage in their own creative expression. Along similar lines in the US, Rebecca Curtin is fighting registration of “Rapunzel” for dolls and has questioned Disney’s control over “Cinderella,” arguing that the value of these figures belongs to the public – that anyone should be allowed to draw on images from folklore to create their own stories, toys, or other works.20 Recognizing that in the US, rejections based on public policy and morality would likely meet First Amendment challenges,21 Curtin would refuse to register cultural symbols on aesthetic functionality grounds – a basis for rejection which, like morality and public policy, is also impervious to cure.22 Admittedly, there is considerable attraction to the idea of keeping cultural symbols in the public domain. It would clearly ensure public access and avoid chilling second-generation creativity. There are, however, several potential problems with the threshold approach reflected in these examples. The US experience in delimiting the subject matter of patent protection provides a cautionary tale.

26.2 the patent law analogue To some extent, the decision to limit the subject matter of patent law stemmed from the same impulse that motivated the EFTA Court in Vigeland. For a variety of reasons, including the entry of universities into the patenting sphere,23 patents had begun to cover upstream scientific advances, such as genes and metabolic pathways, 18

19

20

21 22 23

To be sure, signs may be subject to other legal restrictions, such as those that stem from moral rights or rights of publicity. Martin Senftleben, Vigeland and the Status of Cultural Concerns in Trade Mark Law – The EFTA Court Develops More Effective Tools for the Preservation of the Public Domain, 48 Int’l Rev. Intell. Prop. & Competition L. 683, 707 & 716 (2017). Associated Press, Law Professor Fights Company’s Claim to “Rapunzel” Trademark (June 5, 2018), https://myfox28columbus.com/news/entertainment/law-professor-fights-companys-claimto-rapunzel-trademark. See also Rebecca Schoff Curtin, Zombie Cinderella and the Undead Public Domain, 85 Tenn. L. Rev. 961 (2018). See Iancu v. Brunetti, 139 S. Ct. 782 (2019); Matal v. Tam, 137 S. Ct. 1744 (2017). 15 U.S.C. § 1052(e)(5) and (f ). See, e.g., 35 U.S.C. §§ 200 & ff (the Bayh-Dole Act, giving universities the right to claim patents in federally funded research).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

326

Rochelle C. Dreyfuss

as well as diagnostics and fundamental business practices. As access problems arose, the Supreme Court became concerned that, in Justice Breyer’s words, “sometimes too much patent protection can impede rather than ‘promote the Progress of Science and useful Arts,’ the constitutional objective of patent and copyright protection.”24 In four cases, the Court then emphasized that laws of nature, principles of nature, and abstract ideas must remain in the public domain. In Bilski and Alice, it invalidated patents on risk management techniques; in Myriad, it did the same for patents on isolated genomic DNA associated with early-onset breast cancer; and in Mayo, the Court rejected a patent on a diagnostic that correlated the dosage of a drug with the blood level of its metabolites to determine the appropriate treatment. Subsequent lower court cases invalidated patents on other advances, including a revolutionary method for identifying defects in a fetus which is far safer than amniocentesis.25 From the perspective of access, these decisions were – like Vigeland – highly praised. They were also very effective. The cost of genetic testing declined, diagnostic centers began to compete over quality, and patients could obtain second opinion testing.26 From other perspectives, however, the result was extremely problematic. Because all inventions are based on nature and natural principles, it became difficult to decide when an advance was different enough from nature to be protectable. Inter-examiner variability increased.27 Although the Patent and Trademark Office (USPTO) allowed some claims, these claims may not be offering enough protection to support the investment necessary to bring products to market. The uncertainty has led some inventors to keep inventions as trade secrets, which can be more detrimental to scientific progress than patents.28 Moreover, researchers appear to be avoiding areas where protection is uncertain, not only in the life sciences, but also in engineering, where many advances involve correlating observables (such as crumbling metal) with natural outcomes (such as bridge collapse).29 24

25 26

27

28

29

Lab. Corp. of Am. Holdings v. Metabolite Labs., Inc., 548 U.S. 124, 126–27 (2006) (Breyer, J., dissenting). Ariosa Diagnostics, Inc. v. Sequenom, Inc., 788 F.3d 1371 (Fed. Cir. 2015). Rochelle C. Dreyfuss, Jane Nielsen, and Dianne Nicol, Patenting Nature – A Comparative Perspective, 5 J. Law & Biosciences 550 (2018), https://academic.oup.com/jlb/advance-article/ doi/10.1093/jlb/lsy021/5114251. Mateo Aboy et al., After Myriad, What Types of Claim Amendments Change a Patent Ineligible Isolated Gene Claim into an Eligible Patent Claim that is ‘Markedly Different’ from Nature?, 35 Nat. Biotechnol. 820 (2018). Robert Cook-Deegan et al., The Next Controversy in Genetic Testing: Clinical Data as Trade Secrets?, 21 Eur. J. Hum. Genet. 585 (2013); Bhaven Sampat and Heidi L. Williams, How Do Patents Affect Follow-On Innovation? Evidence from the Human Genome, 109 Am. Econ. Rev. 203 (2019). See, e.g., Jonathon Liddicoat et al., The Effects of Myriad and Mayo on Molecular Test Development in the US and Europe: Interviews from the Frontline, 22 Vand. J. Ent. & Tech L. 785 (2019), http://dx.doi.org/10.2139/ssrn.3381553; David J. Kappos, Oral Testimony Before the U.S. Senate Sub-Committee on Intellectual Property (June 4, 2019), https://perma.cc/GA7CA5KB.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

Cultural Heritage, IP Overlaps & the Public Domain

327

Unsurprisingly, the Director of the USPTO is advocating that the subject matter cases be overruled.30 After two roundtables on patent eligibility, the USPTO issued a report with suggestions for legislatively expanding the category of patent-eligible subject matter,31 and Congress is pondering a bill to eliminate all judicial exceptions to patentability.32 Significantly, a case involving Myriad’s gene patents in Australia did not produce the same results as in the US. There, isolated DNA was held unpatentable because it was regarded as pure information.33 However, the decision produced very little change on the ground. Other advances, including diagnostics, remained protectable on the theory that they constitute applications of information, a human endeavor that is within the scope of patentability.34 At the same time, research incentives appear to be unchanged and there has not been an observable uptick in reliance on trade secrecy. Importantly, patient access was the same before and after the decision. Arguably, Australia can be generous about issuing patents in the life sciences because the legal landscape is more hospitable to public access interests than is US law. Australia has national healthcare to hold prices down on therapeutics; public laboratories that conduct diagnostic testing at low cost; an experimental use exemption to promote research; and a doctrine of crown use, compulsory licensing schemes, and competition law, all of which deal with potential holdups.35 Comparing the US, which lacks all these devices, with Australia demonstrates that dynamic inefficiency need not be the price of ensuring static efficiency; that a good set of exceptions is less disruptive and capable of far more nuance, than a threshold restriction on the availability of protection. Indeed, some of those recommending expansion of patentable subject matter in the US include new defenses to infringement liability in their proposals.36

30

31

32 33 34

35

36

Scott Graham, PTO Director: ‘Something Must Be Done’ to Clarify Section 101, Nat’l L.J. (April 11, 2018). USPTO, Patent Subject Matter Eligibility: Roundtable 1, https://www.uspto.gov/patent/patentsubject-matter-eligibility-roundtable-1 (November 14, 2016) and Roundtable 2, https://www .uspto.gov/patent/initiatives/patent-subject-matter-eligibility-roundtable-2 (December 5, 2016); USPTO, Patent Eligible Subject Matter: Report on Views and Recommendations from the Public, https://www.uspto.gov/sites/default/files/documents/101-Report_FINAL.pdf (2017). The so-called Tillis-Coons proposal is available at https://perma.cc/Y5FK-FW9Q. D’Arcy v. Myriad Genetics, Inc. [2015] HCA 35; Dreyfuss, Nielsen and Nicol, supra note 26. Compare Sequenom, Inc. v. Ariosa Diagnostics, Inc. [2019] FCA 1011 (holding a diagnostic for fetal defects is patentable) with Ariosa Diagnostics, Inc. v. Sequenom, Inc., 788 F.3d 1371 (2015) (invalidating similar claims). See also Dianne Nicol, Rochelle C. Dreyfuss, E. Richard Gold, Wei Li, John Liddicoat and Geertrui Van Overwalle, International Divergence in Gene Patenting, 20 Ann. Rev. Genomics & Hum. Genet. 1 (2019) (noting that the scope of patentability is broader in the EU, China, and Canada. All the countries have safeguards similar to those found in Australia). Banbury Center, Cold Spring Harbor Laboratory, A Proposed Path Forward for Legislatively Addressing Patent Eligibility Law (2016), Comments Submitted at the USPTO Patent Subject Matter Eligibility: Roundtable 2, supra note 31.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

328

Rochelle C. Dreyfuss

26.3 vigeland revisited Will the exclusion of cultural symbols from trademark protection produce the same result that the US is experiencing with patent law limitations? It certainly seems possible. Just as it is proving difficult in the US to decide what is too close to nature, it will be hard to determine which cultural symbols are sufficiently important (or aesthetically functional) to be unregistrable.37 As Senftleben admits, Nike is the Greek goddess of victory:38 should registration of that mark have been refused? Interexaminer variability is likely to be high. Worse, Vigeland implies that every country will safeguard its own cultural symbols. Thus, a refusal to register Nike in Greece would not mean that national registrations would be refused in other countries in Europe or in the rest of the world. And because international law requires countries to protect marks that become well known (Nike being an example), a jurisdiction may be obliged to offer a degree of protection to its local cultural symbols even when it denied them registration on morality, public policy, or aesthetic functionality grounds.39 Thus, uncertainty for trademark applicants will not necessarily be counterbalanced by a guarantee of public access. The bar to registration could also lead traders to rely on IP regimes that are more problematic than trademark registration. Trade secrecy is, of course, unavailable. However, the EFTA Court noted that “refusal of registration of a trade mark does not prevent an applicant from using the sign in question in the course of trade.”40 Is that a better result? Administrative determinations can create a degree of certainty that is absent when traders decide for themselves which symbols to use. Registration provides notice that a mark is in use and thus allows others to avoid confusing consumers or switching costs down the road. Administrative supervision over trademark choice can also ensure that marks are sufficiently distinct from the goods they denote and from one another. As in patent law, limiting subject matter eligibility could also decrease incentives to innovate. While the main goal of trademark law is to ensure the reliability of signals about source, encouraging traders to develop new language and rhetorical devices is a socially beneficial side effect. Yet traders may be discouraged from adopting marks with the potential for significant cultural resonance if they cannot acquire the legal rights associated with registration.41 Furthermore, they may avoid marks that any jurisdiction regards as culturally significant out of concern that it might become difficult to engage in global marketing campaigns or to obtain 37

38 39 40 41

An example is furnished by Brunetti, where the majority and dissent could not agree on what constitutes an “immoral . . . or scandalous” trademark. See Iancu v. Brunetti, 139 S. Ct. 782 (2019). Senftleben, supra note 19, at 700. Paris Convention, art. 6bis; TRIPS Agreement, art. 16(2). Vigeland, } 101. See Matal v. Tam, 137 S. Ct. 1744 (2017), at 1753.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

Cultural Heritage, IP Overlaps & the Public Domain

329

multijurisdictional remedies for infringement.42 This is especially worrisome in a world in which, as Barton Beebe and Jeanne Fromer have demonstrated, the storehouse of available trademarks is shrinking.43 Rejecting certain marks could also threaten copyright incentives. After all, some copyrighted works are created with branding opportunities in mind. In those cases, the profits generated through trademarking are subsidizing creative labor.44 Finally, the practice of restricting subject matter may dilute the perceived need to strengthen defenses to infringement. This is a particularly acute problem under Vigeland. Although the decision did not turn entirely on the lapse of copyright protection, it was limited to copyright: the Court reasoned that there is “no . . . need to safeguard the public domain resulting from works that were primarily created to serve as signs to be registered as trade marks.”45 But that distinction is demonstrably untrue. Trademarks are highly conducive to follow-on creativity; sometimes, even more so than cultural symbols. As I mentioned in an earlier piece, “Betty Crocker has replaced Hestia in the public consciousness.”46 It is Betty Crocker – a trademark – that the public now uses to describe obsessive housekeepers.47 Or consider the artist Richard Prince and his series of paintings based on the Marlboro Man, a trademark.48 Or the use of Barbie – also a trademark – in multiple works.49 While the denotation of a trademark may be fixed (signifying the source of goods), connotations are fluid. Audiences encode marks with their own meanings, and then wish to reference those meanings in further work.50 Trademark owners even promote encoding by vesting their marks with connotations that fill conceptual voids. “Mc” for “safe, clean and inexpensive” is an example.51 Denying registration to works of longstanding cultural significance will therefore not avoid the need for a rich set of reliable defenses to infringement.

42

43

44

45 46 47

48 49

50

51

See, e.g., DHL Express France SAS v. Chronopost SA, C-235/09 (ECJ, April 12, 2011) (refusing to enforce the mark “web shipping” in places where it is generic). Barton Beebe and Jeanne C. Fromer, Are We Running Out of Trademarks? An Empirical Study of Trademark Depletion and Congestion, 131 Harv. L. Rev. 945 (2018). See, e.g., Rupert Maconick, Why Marketers Must Move Beyond ‘Branded Content’ and Create Entertainment, AdWeek (October 28, 2016), https://www.adweek.com/brand-marketing/whynext-step-brands-become-creators-entertainment-174330/ (giving the example of The Lego Movie). Vigeland, } 97. Dreyfuss, Genericity, supra note 6, at 424. Urban Dictionary, Betty Crocker, https://www.urbandictionary.com/define.php?term=Betty %20Crocker. See Guggenheim, Cowboys, https://gagosian.com/exhibitions/2013/richard-prince-cowboys. Examples include Tom Forsythe’s Food Chain Barbie, https://www.tjcenter.org/ArtOnTrial/ copyright.html, and Aqua’s “Barbie Girl,” https://www.youtube.com/watch?v=ZyhrYis509A. See, e.g., David Tan, The Semiotics of Alpha Brands: Encoding/Decoding/Recoding/ Transcoding of Louis Vuitton and Implications for Trademark Laws, 32 Cardozo Arts & Ent. L.J. 225 (2013). See Quality Inns Intern., Inc. v. McDonald’s Corp., 695 F. Supp. 198 (D. Md. 1988).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

330

Rochelle C. Dreyfuss

In the final analysis, the EFTA and US courts made the same mistake. They saw in the absence of effective defenses a reason to curtail subject matter coverage. However, the comparison between the US and Australia suggests that a better way to deal with absence of effective defenses is to strengthen them. Senftleben may be right that defenses grounded in the “freedom of art” (or expression) can be too unpredictable to safeguard access interests and that limitations based on consumer confusion, use in the course of trade, and use as a mark have not withstood manipulation by right holders and the onslaught of dilution law. But these defenses can be invigorated. For example, better use could be made of the parody defense.52 Furthermore, new defenses could be developed. Thus, the emergence of a nominative fair use defense (protecting referential uses) has proved to offer fairly robust protection to a subset of public uses.53 One problem with some of the defenses Senftleben mentioned is that they analyze the language of trademarks in the context of the marketplace for merchandise. They are ineffective because the public’s access interest lies in a mark’s surplus value – its expressive, rather than its competitive, dimension.54 Effective protection for such uses requires the court to focus instead on the marketplace of ideas and the significance of the mark in that arena. The question should be whether, given the connotations that audiences have attached to a particular trademark, its use is crucial to conveying the speaker’s meaning or whether other words will be equally effective. A fully developed defense based on a mark’s expressive significance would provide the safeguards that the Vigeland court was seeking.

26.4 conclusion Intellectual property systems must balance proprietary and public interests. They can do so ex ante, by denying protection to works of great public concern. Or they can do it ex post, with limits and exceptions to infringement liability. There is certainly an attraction to ex ante protection. However, as the US experience in patent law shows, that is a very blunt instrument. Exclusions avoid chilling secondgeneration uses at the expense of discouraging initial creative production. With ex post safeguards, courts can consider the public interest in the context of actual use rather than in the abstract, and can then tailor protection to the contextual need to reference the mark.55 In such a system, follow-on investment can be encouraged 52 53

54 55

See, e.g., Deckmyn v. Vandersteen, C-201/13, ECLI:EU:C:2014:2132 (2014). See, e.g., Mattel, Inc. v. MCA Records, Inc., 296 F.3d 894 (9th Cir. 2002) (Aqua’s “Barbie Girl”); Mattel, Inc. v. Walking Mountain Prods., 353 F.3d 792 (9th Cir. 2003 (Food Chain Barbie). See also Trademark Directive, Recital 27 and Regulation (EU) 2017/1001 of the European Parliament and of the Council of 14 June 2017 on the European Union trade mark, Recital 21 (both referring to right to use third party marks for the purpose of artistic expression). Dreyfuss, Genericity, supra note 6, at 412–418. See also Graeme B. Dinwoodie, Developing Defenses in Trademark Law, 13 Lewis & Clark L. Rev. 99 (2009).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

Cultural Heritage, IP Overlaps & the Public Domain

331

substantively, by expanding the range of defenses and clarifying their applicability; and procedurally, through reliance on declaratory judgment actions. This analysis demonstrates the many ways in which Professor Kur’s work has been right on the mark. She has stressed the benefits of focusing on overlaps among IP regimes (the problem that led to Vigeland) and on interpreting IP regimes coherently (the point of comparing US and Australian patent law to moves in trademark law).56 She has been deeply concerned with ensuring that the prerequisites for obtaining protection are correctly aligned with the limits of exclusive rights regimes (a concern that dominated all the opinions mentioned here).57 This chapter has not considered the compatibility of the defenses it proposed with obligations under international IP law. However, its view of the criticality of exceptions to the proper operation of IP systems underlines the importance of Professor Kur’s efforts to ensure that TRIPS flexibilities – including the notorious exceptions tests – are interpreted generously.58

56

57

58

Annette Kur, Introduction, in The Structure of Intellectual Property Law: Can One Size Fit All? iv (Annette Kur and Vytautas Mizaras, eds. , Edward Elgar 2011). Annette Kur, Gemeinfreiheit und Markenschutz – Bemerkungen zur Entscheidung des EFTAGerichtshofs im “Vigeland,” 119 GRUR 1082 (2017); Annette Kur, Too Pretty to Protect? Trade Mark Law and the Enigma of Aesthetic Functionality, Max Planck Institute for Intellectual Property & Competition Law Research Paper No. 11–16, 9 (2011). Annette Kur and Henning Grosse Ruse-Khan, Enough Is Enough – The Notion of Binding Ceilings in International Intellectual Property Protection, in Intellectual Property Rights in a Fair World Trade System – Proposals for Reform of TRIPS 359 (Henning Grosse Ruse-Khan, Marianne Levin, eds., Edward Elgar 2011).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.036

27 Public Order in the Light of Aesthetic Theory: The Copyright/Trademark Interface after Vigeland Martin Senftleben1

27.1 introduction Annette Kur’s publications on the coherence of the intellectual property system are an incessant source of inspiration for further research and critical reflection. In one of her landmark essays on overlapping intellectual property rights, she develops a general framework for the assessment of cumulative protection. According to this framework, the combination of different intellectual property rights is not problematic per se. In principle, the cumulation of rights is acceptable as long as the individual protection regimes involved are balanced in the sense that the prerequisites for obtaining protection are appropriately aligned with the contents and limits of exclusive rights.2 From this perspective, it is not the cumulation of rights that causes overprotection problems. By contrast, these problems are symptoms of imbalances within the protection systems involved.3 If the prerequisites for obtaining different types of intellectual property rights remain distinct from each other, and the checks and balances in the different protection regimes are sufficient to prevent excessive protection, the overlap is unlikely to obstruct competition and should be deemed permissible.4 If, however, the requirements for obtaining protection in different regimes converge while the contents and limits of protection remain unchanged, overlapping protection raises the problem of “asymmetric convergence” and requires appropriate countermeasures.5

1

2

3 4 5

Ph.D.; Professor of Intellectual Property Law and Director, Institute for Information Law (IViR), University of Amsterdam. A. Kur, Funktionswandel von Schutzrechten: Ursachen und Konsequenzen der inhaltlichen Annäherung und Überlagerung von Schutzrechtstypen, in: G. Schricker/T. Dreier/A. Kur (eds.), Geistiges Eigentum im Dienst der Innovation, Nomos 2001, 23 (45 and 50). Kur, ibid., 45. Ibid., 42–43. Ibid., 43–44.

Downloaded from https://www.cambridge.org/core. 332 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

Copyright/Trademark Interface after Vigeland

333

This assessment standard paves the way for nuanced solutions. Instead of brightline rules, the acceptance of overlaps depends on a thorough scrutiny of the intellectual property systems involved. As long as a sufficient arsenal of legal safeguards against excessive protection is available – safeguards that are aligned with the prerequisites for obtaining protection – it is possible to afford intellectual property owners the opportunity to combine exclusive rights stemming from different protection regimes.

27.2 copyright/trademark protection overlaps In the debate on cumulative protection, overlaps between intellectual property rights with a limited period of protection and indefinitely renewable trademark rights have attracted attention for a long time. Case law sheds light on the underlying problem. In Dastar/Twentieth Century Fox, the U.S. Supreme Court held that overlaps between trademark rights and copyright or patent protection can have a corrosive effect on the “carefully crafted bargain” under which, once the patent or copyright monopoly has expired, the public may use the invention or work at will.6 The Court of Justice of the European Union (CJEU) expressed similar concerns in Hauck/Stokke, where it explained that the functionality doctrine in trademark law served the purpose of preventing the exclusive and permanent right which a trademark confers “from serving to extend indefinitely the life of other rights which the EU legislature has sought to make subject to limited periods.”7 However, the prevention of an artificial extension of rights is not an end in itself. As Annette Kur has emphasized8 – and judges have confirmed9 – an artificial extension of protection via trademark law is undesirable because it may impede competition. The grant of trademark rights may lead to a continuous monopolization of decisive product features which consumers are likely to seek in the products of competitors.10 Despite widespread agreement on this problem statement, intellectual property legislation does not devote the same attention to all overlap scenarios. Instruments to draw a boundary line between copyright and trademark protection, for instance, seem underdeveloped. Character merchandising leads to cumulative copyright and trademark protection of contemporary cultural symbols.11 In principle, it is also possible to acquire trademark rights to cultural heritage symbols once distinctive

6 7 8

9 10 11

U.S. Supreme Court, 2 June 2003, Dastar/Twentieth Century Fox, 539 U.S. 23 (2003), 33–34. CJEU, 18 September 2014, case C-205/13, Hauck/Stokke, paras. 19–20. A. Kur, Too Pretty to Protect? Trade Mark Law and the Enigma of Aesthetic Functionality, Max Planck Institute for Intellectual Property & Competition Law Research Paper No. 11-16, Max Planck Institute 2011, 21. CJEU, supra note 7, paras. 18–19. CJEU, ibid., para. 18; CJEU, 18 June 2002, case C-299/99, Philips/Remington, para. 78. Cf. I. Calboli, Overlapping Trademark and Copyright Protection: A Call for Concern and Action, Illinois Law Review Slip Opinion 2014, No. 1, 25 (29–30).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

334

Martin Senftleben

character is acquired through use in trade.12 The situation differs markedly from the status quo reached in the area of patent/trademark overlaps. At the intersection of these two protection regimes, the doctrine of utilitarian functionality prevents the cumulation of exclusive rights.13 Similarly, the doctrine of aesthetic functionality can be employed to arrive at a far-reaching separation of trademark and industrial design protection.14 Against this background, the question arises whether the copyright/trademark interface is configurated in a way that can be deemed acceptable. Employing Annette Kur’s assessment criteria, it may be asked whether trademark law offers sufficient gatekeeper instruments to prevent the acquisition of cumulative copyright and trademark protection in cases where the overlap would lead to excessive protection in the light of available checks and balances.

27.3 vigeland In this respect, the Vigeland decision of the Court of Justice of the European Free Trade Association States (EFTA Court) is of particular interest. The Court cultivates public order and morality as absolute grounds for refusal that can be invoked in cases where trademark protection is sought for signs that constitute literary and artistic works.15 From a legal-doctrinal perspective, it is consistent to use public order 12

13

14

15

For example, see Court of Appeals of Brussels, 3 October 2013, cases 2013/7132 and 2013/7133, Anne Frank Foundation/ Benelux Office for Intellectual Property (concerning the book titles The Diary of Anne Frank and The Secret Annex), and OHIM Board of Appeal, 31 August 2015, case R 2401/2014-4, Anne Frank Fonds, paras. 32–36 (granting trademark rights to the French indication Le Journal d’Anne Frank); German Federal Court of Justice, 24 April 2008, case I ZB 21/06, Marlene-Dietrich-Bildnis (concerning a portrait photograph of the actress); German Federal Patent Court, 25 November 1997, case 24 W (pat) 188/96, Mona Lisa (concerning the Mona Lisa painting), Gewerblicher Rechtsschutz und Urheberrecht 1998, 1022. For an even broader concept of assuming distinctiveness in cases involving cultural heritage names, see CJEU, 6 September 2018, case C-488/16 P, Bundesverband Souvenir – Geschenke – Ehrenpreise/EUIPO, paras. 65–66, 69. Developments in the EU can serve as an example in this regard. See the decisions CJEU, 18 June 2002, case C-299/99, Philips/Remington, para. 82, and CJEU, 14 September 2010, case C-48/09 P, Lego/Mega Brands, paras. 45, 53–58, which lead to a separation of the two protection regimes even if technical alternatives exist. In the EU, the potential of the doctrine of aesthetic functionality has been enhanced significantly in CJEU, 18 September 2014, case C-205/13, Hauck/Stokke, paras. 19–20, where the Court developed a list of flexible factors for the identification of signs that are to be excluded from trademark protection for this reason. This is not to say that the doctrine of aesthetic functionality has not been criticized. For example, see G. B. Dinwoodie, The Death of Ontology: A Teleological Approach to Trademark Law, Iowa Law Review 84 (1999), 611 (685–691); Kur, supra note 8, 6–12. The central point here, however, is that in practice the application of the functionality doctrine can have the effect of separating the protection regimes to a considerable extent. For case comments, see A. Kur, Gemeinfreiheit und Markenschutz – Bemerkungen zur Entscheidung des EFTA-Gerichtshofs im ‘Vigeland’-Fall, Gewerblicher Rechtsschutz und Urheberrecht 2017, 1082; M. R. F. Senftleben, Vigeland and the Status of Cultural

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

Copyright/Trademark Interface after Vigeland

335

and morality for this purpose. As Annette Kur has demonstrated, the functionality doctrine in trademark law is a subcategory of a refusal based on public order or morality.16 The invocation of functionality leads to a categorical exclusion of the sign concerned – irrespective of whether the sign is distinctive.17 Therefore, the correct basis for explaining the application of the functionality doctrine – in particular under the telle quelle protection rule following from Article 6quinquies (B) No. 3 PC – is public order and morality.18 Given this close connection, it appears consistent to rely on public order and morality as grounds for refusal that supplement the doctrine of aesthetic functionality when it comes to copyright/ trademark overlaps. The field of application of this supplementary rule must not be underestimated. By definition, the functionality doctrine only covers essential product characteristics. Cultural signs, however, need not feature prominently on the product itself. If merely used as a label, a cultural sign can hardly be found to define product characteristics.19 In Vigeland,20 the EFTA Court dealt with an attempt to register several artworks made by Gustav Vigeland – one of the most eminent Norwegian sculptors – as trademarks. Among other works, the underlying trademark application concerned Vigeland’s Monolith (Monolitten) and Angry Boy (Sinnataggen) – popular works that feature prominently in the Vigeland Installation in the centre of Oslo’s Frogner Park. The initiative to acquire trademark rights was taken after the expiry of the term of copyright protection.21 The trademark applicant – the Municipality of Oslo – argued that the rights to Vigeland’s works ought to be due to the city – even after copyright expiry – because of the efforts made to ensure that the works became wellknown.22 The Norwegian Industrial Property Office, however, remained unimpressed and refused the application on the grounds that Vigeland’s works were non-distinctive, descriptive and functional.23 Instead of dispelling these concerns, the Board of Appeal added that the registration of cultural heritage signs may be

16 17

18 19

20 21 22

23

Concerns in Trade Mark Law – The EFTA Court Develops More Effective Tools for the Preservation of the Public Domain, International Review of Intellectual Property and Competition Law 48 (2017), 683. Kur, supra note 8, 20; Kur, supra note 15, 1086. CJEU, 18 June 2002, case C-299/99, Philips/Remington, para. 57; CJEU, 20 September 2007, case C-371/06, Benetton/G-Star, paras. 25–27; CJEU, 14 September 2010, case C-48/09 P, Lego/ Mega Brands, para. 47. Kur, supra note 8, 20. German Federal Court of Justice, 24 April 2008, case I ZB 21/06, Marlene-Dietrich-Bildnis, paras. 19–22. See also the confirmation of this point in German Federal Court of Justice, 31 March 2010, case I ZB 62/09, Marlene-Dietrich-Bildnis II, paras. 20–28. EFTA Court, 6 April 2017, case E-5/16, Municipality of Oslo (“Vigeland”). EFTA Court, ibid., para. 20. Gustav Vigeland died on 12 March 1943. Report for the Hearing, Judge-Rapporteur Baudenbacher, 5 October 2016, case E-5/16, Municipality of Oslo (“Vigeland”), para. 36. EFTA Court, supra note 20, para. 27.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

336

Martin Senftleben

contrary to public policy or accepted principles of morality, and sought the EFTA Court’s guidance on factors that should inform the decision on these grounds for refusal.24

27.4 public order and morality The EFTA Court clearly distinguished the two branches of the ground for refusal: a denial of trademark protection based on grounds of public policy required an assessment of objective criteria, whereas a rejection based on accepted principles of morality implied the examination of subjective values through the prism of consumer perception.25 As to the objective criterion of a conflict with public policy, the EFTA Court noted that “public policy” referred to principles and standards regarded to be of a fundamental concern to the State and the whole of society. A refusal based on public policy required exceptional circumstances, namely “a genuine and sufficiently serious threat to a fundamental interest of society.”26 Nonetheless, the Court saw room for applying this ground for refusal in the context of artwork registrations: An artwork may be refused registration, for example, under the circumstances that its registration is regarded as a genuine and serious threat to certain fundamental values or where the need to safeguard the public domain, itself, is considered a fundamental interest of society.27

Interestingly, the Court stated that a genuine and serious threat to public policy could follow from the “registration” of an artwork as a trademark.28 Hence, not only the nature of the sign at issue but also the registration attempt as such can be sufficient to justify a refusal based on ordre public.

27.5 aesthetic theory To explore the full potential of the approach taken by the EFTA Court, it is advisable to embark on an interdisciplinary analysis. As the Court points out, public order constitutes a ground for refusal that refers to the importance attached to cultural signs in society. With regard to this broader societal perspective, aesthetic theory can offer important insights. The aesthetic theories of Schiller and Adorno can serve as examples.

24 25 26 27 28

Ibid., paras. 28–30. Ibid., paras. 84–86. Ibid., paras. 94–96. Ibid., para. 96. Ibid.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

Copyright/Trademark Interface after Vigeland

337

27.5.1 Friedrich Schiller Disillusioned by the French Revolution, which had culminated in chaos and violence instead of leading to a free and equal society, Schiller wrote his Letters on the Aesthetic Education of Man to explore the possibility of a transition from an absolutist, authoritarian state to a purified, ethical state that is founded on human reason.29 As a precondition for this transition, Schiller emphasizes the need to harmonize human desires with the rules of reason. As it is not the destiny of mankind to renounce its natural senses in favour of moral laws,30 support for an ethical, reasonable state must come from the totality of human dispositions: desire and reason alike. Individuals should not only feel an obligation to follow the rules of reason, they should feel a desire to do so. If human desires are brought in line with the postulates of reason, a revolution will no longer end in chaos and violence. It will lead to the establishment of a moral state instead.31 To align human desires with the rules of reason, however, a catalyst is required that brings moral laws not only to people’s heads but also to their hearts. Schiller solves this problem by positing that art can serve as such a catalyst. Even though being incapable of changing mankind directly, art can point the way to a change for the better by focusing people’s thoughts on the “necessary and eternal,” and make them strive for this ideal.32 Art is predestined to accomplish this task because it satisfies the desire to play and enjoy. Instead of openly criticizing people’s actions and attitudes, art can improve society in a subtle way by making visions of ethical behaviour part of people’s play and pleasure.33 Schiller thus relies on art as a vehicle to let people experience an ideal balance between desire and reason until they finally orient their actions by moral laws instead of following mere physical necessities. A true work of art is capable of evoking an equilibrium between reason and desire, and freedom of the mind through appearances of beauty that neither reflect nor require reality.34 In Schiller’s view, the experience and enjoyment of this ultimate perfection can pave the way for the establishment of a moral society in which individual freedom no longer follows from the restriction of the freedom of others but from a consensus on ethical norms that corresponds with people’s desires, as refined in the aesthetic play.35 An individual driven by physical necessity must first experience beauty – the aesthetical balance between desire and rules of reason – before he can actively and 29

30 31 32 33 34 35

¨ ber die a¨sthetische Erziehung des Menschen, ed. K. L. Berghahn, F. Schiller, U Reclam 2000, 11–14 (Letter 3). Schiller, ibid., 28 (Letter 6). Ibid., 120–121 (Letter 27). Ibid., 36 (Letter 9). Ibid., 37 (Letter 9). Ibid., 111–112 (Letter 26). Ibid., 120–121 (Letter 27).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

338

Martin Senftleben

freely opt for moral norms and moral actions.36 It is thus the task of art to prepare mankind for the transition from the physical state of desire to the moral state of reason.37

27.5.2 Theodor Adorno Theodor Adorno also underlines the societal relevance of art. Against the background of the alienation which the individual faces in a fully rationalized, efficiency-driven world, he warns of the affirmative nature of art. An artwork bringing a conciliatory reflection of enchantment into the disenchanted, empirical reality offers comfort in the rationalized world and supports the unbearable status quo.38 In the light of the inhumanity of the real world, art would make itself an accomplice of present and coming disasters if it sustained positive visions of society and obscured the defects and poorness of reality.39 With the prospect of a better world which, as an ultimate truth,40 shimmers through genuine artworks,41 art may falsely pretend that existing societal conditions are acceptable. Therefore, art is constantly at risk of becoming guilty of supporting the inhuman status quo and fortifying present ideologies.42 However, art must not be condemned altogether because true art is capable of unmasking the negativity of present societal conditions. Showing visions of a better, happier life, art can rouse opposition against the existing reality and contribute to necessary societal changes.43 Art can play a decisive role in society because it generates utopian views of a better life that may become drivers of a change for the better. This role of authentic art defines its social character: art is the “social antithesis” of society.44 Given this delicate position in the social fabric of modern societies, there is a fine line to be walked: the artist must relentlessly expose the inhumanity of reality without offering any prospect of reconciliation. In doing so, the artist creates genuine works which, by their very nature, offer shining visions of a better life and a better society in spite of the hopelessness reflected in the artworks themselves.45 As an antithesis of real-world disasters, art becomes the messenger of an ideal, utopian world.46

36 37 38 39 40 41 42 43 44 45 46

Ibid., 90–91 (Letter 23). Ibid., 92 (Letter 23). T. W. Adorno, A¨sthetische Theorie, ed. G. Adorno/R. Tiedemann, Suhrkamp 1970, 10, 34. Adorno, ibid., 28, 503. Ibid., 128, 196–197. Ibid., 199–200. Ibid., 203. Ibid., 25–26, 56. Ibid., 9–10, 19, 53. Ibid., 127, 199. Ibid., 55–56.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

Copyright/Trademark Interface after Vigeland

339

27.5.3 Importance of a Robust Public Domain These aesthetic theories by no means exhaust the possibilities of describing the role of literary and artistic works in society. Nonetheless, the two examples already show that by presenting alternative visions of society, cultural signs can play a crucial role in the improvement of social and political conditions. For this purpose, however, it is necessary to preserve the genuine, cultural meaning of artworks as a basis for dialogue and discussion in the artistic domain.47 Once a work of art is created, society should seek not to interfere with the artistic discourse. Otherwise, an important impulse for required social and political changes will be lost. Schiller directly points out the merit of leaving the artistic discourse on the meaning and message of works of art intact: as artworks reflect human dignity, they can serve as prototypes for the restoration of an ethical order whenever the principles of morality and reason are lost.48 From this perspective, the preservation of an artwork’s potential for dialogic communication in the artistic domain is an act of preserving expressions of human dignity as a basis for desirable improvements of social and political conditions. The grant of trademark rights in a cultural sign, however, gives one individual speaker more communication power than other participants in the debate surrounding the sign. As trademark rights offer the trademark owner an exclusive position, the definition power which the trademark owner has in the communication process concerning the cultural sign is higher than the degree of control resulting from other forms of commercial use, such as mere use in advertising without trademark rights. The grant of stable, indefinitely renewable rights substantially enhances the risk of impregnating the artwork with commercial messages that may devalue the sign as a reference point in the artistic discourse. In any case, the debate on the artwork is no longer as open and free as it was before. If the trademark right is acquired after the artwork entered the public domain, the resulting imbalance is self-evident. Unlimited freedom of use is sacrificed in favour of a scenario where all speakers, including artists in the cultural domain, must ensure that their contributions to the debate do not amount to trademark infringement. Intentionally or unintentionally, the trademark proprietor will bring trademark claims against those 47

48

For a more detailed discussion of the dialogic process underlying acts of creation, see S. Dusollier, Realigning Economic Rights with Exploitation of Works: The Control of Authors over the Circulation of Works in the Public Sphere, in: P. B. Hugenholtz, Copyright Reconstructed – Rethinking Copyright’s Economic Rights in a Time of Highly Dynamic Technological and Economic Change, Kluwer Law International 2018, 163 (180–183); C. J. Craig, Copyright, Communication and Culture – Towards a Relational Theory of Copyright Law, Edward Elgar 2011, 52–56. Schiller, supra note 29, 35 (Letter 9, English translation taken from: C. W. Eliot (ed.), Literary and Philosophical essays: French, German and Italian. With Introductions and Notes, The Harvard Classics, vol. xxxii, Collier 1910, available at https://sourcebooks.fordham.edu/mod/ schiller-education.asp).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

340

Martin Senftleben

forms of artistic use which, from her point of view, interfere with her own strategy for the development of the message and meaning of the sign. It does not seem decisive whether the trademark proprietor is a commercial enterprise or a public body. Even if the owner of the trademark right is a state, municipality or museum, and even if the objective underlying the acquisition of trademark rights is art custodianship rather than cultural heritage grabbing, the risk of censorship – in the sense of enforcement measures seeking to prevent interferences with the trademark proprietor’s own vision of the sign – cannot be ruled out.49 Considering this problem scenario, it is of particular interest that Adorno explicitly warns of the risk of artworks losing their subversive character through efforts to submerge their critical meaning and popularize them in commercial products.50 Once art is consumed and enjoyed, it inevitably loses its value as a mirror of defects in society and a provocation that is capable of triggering necessary changes. Instead, it becomes a means of pretending that the present situation is sufficient and acceptable.51 From this perspective, it seems necessary to shield artworks from the corrosive effect of use in product marketing to safeguard the potential of art to expose flaws in society. 27.5.4 Vigeland Criteria A line between these considerations and the public order criteria developed by the EFTA Court can easily be drawn. As explained, the Court saw room for a refusal based on public policy where the trademark registration would pose a genuine and sufficiently serious threat to “certain fundamental values,”52 “a fundamental interest of society”53 or “a fundamental concern to the State and the whole of society.”54 Following Schiller’s and Adorno’s theories, it can be posited that the value at stake in cases where trademark rights are sought in relation to literary and artistic works, is of fundamental importance indeed. The use of artworks in product marketing and the attachment of commercial messages can blur a sign’s genuine cultural meaning and devalue the sign as a basis for dialogue and discussion in the artistic domain.55 It may deprive a literary and artistic creation of its potential to reveal shortcomings of 49 50 51 52 53 54 55

Cf. Senftleben, supra note 15, 688–689. Adorno, supra note 38, 33–34. Ibid., 32. EFTA Court, 6 April 2017, case E-5/16, Municipality of Oslo (“Vigeland”), para. 96. Ibid., para. 95. Ibid., para. 94. Cf. M. R. F. Senftleben, Der kulturelle Imperativ des Urheberrechts, in: M. Weller/N. B. Kemle/T. Dreier (eds.), Kunst im Markt – Kunst im Recht, Nomos 2010, 75 (101–102); K. Assaf, Der Markenschutz und seine kulturelle Bedeutung: Ein Vergleich des deutschen mit dem US-amerikanischen Recht, Gewerblicher Rechtsschutz und Urheberrecht – Internationaler Teil 2009, 1 (2–3); A. Wandtke/W. Bullinger, Die Marke als urheberrechtlich schutzfähiges Werk, Gewerblicher Rechtsschutz und Urheberrecht 1997, 573 (578).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

Copyright/Trademark Interface after Vigeland

341

present society and encourage changes for the better. In other words: the grant of trademark rights may conflict with society’s fundamental interest in art that can serve as an engine for the improvement of social and political conditions. This line of reasoning is only one example of applying the EFTA Court’s public order concept in the light of aesthetic theory. Further arguments can easily be found. The public domain of cultural expressions also serves as a source of inspiration and basis for new acts of creation. Artists often fulfil their fundamental societal role on the basis of a rich and diverse reservoir of works in the public domain. As preexisting works on which they are built, new cultural productions may mirror social and political shortcomings and offer the prospect of a better society. The grant of trademark rights, however, offers the trademark proprietor the opportunity to influence – or even control – the communication surrounding the protected sign. Even if trademark law provides for defences, such as due cause, non-distinctive use and referential use, the risk of trademark infringement is hanging above unauthorized use in the artistic domain like the sword of Damocles. Given the deterrent effect of trademark enforcement measures on risk-averse artists with limited financial means for a lawsuit, trademark protection may impede or even stifle the dialogic remix and reuse of cultural resources in the artistic domain.56 This corrosive effect can lead to a conflict with public order because it compromises society’s fundamental interest in cultural expressions that can serve as a basis for new artworks and fresh reflections of societal conditions. This second argument refers to an additional point which the EFTA Court discussed: a refusal of trademark rights for public policy reasons may be necessary to safeguard the public domain. In the light of Schiller’s and Adorno’s theories, it may be said that the need to preserve and cultivate the public domain, itself, constitutes a fundamental interest of society that should be safeguarded by invoking public order as a ground for refusal.57 Aesthetic theory can thus serve as a basis for arguments that broaden the ambit of operation of public order as a ground for refusal in cases that concern artworks, such as Gustav Vigeland’s creations. The EFTA Court expressed the view that “registration of a sign as a trade mark may only be refused as contrary to public policy in accordance with [EU trademark legislation] in exceptional circumstances.”58 In light of the societal values, interests and concerns which aesthetic theory reflects,

56 57

58

Cf. Senftleben, supra note 15, 699–705. EFTA Court, supra note 52, para. 96. As to the debate on the legal status of the public domain, see S. Dusollier, A Positive Status for the Public Domain, in: D. Beldiman (ed.), Innovation, Competition and Collaboration, Edward Elgar 2015, 135; W. Nordemann, Mona Lisa als Marke, Wettbewerb in Recht und Praxis 1997, 389 (391); D. Lange, Recognizing the Public Domain, Law and Contemporary Problems 44 (1981), 147 (150). As to the removal of literary and artistic works from the public domain for a limited period of revived copyright protection, see U.S. Supreme Court, 18 January 2012, Golan/Holder, 565 U.S. 302, 315–321 and 328–329 (2012). EFTA Court, ibid., para. 96.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

342

Martin Senftleben

this threshold of exceptionality need not be high. In particular, a sign need not have “high symbolic value.”59

27.6 conclusion In sum, aesthetic theory offers starting points for the transformation of public order criteria stemming from the Vigeland decision into grounds for refusal that can fulfil a central function with regard to copyright/trademark overlaps – a function that is similar to the role which utilitarian and aesthetic functionality plays vis-à-vis patent/ trademark overlaps and cumulative industrial design and trademark protection. However, it also becomes clear that the Vigeland decision – when combined with arguments derived from aesthetic theory – adds a further category of values. Annette Kur has developed a theoretical framework that focuses on the prevention of excessive protection to safeguard undistorted competition. Her approach to cumulative protection allows competition-based arguments to override considerations of aesthetic appeal and attractiveness: An absolute and permanent ban against trade mark protection for attractive shapes, which excludes any consideration of the fact that the aesthetic appeal may eventually become inferior to the message it conveys about commercial origin – meaning that it is primarily bought for the brand, and not for its design – does not make sense. At least, it can hardly be the attractiveness displayed at the early stages of marketing which accounts for the everlasting nature of that ban. It is yet another question, whether the shape – beyond the information it has come to convey about commercial origin, and unrelated to the goodwill it represents – still confers such strong competitive advantages to the prospective right holder that to keep it free from monopolisation appears “essential” for others to engage in meaningful competition.60

Once the public order criteria following from the Vigeland decision are charged with the outlined arguments derived from aesthetic theory, they constitute an independent, alternative basis for a categorical exclusion of cultural signs from trademark protection: the ban on the acquisition of trademark rights need no longer follow from economic concerns about undistorted competition. Alternatively, trademark rights may be denied because of cultural concerns about the blurring of the cultural meaning of artworks, an impediment to dialogic remix and reuse in the artistic domain and, ultimately, the loss of alternative visions of society. In the analysis, competition-based and cultural concerns would thus have the same weight. The final decision on this equal status, however, depends on whether other checks and balances in trademark law, such as the gatekeeper requirement of

59 60

Cf. Article 4(3)(b) EU Trade Mark Directive 2015/2436. Kur, supra note 8, 18.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

Copyright/Trademark Interface after Vigeland

343

use as a trademark61 and defences against infringement claims, for example the referential use defence, the non-distinctive use defence and the due cause defence,62 are sufficient to dispel concerns about an impediment of cultural follow-on innovation. Applying Annette Kur’s criteria by analogy, it may be concluded: only if these checks and balances are sufficient to prevent excessive protection that endangers cultural values, is the cumulation of rights acceptable.63

61

62

63

As to the debate on the desirability of solutions based on this gatekeeper requirement, see S. L. Dogan/M. A. Lemley, The Trademark Use Requirement in Dilution Cases, Santa Clara Computer & High Technology Law Journal 24 (2008), 541 (542); G. B. Dinwoodie/M. D. Janis, Confusion Over Use: Contextualism in Trademark Law, Iowa Law Review 92 (2007), 1597 (1657–1658). With regard to the EU, see A. Kur, Confusion Over Use? Die Benutzung ‘als Marke’ im Lichte der EuGH-Rechtsprechung, Gewerblicher Rechtsschutz und Urheberrecht International 2008, 1 (11), who warns of limiting trademark protection from the outset on the basis of a restrictive notion of trademark use, in particular with regard to European Union Trade Marks. For a more detailed discussion of the role of these defences in cases involving artistic expression, see A. Kur, ‘Yellow Dictionaries, Red Banking Services, Some Candies, and a Sitting Bunny: Protection of Color and Shape Marks from a German and European Perspective’, in: I. Calboli/M.R.F. Senftleben (eds.), The Protection of Non-Traditional Trademarks – Critical Perspectives, Oxford: Oxford University Press 2018, 89 (101–104). A. Kur/M. R. F. Senftleben, European Trade Mark Law – A Commentary, Oxford University Press 2017, paras. 5.265–5.268, 6.59–6.68; J. Schovsbo, “Mark My Words” – Trademarks and Fundamental Rights in the EU, UC Irvine Law Review 8 (2018), 555 (575–580); W. Sakulin, Trademark Protection and Freedom of Expression – An Inquiry into the Conflict between Trademark Rights and Freedom of Expression under European Law, Kluwer Law International 2010, 85–87; C. Geiger, Trade Marks and Freedom of Expression – The Proportionality of Criticism, International Review of Intellectual Property and Competition Law 38 (2007), 317. For a critical assessment of these checks and balances, see Senftleben, supra note 15, 699–705.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.037

28 Separability as Channeling: A Cautionary Tale Mark P. McKenna1

One of the most prominent themes in Annette Kur’s influential scholarship has to do with cumulation of intellectual property rights, especially in relation to design. Professor Kur is much more comfortable with that cumulation than I have been, at least in part because she is less confident that subject matter can neatly be separated.2 This chapter takes up one aspect of the cumulation question – the effectiveness of separability as a channeling doctrine with respect to industrial design. Though I remain more optimistic than Professor Kur about the benefits of channeling even when it is imperfect, the American experience with separability is a cautionary tale that highlights the challenges of channeling by subject matter.

28.1 applied art and industrial design According to the U.S. Copyright Act, copyright protection subsists “in original works of authorship fixed in any tangible medium of expression.”3 Works of authorship “include” those that fall within the eight categories identified by the statute, one of 1

2

3

John P. Murphy Foundation Professor of Law, Notre Dame Law School; Director, Notre Dame Technology Ethics Center. See, e.g., Annette Kur, Protection for Fashion: The European Experience, in Intellectual Property at the Edge – The Contested Contours of IP (Rochelle Cooper Dreyfuss and Jane C. Ginsburg, eds., 2014); Annette Kur, What to Protect, and How? Unfair Competition, Intellectual Property, or Protection Sui Generis, in Intellectual Property, Unfair Competition and Publicity (Nari Lee et al., eds., 2014); Annette Kur, Too Pretty to Protect? Trade Mark Law and the Enigma of Aesthetic Functionality, in Technology and Competition: Contributions in Honour of Hanns Ullrich (Josef Drexl, Reto M. Hilty, Laurence Boy, Christine Godt and Bernard Remiche, eds., 2009); Annette Kur, Cumulation of IP Rights Pertaining to Product Shapes – An ‘Illegitimate Offspring’ of IP Law?, in Intellectual Property and Market Power (G. Ghidini and L. M. Genovesi, eds., 2008). 17 U.S.C. § 102 (2012).

Downloaded from https://www.cambridge.org/core. 344 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.038

Separability as Channeling: A Cautionary Tale

345

which is “pictorial, graphic, and sculptural works” (PGS works).4 Congress added PGS works as a category in 1976, replacing the former category of “works of art,” and it defined PGS works to “include two-dimensional and three-dimensional works of fine, graphic, and applied art, photographs, prints and art reproductions, maps, globes, charts, diagrams, models, and technical drawings, including architectural plans.”5 Copyright protection for the subset of PGS works that qualify as “useful articles” is then importantly limited: [PGS works] shall include works of artistic craftsmanship insofar as their form but not their mechanical or utilitarian aspects are concerned; the design of a useful article, as defined in this section, shall be considered a pictorial, graphic, or sculptural work only if, and only to the extent that, such design incorporates pictorial, graphic, or sculptural features that can be identified separately from, and are capable of existing independently of, the utilitarian aspects of the article.6

The Copyright Act’s legislative history suggests that Congress borrowed the statutory language dealing with useful articles “from Copyright Office Regulations promulgated in the 1940’s,” as well as language added to the regulations in the 1950s to implement the Supreme Court’s decision in Mazer v. Stein, 347 U.S. 201 (1954 ).7 According to those implementing regulations, [i]f the sole intrinsic function of an article is its utility, the fact that it is unique and attractively shaped will not qualify it as a [copyrightable] work of art. However, if the shape of a utilitarian article incorporates features, such as artistic sculpture, carving, or pictorial representation, which can be identified separately and are capable of existing independently as a work of art, such features will be eligible for [copyright].8

The statutory language defining the scope of copyright in useful articles was intended “to draw as clear a line as possible between copyrightable works of applied art and uncopyrightable works of industrial design.”9 Congress meant for the limitations on the scope of copyright in useful articles to draw that clear line because it did not want copyright to serve as a general industrial design protection system. Indeed, during most of the time period in which it was drafting the 1976 Act, Congress was also considering a new form of design protection. The distinction between applied art and industrial design was meant to help

4 5 6 7 8 9

Id. at § 102(5). Id. at § 101. Id. at § 101. H.R. Rep. No. 94-1476, at 54 (1976), reprinted in 1976 U.S.C.C.A.N. 5659. 37 C.F.R. § 202.10(c) (1959). H.R. Rep. No. 94-1476, at 55; see also id. at 62–63 (“The Committee has added language to the definition of ‘pictorial, graphic, and sculptural works’ in an effort to make clearer the distinction between works of applied art protectable under the bill and industrial designs not subject to copyright protection . . .”).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.038

346

Mark P. McKenna

differentiate the subject matter of copyright from that of the new design protection. And although that separate design protection never materialized, Congress maintained the distinguishing language to make clear that copyright protection was not available for industrial design.10 But if the copyright Act’s separability language was meant to differentiate applied art and industrial design, what, exactly, did Congress contemplate would qualify as separable applied art? According to the House Report, “[a] two-dimensional painting, drawing, or graphic work is still capable of being identified as such when it is printed on or applied to utilitarian articles such as textile fabrics, wallpaper, containers, and the like. The same is true when a statue or carving is used to embellish an industrial product or, as in the Mazer case, is incorporated into a product without losing its ability to exist independently as a work of art.”11 By contrast, “although the shape of an industrial product may be aesthetically satisfying and valuable, the Committee’s intention [was] not to offer it copyright protection under the bill.”12 As a result, “[u]nless the shape of an automobile, airplane, ladies’ dress, food processor, television set, or any other industrial product contains some element that, physically or conceptually, can be identified as separable from the utilitarian aspects of that article, the design would not be copyrighted . . .”13 The House Report makes clear that Congress understood features to be separable when they could exist as a copyrightable work apart from the useful article to which they have been applied. The configuration of a useful article, however attractive, is not protectable because there are no features to separate. Importantly, because the question is separability and not aesthetic value, it does not matter whether the features of a useful article could have been designed differently or whether they were designed “independently” of functional considerations. “[I]ndependence from ‘the utilitarian aspects of the article’ does not depend upon the nature of the design – that is, even if the appearance of an article is determined by esthetic (as opposed to functional) considerations, only elements, if any, which can be identified separately from the useful article as such are copyrightable.”14 The concept is perhaps abstract, but Congress offered several examples to illustrate the distinction. “[C]arving[s] on the back of a chair,” or “floral relief design[s] 10

11 12 13 14

Separate protection for industrial design has repeatedly been proposed, but Congress has always declined to add such a provision to the Copyright Act. See Duraco Products, Inc. v. Joy Plastic Enterprises, Ltd., 40 F.3d 1431, 1446–47 (3d Cir. 1994) (recounting the history of congressional rejection); In re Nalbandian, 661 F.2d 1214, 1219 (C.C.P.A. 1981) (Rich, J., concurring) (explaining that design protection was initially included in the 1976 Act, but “was jettisoned to facilitate passage of the main bill”). H.R. Rep. No. 94-1476, at 55. Id. Id. Id. (emphasis added).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.038

Separability as Channeling: A Cautionary Tale

347

on silver flatware” are separable because those designs could just as easily have been depicted on a poster or in a painting.15 The same is true of the designs identified separable by the Copyright Office in its Compendium of U.S. Copyright Practices – “[a]n engraving on a vase,” “[a]rtwork printed on a t-shirt,” “[a] colorful pattern decorating the surface of a shopping bag,” ‘‘[a] drawing on the surface of wallpaper,” and “[a] floral relief decorating the handle of a spoon.”16 What all of those examples have in common is that they treat physical separability as the core and conceptual separability simply as a coda to physical separability.17 Specifically, they all involve features that one could easily view as copyrightable works if features were physically extracted from the the articles in which they were fixed (say, taking the artwork printed on a t-shirt off of the shirt and printing it on a poster), but it happens that physically extracting the designs may not be possible because doing so would destroy the physical objects in which they were fixed. Take, for example, an artistic design printed onto a piece of paper and then veneered onto the back of a chair. That design might be separable in the ordinary physical sense. By contrast, the same artistic design carved into the back of the chair might not be removable without destroying the chair. The design is nevertheless “conceptually” separable: it would be the same graphic work if it were printed on a poster or fixed in any other medium of expression. The design can exist as a copyrightable work apart from the underlying useful article. Similarly, while the plaster-of-Paris sculpture encasing the electrical cords and wiring of the lamp in Mazer v. Stein could not be physically separated from the cords and wiring without destroying the lamp (because of the materials from which it was made), the sculpture was conceptually separable because it would have been the same sculptural work without the wiring – it just would not have been a lamp.18 These examples suggest that, in the view of Congress, potentially copyrightable applied art differs from industrial design in that applied art can stand apart from a useful article and be recognized as a copyrightable work. Aesthetic features of a useful article are “capable” of existing separately if they can be removed or extracted from the useful article and the removed features would constitute a PGS work, but not the design of a useful article. By contrast, the industrial design features Congress meant to exclude from copyright protection are integrated into works of industrial design and are difficult to conceive of apart from those useful articles. Not incidentally, features that are inseparable from the article’s function

15 16

17

18

Id. U.S. Copyright Office, Compendium of U.S. Copyright Office Practices § 924.2(B) (3rd. ed. 2014). This paragraph, and the two that follow, are adapted from the Brief of Amici Curiae on Behalf of Intellectual Property Professors in Support of Petitioner, Star Athletica L.L.C. v. Varsity Brands, Inc., No. 15-866 (S. Ct. 2016). I was a co-author on that Brief with Mark Lemley, Chris Sprigman, and Rebecca Tushnet. See Mazer v. Stein, 347 U.S. 201 (1954).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.038

348

Mark P. McKenna

can readily be protected through the law of design patent, and as a result, maintaining the distinction Congress intended serves the purpose of channeling protection of industrial design to design patent.19

28.2 the imperfection of the applied art/industrial design distinction The Supreme Court missed an opportunity in Star Athletica v. Varsity Brands to put the law of separability on the course Congress intended because it failed to work outward from the specific examples that Congress and the Copyright Office used and therefore lost sight of the applied art/industrial design distinction.20 Because the Court abstracted away from that animating distinction, it ignored the possibility that some features, when imagined separately, “would necessarily bring along the underlying useful article.”21 That is a problem because, in light of copyright’s capacious originality standard, any feature of a useful article, described with sufficient particularity, could be “imagined” as a separate work.22 That rule excludes little from copyright, and in particular it accepts considerable overlap between copyright and design patent. But of course, even if the Court had been true to Congress’s purposes, the distinction between copyrightable applied art and uncopyrightable industrial design would only partially have channeled design patent subject matter out of copyright, because that distinction does not track very closely any definition of design patent subject matter. From very early on, design patent subject matter has included both surface ornamentation and configuration.23 Channeling industrial design out of copyright roughly addresses one category of design patent subject matter, namely configuration. But that mapping is highly imperfect – even if the examples of industrial design Congress used to illustrate separability were largely configurations, not all configurations are the sorts of “industrial products” Congress identified. Take, for example, Walt Disney’s design patent on a toy shaped like Mickey Mouse: 19

20

21 22

23

See Mark P. McKenna and Katherine J. Strandburg, Progress & Competition in Design, 17 Stan. Tech. L. Rev. 1 (2013) (focusing on the integration of form and function to define design patent subject matter). Star Athletica, L.L.C. v. Varsity Brands, Inc., 137 S. Ct. 1002 (2017). For a longer criticism of Star Athletica, see Mark P. McKenna, Knowing Separability When We See It, 166 U. Penn. L. REV. Online 131 (2017). Star Athletica, 137 S. Ct. at 1033 (Breyer, J., dissenting). See Feist Publ’ns, Inc. v. Rural Tel. Co., Inc., 499 U.S. 340, 345 (1991) (“Original, as the term is used in copyright, means only that the work was independently created by the author (as opposed to copied from other works), and that it possesses at least some minimal degree of creativity . . . To be sure, the requisite level of creativity is extremely low; even a slight amount will suffice. The vast majority of works make the grade quite easily, as they possess some creative spark, ‘no matter how crude, humble or obvious’ it might be.” (citation omitted)). See Sarah Burstein, The “Article of Manufacture” Today, 31 Harv. J. Law & Tec. 781, 785–86 (2018).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.038

Separability as Channeling: A Cautionary Tale

figure 1

24

24

U.S. Patent No. D82,802 (issued Dec. 16, 1930).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.038

349

350

Mark P. McKenna

Toys are, without question, articles of manufacture, the designs of which are subject to design patent protection.25 And this design patent claims the overall configuration of the toy. Still, the configuration of a Mickey Mouse toy seems quite different than the shape of an automobile, a lady’s dress, or an airplane. Overall configuration is simply not co-extensive with industrial design. The applied art/industrial design distinction fails more dramatically as a channeling principle with respect to surface ornamentation. Nearly all of Congress’s applied art examples fit neatly in the category of surface ornamentation. The engraving on the back of a chair, the floral design on silverware, patterns on china, and the design on the front of a t-shirt all could exist as separate PGS works without “bringing along” the underlying useful articles – and that is true precisely because those things merely decorate the surfaces of the articles they adorn. As a result, those features pass the separability test, making them potential copyright subject matter, even while they also seem like core design patent subject matter. All of this is to say that the applied art/industrial design distinction, while tractable and more limiting of copyright subject matter than the Supreme Court recognized in Star Athletica, only weakly channels design patent subject matter out of copyright. There is, of course, nothing inevitable about the applied art/industrial design distinction as the dividing line between copyright and design patent. Nor is there anything immutable about design patent subject matter. We could, for example, redefine design patent subject matter to focus on designs that integrate form and function in a meaningful way.26 That way of understanding design patents would probably rule out most surface ornamentation, at least when the ornamentation is not, by necessity, combined with configuration. Under that approach, separability would be a better approximation of design patent eligibility. But the operation of separability doctrine against the background of Congress’s intention in developing the applied art/industrial design distinction is a cautionary tale about channeling doctrines more generally. To do channeling well, policymakers developing channeling rules need a clear idea of the subject matter they intend to keep out of one regime, and that can only happen when they have a clear idea of what belongs in some other regime.27 And there’s a strong tendency for one regime’s understanding of what belongs in other regimes to become stylized and disconnected from the actual operation of those other regimes. It’s far from clear, for example, that trademark law’s understanding of the types of utilitarian function that should be excluded from trademark protection because they “belong” to patent law corresponds to the operative subject matter rules in utility patent. For that reason, we 25

26 27

35 U.S.C. § 171 (“whoever invents any new, original and ornamented design for an article of manufacture may obtain a patent therefor . . .”). McKenna and Strandburg, supra note 19 at 48–51. See generally Mark P. McKenna & Christopher J. Sprigman, What's In, and What's Out: How IP's Boundary Rules Shape Innovation, 30 Harv. J. Law & Tec. 491 (2017).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.038

Separability as Channeling: A Cautionary Tale

351

ought to think carefully about how much work we expect channeling doctrines to do in preventing overlap. We might be better off in many circumstances focusing on subject matter limitations that don’t require external reference. Doing so may reduce the instance of overlap and take pressure off these imperfect channeling doctrines. In copyright, that would require rethinking the low originality threshold or giving idea/expression more bite. It is precisely the growth of copyright subject matter and the infrequent application of other limits with respect to PGS works that makes separability so salient in these cases. The reminder to be attentive to internal limitations is a fitting tribute to Professor Kur and her substantial contributions to our field.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.038

29 Novelty, Idea or New Meaning as Criteria for Copyright Protection? Transitions in Swedish Design Law Jan Rosen1

29.1 introduction While the Court of Justice of the European Union (CJEU) has proclaimed a harmonized concept for the very object of copyright protection – an original work – in a fairly comprehensive bulk of decisions,2 by mainly defining a protected work as being the result of an author’s own intellectual creation, as expressed in three different EU Directives,3 some recent signs of a Swedish fall-out from that concept are in view, as will be discussed in this text, concerning artistic works and applied art. It is written in the hope that Professor Annette Kur, being a long-time observer of Nordic IP law and an esteemed link between Nordic and Continental IP law, may take some interest in those possible transitions in Swedish design law.4 1 2

3

4

Professor of Private Law, Stockholm University. The (still) most prominent case is C-5/08, ECLI:EU:C:2009:465, Infopaq, followed by C-393/ 09, ECLI:EU:C:2010:816, BSA; joined cases C-403/08 and C-429/08, ECLI:EU:C:2011:631, Premier League; C-145/10, ECLI:EU:C:2013:138, Painer; C-406/10, ECLI:EU:C:2012:259, SAS Institute; C-173/11, ECLI:EU:C:2012:642, Football Dataco; C-355/12, ECLI:EU:C:2014:25, Nintendo; C-310/17, ECLI:EU:C:2018:618, Levola Hengelo, and others. References were made, basically, to the Software Directive 2009/24/EC (uncodified version 91/ 250/EC), the Database Directive 96/9/EC and the Term Protection Directive 2006/116/EC, and the almost identical phrasing in those Directives of the contours of an original work. The CJEU finds reason to note that Article 1.3 of the Software Directive says: “A computer program shall be protected if it is original in the sense that it is the author’s own intellectual creation. No other criteria shall be applied to determine its eligibility for protection.” Article 3.1 of the Database Directive, as far as copyright protection for databases is concerned, says that such protection may come about if they “by reason of the selection or arrangement of their contents, constitute the author’s own intellectual creation . . . No other criteria shall be applied to determine their eligibility for that protection.” Article 6 of the Term Protection Directive (uncodified original version, Directive 93/98/EC) says the following: “Photographs which are original in the sense that they are the author’s own intellectual creation shall be protected in accordance with Article 1. No other criteria shall be applied to determine their eligibility for protection.” Professor Annette Kur was appointed honorary doctor of law at Stockholm University in 2012.

Downloaded from https://www.cambridge.org/core. 352 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

Transitions in Swedish Design Law

353

As just indicated, the CJEU has basically borrowed the said terminology from the Computer Programs (Software) Directive and some other Directives, but it can be traced back to the Berne Convention’s article on the protection of collections, for example anthologies, as literary works in their own right. In the eyes of the CJEU a work is an original intellectual creation of the author on condition that it is “reflecting his personality and expressing his free and creative choices in its production”. In the landmark Infopaq decision, the CJEU stresses, at no. 35, that the reproduction rights which were focused on in the case were valid merely for a work original in the sense that it is the author’s own intellectual creation, thus as a result of creative choices – what has eventually become normative for all kinds of works, according to the CJEU. What seems clear from the Court’s case law is that an intellectual creation easily qualifies as original. Obviously, in many copyright laws, the work of authorship has already become a vessel that accommodates a very broad array of works of the mind, from high art to low art, from the purely aesthetic to the (almost) predominantly functional or technical. The standard for protection in many jurisdictions has evolved to the point where original and creative sometimes seem to be synonymous terms, both meaning little more than not directly copied or resulting from a modicum of freedom of choice. It is important to note, though, that the CJEU’s recent case law certainly doesn’t do away with the creative element of an author, absolutely being a physical person (or a group of physical persons), and the amount of his or her factually available choices facilitating the form of a work. Hence, it is not possible among the Member States to uphold, for example, stricter originality terms for certain types of works, as some States probably have done. A more doubtful possibility, however, is if a fully harmonized concept for originality is valid also for designs, or objects of applied art, although the CJEU may indicate a general and cohesive application within the whole field of copyright objects. According to the Designs Directive,5 Article 17, a registered design right in or in respect of a Member State in accordance with the Directive, shall also be eligible for protection under the law of copyright in that State. Each and every Member State thus decides “the extent to which, and the conditions under which, such protection is conferred, including the level of originality required”.6 Therefore, one could argue that the substance of originality would (still) be an internal matter for the individual Member State where designs are concerned. But one should bear in mind that the CJEU generally prefers interpretations and applications best suited for the most profound EU harmonization and integration,

5

6

Directive 98/71/EC of the European Parliament and of the Council of 13 October 1998 on the legal protection of designs. Cf. C-683/19, ECLI:EU:C:2019:721, Cofemel; and C-833/18, ECLI:EU:C:2020:461, Brompton Bicycle.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

354

Jan Rosen

namely those that drive in the direction of a harmonized originality concept also valid for designs. This was confirmed several years ago by the Swedish Supreme Court, concerning an internationally well-known design, that of the Mini Maglite torch. In that case, the Court set out in detail the prerequisites for copyright protection of a work of applied art well in line with what eventually was to come out of the CJEU’s own judgment practice.7 The Supreme Court underlines that works of applied art must show the same amount of originality as other copyright works, and that there was no reason to apply a higher or specific originality request for a work of applied art, different from other copyright works, although a work of applied art may be very bound by its inherent functional elements.8 The Supreme Court proclaimed that to allow copyright protection to a design of applied art the work should (i) distinctly rise above being banal, (ii) not be the result of a routine effort, and (iii) show a modicum of originality or independence as compared to works of others of the same kind.9 However, three recent decisions of Swedish courts seem not to be in line with those findings of the CJEU contoured above. Those cases form the basis for this study. I will insert here that, as of 1 September 2016, there is a new Swedish court structure specially for intellectual property and market/competition law cases, making the Stockholm district court of first instance a special court for all IP and market law cases in Sweden, called PMD,10 and the Svea Appeal Court (also in Stockholm) the final special appeal court, now called PMÖD.11 An appeal to the Supreme Court of Sweden will still be possible, but only if the PMÖD so decides. The aim of this reform is to strengthen the knowledge and expertise of the judges handling complex intellectual property and market law phenomena and, naturally, to make the procedures before the courts more speedy than before. Thus, in IP law, Sweden may currently be said to have two precedent courts, the Supreme Court and the PMÖD. But already the PMD, certainly intended to be a court of specialists, has produced a fairly large number of judgments that demand a certain respect and attention, as decisions of a special court probably should, and fall in line with the actual aim of the IP court reform in Sweden.

7 8 9

10 11

See NJA 2009 p. 159, Mini Maglite. Analysis by Rosén available at www.InfoTorg.se. Hereby referring to prop. 1969:168 s 124 et seq. and 135. Cf. the judgment of the German BGH, I ZR 143/12, 2013.11.13, Geburtstagszug [Birthday Train], whereby the BGH confirmed that after the 2004 reform of the Geschmacksmustergesetz the traditional distinction between “works of applied art” and “works of purpose free art” has been abandoned. We may immediately note, that those quite distinctive prerequisites of the Swedish Supreme Court should, in my mind, have barred protection to the banal design, or concept, of the anchor bracelet, as well as that of the Daniel Wellington clock, described under sections 29.2 and 29.3 below. Patent- och Marknadsdomstolen (PMD). Patent- och Marknadsöverdomstolen (PMÖD).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

Transitions in Swedish Design Law

355

All the more interesting then to check out below the PMD’s “anchor bracelet” decision,12 the PMÖD’s “clock” decision,13 as well as a fairly recent judgment of the Supreme Court on the effects of a “new meaning” of a painting.14

29.2 the anchor bracelet design A few years ago a type of decorative bracelet based on marine features became very popular on the market, and was sold in many shops, on the internet and elsewhere, at low prices. The concept for most of these bracelets, at least for those being tried by the PMD, was easy to identify: (i) a standard marine cord of polyester, to be found in any marine shop; (ii) both ends tied in a loop by a thin wound string, as found on marine ropes or moorings; and (iii) a mini log anchor (in typical design, but with the Tom Hope TM stamped in the metal) holding the loops together, the size of it all adjusted to fit a person’s wrist (see below). Tom Hope Ltd, a Swedish entity designing and marketing low-price jewelry, claimed before the PMD that they owned the design of such anchor bracelets, it being an original work of the company’s own (named) designer, and that Ur & Penn AB,15 by marketing a series of anchor bracelets, varied somewhat as to coloring from those put on the market by Tom Hope, and featuring the same or almost identical forms, [Jan, this sentence needs to be completed at proof stage]. The bracelets of Ur & Penn AB were produced in China to order, but were merely based on Ur & Penn’s request to receive “anchor bracelets”, à la mode worldwide at the time. The PMD in its judgment starts off by noting that the construction of a bracelet tied together with an anchor, in particular in the shape of a “classical” log anchor, and otherwise consisting of marine strings, was nothing new to the world when Tom 12 13 14 15

PMD Mål nr PMT 1803-17, 2017.12.21, Ankararmband. PMÖD Mål nr PMT 5885-18, 2019.03.22, Daniel Wellington. NJA 2017 p. 75, Svenska syndabockar (Swedish Scapegoats). Ur & Penn AB, a Swedish entity, selling since the early 1940s clocks and cheap jewelry via its well over 100 shops throughout Sweden.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

356

Jan Rosen

Hope introduced its product. However, the court declared that an original form may come about by the combination of known/available objects: “previously known form elements pulled together to constitute a product, doesn’t prevent them from being combined in such a way so as to allow the overall expression to be afforded copyright protection” (my translation). Naturally, this is per se undisputable. The court, thus finding no originality as to the shape of the little anchor, the mustered marine strings or those wound string loops, continued to state that the bracelet(s) of Tom Hope, as a whole, met the demands of originality, “though with a not so high margin”. Further, in spite of the fact that Ur & Penn sold a bracelet with a slightly differently designed anchor and with somewhat different winding of the loops, hence no plain plagiarism, the PMD found its anchor bracelet to violate the rights of the author of Tom Hope’s design. This outcome of the PMD’s judgment must be considered as quite remarkable, as it obviously ignores the shape of the product emanating almost directly from the basic idea of an anchor bracelet, composed by those three “standard” elements described above, and realized basically in one and the same way. The overall impression of Tom Hope’s bracelet is thus given already by the idea of the anchor bracelet – to make a bracelet with those standard components – particularly as it showed nothing declaring some individuality. In the individual case, the strings or the anchor could of course have been shaped differently and in a more original way, in which case they could have evoked an overall impression of an original quality – but quite obviously, in my eyes, this was not the case here. Anyhow, those features of a protected design, as stated by the Swedish Supreme Court, noted above – namely that a design should (i) distinctly rise above being banal, (ii) not be the result of a routine effort, and (iii) show a modicum of originality and independence compared to works of others of the same kind – are simply not met by Tom Hope’s bracelet. To conclude, Tom Hope’s anchor bracelet must almost per definition be called banal, lacking any individual independency relative to works of the same kind. It might have been clearly à la mode at the time, hence “new” and in that sense adequate on the market. But I would still claim that it stems from a distinctive and quite strict idea, making it non-eligible for copyright protection according to contemporary EU standards.16

16

In Birthday Train, see note 9 supra, the German Court of Appeal, to which the case was remanded back, also argued that the idea of a “birthday train” was not protected and that the train itself (without the candles and numbers which were discounted under the idea/expression dichotomy) was not sufficiently different from other toy trains. One may observe, though, that in the design world arguably the idea and expression tend to merge into a “concept”; see Antoon Quaedvlieg, Overlap/Relationships between Copyright and Other Intellectual Property Rights, in Research Handbook on the Future of EU Copyright, ed. Estelle Derclaye, 480 (Edward Elgar Publishing, 2009).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

Transitions in Swedish Design Law

357

29.3 the daniel wellington wristwatch design In this case (see note 13 supra), Daniel Wellington AB firstly claimed that their design of a wristwatch, the DW clock pictured below, was an artistic work, either by itself or in combination with a so-called NATO textile bracelet (in various coloring combinations) with which it was put on the market. The Daniel Wellington entity was not claiming any rights to the bracelet as such, known to have been put on the market by others since the 1960s, but to the combination of it with their watch design. Secondly, the company claimed inter alia that the design of a wristwatch, marketed under the William Gregor trademark by (again) Ur & Penn AB (the WG watch, see picture below), was using the protected design of the DW clock. The PMÖD noted that the DW clock itself consisted of elements often used in wristwatch design going back a long time, but that nevertheless the chosen combination of such elements may, seen as a whole, render the designer copyright protection for the DW clock. And there were indeed, the PMÖD observed, quite a large number of watches on the market, well before the DW made its commercial entry, Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

358

Jan Rosen

pretty much having the same “look” as the DW clock, as was shown by the defendant in the case. However, the court observed especially that (i) the edging had the same color as the dial plate, which made the watch look bigger than it actually was, and (ii) the flat glass on the clock face and the beveling of the watch’s edge gave a thin and flat expression to the watch, which was in line with the intentions of the designer, as told to the court when he was heard as a witness. Hereby, the designer of the DW clock had, according to the PÖMD, executed his own free and creative choices, the design thus being his own intellectual creation. Hence, the PMÖD concluded, the DW clock itself (with or without the bracelet) was an original work, and not merely a use of elements from “the common form storage” for watches, as claimed by the defence. As for the WG clock, the PMÖD found it to be “almost identical” to the DW clock, although it noticed that the WG clock was slightly bigger than the DW clock, that the color of the WG clock face was of a different shade from that of the DW, Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

Transitions in Swedish Design Law

359

and that the WG had a second-hand. Other factual differences, such as the design of the tiny horns holding the bracelet, were not observed by the court as being of any importance in this context. For myself, I am unable to feel any enthusiasm for the outcome of the decision by the PMÖD. Not that the design of the two compared objects does not offer the same impression – they do, of course! But this is because both resemble an ocean of identical or almost identical wristwatches, easily found on the market for decades. As Ur & Penn had exhaustively showed in its defence, there were many more or less identical precedents to the DW clock, such as, to mention just a few, the Patek Philippe model Calatrava (ref. 3537) from the 1930s, the Seiko Cronos model (2008), the Omega Genève (1969) and Certina New Art (1970). But many others are easy to find in the repertoires of a number of well-known clock producers – in particular if the overall impression is sought for, disregarding differences in details, as the PMÖD actually did. In my eyes, the design efforts behind the DW clock are restricted merely to the choice of an already existing model or, rather, appearance of a watch, considered to have a chance of commercial success at a particular time, in other words to matching what’s in fashion by picking something out of the available common storage of wristwatch design. Such a choice can’t fall in line with what the criterion of “creative choice” stands for, as stamped by the CJEU. Quite another issue is Ur & Penn’s very probable decision to follow up on the commercial success of Daniel Wellington by copying the appearance of the DW watch design. Probably, the PMÖD was very eager to attack plagiarism, generally speaking, thus to offer protection against commercial actors prone to let a Chinese factory copy and produce for them any bestselling item on the Swedish market. But in this case the delicate instrument of copyright was not at hand, I would claim, notwithstanding that, for example, trademark law or unfair competition law might have been available instruments.

29.4 “new meaning” of an artistic work A judgment of the Swedish Supreme Court, also from 2017, being an old enough case to have fallen out of the new procedural order of the PMD and PMÖD, adds to the confusion of what prerequisites may constitute a work protected within the frames of copyright law. The basic story was the following. For a couple of days a photographer followed and took pictures of Christer Pettersson, who had been found guilty of the murder of Swedish Prime Minister Olof Palme, but was eventually acquitted. A close-up portrait of Mr Pettersson was, after some technical touching up, frequently seen in the media. Without permission of the photographer, a professional painter, MA, used that portrait photo much later as a model for an oil painting on canvas called Scapegoats of Sweden. The painting displays Mr Pettersson, as he appears in the photographical picture, but in a deserted Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

360

Jan Rosen

landscape together with a goat with a blue and yellow (the colors of the Swedish flag) ribbon around his horns (see picture below). The painter exhibited the painting at the Modern Arts Museum of Stockholm, exposed it on his public website and sold copies of it in the form of posters. The photographer later claimed before the courts that MA had violated his copyright in the work by making copies of it and communicating it to the public. All courts trying the case found the photographical portrait to be an artistic work, independently created and original enough to be afforded full protection by copyright proper, not merely protected as any photographical picture.17 The pertinent issue of relevance in the case was, in short, how to draw the line between an adaptation, protected as such but depending on the protection of the underlying work (the photographic portrait), and a new original work, containing or using another work but not carrying its protected/original features.18 The court of first instance found that MA had not created a new work by himself but had made copies of the photographer’s work, while the appeal court came to the opposite conclusion.19 The appeal court stressed that the painting was an original work, not an adaptation of the artistic/photographical work, but made in a free connection to that work by a reference to it. The Supreme Court started off by pointing out what was said to be the decisive factor in this case – whether the work being used as a model had been changed or

17

18 19

The Swedish Copyright Act distinguishes between photographic works, being original and therefore treated just as other forms of original artistic works, like original paintings, on the one hand, and mere photographic pictures, not showing originality, but protected like other accomplishments as neighboring or related rights, on the other. Cf. Article 4, paragraph 1 and 2, of the Swedish Copyright Act. We may note here that the Swedish Copyright Act, ibid., distinguishes between an original work and an independent adaptation as well as dependent adaptation of it. An adaptation that is independent of the original form of the model work is rendered full protection, whereas the adaptation lacking that element of independence is also offered protection as a work, but this cannot be used without consent from the owner of rights to the underlying original work.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

Transitions in Swedish Design Law

361

reshaped enough to emerge as a new and independent work by the painter, or if MA’s work was an adaptation, still in dependence on the original photo. From this position the Supreme Court first concluded that merely the portrait as such was to be seen as an adaptation of the photographer’s work as it was used in the painting, although the Court conceded that this was a borderline issue. Generally, by this finding of the Court, the conclusion would pretty much have been a given – the painter had used a work by someone else. This is because someone else’s original work, the photographical rendition of Pettersson’s face, had factually been used in the painting, thus reproduced, and copies were indeed made thereof. There were some differences between the painting and the photography, mostly of a technical nature, due to the transformation of the photography to the artistic form of the painting and because of other added items in the painting. But the Supreme Court made a statement that must be considered as a reduction of the realm of a protected work, disregarding the fact that the painting contained at least a dependent adaptation of the photographic picture: “One should not look merely at the photographic work as such, as it is actually seen in the oil painting, but take in what the whole painting is showing!” The composition of the painting dominated over the photographic picture, as the former was made in dark, drab colors, displayed a deserted landscape and, in particular, the emblematic and thus important (scape)goat. Those features afforded the painting a totally different meaning than that of the portrayal photography, the Supreme Court concluded: instead of a photographically strong portrait, the painting displayed an allegory suggesting criticism of a media craving for scapegoats, and displaying Mr Pettersson as a scapegoat; the painting was a comment on society and the prevailing times, whereby Pettersson, as a phenomenon, not a person, together with the goat and the painted background, conveyed a symbolic message. What strikes us here is the fact that the Supreme Court finds the photographic work, although said to be protected as a (dependent) adaptation when appearing in the painting, to have undergone a “change” as used in the new context or “totality” of the painting. The protection of the photographic (artistic) work is hereby consumed by another artist’s work, letting an overall assessment lead to the painter having created a new and independent original work. This way of reasoning seems to be meant as a fair gesture to support new thoughts and ideas of artists, even if such acts would factually be using the works of other artists in a narrow copyright sense. But this is not really in line with the traditional view in copyright law, and it might result in future challenges. Millions of works are used on a daily basis, particularly in the digital environment, and presented in very different contexts. Hereby, they may be part of something possibly having a “different meaning”, based upon an overall evaluation of a new context as compared to the context from which they originated – a classic issue of “transformative use”. Following the decision of the Supreme Court, an original work, in full protection Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

362

Jan Rosen

by its actual form, would easily lose that protection if a user decided to “transform” it merely by presenting it in a way so that the “meaning” of the original, whatever that might have been, emerges as “different” from that of the new work.

29.5 some conclusions The protection of designs generally and, not least, items of applied art, has had a fairly rocky road to follow during the last 100 years before it was granted full protection within the framework of IP law. The Designs Directive is a fairly new instrument for EU-wide protection, and national protection has a recent, yet complex and not particularly coherent, history among EU Member States and worldwide.20 Maybe the same can be said about designs and applied arts protection by copyright law proper; authorian works, creations by authors as original artistic objects, were relatively late in being provided with copyright protection by national legislators (e.g. about 1970 in Sweden) and the CJEU (during this century).21 Nevertheless, how to set the contours of an original work, be it something called applied art or a phenomenon of a literary or artistic nature, has now become an issue in EU law, at least according to the CJEU, hence a fully harmonized legal figure within the EU. The CJEU has set out some often-observed criteria, namely: (i) originality; (ii) something being the author’s own creation; and (iii) no or minimal risk for independent double creation, etc. We may also note the CJEU’s observations (iv) that the author’s own creation shall reflect his or her personality or personal touch,22 or (v) that the author has expressed his or her creative capacity by factually making creative choices,23or (vi) that selection and compilation may be expressions for creative freedom,24 or, lastly, (vii) that creativeness and skill have been demonstrated by the author coming to the same functions and purposes as the original specimen, but not by using the same source or object codes.25 But none of those criteria are really to be found in the judgments by Swedish courts analyzed above; or, rather, none have been reasonably applied by the courts. Still, those cases cannot be observed merely as a few mishaps in the mighty stream of (adequate) IP decisions. The first two were decided by the new specialist IP court of 20

21

22 23 24 25

There are different legal positions on this in different countries, though; designs were only afforded protection as works of authorship at a relatively late stage, at least by some national legislators. By contrast Germany, for example, has protected works of applied art since 1907. It still seems proper to be careful with this particular issue, even though works of applied art are no doubt protected under EU law (see the BSA case, C-393/09), and, as confirmed in the Cofemel and Brompton Bicycle cases (see note 6 supra), such harmonized EU law on the protection of works equally embraces (functional) design objects and works of applied art. Cf. the Painer case in note 2 supra. Cf. the Painer and, e contrario, Premier League cases in note 2 supra. Cf. the Dataco case in note 2 supra. Cf. the SAS Institute case in note 2 supra.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

Transitions in Swedish Design Law

363

Stockholm (PMD, in 2017) and the Appeal Court (PMÖD, in 2019) respectively, the third by the Supreme Court of Sweden, in 2017. Obviously, neither the Tom Hope design of the anchor bracelet nor the DW wristwatch were really tested on the basis of any of those criteria set forth by the CJEU, though the courts made references to relevant CJEU decisions; rather, protection was based on a typical “no no” in copyright law, namely paying homage to the idea or concept of an object. The mere use of the three conceptual items of an anchor bracelet – a standard marine cord of polyester, tied in a loop at both ends by a thin wound string, with a mini log anchor holding the loops together – and, much less, picking a clean-looking clock design from the “storage”, simply cannot demonstrate an author’s own creation, unless showing considerable individualized features and form elements adding to the used concepts. To afford those designs, or rather concepts, copyright protection for about the next 100 years, I find grossly exaggerated; if anything (any idea) is protected, then nothing factually is! This “generosity” of the court does not really help authors, apart from the claimant in those (hopefully) isolated cases. I do not think, however, that those decisions came out of “rebellion” against the CJEU – which has, after all, created a concept of work out of thin air – or preoccupation with good old national traditions (certainly not!), but maybe out of sheer negligence or, more probable, eagerness to fight plagiarism and commercial rip-off in the design industry and to open up for artistic freedom in the field of pure art. As for the decision of the Supreme Court, it is dualistic in its reduction of the scope of per se undenied protection of a work, namely that of an original photographic portrayal of a person, on the one hand, and its “generous” protection of an author’s work which incorporates another work, because the new context gives it all a “new meaning”, on the other. The latter formula of the Supreme Court was probably construed in some haste, and it might open the door for massive (re)uses of protected objects without permission being needed. Applications in the world of AI demonstrate such fears, for example by production of collections of protected works resulting in new “creations” which have a “meaning” different from the individual works, which are being exploited by the running of the machine.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.039

30 Examining Functionality Mark D. Janis1

30.1 introduction United States design law – that is, the collocation of design patent law and designrelated aspects of copyright and trademark law – seems to be moving ever further from a state of reasonable equilibrium. Functionality doctrines exist at the very hub of this upheaval,2 and the task of improving coherence among these doctrines, internally and trans-substantively, has become critical.3 In this chapter, I select one functionality doctrine (trade dress functionality)4 and assess one source of incoherence: the mismatch between the modern substantive standard for trade dress functionality and the procedural constraints of ex parte examination practice. By asking how trade dress functionality does (and should) work in the context of ex parte examination, I seek to call attention to a larger 1

2

3

4

Robert A. Lucas Chair of Law, Indiana University Maurer School of Law (Bloomington). Thanks to LaShaila Spivey, Ying Zhu, and Allison Strong for excellent research assistance. See, e.g., Christopher Buccafusco, Mark A. Lemley, and Jonathan S. Masur, Intelligent Design, 68 Duke L.J. 75 (2018), 82–86 (commenting on the centrality of functionality doctrines); but cf. Perry J. Saidman, The Demise of the Functionality Doctrine in Design Patent Law, 92 Notre Dame L. Rev. 1471 (2017) (asserting that functionality no longer plays a useful role in either validity or infringement analysis, and should be discarded from design patent law). See, e.g., Christopher Buccafusco and Mark A. Lemley, Functionality Screens, 103 Va. L. Rev. 1293 (2017) (assessing the operation of functionality and analogous doctrines across design patent, trademark, and copyright regimes, and suggesting reforms). Faced with a problem of incoherence in design law in Europe, Professor Annette Kur and colleagues responded with a proposal for autonomous legislation that presaged the creation of the Community Design system. See Annette Kur and Marianne Levin, The Design Approach Revisited: Background and Meaning, in The EU Design Approach: A Global Appraisal 1 (Annette Kur, Marianne Levin, and Jens Schovsbo, eds.) (Edward Elgar, 2018). In this chapter, part of a collection honoring Professor Kur’s work, it seems fitting to begin with a difficult problem of design law’s incoherence. I am confining my remarks here to utilitarian functionality. Aesthetic functionality presents related, but slightly different, challenges in ex parte examination.

Downloaded from https://www.cambridge.org/core. 364 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

Examining Functionality

365

problem in which judges lay out IP protectability (or registrability) standards with a vision about how those standards will operate in contested proceedings, but with too little regard for the difficulties of administering those standards in ex parte examination. That analysis in turn raises several broader questions: should the ex parte trademark examination process be reformed to empower examiners to play a more prominent role in enforcing compliance with substantive conditions of registrability? If so, what are the key institutional impediments, and how should they be addressed? Alternatively, should courts reformulate substantive trademark registrability doctrines to make them more compatible with the constraints of the existing ex parte examination process?

30.2 substantive trade dress functionality doctrine, reconfigured for ex parte proceedings Trade dress functionality is a doctrine of grandiose aspirations, purporting to “prevent . . . trademark law, which seeks to promote competition by protecting a firm’s reputation, from instead inhibiting legitimate competition,”5 and striving to do so by channeling protection of useful features away from the trademark regime and to the utility patent regime.6 But because non-functionality is a condition for registering trade dress on the Principal Register,7 it must be applied in the most humble of settings: ex parte examination of registration applications at the U.S. Patent and Trademark Office (PTO). To carry out that task, the PTO must penetrate the mysteries of the U.S. Supreme Court’s TrafFix opinion,8 and attempt to reconcile them with post-TrafFix pronouncements from the U.S. Court of Appeals for the Federal Circuit, the PTO’s reviewing court. This is no small endeavor. Under the TrafFix standard, trade dress is functional (in the utilitarian sense) when “it is essential to the use or purpose” of the associated article, or if “it affects the cost or quality of the article.”9 According to 5

6

7

8 9

Qualitex Co. v. Jacobson Prods. Co., Inc., 514 U.S. 159, 164 (1995). See Robert G. Bone, Trademark Functionality Reexamined, 7 J. Leg. Analysis 183, 230–242 (2015) (assessing whether functionality doctrine is well tailored to promote competition). Id. (“It is the province of patent law, not trademark law, to encourage invention by granting inventors a monopoly over new product designs or functions for a limited time . . .”). See, e.g., Mark P. McKenna, (Dys)functionality, 48 Hous. L. Rev. 823, 827 (2011) (remarking on the channeling function). Congress added Lanham Act § 2(e)(5), 15 U.S.C. § 1052(e)(5), as part of the Trademark Law Treaty Implementation Act, PL 105-330 (Oct. 30, 1998), 112 Stat. 3064 (Title II, Sec. 201 – technical corrections). Before that time, even though there was no explicit statutory basis for a functionality bar to registration, the U.S. Patent and Trademark Office (PTO) was issuing functionality refusals on the ground that the subject matter sought to be registered failed to function as a mark. See Trademark Manual of Examining Procedure 1202.02(a)(i) (Oct. 2018) (hereinafter “TMEP”). TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 (2001). Id. at 32.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

366

Mark D. Janis

TrafFix, decision-makers should treat utility patent evidence as critical in this inquiry,10 but they should not indulge in “speculation about other design possibilities” that would achieve the same functional purpose where the trade dress has been deemed functional under the TrafFix test.11 Throughout, the TrafFix opinion distinguishes between “arbitrary, incidental, or ornamental” features (which are potentially protectable) and all others (which are evidently at risk of being deemed functional).12 By contrast, the Federal Circuit’s pre-TrafFix functionality cases had adopted a four-factor balancing test (the “Morton-Norwich factors”) 13 that treated utility patent evidence as merely one factor to be considered, but that also emphasized reliance on evidence of alternative designs on the ground that “competitive necessity” should be the touchstone for assessing functionality. 14 Moreover, the Federal Circuit had highlighted the distinction between de facto functionality and de jure functionality to warn the PTO against basing functionality refusals on mere evidence that a trade dress feature was useful for something. 15 Post-TrafFix, the Federal Circuit has retained its Morton-Norwich factor test,16 while simultaneously indicating that the PTO may truncate the test to omit the alternative designs factor if patent and advertising evidence (Morton-Norwich factors 1 and 2) establish functionality.17 The Federal Circuit also continues to invoke the distinction between de facto and de jure functionality, commenting that de facto functionality “simply means that a design has a function” and observing that de facto

10

11

12 13

14 15 16 17

Id. at 29. According to the Court, a prior utility patent is “strong evidence” of functionality if it claims the trade dress features. Id. In addition, the functionality inquiry can “be aided by going beyond the claims and examining the patent and its prosecution history to see if the feature in question is shown as a useful part of the invention.” Id. at 34. Id. at 29, 33 (rejecting “competitive necessity” as a “comprehensive definition” of functionality). Nonetheless, some courts have continued to rely on evidence of alternative designs in assessing utilitarian functionality, using various strategems to justify doing so. See, e.g., MoldexMetric, Inc. v. McKeon Prods., Inc., 891 F.3d 878 (9th Cir. 2018); Valu Eng’g, Inc. v. Rexnord Corp., 278 F.3d 1268 (Fed. Cir. 2002). Supra note 8 at 34. In re Morton-Norwich Prods., Inc., 671 F.2d 1332 (CCPA 1982). The Morton-Norwich test involves balancing the following factors: (1) the existence of a utility patent that discloses the utilitarian advantages of the design sought to be registered; (2) advertising by the applicant that touts the utilitarian advantages of the design; (3) facts pertaining to the availability of alternative designs; and (4) facts pertaining to whether the design results from a comparatively simple or inexpensive method of manufacture. Id. at 1340–41. Id. at 1339. Id. at 1338. In re Becton, Dickinson & Co., 675 F.3d 1368, 1374–75 (Fed. Cir. 2012). Becton Dickinson, 675 F.3d at 1376. See also Valu Eng’g, Inc., 278 F.3d at 1276 (ruling that “once a product feature is found functional based on other considerations there is no need to consider the availability of alternative designs”).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

Examining Functionality

367

functionality “is irrelevant to the question of whether a mark as a whole is functional so as to be ineligible for trademark protection.”18 The PTO, via its Trademark Manual of Examining Procedure (TMEP),19 has gamely attempted to synthesize all of this while also adding its own gloss.20 Two elements of the PTO’s instructions are especially notable. First, the PTO adheres to the Morton-Norwich factors as the initial organizing rubric for utilitarian functionality analysis,21 but it seizes on opportunities to truncate it. Indeed, the PTO arguably goes well beyond the Federal Circuit in that the TMEP encourages examiners to reduce the Morton-Norwich test to a single-factor inquiry based on utility patent evidence, instructing that “[e]vidence that the proposed mark is the subject of a utility patent that discloses the utilitarian advantages of the configuration at issue can be sufficient in itself to support a functionality refusal.”22 This is especially significant when coupled with the rules that (1) any part of the patent document (claims or disclosure) may be probative;23 (2) the patent’s prosecution history may be consulted;24 and (3) there is no requirement that the utility patent be owned by the trade dress applicant; third-party patents may form the basis for a functionality refusal.25 Second, on the critical conceptual question of distinguishing de facto from de jure functionality, the TMEP offers contradictory advice. It acknowledges that prior Federal Circuit and PTO decisions had relied on the distinction, but recommends that “in general, examining attorneys no longer make this distinction in Office actions that refuse registration based on functionality” given that the distinction does not appear either in TrafFix or in the Lanham Act.26 However, having put the distinction to rest, the TMEP then immediately resurrects it, reciting the rule that “[d]e facto functionality is not a ground for refusal.”27 The complexities that I have noted are difficult enough to manage in the course of a full-fledged adversarial process involving economically motivated litigants. The difficulty is only magnified when the functionality standard must be adjudicated in the ex parte process, in which examiners operate under severe time pressure, lack the means to compel discovery or retain experts,28 and nonetheless bear the ultimate 18

19

20 21 22 23

24 25 26 27 28

Becton Dickinson, 675 F.3d at 1373–74. The court drew from its pre-TrafFix precedent, including In re R.M. Smith, Inc., 734 F.2d 1482 (Fed. Cir. 1984). The TMEP does not have the force of law but is authoritative as a guide to actual PTO practice. See generally TMEP § 1202.02(a). Id. at 1202.02(a)(v). TMEP § 1202.02(a)(v) (citing TrafFix and decisions from the TTAB). TMEP § 1202.02(a)(v)(A) (“the specification and/or drawings may also provide strong evidence [of functionality], even if the feature is not claimed as part of the invention.”) Id. Id. TMEP § 1202.02(a)(iii)(B). Id. (citing pre-TrafFix decisions of the PTO’s Trademark Trial and Appeal Board). See generally 37 CFR §§ 2.61–2.69 (“Examination of Application and Action by Applicants”); TMEP Ch. 0700 (“Procedure for Examination”). In the ordinary case, the ex parte trademark

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

368

Mark D. Janis

burden of proving that registration is barred.29 Of course, this problem arises in the course of many other aspects of the PTO’s ex parte examination responsibilities in both patent and trademark, and the PTO sometimes responds by deploying procedural mechanisms that attempt to shift costs (by way of shifting burdens of production) back to applicants. The PTO has done this for the ex parte examination of trade dress functionality, but the procedural mechanisms present problems of their own.

30.3 procedural mechanisms for adjudicating functionality ex parte Even before TrafFix, the PTO and the courts recognized the mismatch between the demands of substantive functionality analysis and the limitations of the ex parte process. In response, the PTO has adopted a simple and well-worn expedient: the prima facie case. The examiner need only make a prima facie case of functionality,30 and the burden of production then shifts to the applicant, who must produce “competent evidence” rebutting the prima facie case.31 In concept, the PTO’s maneuver is unremarkable.32 Moreover, the idea of shifting the burden of production might be said to line up well with the Lanham

29

30

31

32

examination process is based primarily on written exchanges between the applicant and the examiner. The applicant has the opportunity to submit expert evidence in affidavit form, but the examiner has no opportunity to retain an opposing expert or interrogate the applicant’s expert. This is significant because functionality is a complex inquiry that seems to invite proof through expert evidence, and the use of experts seems to be routine when functionality is litigated in court. See, e.g., Schutte Bagclosures Inc. v. Kwik Lok Corp., 193 F.Supp.3d 245 (S.D.N.Y. 2016) (relying on functionality testimony from alleged infringer’s plastics engineering expert); Keystone Mfg. Co., Inc. v. Jaccard Corp., 394 F.Supp.2d 543, 562 (W.D.N.Y. 2005) (noting the existence of competing expert reports on functionality that precluded the grant of summary judgment). See, e.g., In re St. Helena Hospital, 774 F.3d 747, 750 (Fed. Cir. 2014) (observing that the PTO is the burden holder according to the language of Lanham Act § 2). TMEP 1202.02(a)(iv) (instructing that “[t]he examining attorney must establish a prima facie case that the proposed trade dress mark sought to be registered is functional in order to make and maintain the §2(e)(5) functionality refusal”) (citations omitted). The Federal Circuit’s postTrafFix case law endorses this approach. Becton Dickinson, 675 F.3d at 1374 (asserting that “[t]o support a functionality rejection in proceedings before the Board, the PTO examining attorney must make a prima facie case of functionality . . .”). Becton Dickinson, 675 F.3d at 1374 (indicating that if a prima facie case is established, the applicant must rebut it by “competent evidence.” In re Teledyne Indus., 696 F.2d 968, 971 (Fed. Cir. 1982), which requires “proof by preponderant evidence”). See also TMEP § 1202.02(a) (iv) (same). See, e.g., In re the Newbridge Cutlery Co., 776 F.3d 854 (Fed. Cir. 2015) (presuming a goods/ place association under specified conditions, for purposes of assessing whether a mark is geographically descriptive). For an example from patent law, consider the obviousness inquiry under 35 U.S.C. § 103 for utility patents – like functionality, another protectability condition that is of enormous policy sensitivity, resting heavily on technical facts and complicated, casedriven legal rules. See In re Oetiker, 977 F.2d 1443, 1445 (Fed. Cir. 1992) (invoking the prima

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

Examining Functionality

369

Act’s burden-shifting rule for functionality determinations for unregistered trade dress, although that rule shifts the burden of proof, not merely the burden of production.33 What warrants scrutiny is the Federal Circuit’s (and PTO’s) definition of the content of the prima facie case of functionality. In its early decision in Teledyne,34 the Federal Circuit seemed to adopt the position that to establish prima facie functionality in ex parte examination, the PTO need do nothing more than identify possible de facto functions of the configuration sought to be registered. That approach, the court made clear, was a pragmatic response to the limitations of ex parte examination. Indeed, the court was uncharacteristically frank: “We recognize that in most cases such as this, the best the PTO can probably do is to analyze a design from the standpoint of its de facto functional features, perhaps with support from technical articles, patent disclosures, or the applicant’s own advertisements.”35 This approach is premised on the assumption that the applicant is best suited to identify rebuttal evidence. As the court put it in Teledyne, the “[d]etermination that the design as a whole is not de jure functional may well be possible only in light of evidence more readily available to, or uniquely in the possession of, the applicant.”36 The prosecution history of the application at issue in Becton Dickinson illustrates how low the bar is for establishing a prima facie case of functionality. In the initial office action, the examining attorney entered a functionality rejection. The extent of the examiner’s reasoning was as follows: The proposed mark is the “configuration of a closure for medical collection tubes,” and the goods are “closures for medical collection tubes.” The particular feature of this proposed mark, namely, the circular closure, is functional for the goods because the closure is utilitarian – it serves a purpose. More specifically, the closure acts as a lid to keep substances within a medical collection tube.37

The TMEP embraces this approach to prima facie functionality and goes even further in facilitating refusals. According to the TMEP, “[i]n applications where there is reason to believe that the proposed mark may be functional, but the evidence is lacking to issue the §2(e)(5) refusal in the first Office action,” the examiner is to “obtain information from the applicant” through a request for

33 34 35 36 37

facie case of obviousness for use in ex parte utility patent examination); see also In re Jung, 637 F.3d 1356, 1362 (Fed. Cir. 2011) (prima facie case used during ex parte patent examination of obviousness “is merely a procedural device that enables an appropriate shift of the burden of production” from the PTO to the applicant, who must come forward with rebuttal evidence when the examiner makes out a prima facie case of obviousness). Lanham Act § 43(a)(3). In re Teledyne Indus. Inc., 696 F.2d 968 (Fed. Cir. 1982). Id. at 969–970 (citing Morton-Norwich). Id. Office Action, Oct. 17, 2007, U.S. App. Ser. No. 77254637. By the time of the appeal, the application claimed the trade dress in greater detail, calling out eight specific features. Becton Dickinson, 675 F.3d at 1370.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

370

Mark D. Janis

information pursuant to 37 C.F.R. §2.61(b).38 This request for information, which “should be issued for most product design marks,”39 calls for the applicant to supply the information needed to make a case of functionality under each of the MortonNorwich factors.40 The TMEP does specify that at the initial stage of examination, the trademark examining attorney should conduct “independent research to obtain evidentiary support for the refusal.”41 But the availability of the request for information regulation gives reason to doubt whether examiners will bother to conduct their own research in the typical case. Practitioner anecdotes seem to confirm this. According to one source, in making out a prima facie case, “[t]he examiner will typically make a minimal search of available information, and if there is reason to believe the proposed mark may be functional, for most product design applications, will issue an office action requesting further information from the applicant.”42 After the applicant supplies the information, a functionality refusal “frequently ensues.”43 What does all of this prove? I offer some concluding thoughts below.

30.4 implications One obvious consequence of the PTO’s ex parte practices regarding trade dress functionality is that, notwithstanding the Lanham Act’s assignment of the burden of proof, in no meaningful sense does the PTO actually bear the burden of proving functionality. The costs of producing evidence negating functionality are borne by the applicant. As the Federal Circuit’s cases imply, this is probably efficient. Indeed, it may be the only realistic way to allow for at least some limited scrutiny of functionality in the ex parte context.44 But it does diverge from the statute. Another consequence is that there is distance between the substantive functionality test of TrafFix and the test actually being applied in ex parte examination. How much difference, and whether that difference matters, depends upon one’s reading of TrafFix. If it is right to read TrafFix as having dramatically transformed the functionality doctrine into a mere physicalist accounting of whether features are useful or incidental, then there may be little distance between it and the PTO’s use 38 39 40 41

42

43 44

TMEP § 1202.02(a)(iv). Id. Id. TMEP § 1202.02(a)(iv). This independent research is supposed to include scrutiny of “applicants’ and competitors’ websites, industry practice and standards,” and, optionally, patent records. TMEP § 1202.02(a)(v). Margaret McHugh, Overcoming Functionality Refusals, 136 The Recorder (No. 7) (Oct. 22, 2012). Id. This reasoning might also be extended to other registrability conditions, presumably an argument that would appeal to those who favor a substantively more aggressive registration process.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

Examining Functionality

371

of a de facto functionality model as the basis for prima facie functionality. If, as I think, that view of TrafFix is faulty (because it is based on a substantial overread of TrafFix dicta, for example), then the gulf between the TrafFix standard and the PTO’s implementation of it may appear wide. That is, in ex parte examination, it may appear unlikely that the PTO is investigating whether trade dress features are “essential to the use or purpose” of the article or meaningfully interrogating whether those features “affect the cost of quality” of the article. Indeed, it could be argued that the TrafFix functionality analysis has become enslaved to the pragmatic limitations of the ex parte examination process, and has morphed into something akin to a blunt form of de facto functionality, cloaked in the rhetoric of a truncated version of the Morton-Norwich analysis that focuses on utility patent and advertising evidence. While this is an indictment of doctrinal coherence – and that matters because it suggests uncertainties that drive up the cost of administering functionality doctrine – does the PTO’s variant approach to functionality in fact change registrability outcomes?45 That question warrants empirical investigation, but, short of that, one can readily spin out competing plausible narratives. One possibility is that the incidence of false-negative errors (erroneously failing to register trade dress) might be considerably higher than previously appreciated. This supposes that the prima facie case of functionality based on mere de facto functionality frames the analysis, biasing the examiner towards refusal notwithstanding the applicant’s subsequent production of rebuttal evidence. Indeed, it presumes that the examiner’s “analysis” of the proffered rebuttal evidence is primarily an exercise in hunting for nuggets that would support the examiner’s ex ante conclusion of functionality.46 Another possibility is that the PTO’s approach presents the risk of many falsepositive errors. This narrative assumes that examiners are generally supine in regard to functionality analysis. More particularly, it assumes that even if examiners dutifully go through the motions of issuing functionality refusals (based on de facto functionality) and requesting information from the applicant, they then may capitulate in the face of the applicant’s submissions, especially to the extent that those submissions include expert affidavits to which the examiner cannot reasonably 45

46

While I refer to registrability outcomes, there may also be an effect on the mark owner’s strategic decision-making about whether to apply for registration, as opposed to asserting unregistered rights under Section 43(a). Scholars who favor a per se bar to registering product configuration trade dress might find the first narrative acceptable as a happy, if unintended, consequence of the current system. For relevant scholarship, see Caitlin Canahai and Mark P. McKenna, The Case Against Product Configuration Trade Dress, in Trademark Law and Theory: Reform of Trademark Law (Graeme B. Dinwoodie and Mark D. Janis, eds.) (Edward Elgar, forthcoming 2020); Glynn S. Lunney, Trademark Monopolies, 48 Emory L.J. 367 (1999) (arguing, inter alia, that Congress never intended to offer registration for product design trade dress on the Principal Register). If, in fact, the PTO is doing nothing more than applying a thinly veiled version of de facto functionality analysis that should toss out virtually all product configuration claims, one might wonder whether a per se subject matter exclusion would achieve the same end more transparently.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

372

Mark D. Janis

respond. Indeed, this might reflect the view that the ex parte examination process should strive merely to force the applicant to produce its proofs on the record, leaving close assessments of registrability conditions to the opposition and cancellations processes. Examiner sensibilities are not monolithic, and so it seems probable that at any given moment, the practices described in either or both of these narratives may be occurring. Regardless, both are unpalatable, at least to those who see TrafFix as standing for something other than an invitation to engage in an assessment of mere de facto functionality. Accordingly, what would happen if Congress did away with this mismatch by eliminating ex parte examination for functionality? For example, what if Congress deleted Lanham Act § 2(e)(5) and instead provided that functionality could be raised only in the context of an opposition or cancellation proceeding?47 In pursuing this line of thinking, I am taking a cue (methodologically, if not normatively) from others who have asked hard questions about the role that registration schemes should play in a trademark system.48 I doubt whether such a move would open the floodgates to product configuration trade dress registration (and so I doubt that we would see a spike in anti-competitive strike suits founded on registered trade dress). Functionality could still be raised as a basis for opposition or cancellation. Some claims of trade dress, perhaps many, would fail on the grounds of distinctiveness in any event.49 Thus, the most notable consequence of eliminating the Section 2(e)(5) bar is that it would amplify the pressure on examiners to assess whether product configuration trade dress complies with the requirement for secondary meaning. That leads to further questions worth exploring: to what extent do the constraints of the ex parte process impede examination of secondary meaning, and does this also lead to concerns about a mismatch between the substantive standard and the capacities of ex parte examination? Those are complex questions that go to the core of the trademark registration scheme, and warrant further investigation.

47

48

49

This is the current role of the dilution concept in registration: it is a ground of opposition or cancellation but not a bar to registration in ex parte examination. Lanham Act § 2(f ). See, e.g., Rebecca Tushnet, Registering Disagreement: Registration in Modern American Trademark Law, 130 Harv. L. Rev. 873, 932 (2017) (arguing for a “substantive” approach to registration, which, among other things, would recognize the role of doctrines such as functionality in defining the boundaries of trademark rights). The foregoing review of the operation of functionality in the ex parte context suggests that efforts to design a more “substantive” registration scheme should focus on inter partes proceedings with respect to enhancing the role of functionality. See, e.g., In re 3M Co., 2012 WL 1881484 (TTAB May 10, 2012) (non-precedential) (reversing the examiner’s functionality refusal but affirming on distinctiveness); In re Brayco Prods., Ltd., 2009 WL 4329104 (TTAB Nov. 16, 2009) (non-precedential) (same).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.040

31 Substantial Value and the Concept of Shapes Antoon A. Quaedvlieg1

This chapter discusses the substantial value exclusion for shape marks and its relation to the anti-monopoly ratio.

31.1 the substantial value exclusion According to the identical provisions of Article 4(1)(e) of the Trade Mark Directive2 and Article 7(1)(e) of the EUTM Regulation,3 a trade mark must be refused or declared invalid in the case of: (e) signs which consist exclusively of: (i) the shape, or another characteristic, which results from the nature of the goods themselves; (ii) the shape, or another characteristic, of goods which is necessary to obtain a technical result; (iii) the shape, or another characteristic, which gives substantial value to the goods. These grounds for refusal or invalidation, further referred to as ‘exclusions’, apply regardless of whether the shape, following the use which has been made of it, has

1

2

3

Professor of private law, especially industrial property and copyright law, at the Radboud University of Nijmegen, the Netherlands. Directive (EU) 2015/2436 of the European Parliament and of the Council of 16 December 2015 to approximate the laws of the Member States relating to trade marks (Recast), OJ 2015 L 336/1, entry into force 15 January 2016 (some Articles 15 January 2019). See, in old versions of the Directive: Articles 3(1)(e) of TM Directive 2008/95 of 22 October 2008 and of the first TM Directive 89/104 of 21 December 1988. Regulation (EU) 2017/1001 of the European Parliament and of the Council of 14 June 2017 on the European Union trade mark, OJ 2017 L 154/1.

Downloaded from https://www.cambridge.org/core. 373 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

374

Antoon A. Quaedvlieg

acquired a distinctive character.4 The three grounds for exclusion of shape marks find their origin in Article 1, second sentence, of the former Uniform Benelux Trade Mark Law (UBTML) of 1 January 1971,5 the text of which had been drafted in the 1950s and early 1960s.6 This chapter concentrates on the third category in the Directive, the ‘substantial value’ exclusion. The UBTML excluded shapes which ‘affect the essential value of the good’, in a provision that was probably inspired by the US doctrine of aesthetic functionality.7 The interpretation of this notion proved all but obvious. The Common Governmental Comment of the three Benelux countries that accompanied the UBTML8 did not reveal a proper rationale for the essential value exclusion. It only announced that it aimed at restricting ‘to some extent’ the possibility of concurrent protection by trade mark law and copyright or design law. The example was given of a crystal service: ‘the artistic form given to a crystal service will not be eligible for trademark protection in the crystal industry, as such a product does not find its essential value only in the material used, but also and maybe even in particular in the beauty of the shape’.9 By contrast, in the foodstuff industry the shape would not really affect the intrinsic value of the delicacy. Such a shape would therefore be able to serve as a mark. It was not until 1986–1992 that cases concerning the essential value exclusion reached the Benelux Court of Justice and revealed the exclusion’s dreadful complexity.10 How to establish whether a shape had essential value? The taste of the 4

5

6

7

8

9

10

Article 3(4), providing that a trade mark shall not be refused registration or declared invalid if it has acquired a distinctive character, does not apply to the exclusions in paragraph 1(e). Cf. ECJ, 20 September 2007, C-371/06 G-Star v. Benetton Group, 2007 ECR I-7709. As established by Annette Kur, Formalschutz dreidimensionaler Marken – Neue Aufgaben für die Markenabteilung des Deutschen Patentamts, in: DPA. 100 Jahre Marken-Amt. Festschrift, Wila Verlag 1994, pp. 175–197; Benoît Strowel, Benelux: A Guide to the Validity of Three-dimensional Trade Marks in Europe, [1995] 3 EIPR 154; Thomas Helbling, Shapes as Trade Marks? The Struggle to Register Three-dimensional Signs: A Comparative Study of United Kingdom and Swiss Law, IPQ 1997, 413–449, 420. The Benelux groups of AIPPI started to work on a draft as from 1948. As from 1959 the Benelux took over the initiative. The text of the Uniform Benelux Law was published in Holland in the Tractatenblad 1962, 58 and approved by the Dutch parliament on 25 April 1963, Staatsblad (Stb.) 1963, 221. Annette Kur, Too Pretty to Protect? Trade Mark Law and the Enigma of Aesthetic Functionality, Max Planck Institute for Intellectual Property and Competition Law Research Paper No. 11-16, September 2011, at pp. 3 ff, p. 8. Gemeenschappelijk Commentaar der Regeringen, Tractatenblad 1962, 58. As to the essential value exclusion, see pp. 53–54. Professor Ansgar Ohly kindly pointed out to me the resemblance with the US decision in Pagliero v. Wallace China Co. [1952] 198 F.2d 339. Regarding Pagliero, see also Justin Hughes, Non-Traditional Trademarks and the Dilemma of Aesthetic Functionality, in: Irene Calboli, Martin Senftleben, The Protection of Non-Traditional Trademarks: Critical Perspectives, Oxford University Press 1998, pp. 107–124, with further references. See, in particular, Benelux Court of Justice 23 December 1985, case A 83/4, Adidas Three Stripes, GRUR Int. 1987, 707–711, Anm. Eva Marina Bastian; Benelux Court of Justice 14 April

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

Substantial Value and the Concept of Shapes

375

public is no reliable criterion: it is subjective and it varies greatly according to the time,11 whereas an absolute exclusion should be stable. Looking at the commercial value that the shape lends to the goods is no real alternative: it depends no less on the taste of the public. The view of educated people and specialists offers no guarantee of being objective either.12 To complicate the matter, the mere popularity of a design as a trade mark increases its commercial value but does not count as essential value. In 1989, the First European Trade Mark Directive introduced three shape exclusions based on the Benelux model. The term ‘substantial’ replaced the term ‘essential’. Although criticisms of the substantial value exclusion grew ever louder,13 it survived the European trade mark reform package of 2015. In 2014, the Court of Justice of the European Union (CJEU) was called upon to elucidate the notion. In Hauck v. Stokke,14 it held that functional characteristics as well as aesthetic elements may give substantial value to the shape (32), whereas assessment criteria may be taken into account, such as: the nature of the category of goods concerned; the artistic value of the shape in question; its dissimilarity from other shapes in common use on the market concerned; a substantial price difference in relation to similar products; and the development of a promotion strategy which focuses on accentuating the aesthetic characteristics of the product in question (35).

11

12

13

14

1989 case A 87/8, Superconfex v. Burberry’s, (‘Burberry’s I’), GRUR Int. 1990, 863; IIC 1991, no. 4, 567–570; see also Kamperman Sanders in: EIPR 1995, 68; Frauke Henning-Bodewig and Heijo Ruijsenaars, Designschutz qua Markenrecht? Das ‘Burberrys’ Urteil des BeneluxGerichtshof im Rechtsvergleich, GRUR Int. 1990, 821–831, at p. 826; Benelux Court of Justice 16 December 1991, case A 90/4, Burberry’s v. Bossi (‘Burberry’s II’) [1992] 7 EIPR D-140; GRUR Int. 1992, 552 with article by Heijo E. Ruijsenaars, Neue Entwicklungen im Muster- und Markenrecht der Benelux-Länder: Hat das Benelux-Geschmacksmustergesetz überlebenschancen?, at p. 505. Max Planck Institute for Intellectual Property and Competition Law, Study of the Overall Functioning of the European Trade Mark System, 2011, p. 73 para. 2.33: the substantial value essentially depends on the public’s apprehension, which involves subjective evaluations which moreover change with time. Commented on in more detail in: A. A. Quaedvlieg, Shapes that Give Substantial Value to the Goods: Towards a Systematic and Homogeneous Protection of Designs in the EU’, in: MarieChristine Janssens and Geertrui van Overwalle (eds.), Harmonisation of European IP Law: From European Rules to Belgian Law and Practice – Contributions in Honour of Frank Gotzen, Bruylant/Larcier 2012, pp. 175–218. Cf. Annette Kur, Harmonization of the Trade Mark Laws in Europe – An Overview, IIC vol. 28, 1997, pp. 1–23, p. 17 note 106: ‘The exact meaning and scope of application of the clause relating to the addition of “substantial value” could hardly be more obscure. In my opinion, it was a mistake to include this provision in the European legislation, and the best solution would be to disregard it in practice – which would be all the more justified as the provision is probably not in accordance with Art. 15(2) of the TRIPs Agreement’; ibid., in: DPA: 100 Jahre MarkenAmt, Wila Verlag 1994, pp. 175–196, at pp. 192–193. See also C. E. F. M. Gielen, Substantial Value Rule: How it Came into Being and Why it Should be Abolished, in: Festschrift Willem Hoyng: Litigator, deLex 2013, pp. 178–193 and in EIPR 2014, pp. 164–169. CJEU 18 September 2014, C-205/13, Hauck v. Stokke, GRUR 2014, 1097, with note A. Kur, para. 18.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

376

Antoon A. Quaedvlieg

This large list of criteria, which is moreover an open one (‘such as’) did not bring great clarity in the matter. Does this mean that we are irrevocably stuck with an inexplicable provision, an enigmatic hieroglyph unthinkingly copied from the funeral chamber of the ancient Benelux trade mark law? Several authors have rightly advocated that the problem should be approached from the angle of the competitive interests and needs involved.15 My starting point will therefore be the anti-monopoly rule developed by the CJEU in trademark law (31.2). Copyright law provides a clue as to when aesthetic shapes may produce monopolistic effects (31.3). Those results may in turn help to establish what ‘substantial value’ in trade mark law means (31.4).

31.2 the anti-monopoly rationale Philips v. Remington According to the CJEU in Philips v. Remington,16 ‘the rationale of the grounds for refusal of registration laid down in Article 3(1)(e) of the Directive is to prevent trade mark protection from granting its proprietor a monopoly on technical solutions or functional characteristics of a product which a user is likely to seek in the products of competitors’. Competitors should remain free to offer for sale products incorporating such characteristics. This anti-monopoly rationale was confirmed in Hauck v. Stokke.17 The CJEU regularly emphasizes the ‘anti-perpetuation rationale’: trade mark law must not be used as an instrument to perpetuate intellectual property protection after the expiration of the term of copyright protection.18 However, the absolute exclusions apply as from the date of creation of the shape, not only after the expiration of other IP rights. They concern intrinsic boundaries of the trade mark domain. Not only after expiration, but also during the protection by patent or copyright, parallel protection of a shape as a trade mark may be considered undesirable for many reasons. It could result in a scope of protection that is undesirable. Trade mark protection must not protect shapes that copyright destines as a matter of 15

16

17

18

See, inter alia, Kur, supra note 7, at pp. 7, 21, referring to G. Dinwoodie, The Death of Ontology: A Teleological Approach to Trade Mark Law, Iowa Law Review vol. 84, 1999, pp. 611–685; Andreas Sattler, In bad shape? – Der Schutz dreidimensionaler Registermarken, GRUR 2018, 565–574. CJEU 18 June 2002, C-299/99 Philips v. Remington, Jur. 2002, I, pp. 5475 ff.; IIC 2002, 849; GRUR Int. 2002, 842; GRUR 2002, 804; NJ 2003, 481, note Spoor; cf. further Suthersanen, ECJ in Philips/Remington – Trademarks and Market Freedom, IPQ 2003, 257; Karl-Heinz Fezer, Entwicklungslinien und Prinzipien des Markenrechts in Europa, GRUR 2003, 457–469, at p. 468. Para. 78 is quoted; see also paras. 80 and 82. CJEU 18 September 2014, C-205/13, Hauck v. Stokke, GRUR 2014, 1097, with note A. Kur, paras. 18, 20. Cf. Philips, with regard to technical solutions (82), Lego (45) and Hauck (19; 31).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

Substantial Value and the Concept of Shapes

377

principle for the free domain. Goodwill protection for a shape mark should not fill the gaps caused by omissions in the patent claims. In Lego Juris v. Megabrands,19 the Court of Justice of the European Union (CJEU) explained why the availability of alternative shapes, with other dimensions or another design, does not make a difference for the exclusion. Similar shapes could still fall under the scope of protection with the result that ‘a significant number of alternative shapes might . . . become unusable for the proprietor’s competitors’ (56). If various purely functional shapes were registered at the same time, that might completely prevent other undertakings from manufacturing goods with the same function (57).20

31.3 concepts of shapes in copyright The exclusion of shapes or other characteristics which give substantial value includes aesthetic shapes. When can aesthetic shapes be said to produce monopolistic effects? Copyright law provides an indication. Only the concrete personal expression is eligible for copyright, whereas ideas, style and functional shapes are excluded.21 This universal limitation is explained in different ways, however the desire to avoid monopolies unquestionably forms a prominent rationale of it. The effects of protection of ideas and style, of concepts and functional subject matter, would stretch much further than the protection of individual works. Ideas and style are fit for use in multiple creations. An exclusive right would prevent potential productions from coming about; it would hamper progress and limit variety on the market. The exclusion of ideas and style does not, in principle, prevent the protection of any concrete shape of a product; it will merely limit the scope of protection. Nonetheless, in sporadic cases the scope of that limitation may be extremely large. These cases are of special interest. They occur when shapes which are new and unique strongly integrate individual expression as well as excluded elements of idea, style and/or function. Thus, the shape of the concrete product at once incorporates its ‘concept’.22 A shape concept creates a new basic form: a ‘type’ or ‘category’ of a useful object. That ‘type’ of 19 20

21

22

CJEU 14 September 2010, C-48/09 P, Lego Juris v. Megabrands, paras. 54–60. Cf. also the reasoning in CJEU 6 May 2003, C-104/01, Libertel. The company of that name wished to register a colour orange for telecommunications goods and services. The Court considered that the number of colours actually available is limited. A small number of trade mark registrations could exhaust the entire range of the colours available, which monopoly would be incompatible with a system of undistorted competition (54). See, inter alia, Schricker/Loewenheim, Urheberrecht Kommentar, 5th ed., C. H. Beck 2017, pp. 145–146 paras. 73–74; Andre´ Lucas, Agne`s Lucas-Schloetter, Carine Bernault, Traıˆte´ de la proprie´te´ litte´raire et artistique, 5th ed., LexisNexis 2017, pp. 81–86 paras. 60–64. The American doctrine gave it the name of the ‘idea-expression dichotomy’. One could think of the design of the Tripp Trapp chair or the fashion doll Barbie; the Tuliplu (an umbrella shaped and coloured as a tulip); the ‘dented cup’ (an avant garde design consisting of an average disposable coffee cup, dented as if thrown away, and in that form

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

378

Antoon A. Quaedvlieg

product can be realized in numerous different ways. If these are all covered by the scope of protection of such a shape, that scope will exceed that of a normal design and will approach that of a concept or an ‘idea’. Conversely, if one wants to prevent this anti-competitive effect by applying the exclusion of protection of ideas and style, the effect will be that copyright protection will be severely limited or even totally withheld for a concrete shape. This phenomenon has been the object of fierce debates in the Netherlands as from a remarkable decision of the Dutch Supreme Court23 in 1995. The cassation court endorsed a judgment of the Court of Appeal of Amsterdam in which it was found that the essential features of an iconic design for a billboard were, as elements of style, in the free domain. They were not eligible for copyright protection. Therefore, protection was refused to a design which, considering its aesthetic merits, might instead have justified a large scope of protection. The same dilemma came up in 2013 with regard to the iconic design of the Tripp Trapp children’s chair. The appearance of the Tripp Trapp chair is dominated by the two L-shaped standards which form its side posts and basis. The side posts have horizontal slits at the inside, in which the seat and feet board can horizontally be slid in, resulting in the ‘floating appearance’ of the Tripp Trapp. The slits also make the chair continuously adjustable to the size of the growing child, in a way taking into account certain ergonomic goals. Beauty and function find themselves fully merged in one encompassing concept. Dutch courts used to grant full copyright in the Tripp Trapp, until in 2007 the District Court of the Hague24 analysed it as a product essentially composed of free elements of either technical utility or style. Only the choice for the L-shaped standards was considered original.25 This limited the scope of protection. The same approach was followed by the Court of Appeal of Amsterdam in 2011 in another case that also concerned the Tripp Trapp chair.26

23

24

25

26

offered as chinaware); the Dyson Dual Cyclone bagless vacuum cleaner, at the time of its introduction. HR 29 December 1995, Decaux v. Mediamax, NJ 1996, 546, note D. W. F. Verkade; Informatierecht, AMI 1996, pp. 199–207, note Quaedvlieg pp. 195–198; BIE 1997, no. 66, p. 343. See also A. A. Quaedvlieg, ‘Style is Free’: Designs Beware (Opinion), [2001] EIPR 445–453. District Court of the Hague 7 February 2007, Stokke v. Fikszo, AMI 2007, no. 13, pp. 99–103, note Dirk Visser; IER 2007, no. 74, p. 265, note F. W. Grosheide. The two posts at the sides between which the back is inserted, the horizontal slits, the horizontal position of the seat and feetboard, a certain thickness of the material, and the choice of beech wood as combining good mechanical qualities with a low price, were considered by the court as being inspired by a technical or functional motive. From the ‘sober Scandinavian style’ resulted the austere lines, the white finish, the sober finish by the use of metal parts, the absence of coloured elements, and the fact that the ‘visualised technical approach’ had only modestly been furnished with decorative aspects. Court of Appeal Amsterdam 15 March 2011, ECLI:NL:GHAMS:2011:BQ3808, Stokke/H3. The appeal in cassation was rejected by the Dutch Supreme Court on 22 February 2013, NJ 2013, no. 503, note P. B. Hugenholtz.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

Substantial Value and the Concept of Shapes

379

The Court of Appeal of The Hague, however, boggled at the perspective of reducing protection for a design considered highly meritorious.27 Although it limited copyright protection to only two features of the design,28 it granted those features a large scope of protection, considering that the Tripp Trapp chair was a ‘revolutionary’ design with a high degree of originality. On this ground, it sufficed that only one of the two protected features would need to have been copied – the Lshape – to find infringement of the whole chair. This L-shape, however, mainly or entirely represented a combination of free elements of idea, style and function. Weak copyright elements29 thus served to found strong copyright protection.30 Copyright therefore shows that the protection of aesthetic shapes can have anticompetitive effects where the shape of a concrete product incorporates the concept of the design of the product.

31.4 design concepts in trade mark law We turn to trade mark law once again, in order to assess whether these developments in copyright could help to clarify the substantial value exclusion. Indeed, the protection of shape concepts by trade mark law could have the same anticompetitive effects. Firstly, unique shape concepts tend to be highly distinctive and would correspondingly enjoy a vast scope of protection. Copying the shape concept will most likely create confusion (including the likelihood of association) between the sign and the trade mark with the public. Where it concerns signs with a reputation, it will suffice that the public establishes a mere link between the sign and the mark to acquire the protection granted by Article 10(2)(c) of Directive 2015/2436.31 It will not be difficult to claim that the use of the shape takes unfair advantage of its distinctive character and repute as a trade mark. Secondly, shape concepts are scarce. It is difficult for competitors to find substitutes. The number of realistic alternatives is small and a company could monopolize the market by registering several shape marks.

27

28

29

30

31

Court of Appeal of The Hague 30 May 2009, Stokke/Fikszo c.s., IER 2009, no. 77, pp. 312–318, note F. W. Grosheide; BIE 2010, no. 25, p. 152, note A. A. Quaedvlieg; AMI 2009, pp. 211–212. Those elements being the L-shaped sideposts as well as, to be distinguished from these, the two slanting sideposts in which the back, seat and feetboard of the chair were incorporated. One wonders whether this really constitutes an independent element. Weak, for it would be hard to deny that the element of the L-shape had a strongly functional character. The Dutch Supreme Court also upheld this finding, in my view wrongly treating the scope of the technical exclusion as a question relating to the facts. Supreme Court of the Netherlands 12 April 2013, Stokke/Fikszo, NJ 2013, no. 502, note Hugenholtz under 503; ECLI:NL:HR:2013: BY 1532, para. 6.3.2. CJEU 23 October 2003, C-408/01, Adidas v. Fitnessworld, para. 29.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

380

Antoon A. Quaedvlieg

Thirdly, on top of these arguments, the potentially eternal duration of the trade mark protection makes its competitive footprint even bigger than in copyright. Thus, the protection of such shape ‘concepts’ would be permanently and intrinsically anti-competitive. This is what the permanent absolute exclusions in Article 4 (1)(e) aim to prevent. Conversely, some elements mentioned in the open list given in the 2014 Hauck decision of the CJEU referred to in 31(1) might be less appropriate criteria for the exclusion of protection, such as the artistic value of the shape or a promotion strategy which focuses on accentuating the aesthetic characteristics of the product. Nothing could be as contradictory and confusing as to make the rigid substantial value exclusion depend on a value judgement that itself is variable and inconstant.32 The free elements of idea, style and function identified in the context of copyright law are not relevant for trade mark law as long as they do not concern all essential characteristics of the shape.33 Such cases will be rare. That, however, does not make the exclusion redundant. There still is a semantic question. The substantial value of the shape has always been understood as a reference to the aesthetic or artistic qualities of its design. Can it be reinterpreted as referring to shapes which give substantial value from a competitive point of view? The wording as such of Article 4(1)(e)(iii) could not form an obstacle. It uses neither the term ‘aesthetic’ nor ‘artistic’. From a teleological point of view, a change of perspective might even be welcomed. The traditional ‘aesthetic’ approach in essence conveys the message that true art must not serve commercial purposes. The drafters of the Benelux law may have felt this to be so seventy years ago, but it is an approach that will probably find little support in contemporary thinking. By contrast, the prevention of undue monopolistic effects of trade mark protection will. By holding, in Hauck v. Stokke, that functional as well as aesthetic elements may give substantial value (32), the Court in essence had already abandoned the aesthetic approach.

31.5 conclusion What does the substantial value exclusion in trade mark law aim at? As a starting point, we took the anti-monopoly approach the CJEU considers as the common basis of all three ‘absolute’ exclusions of shapes. As far as aesthetic shapes are concerned, it was argued that competitive interests of third parties are a prominent reason for the exclusion of ideas and style from copyright protection. In exceptional cases, this exclusion may have consequences 32

33

Annette Kur, Funtionswandel von Schutzrechten: Ursachen und Konsequenzen der inhaltlichen Annäherung und Überlagerung von Schutzrechtstypen, in: Gerhard Schricker/Thomas Dreier/Annette Kur (eds.), Geistiges Eigentum im Dienst der Innovation, Nomos 2001, pp. 23–39, at p. 37. Lego 52, 72; Hauck 21–23 and 36.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

Substantial Value and the Concept of Shapes

381

for the protection of concrete shapes as well. That may happen where ideas and style, apart from useful elements, define the concept itself of a shape; the shape will represent a new ‘type’ of a product. In that case, granting an exclusive right in the shape may amount to protecting the concept of a product, or approach that effect. This would mean a restriction on competition exceeding the scope of protection copyright is meant to provide. Aligning the enigmatic substantial value exclusion in trade mark law with these rare cases in copyright would clarify its purpose and enhance the consistency and the transparency of the system at large.34 Annette Kur, to whom this contribution is devoted, has enriched countless aspects of intellectual property with her expert view. It is time to celebrate a colleague of great stature. Hopefully, we will see many more of Annette’s invaluable contributions!

34

Quaedvlieg, supra note 12, at p. 215.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.041

32 Copyright and Patents on Software: The UPC’s Answer to an Old Problem of Intellectual Property Overlaps Axel Metzger1

32.1 introduction Whereas most creations of the mind, such as works of literature, music and art, technical inventions and typical marks capable of distinguishing goods or services, are subject matters eligible for only one particular intellectual property right, some artefacts pose difficulties in terms of characterisation. The three-dimensional shape of goods may be protected by design, copyright or trademark law. Innovative plants can meet the requirements of either plant variety protection or that of biotechnological inventions. Computer programs may be protected as literary works under copyright law or as technical inventions under patent law. Annette Kur has worked intensively on those intellectual property overlaps, or cumulation, in the last twenty years since her seminal work published in a conference volume2 that gathered the presentations given on the occasion of the opening of the new building of the Max Planck Institute in Munich in the year 2000 – the first conference which the author of these lines ever attended at the Institute and which left a deep impression on him. In this early paper Annette Kur had already developed her general theory of intellectual property overlaps which she later explored in more detail for the field of trademark and design law.3 According to her theory, cumulation of intellectual 1

2

3

Prof. Dr. iur., LL. M. (Harvard), Professor of Civil and Intellectual Property Law, HumboldtUniversity, Berlin. Annette Kur, Funktionswandel von Schutzrechten, in: Gerhard Schricker/Thomas Dreier/ Annette Kur (eds.), Geistiges Eigentum im Dienst der Innovation (Nomos, 2001), 23–50. Annette Kur, Probleme im Überschneidungsbereich von Marken und Designs, in: MarkenR – Zeitschrift fu¨r deutsches, europa¨isches und internationales Kennzeichenrecht 2017, 185–194; id., Cumulation of Rights with Regard to Three-dimensional Shapes – Two Exemplary Case Studies, in: Severine Dusollier/Alexandre Cruquenaire (eds.), Le Cumul des Droits Intellectuels (Larcier, 2009), 155–175; id., Cumulation of IP Rights Pertaining to Product Shapes – An “Illegitimate Offspring” of IP Law?, in: Gustavo Ghidini/Luis Mariano Genovesi (eds.), Intellectual Property and Market Power – ATRIP Papers 2006–2007 (Eudeba, 2008), 613 – 632.

Downloaded from https://www.cambridge.org/core. 382 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

IP Overlaps – Copyright and Patents on Software

383

property rights does not pose a problem per se as long as (1) the requirements of protection under the different legal regimes are formulated and applied so as to fully justify the protection granted, and (2) each one of the rights applied is adequately balanced in itself, in particular as regards the relationship between the requirements, the general scope of protection and the limitations. By contrast, if in the application of the different regimes, their protection requirements become more and more blurred whereas the differences are maintained in respect of the scope of protection and the limitations, the balance of the rights in question may be affected negatively, a phenomenon labelled by Annette Kur as “asymmetric convergence”,4 an expression that has been picked up by many later works on the subject.5 This chapter focuses on the overlap of copyright and patent protection in the field of computer programs. The issue is not a new one. It was addressed in Annette Kur’s first paper on the cumulation problem. What is new is the European legislator’s approach to the problem in Article 27 lit. k) of the United Patent Court (UPC) Agreement, which aligns, at least in part, the limitations and exceptions of copyright and patent law with regard to computer programs.

32.2 overlap of copyright and patent protection for software At first glance, one could come to the conclusion that copyright and patent protection of computer programs relate to different subject matters and as such do not represent an example of a genuine cumulation of intellectual property protection.6 Under copyright law, computer programs are protected as literary works.7 The main point of reference for the protection is the individual intellectual creation of the programmer as it is laid down in the source code and the compiled object code of the program. Ideas and principles which underlie the computer program, including those which underlie its interfaces, are not protected by copyright law. By contrast, patent protection is granted for the technical solution of a given problem in the form of a condensed, abstract description of the invention in the patent claims. Those claims describe the functionality of the “computer-implemented invention” but do not comprise a concrete implementation in source or object code. However, such a delimitation of code versus functionality does not capture the actual dimension of 4 5

6

7

Kur, Funktionswandel von Schutzrechten, supra fn. 2, 23, 34. See e.g. Estelle Derclaye/Matthias Leistner, Intellectual Property Overlaps: A European Perspective (Hart, 2011), 93; Alexander Fiedler, Der Computerprogrammschutz und die Schutzrechtskumulation von Urheber- und Patentrecht (Nomos, 2013), 43; Artur Geier, Schutzkumulationen (Mohr Siebeck, 2015), 4. See e.g. Malte Gru¨tzmacher, Commentary on § 69g, in: Artur Wandtke/Manfred Bullinger, Praxiskommentar zum Urheberrecht, 5th ed. (C. H.Beck, 2019), N 1. This equation is codified on an international level in Article 10 para. 1 TRIPS. See also Article 1 para. 1 Directive 2009/24/EC on the legal protection of computer programs.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

384

Axel Metzger

the overlap of the two protection regimes. On the one hand, copyright protection is not restricted to the source code of a computer program. As made clear in Article 1 para. 1 of the Computer Program Directive 2009/24/EC, “preparatory design material” is also eligible for copyright protection. Obviously there seems to be a form of a program which is not yet put in the form of a concrete source code but which is still more concrete than a mere idea or principle underlying the program. The copyright protection of this “preparatory design material” may be infringed by a program which is not based on the identical source code but copies the structure and organisation of the implemented algorithm(s).8 On the other hand, patents on computer-implemented inventions are not restricted to the protection of the functionalities of a program but are broader in scope. In fact, a typical patent claim covers all different concrete programs with different – copyright protected – source codes which implement the protected functionalities.9 Therefore, it is misleading to present software patents as if these patents would protect something which is not protected by copyright. In fact, they protect the same subject matter but on a different level of abstraction.10 Moreover, concrete implementations may be used as examples in the description to provide a sufficient disclosure of the invention in the sense of Article 83 of the European Patent Convention.11 The European Patent Office may even permit inclusion of extracts from the source code of a concrete embodiment of the invention in the patent description.12 In such a case, it becomes evident that a patent on a computer-implemented invention may cover copyrightprotected programs as embodiments falling within the ambit of the claim.

32.3 lack of coherence of the two regimes The protected subject matter of copyright and patents with regard to innovative computer programs may thus overlap to a significant extent. Nevertheless, the attributes of copyright and patent protection in the field have not been adapted to this cumulative scheme of protection. As a result, legislative decisions taken in one protection regime are undermined by the application of the overlapping second regime. Here it may suffice to mention some examples for this “asymmetric convergence” – besides the issue of limitations and exceptions, which will be explored below with more details. The most obvious example is the term of protection, which 8

9 10

11 12

See with more details Fiedler, supra fn. 5, 34–36. See also Grützmacher, supra fn. 6, Commentary on § 69c, N 9–13. See e.g. EP0717354: Data transfer with expanded clipboard formats. Cf. Fiedler, supra fn. 5, 117–118; Geier, supra fn. 5, 37; Valerie Schweppe, Mehrfache Schutzrechtsverletzungen in Fa¨llen von Doppelschutz, Kumulation und verbundenen Schutzrechten (Carl Heymans, 2017), 100–101. See Rule 42 lit. e) Implementing Regulation. See EPO Guidelines for Examination, Part F, N 4.12: “Short excerpts from programs written in commonly used programming languages can be accepted if they serve to illustrate an embodiment of the invention.”

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

IP Overlaps – Copyright and Patents on Software

385

runs in European patent law for twenty years counted from the filing date,13 whereas copyright protection is granted for seventy years after the death of the author, namely the programmer.14 If one understands the twenty-year term of patent law as a balance between the interests of the patent holder, who needs exclusivity for his return of investment, and the public, which accepts the granting of such exclusivity for a certain period of time to incentivise technical innovation, it is questionable whether the protection of concrete implementations of the invention should endure for many more decades. As a second example of lacking coherence, ownership of copyrights and patents with regard to computer programs may be highlighted. According to Article 2 of the Computer Programs Directive 2009/24/EC, the natural person creating a program shall be considered as the author and, at least in principle, also as the rightholder of the copyright in the program. However, as made clear by Article 2 para. 3, for programs created by “an employee in the execution of his duties or following the instructions given by his employer, the employer exclusively shall be entitled to exercise all economic rights in the program”. Case law however has shown that the rule is less straightforward as it may appear. Whenever an employee is not engaged as a programmer and does not follow the employer instructions, the economic rights remain with the programmer.15 In patent law, the employer’s rights are more comprehensive. According to the German Employee Inventions Act, inventions made by employees are also covered by the employer’s right to exploit the invention, if the invention originates from the employee’s experience at the company, irrespective of whether the employee was engaged as a programmer or followed the employer’s instructions. Similar provisions are to be found in the French Intellectual Property Code, Article L. 611-7 para. 2.16 UK patent law follows a more restrictive approach – see Section 39 UK Patents Act. This lack of coherence of copyright and patent law may result in a divergent and even conflicting allocation of rights.17 Further examples of inconsistencies may be picked from the enforcement stage of intellectual property rights.18 How should a court deal with 13 14

15

16

17

18

See Article 63 European Patent Convention. See Article 1 para. 1 Directive 2006/116/EC on the term of protection of copyright and certain related rights. See Regional Court Munich, Computer und Recht (CR) 1997, 351; for the situation in France see Derclaye/Leistner, supra fn. 5, 147. See Derclaye/Leistner, supra fn. 5, 146 et seq.; Christian Le Stang, Interferences entre droit d’auteur et droit des brevets quant à la protection du logiciel: approche Française, in: Surinder Kaur Verma/Raman Mittal (eds.), Intellectual Property Rights: A Global Vision, ATRIP Papers 2002–03 (Indian Law Institute, 2004), 162, 164 et seq. See Christophe Caron, La coexistence du droit d’auteur et du droit des brevets sur un même logiciel, in: J. L. Goutal (ed.), Brevetabilite´ des logiciels, travaux du CUERPI: droit des technologies avance´es (Hermes Science Publications, 2002), 184, 199; Fiedler, supra fn. 5, 229 et seq. See on the following Mary-Rose McGuire, Kumulation und Doppelschutz: Ursachen und Folgen des Schutzes einer Leistung durch mehrere Schutzrechte, in: Gewerblicher Rechtsschutz und Urheberrecht (GRUR) 2011, 767; Schweppe, supra fn. 10, 115.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

386

Axel Metzger

a claim-based patent infringement if another lawsuit is already pending which is based on copyright infringement? Is this a case of lis pendens? Is it a case of abuse of rights if the rightholder is splitting up the case for strategic reasons? Does the doctrine of res iudicata prevent the rightholder from bringing its action if an infringement case involving certain, concrete products has already been decided on the basis of the other right? How should a court calculate the damages if it decides to hear the second case? The practical problems caused by an overlap of patent and copyright protection for computer programs are far from being resolved.

32.4 limitations and exceptions 32.4.1 The Way to Article 27 lit. k) UPC Agreement The limitations and exceptions of copyright and patent law follow different patterns, especially with regard to computer programs. Articles 5 and 6 of the Computer Programs Directive 2009/24/EC provide a specific list of limitations for copyrightprotected software. At present, exceptions and limitations in the European national patent Acts do not comprise comparable provisions. The UPC Agreement is now following a different approach in Article 27 lit. k), according to which the rights conferred by a patent shall not extend to “the acts and the use of the obtained information as allowed under Articles 5 and 6 of Directive 2009/24/EC, in particular, by its provisions on decompilation and interoperability”. The Commission’s Proposal for a Council Regulation on the Community Patent of August 1, 200019 did not include the provision. It was only included later during the negotiations with the Council and the European Parliament and appeared, apparently for the first time, in a Council document of April 200320 without any explanation. A bit more background information may be found in the legislative history of the later failed “Proposal for a directive on the patentability of computerimplemented inventions”, which was negotiated at the same time. Article 6 of the Commission’s proposal for that Directive contained an unspecified reference to “acts permitted under Directive 91/250/EEC on the legal protection of computer programs by copyright, in particular provisions thereof relating to decompilation and interoperability”, which was later clarified by the European Parliament (now as Article 6a) into a wording which is close to the UPC Agreement’s wording: “The rights conferred by patents granted for inventions within the scope of this Directive shall be without prejudice to acts permitted by way of exception under Directive 91/ 250/EEC on the legal protection of computer programs by copyright, in particular the acts specified and described in the closed list set out in Articles 5(2) and (3) and 19 20

OJ C 337 E, 28.11.2000, 278. Proposal for a Council Regulation on the Community patent – Text revised by the Presidency, Document 8539/03, 16.4.2003, Article 9 lit. i).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

IP Overlaps – Copyright and Patents on Software

387

6 of Directive 91/250/EEC”.21 It appears likely that the provision in Article 27 lit. k) UPC Agreement was inspired by the more specific instrument on computer programs.22 Interestingly though, the reference to Articles 5 and 6 is drafted now in a more generous way, comprising all limitations and exceptions from the Computer Programs Directive. Still, the text of the UPC Agreement, as do the earlier texts, specifically emphasizes the importance of the provisions on decompilation and interoperability without being restricted to those provisions.23 32.4.2 Acts Covered by Articles 5 and 6 Computer Programs Directive and Article 27 lit. k) UPC Agreement Acts covered by Articles 5 and 6 Computer Programs Directive should therefore no longer cause “asymmetric convergence” issues if the UPC Agreement is applied. They are privileged symmetrically under both regimes by way of a statutory reference in Article 27 lit. k) UPC Agreement to the requirements and effects of Articles 5 and 6 Computer Programs Directive.24 Since Articles 5 and 6 Computer Programs Directive do not leave any room for manoeuvre for EU Member States with regard to their implementation, the difficult question of whether such reference leads to the incidental application of the provisions of the Directive or of a national implementation can be neglected here.25 Under Article 5 para. 1, reproduction and translation of a program is permitted without a license contract, if it is “necessary for the use of the computer program by the lawful acquirer in accordance with its intended purpose, including for error correction”. Article 5 para. 2 allows “the making of a back-up copy by a person having a right to use the computer program”. Article 5 para. 2 is mandatory (“may not be prevented by contract”). Article 5 para. 3 entitles the lawful user of a program “to observe, study or test the functioning of the program in order to determine the 21

22 23

24

25

Proposal for a Directive of the European Parliament and of the Council on the patentability of computer-implemented inventions, COM (2002) 92, OJ C 151, 25.6.2002, 129, and amendments by the European Parliament, 18.6.2003, C5–0082/2002 – 2002/0047(COD), Article 6. See Fiedler, supra fn. 5, 211. See also the Justification given by the European Parliament’s report, supra fn. 21, for Article 6a: “Unlimited patent protection for software could make it illegal under patent law to engage in reverse engineering practices employed by software developers to achieve interoperability as currently permitted under the exceptions in the Software Copyright Directive. Therefore future EU-legislation related to software patents must include an explicit exception to patent rights in order to ensure that developers of software can continue to engage in the same acts to achieve interoperability under patent law as they are allowed to today within the limits of copyright law. The Council’s common approach of 8 November 2002 is supported and clarified by a reference to Articles 5 and 6 of Directive 91/250/EEC.” A reference of this kind is described as “Rechtsgrundverweisung” in German terminology – see Rainer Wo¨rlen/Sabrina Leinhas, Rechtsfolgen- und Rechtsgrundverweisungen im BGB (Juristische Arbeitsblätter, 2006), 22. For a direct application of the Directive see Maximilian Haedicke, Rechtsfindung, Rechtsfortbildung und Rechtskontrolle im Einheitlichen Patentsystem, GRUR Internationaler Teil 2013, 609, 612.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

388

Axel Metzger

ideas and principles which underlie any element of the program if he does so while performing any of the acts of loading, displaying, running, transmitting or storing the program which he is entitled to do”. The Court of Justice of the European Union (CJEU) has decided that the rightholder may not exclude those rights by restrictive license terms.26 Given the clear reference in Article 27 lit. k) UPC Agreement, courts can avoid entering into the analysis of the question whether mere use of a program, back-up copies or the studying of a program can also be justified under general limitations and exceptions of patent law, such as the doctrine of exhaustion, private use or the research exemption. According to Article 6 Computer Programs Directive, the authorization of the rightholder shall not be required where reproduction of the code and translation of its form are indispensable to obtain the information necessary to achieve the interoperability of an independently created computer program with other programs, provided that decompilation is carried out by a licensee or other legal user and interoperability information has not been made available before by the rightholder. Information obtained by decompilation may be used commercially and given to others as long as this is necessary to achieve interoperability. The provision has not been controversial in court cases,27 but has given rise to disparate views in academic writings.28 Some have argued that the provision is too narrow; decompilation should be allowed in general and not just for interoperability purposes,29 a position that receives further support by Article 3 para. 1 lit. b) of the Trade Secret Directive 2016/943/EU.30 Others have argued that the general interoperability rule of Article 6 may be too unspecific, since “its optimal degree, and the optimal path will differ depending on the technological context and the market environment”.31 Still, the rule has been accepted in legal practice as the result of a hard-fought battle between different interest groups.32 The last half-sentence of Article 27 lit. k) UPC Agreement (“in particular, by its provisions on decompilation and interoperability”) 26 27

28

29

30

31

32

See CJEU, C-406/10, 2.5.2012 – SAS Institute/World Programming. In CJEU, C-406/10, ibid., the defendant achieved interoperability apparently by mere observation of the functioning of the program. For a comparative overview see John Abbott, Reverse Engineering of Software: Copyright and Interoperability, 14 (2003) JlLawInfoSci 7; Florian Schweyer, Die rechtliche Bewertung des Reverse Engineering in Deutschland und den USA (Mohr Siebeck, 2012), passim; Pamela Samuelson, The Past, Present and Future of Software Copyright Interoperability Rules in the European Union and United States, EIPR 2010, 229. See Andreas Wiebe, Interoperabilität von Software: Art. 6 der Computerprogramm-Richtlinie aus heutiger Sicht, 2 (2011) JIPITEC 89. Cf. Thomas Hoeren, The EU Directive on the Protection of Trade Secrets and its Relation to Current Provisions in Germany, 9 (2018) JIPITEC 138 para. 17; Ansgar Ohly, Das neue Geschäftsgeheimnisgesetz im Überblick, GRUR 2019, 441, 447–48. See Wolfgang Kerber/Heike Schweitzer, Interoperability in the Digital Economy, 8 (2017) JIPITEC 39, 58. See e.g. Thomas Dreier, Commentary on § 69c, in: Thomas Dreier/Gernot Schulze, Urheberrechtsgesetz, 6th ed. (C. H. Beck, 2018), N 2; Malte Grützmacher, supra fn. 6, N 2.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

IP Overlaps – Copyright and Patents on Software

389

demonstrates the legislator’s clear intention to permit decompilation for the purpose of achieving interoperability. Without Article 27 lit. k), one could still discuss whether the decompilation as such is covered by the general research exemption in patent law, depending on how this exemption is interpreted.33 However, a commercial use of patent-protected interoperability information is certainly not covered by the research exemption but only by Article 27 lit. k) UPC Agreement and Article 6 Computer Programs Directive.

32.4.3 Private and Experimental Use Only for Patents? Article 27 lit. k) UPC Agreement tries to solve deficiencies caused by incoherent limitations and exceptions of copyright and patent law, but it stops half way. The regular limitations and exceptions of patent law are still not reflected in the Computer Programs Directive and its national implementations. If both rights were applied cumulatively at the choice of the rightholder, copyright protection could be used to prevent users from profiting from the patent law limitations and exceptions. This raises the question whether one can achieve an interpretation of the two protection regimes modelled on the basis of Article 27 lit. k) UPC Agreement which safeguards the policy choices behind patent law limitations and exceptions. The first patent law limitation that deserves closer scrutiny in this context is the private use exception. Article 27 lit. a) UPC Agreement and national patent Acts34 provide that the rights conferred to the patent owner shall not extend to “acts done privately and for non-commercial purposes”. If taken literally, Article 27 lit. a) would justify, for example, the reproduction and use of a computer program on multiple privately used hardware devices, the reproduction and sharing of a computer program with friends and family as long as this remains in the private sphere, or the modification and private sharing of computer programs, including “mods” of video games (presumed the technology used for the video game is protected by patent law). If the private use exception of patent law was interpreted as described and if it was applied both to patent and copyright protection of computer programs, the software industry’s worst nightmare would come true. Indeed, it seems hardly compatible with Article 30 TRIPS to justify identical reproductions of the program for private purposes. Such a freedom to copy would conflict with the normal exploitation of the program. On the other hand, it seems worth considering whether the act of modifying the program for private purposes should be allowed on the basis of the private use exception in case the rightholder decides to register a patent for the program.

33 34

See the following section. See Section 11 N 1 German Patent Act, Section 60 para. 5 lit. a) UK Patents Act 1977, Article L613–5 lit. a) French Intellectual Property Code.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

390

Axel Metzger

An even more difficult case is posed by the research exceptions in Article 27 lit. b) UPC Agreement and national patent Acts,35 according to which “acts done for experimental purposes relating to the subject-matter of the patented invention” are privileged. It is not yet clear how the UPC will interpret the research exemption given that national courts have developed quite different concepts, especially with regard to medical trials.36 However, it should be clear that any examination of the invention37 with the purpose of developing new technology or improving the state of the art constitutes “research” in the sense of Article 27 lit. b) UPC Agreement. Applying the research exemption to the field of software development would be of great interest for downstream developers and for the innovation process of the whole industry.38 Developers could observe, examine and reverse engineer computer programs of their competitors and develop new programs based on the findings from that research. However, if patent and copyright protection are applied cumulatively at the choice of the rightholder, he or she could avoid any research beyond the tight limits of Articles 5 and 6 Computer Programs Directive. It is highly questionable whether such a result is appropriate. Considering that the developer (or the company who owns the invention) has discretion to apply for a patent or not, it is only consequent to grant the patent with all the rights and limits provided for by patent law, irrespective of whether the implementation of the invention is protected by copyright law. Moreover, applying the research exemption to patent-protected programs would be in full conformity with Article 30 TRIPS. It would neither conflict with the normal exploitation of the program nor would it unreasonably prejudice the legitimate interests of the rightholder. This may sound like a daring theory to some. But Article 27 lit. k) UPC Agreement proves that synchronisation of limitations and exceptions in the field of software inventions is within the realm of possibility for legislators. Why should it not be for those who develop the law by interpretation and gap-filling?39 Article 27 lit. k) UPC Agreement has opened the door for the development of a coherent system of protection for innovative software. This chapter argues that lawmakers, courts and academics should not stop here but continue to seek for a more symmetric convergence of patent and copyright protection in this area. 35

36

37

38

39

See Section 11 N 2 German Patent Act, Section 60 para. 5 lit. b) UK Patents Act 1977, Article L613–5 lit. b) French Intellectual Property Code. See Court of Appeal (England and Wales), Monsanto Co v. Stauffer Chemical Co (1985) RPC 515, and German Federal Supreme Court, NJW 1997, 3092 – Klinische Versuche II. The issue of medical trials is solved in Article 27 lit. d) UPC Agreement. Research with the help of the invention is not necessarily covered – see Christine Godt, Eigentum an Information (Mohr Siebeck, 2007), 183 et seq.; Esther Pfaff, “Bolar” Exemptions – A Threat to the Research Tool Industry in the U.S. and the EU?, International Review of Intellectual Property and Competition Law 2007, 258. See Pamela Samuelson/Suzanne Scotchmer, The Law and Economics of Reverse Engineering, 111 (2001) Yale L.J. 1575. In the same direction Caron, supra fn. 17, 201.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

IP Overlaps – Copyright and Patents on Software

391

A possible next step could be to recognise that some of the core limitations and exceptions of patent law should remain applicable irrespective of the parallel copyright protection of concrete source codes, in particular the research exception in Article 27 lit. b) UPC Agreement and national patent Acts. Such an interpretation of patent and copyright law is compatible with Article 30 TRIPS and could help to accelerate innovation in the software industry.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.042

33 Chopping off Hydra’s Heads: Spare Parts in EU Design and Trade Mark Law Anna Tischner1

33.1 introduction Establishing a legal framework for the protection of interests with respect to spare parts has been a hot topic in EU intellectual property law for years, marked on the one hand by the economic policy of the European Union2, aiming at market liberalization and free competition, and on the other by the economic interests of individual Member States, not always concerned with the liberalization of spare parts markets3. The problem was solved in Europe partially at the stage of creating the principles of EU design protection in the form of a repair clause introduced only into the unitary design regime in Article 110(1) CDR4 and the compromise obligation set forth in Article 14 DD5 to maintain the status quo or alternatively to liberalize the national law in that regard (freeze plus clause).

1

2

3

4

5

(dr hab.) Professor at the Jagiellonian University in Krakow, Poland. This paper was completed as a part of research undertaken during the author’s stay at the Max Planck Institute for Innovation and Competition (MPI) in Munich in 2018. The author expresses gratitude to Annette Kur and Ansgar Ohly for enlightening discussions, as well as to Irene Calboli, Dev S. Gangjee and Martin Senftleben – members of a panel at the IP Researchers Europe Conference in Geneva on 29 June 2018 – for their comments. Which harmonizes with the policy of those Member States which are interested in liberalization of the spare parts aftermarket. Especially the countries with strong automotive industries. The introduction of a repair clause into § 40a of the German Design Act (as adopted on 10 September 2020) may cause a domino effect in Europe in the countries which have no repair clause, i.e. Austria, France, Lithuania, Portugal, Slovenia, Sweden, Czech Republic and Denmark. Council Regulation (EC) No 6/2002 of 12 December 2001 on Community designs, OJ 2002 L 3/1 (hereinafter CDR). Directive 98/71/EC of the European Parliament and of the Council of 13 October 1998 on the legal protection of designs, OJ 1998 L 289/28 (hereinafter DD).

Downloaded from https://www.cambridge.org/core. 392 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

Spare Parts in EU Design and Trade Mark Law

393

While the EU lawmakers’ attempt to resolve the problem of spare parts uniformly at the EU level resembles a Sisyphean task,6 the EU Court of Justice judgments in cases pertaining to this problem look like one of the labours of Heracles, namely cutting off the Hydra’s heads. The regrowing heads show different faces of the evolving problem of spare parts in European intellectual property and competition law. The inclusion of the repair clause into the design regime, categorically recommended in the Legal Review on the Industrial Design Protection7 as a necessary EUwide solution, may be compared to the cutting off of the Hydra’s immortal head carried out by Heracles with a golden sword given to him by Athena. The importance of this desirable step should not however be overestimated, as the lack of horizontal effect of this regime-specific limitation threatens to bear more unexpected Hydra’s heads emerging from other IP regimes and endangering the area of freedom guaranteed by the repair clause.

33.2 the first hydra’s head cut off – spare parts and competition rules The problem of the design right holders’ refusal to grant a licence to independent spare parts manufacturers enabling manufacture and offer of spare parts was referred to the ECJ for the first time in the 1980s, when no harmonized rules of IP protection of spare parts existed in Europe. In its landmark cases Volvo v. Veng8 and Circa/ Mexicar v. Renault,9 the ECJ confirmed that neither the enforcement of design rights nor the refusal to grant a licence amounted to an abuse of the dominant position in the meaning of the Treaty.10 Despite the serious arguments put forward by the parties and AG Mischo, who pointed out the problem of double reward of the design right holders in such a case,11 the Court’s uncompromising position in both cases resembles very much the cutting off of the Hydra’s head, or rather – less brave – the turning away from her.12 Although after these judgments the Court had a few opportunities to discuss the interface between intellectual property rights and 6

7 8 9

10

11 12

See the Commission legislative proposals: of 1993 limiting the period of protection of the parts of a complex product to three years [COM(93)344 final]; of 1997 introducing the remuneration system with regard to spare parts [COM(96)66 final]; and of 2004 introducing a repair clause into DD [COM(2004)582 final]. None of them succeeded, and the latter one was withdrawn after ten years as obsolete, OJ 2014 C 153/3. MARKT2014/083/D, Brussels 2016, 135–151. ECJ, 5.10.1988, case C-238/87, AB Volvo v. Eric Veng (UK ) Ltd, ECLI:EU:C:1988:477. ECJ, 5.10.1988, case C-53/87, CICRA and Maxicar v. Régie nationale des usines Renault, ECLI:EU:C:1988:472. See ECJ case C-238/87, AB Volvo, paras. 8–9, and ECJ case C-53/87, CICRA and Maxicar, paras. 11, 16. Opinions of AG Mischo, 21.6.1988, in case C-238/87 and case C-53/87. See critical remarks by A. Kur, Limiting IP Protection for Competition Policy Reasons – A Case Study Based on the EU Spare-Parts-Design Discussion, in: Josef Drexl (ed.), Research Handbook on Intellectual Property and Competition Law (Edward Elgar, 2008), 321.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

394

Anna Tischner

competition law and showed a new approach towards this topic,13 the core problem of spare parts in competition law did not return to the Court which, as a result, had no further opportunities to reconsider its short-sighted approach presented in these cases. In the meantime, a legislative step was taken and a repair clause was introduced to the CDR and in some European legislations.

33.3 the second hydra’s head cut off – spare parts and trade mark protection A new systemic perspective of a repair clause emerged in the Ford case.14 The repair clause as set forth in Article 110 CDR and as provided for in selected national legislations, allowing the manufacture and sale of the design of a component part for the purpose of the repair of a complex product in order to restore its original appearance, does not constitute a horizontal limitation in the intellectual property law. The omnipresent tendency to accumulate different forms of IP protection (including for products such as spare parts) makes the full implementation of the repair clause dependent on the reaction of the other overlapping regimes to its existence. The horizontal impact of the design repair clause on trade mark law was the first important aspect of this limitation the Court of Justice of the European Union (CJEU) had the opportunity to deal with. It did so in the Ford ruling. The very concise order15 of the Court in this case resembles, however, cutting off the next regrowing Hydra’s head. The Court referred to the problem of uncoordinated limitations to trade mark and design rights very shallowly, assuming that the repair clause provided for in the design regime cannot be applied per analogiam in trade mark law, because the EU harmonized trade mark law has its own catalogue of limitations which should not be exceeded by national courts.16 Consequently, the trade mark cannot be reproduced as an element of the spare part’s appearance without the consent of the right holder,17 although in the past in the Dior case18 the Court had had the courage to apply the principles of trade mark law exhaustion to

13

14

15

16 17 18

ECJ, 1995, joined cases C-241/91 P and C-242/91 P, Radio Telefis Eireann (RTE) & Independent Television Publications (ITP) v. Commission, ECLI:EU:C:1994:21; and ECJ, 29.4.2004, case C418/01, IMS Health v. NDC Health, ECLI:EU:C:2004:257. CJEU, 6.10.2015, case C-500/14, Ford Motor Company v. Wheeltrims srl, ECLI:EU:C:2015:680 (hereinafter the Ford case). According to the CJEU the answer to the question referred for a preliminary ruling raised no reasonable doubt hence the Court decided to rule by reasoned order pursuant to Article 99 of its Rules of Procedure. CJEU, case C-500/14, Ford , para. 44. Ibid., para. 45 and the operative. ECJ, 4.11.1997 r., case C-337/95, Parfums Christian Dior v. Evora, ECLI:EU:C:1997:517, para. 58.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

Spare Parts in EU Design and Trade Mark Law

395

copyright law by analogy,19 and recently, in the Nintendo case,20 it showed considerable flexibility when interpreting the design right limitations (a design citation in form of referential use21). It seems that in the Ford case the Court was focused on the internal order in the system of intellectual property rather than on the essence and aim of this exceptional limitation.

33.4 the third hydra’s head cut off – spare parts and design protection The new Hydra’s head has regrown quickly. Homework not done properly by the Court in the Ford case, namely in-depth analysis of the scope of a repair clause and correlation between trade mark and design regimes in the context of spare parts, reappeared as a rebound of the Ford decision in the Acacia cases.22 This time, however, the Court, supported by the in-depth and progressive opinion of AG Saugmandsgaard Øe,23 showed greater understanding of the special role of a repair clause in the realization of the idea of a common market. The need for the extensive liberalization of the spare parts market greatly influenced the all-inclusive definition of a spare part adopted in the Acacia decision.24 For a long time, an additional condition of the design dependence upon the appearance of the complex product25 was a bone of contention for those interpreting the scope of the repair clause. As clarified in the Acacia judgment, rejecting this requirement,26 “the purpose of the ‘repair’ clause is to avoid the creation of captive markets in certain spare parts and, in particular, to prevent a consumer who has bought a long lasting and perhaps expensive product from being indefinitely tied, for the purchase of external parts, to the manufacturer of the complex product”.27 Nonetheless, the repair clause may benefit only those parts with a dual nature (i.e. present on the spare parts market and on the accessories market purchased for 19 20

21

22

23

24 25 26 27

A. Kur, Ersatzteilfreiheit zwischen Marken- und Designrecht, GRUR 2016, 24. CJEU, 27.09.2017, joined cases C-24/16 and C-25/16, Nintendo Co. Ltd v. BigBen Interactive GmbH i BigBen Interactive SA, ECLI:EU:C:2017:724, para. 86. Levelling down the threshold of copyright protection of works of applied art in the EU (see AG Szpunar’s opinion of 2.5.2019 in case C-683/17, Cofemel v. G-Star, ECLI:EU:C:2019:363) may soon result in the same problems under the copyright regime, i.e. with atypical quotation of such works in the form of referential use in commercial communication. In this case the Nintendo flexibility will be welcomed and needed. CJEU, 20.12.2017, joined cases C-397/16 and C-435/16, Acacia Srl v. Pneusgarda Srl and Audi AG, and Acacia Srl and Rolando D’Amato v. Dr. Ing. h.c.F. Porsche AG, ECLI:EU:C:2017:992 (hereafter the Acacia case). Opinion of AG Saugmandsgaard Øe, 28.9.2017, in cases C-397/16 and C-435/16, Acacia, ECLI: EU:C:2017:730. Ibid., paras. 75–76. Mentioned only in Recital 13 CDR. CJEU, C-397/16 and C-435/16, Acacia , para. 53. Ibid., para. 50.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

396

Anna Tischner

purposes other than repair, e.g. stylistic) that are functionally linked to the repair of a complex product.28 In fact, it is the “repair” that makes a part of a complex product a spare part in the meaning of Article 110(1) CDR. The purpose of repair (i.e. restoring the original appearance of a complex product) may be achieved by parts which are visually identical to original parts.29 The identical appearance is objectively necessary for spare parts to perform their function properly. Lack of identity, which should be assessed with the apparatus of the design regime, namely by considering all the factors creating the appearance of the product (i.e. features of the lines, contours, colours, shape,30 among which trade mark is not an immaterial detail in the case of many parts), takes such non-identical parts out of the scope of the repair clause. However, in the light of both CJEU decisions, the situation of spare parts with a trade mark constituting an integral feature of their appearance resembles squaring the circle. On the one hand, in line with the Ford authority, the trade mark cannot be used in such a case as a repair clause is provided for in the design regime only. On the other hand, however, as underlined in the Acacia ruling, a change of the product’s appearance, including by affixing a different trade mark or removing it completely, may make the part unsuitable to perform its function as a spare part properly. This stalemate situation proves that using a trade mark as an element of the product appearance in many cases is a must. The prohibition of placing such parts on the market in the circumstances permitted by the repair clause means that manufacturers of these parts cannot offer a full range of products to users, and are required to enter into licensing agreements to reproduce the original appearance of a complex product faithfully. Noteworthy is the fact that a wide scope of a repair clause as outlined in the Acacia judgment is counterbalanced by the extended precautionary measures demanded from its beneficiaries. A manufacturer or supplier of the replacement parts who intends to rely on this limitation is obliged to ensure a strict compliance of its downstream users with the conditions of the repair clause. The duty of diligence with regard to further stages of distribution consists of a set of very burdensome duties, among which information duties – besides appropriate contractual instruments – are intuitively the first (i.e. clear and visible information on the product, on the packaging, in catalogues and commercial documents, stating that the product 28 29 30

Ibid., paras. 69–70. Ibid., paras. 69–70. The assessment of identity of a spare part should not be mixed up with the assessment of the same overall impression created by designs (Art. 5 DD). Under the design regime designs are identical if their features differ only in immaterial detail (Art. 4 DD). The AG adopted a different line of interpretation on this stalemate issue, compromising on the identity criterion. His statements that affixing of a different trade mark does not imply that the appearance of the replacement part differs from that of the original part to which a trade mark owned by the Community design holders was affixed are rather speculative (opinion of AG Saugmandsgaard Øe in C-397/16 and C-435/16, Acacia, para. 108).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

Spare Parts in EU Design and Trade Mark Law

397

reproduces the protected design of another entity and is intended only for the purpose of repairing a complex product in order to restore its initial appearance).31 Interestingly, a draft Europe-wide repair clause as proposed by the Commission in 2004 included the parallel obligation to “ensure that consumers are duly informed about the origin of spare parts so that they can make an informed choice between competing spare parts”.32 Finally, a diligent manufacturer or supplier is obliged to refrain from selling these component parts if it possesses the knowledge, or, in the light of all the relevant circumstances, ought reasonably to know, that parts will not be used by the downstream users in accordance with the conditions of the repair clause.33 One may be concerned whether such burdensome duties and a risk of indirect liability will weaken the interest of the stakeholders in a repair clause in future. Nonetheless, a new approach to a repair clause which “does not constitute a derogation from a fundamental principle of EU law . . . but rather a provision which seeks to maintain the balance between legitimate yet competing interests”34 can be fraught with consequences, not only within the design regime. The line of interpretation established by the Acacia judgment shows that a repair clause may be used in very special circumstances and the purchasers of these parts are a particular, wellinformed and well-oriented, group of consumers. Undoubtedly factors shaping the scope of this limitation, so precisely described in the Acacia ruling, create the standards of fair commercial practices on the European spare parts market.

33.5 ripple effect of a repair clause in the trade mark regime The CJEU's uncompromising position in the Ford case and the gate opened by the Acacia ruling make it necessary to pursue the CJEU line of reasoning based on systemic order and search within the trade mark regime itself for a solution which will take into account the rationale and important interests embodied in a repair clause.35 The special status of a repair clause within EU legislation, harmonizing with EU economic policy, supports the view that despite being embedded in the design regime only, it may have a very special radiating effect on other, neighbouring IP regimes.

31 32 33

34 35

CJEU, C-397/16 and C-435/16, Acacia, paras. 86–87. Art. 14(2) of the amended DD draft, COM(2004)582 final. Supra n. 31, para. 88. This is in line with the tendency of adopting a duty of care approach in cases considered under the umbrella of secondary liability. See further A. Ohly, The Broad Concept of “Communication to the Public” in Recent CJEU Judgments and the Liability of Intermediaries: Primary, Secondary or Unitary Liability?, GRUR Int. 2018, 517. Opinion of AG Saugmandsgaard Øe in C-397/16 and C-435/16, Acacia, para. 117. The solution suggested also by Kur, supra n. 19, 25.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

398

Anna Tischner

First and foremost, the strong pro-competitive orientation of a repair clause outlined in the Acacia ruling encourages a “repair clause friendly” interpretation of the general trade mark infringement criteria. These preliminary criteria as developed by the CJEU (in particular the quite unpredictable interpretation of the requirement of use in respect of goods,36 and of use which has an adverse effect on the protected functions of the trade mark37) make it prima facie difficult to argue for the legality of use of a trade mark as a necessary aesthetic element of a spare part. Nevertheless, the use of a trade mark as a necessary feature of the appearance of a legally manufactured spare part offered within the strict limits of a repair clause (i.e. within controlled distribution chains) and in circumstances excluding confusion (e.g. in appropriate packaging and with sufficient information about its source) does not have to be automatically perceived as an indication of the commercial origin of these parts by their customers.38 The trade mark effect in such cases may be neutralized.39 A very flexible trade mark infringement gatekeeper of use in relation to goods compels however the next step, namely challenging the function theory with regard to the use of a trade mark to the extent necessary to benefit from a repair clause. The criterion of the adverse effect on the functions of a trade mark should not be regarded merely as a proxy for the CJEU broad concept of trade mark infringement, as it also acts as an additional balancing instrument in infringement cases.40 Protection of the trade mark functions “must respect overriding other interests” – this being excellently expressed by AG Kokott who underlined that not every adverse effect on the functions of the mark should result in an infringement.41 In the case of spare parts, the interests of consumers and competitors may outweigh the interests of the proprietor embodied in the protected functions. The tangible effect of the recent EU trade mark law reform is the extension of the catalogue of limitations of trade mark rights. As such the resort to the function theory to restrict trade mark rights in view of the use included in that catalogue may come less justified and more rarely practiced.42 However, the question is to what extent this balancing exercise will still be carried out within the framework of trade mark functions in cases not covered by

36 37 38 39

40

41 42

ECJ, 25.1.2007, case C-48/05, Adam Opel AG v. Autec AG, [2007] ECR I-01017, para. 20. CJEU, 18.6.2009, case C-487/07, L’Oréal v. Bellure, [2009] ECR I-05185, para. 58. Cf. opinion of AG Colomer in case C-48/05, supra n. 36, paras. 29, 32. So far the case law is not inclined to conclude that such use has no effect on the origin function. See the German Federal Supreme Court decisions of 12.3.2015, I ZR 153/14 (BMWEmblem), IIC 2016, 114 and 7.3.2019 r., I ZR 61/18 (Radiator Grille), IIC 2020, 113, and the decision of the Polish District Court in Warsaw of 27.8.2018 r., XXII GWzt 6/17 relating to the similar form of TM infringement by a logo fixture in the radiator grilles. As stressed by A. Kur, Trade Marks Function, Don’t They? CJEU Jurisprudence and Unfair Competition Principles, IIC 2014, 442. Opinion of AG Kokott, 7.4.2011, in case C-46/10, Viking Gas, para. 59. A. Kur, M. Senftleben, European Trade Mark Law (Oxford University Press, 2017), 293.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

Spare Parts in EU Design and Trade Mark Law

399

this catalogue but justified by other important interests, as may be the case with spare parts. Then, as regards the origin function, relevant confusion (which may be effectively eliminated if a trade mark is used strictly within the limits of a repair clause) should not be confused with possible confusion triggered by third parties subsequent to the purchase of a spare part in order to repair a complex product, for example when a spare part is fitted to a vehicle.43 As regards parts which are distributed within the limits of a repair clause, the relevant public is a particular and very restricted group of downstream users of the repair clause (usually car-repair professionals).44 Their perception of the trade mark cannot be aligned with that of the end users of the complex products in which the parts are assembled. This may result in diversification of the level of attention of both groups which, if higher, may limit or eliminate the risk of confusion.45 Moreover, there are no grounds to extend the essential function of trade marks to guarantee to purchasers of complex products on the aftermarket that all their components come from their original manufacturer. The quality function is being exhausted in this respect on the primary market. It is rather the role of product safety regulations and technical approvals to guarantee the quality and safety of spare parts installed in a complex product. In the case of double identity and trade marks with a reputation, side effects of this type (inaccurately referred to as post-sale confusion) are taken into account in the assessment of a negative impact on other accessory functions (communication, advertising, investment).46 Such outcomes in case of “blurring”, “tarnishment” and “free-riding” on a trade mark with a reputation require evidence of a change in the economic behaviour of the average consumer,47 the detriment to the trade mark’s power of attraction48 or its exploitation,49 the latter effect visible in the Ford case where higher prices were achieved from the sale of parts with the Ford trade 43

44 45

46

47

48 49

This difference in practice is often blurred; e.g. in the German Federal Supreme Court decision of 7.3.2019 r., I ZR 61/18 (Radiator Grille), IIC 2020, 118, para. 29, it was assumed that after mounting in the vehicle, the general public would see the mounting fixture similar to the trade mark as an indication of the origin of the vehicle and also of the radiator grille. Cf. in the same vein its decision of 12.3.2015, I ZR 153/14 (BMW-Emblem), GRUR 2015, 1011, para. 34. CJEU, C-397/16 and C-435/16, Acacia, paras. 82, 85–87. Cf. GC, 12.7.2019, case T-792/17, MAN Truck & Bus AG v. EUIPO, ECLI:EU:T:2019:533, para. 42 (“even if spare parts are priced relatively modestly, the relevant public will have a level of attention that is higher than average”), as pointed out by Peter Ling in his comment of 20.8.2019: Four Rings to Rule Them All – German Federal Court of Justice Finds Trademark Infringement in Radiator Grille with Audi-Logo-Shaped Mounting Fixture, https://ipkitten .blogspot.com/2019/08/four-rings-to-rule-them-all-german.html. A. Ohly, Post-Sale Confusion?, in: W. Büscher et al. (eds.), Marktkommunikation zwischen Geistigem Eigentum und Verbraucherschutz: Festschrift fu¨r Karl-Heinz Fezer zum 70. Geburtstag (C. H. Beck, 2016), 625. CJEU, 27.11.2008, case C-252/07, Intel Corporation v. CPM United Kingdom, ECLI:EU: C:2008, para. 77. CJEU, case C-487/07, L’Oréal v. Bellure, ECLI:EU:C:2009:378, para. 40. Ibid., para. 41.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

400

Anna Tischner

mark than from the parallel sale of no-name parts.50 Notwithstanding the above, a repair clause as outlined in the Acacia judgment is a strong due cause factor deriving from the freedom of competition and the freedom to conduct a business,51 and constitutes an objectively overriding reason which in many cases justifies the use of a trade mark as a necessary element of a spare part in a way enabling the full benefit of this design right limitation.52 As the conflicting interests are considered with reluctance within the trade mark infringement analysis, trade mark rights limitations are better suited to balancing the interests of the trade mark owner and other, competing interests of spare parts manufacturers and users.53 First of all, the standard of due diligence as established in the Acacia decision is relevant for a safety valve factor of compliance with honest practices54 – a final check of all trade mark rights limitations. The recent reform of trade mark rights limitations in the EU, rooting them explicitly in fundamental rights and freedoms,55 is particularly powerful in the context of spare parts. The procompetitive effect of the clause, as excellently described in AG Saugmandsgaard Øe’s opinion and the CJEU’s judgment in the Acacia case, is in line with the rationale for trade mark protection in the EU for which freedom to conduct business activity enshrined in Article 16 of the Charter of Fundamental Rights and its emanation – freedom of competition as recognized in Article 119(1) and (3) of the Treaty on the Functioning of the European Union, are constitutive.56 Despite the absence of a limitation corresponding to a repair clause in trade mark law, the use of another person’s trade mark constituting an immanent element of a spare part’s appearance may be considered as a form of descriptive use. The alternative implied in Article 14(1)(b) TMD between marks and indications which are non-distinctive57 and those which relate to various product characteristics suggests that this limitation may be also applied in cases where a sign has a distinctive character (it is not merely descriptive) but in particular circumstances it indicates the characteristics of the product. The question is whether a faithful reproduction of 50 51 52

53 54

55

56 57

CJEU, case C-500/14, Ford , para. 22. CJEU, 22.9.2011, case C-323/09, Interflora v. Marks & Spencer, ECLI:EU:C:2011:604, para. 91. A possible outcome of that line of reasoning is the change in the design registration practices by way of more frequent use of disclaimer techniques with regard to trade marks. Kur and Senftleben, supra n. 42, 408. Cf. in that regard the opinion of AG Bot, 1.3.2017, in joined cases C-24/16 and C-25/16, Nintendo Co. Ltd v. BigBen Interactive GmbH i BigBen Interactive SA, para. 82, pointing at Directive 2005/29/EC as a uniform benchmark of commercial practices fairness in the EU based on the requirements of professional diligence, which essentially links unfairness to the risk of confusion and operates on the model of the average consumer. ˙ Recital 21 TMD and EUTMR. See Ł. Zelechowski, Invoking Freedom of Expression and Freedom of Competition in Trade Mark Infringement Disputes: Legal Mechanisms for Striking a Balance, ERA Forum 2018, No. 19, 117. Explanations relating to the Charter of Fundamental Rights, [2007] OJ C 303/17. Which may legitimize the use of a shape as an element of a combined mark which is devoid of distinctive character. See Kur and Senftleben, supra n. 42, 415, 418.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

Spare Parts in EU Design and Trade Mark Law

401

a part’s appearance (i.e. including a trade mark as its necessary visual feature or the image of the trade mark reproduced for technical purposes58) constitutes ‘use which concerns the characteristics of the goods’ within the meaning of Article 14(1)(b) TMD. The CJEU is very reluctant to decide on the legality of the use of a trade mark on this basis. Some inconsistent interpretative guidance on the content of this limitation was given in the past in the Opel case, where the Court finally ruled out the purpose of this provision in the case of the reproduction of the Opel logo on miniature models, not being an indication of the type, quality or other characteristics of the reduction models, but merely an element of a faithful reproduction of the original vehicles.59 The arguments of AG Colomer in this case, who stressed the balancing effect of that provision, seem to be more insightful in this respect.60 In the light of this opinion, it may therefore seem less risky to argue that, since the manufacture of spare parts consists in faithful and detailed copying of their appearance, where the trade mark is an integral part of the original part, its reproduction on that part may be regarded as the use of “another feature” of the product referred to in Article 14(1)(b) TMD, justified by the interest of other operators to compete on the same terms and by the interest of the user of spare parts in having access to a wider choice of those parts. As a result of trade mark law reform in Europe, also the scope of limitation for referential use in Article 14(1)(c) TMD has been significantly extended to all forms of reference to the goods or services of the entitled trade mark, for example in order to provide information about an alternative offer. The right to market alternative products provided for in the form of a repair clause should have an impact on the interpretation of this limitation but also on the reading of Article 4(g) of Directive 2006/114/EC on comparative and misleading advertising,61 which prohibits the presentation of goods or services as imitations or replicas. Since the repair clause is even more explicit than an unwritten principle of freedom of imitation beyond the limits of the IP protection, the freedom to carry out comparative advertising of 58

59

60

61

This may be the case also with regard to a particular group of spare parts (e.g. radiator grilles) on which a mounting fixture for a logo is affixed, which resembles the trade mark of the car. However, in the German Federal Supreme Court decision of 7.3.2019 r., I ZR 61/18 (Radiator Grille), IIC 2020, 119–120, only the limitation for referential use was considered in such a case but ultimately rejected. For the Court, the form of logo fixture was not “necessary” to indicate the intended purpose of the radiator grille. ECJ, 25.1.2007, case C-48/05, Adam Opel AG v. Autec AG, ECLI:EU:C:2007:55, paras. 40–45. In that dispute, however, the Commission took the view that, in other factual circumstances in which the reduction models were intended for collectors, an identical reproduction of every detail of the original vehicle could possibly constitute an essential characteristic of that category of goods, with the result that the limitation for descriptive use then also covers the faithful reproduction of the trade mark. Opinion of AG Colomer in case C-48/05, Adam Opel AG v. Autec AG, ECLI:EU:C:2006:154, paras. 50, 51. OJ L 376, 27.12.2006, 21–27.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

402

Anna Tischner

legally commercialised spare parts (which by their nature imitate original parts) seems to be logical. In the context of the commercialization of spare parts, the use of someone else’s trade mark is usually necessary to provide information on the intended use or compatibility of such parts. Only in the case of this form of referential use the “necessity” factor, which incorporates the principle of proportionality, determines the legitimacy of such reference (especially its form). However, in many cases a particular form of reference is the one needed and useful but not the only one necessary and possible,62 which means such references are rarely accepted.63 In fact, this factor leaves room for balancing conflicting interests, but should not, in the case of commercialisation of spare parts, lead to the conclusion that referential use of trade marks to inform about alternative offers is subject to less restrictive evaluation rules than a reference to inform about the intended use of the product as a spare part. Indeed, non-original spare parts are usually market alternatives to original parts. Moreover, it seems that nowadays the changes in customs and practices in market communication, especially the dominance of image over text, affect which forms and instruments of this communication are considered both necessary and justified as honest commercial practices.64

33.6 conclusions The expansion of intellectual property protection, which adapts to the constant transition of the world, forces us to look for solutions to ensure systemic coherence, as it is not feasible to cut off all problems arising from the complexity of this system. 62 63

64

CJEU, 23.2.1999, case C-63/97, BMW v. Deenik, ECLI:EU:C:1999:82, para. 60. Cf. the decisions of the German Federal Supreme Court of 14.4.2011 – I ZR 33/10 (GROSSE INSPEKTION FÜR ALLE), GRUR 2011, 1135 and 28.6.2018 – I ZR 236/16 (keine-vorwerkvertretung), GRUR 2019, 165, and Marco Bellia, Comment on “Vorwerk Folletto”: An Important Decision Concerning Trade Mark Reproduction on Replacement Components, IIC 2015, 620. There is a fundamental doubt as to whether information about compatible or accessory products must be provided exclusively with the help of a word mark, which for years has been treated as the only acceptable form falling within the limits of the criterion of necessity of referential use. Visualization of the mark will in many cases be an appropriate and effective tool to inform consumers about the intended use of the product, not the only one possible but the one best suited to this end. Cf. German Federal Supreme Court decision of 15.7.2004 – I ZR 37/01 (Aluminiumräder), GRUR 2005, 165, progressive in that regard (“It also cannot be ruled out that the defendant’s product may benefit from being depicted in the brochure and advertisement as an accessory of a sports car of high prestige value. However, this has not been done unfairly without due cause”), and more reactionary decision of 14.4.2011 – I ZR 33/10 (GROSSE INSPEKTION FÜR ALLE), GRUR 2011, 1138, para. 26 (“However, the use of a word mark will normally affect the legitimate interests of the trade mark proprietor less severely than the use of his figurative mark because the word mark is primarily suitable for describing the intended purpose of the services.”); and para. 27 (“[T]he eye-catching depiction of the wellknown figurative mark and the associated unfair exploitation of the repute of that mark is contrary to the legitimate interests of the claimant.”).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

Spare Parts in EU Design and Trade Mark Law

403

The twentieth century was certainly the age of strengthening the protection of intellectual property; the twenty-first century promises to be rather a good time to work on its limitations, which should not be interpreted “broadly or narrowly, but simply correctly”.65 The ripple effect of the repair clause on the trade mark regime proposed here is an example of an effort to trim the IP protection for quite a narrow set of situations and for a good reason. Annette Kur, being the leading expert in this way of thinking, might agree. A repair clause is certainly Annette’s topic, not only because it vents her erudition. The search for well-balanced solutions and horizontal focus on the phenomena in intellectual property law are enduring in her oeuvre. Dealing with a topic that is highly controversial and involves substantial economic interests reveals the great value of academic autonomy and independent research. In the light of this value, Annette’s acquis and attitude, in particular with regard to spare parts issues, are of exceptional importance for the IP environment and for the processes of law-making and legal interpretation.

65

A pertinent phrase attributed to Matthias Leistner.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.043

b

IP Overlaps

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

34 Geographical Indications as Intellectual Property Rights – Beyond Transition and Coherence Roland Knaak1

Geographical indications (GIs) are a subject matter that has not been in the focus of Annette’s research on her big themes – transition and coherence in IP law. This leads to the question whether GI protection is a topic beyond transition and coherence.

34.1 legal nature of geographical indications According to Article 22(1) of the TRIPs Agreement, GIs are indications which identify a good as originating in the territory of a Member, or a region or locality in that territory, where a given quality, reputation or other characteristic of the good is essentially attributable to its geographical origin. The core elements of this definition are almost identical with the requirements for designations of origin and geographical indications under Article 5(1) and (2) of EU Regulation No 1151/2012.2 GIs within this meaning are intellectual property rights (IPRs). They belong, with copyrights, trademarks, industrial designs and patents, to the IPRs for which the TRIPs Agreement has established minimum rules of protection. Pursuant to Article 1(2) of TRIPs, the term intellectual property refers to all categories that are the subject of sections 1 through 7 of Part II, among them GIs as a subject matter of Section 3. Also, the European Union legislator leaves no doubt that IPRs include GIs. As an objective of the scheme for sui generis protection of GIs, Article 4(b) of Regulation No 1151/2012 refers explicitly to ensuring Union-wide protection for protected 1

2

Dr. jur., Attorney-at-Law in Munich, Affiliated Research Fellow at the Max Planck Institute for Innovation and Competition, Munich. Regulation (EU) No 1151/2012 of 21 November 2012 on quality schemes for agricultural products and foodstuffs, OJ L 343/1.

Downloaded from https://www.cambridge.org/core. 407 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

408

Roland Knaak

designations of origin and protected geographical indications as intellectual property rights. Ten years earlier the Court of Justice of the European Union (CJEU) had already recognized in its case law that GIs protected under EEC Regulation No 2081/923 fall within the scope of industrial and commercial property rights.4 The Court based this view mainly on the idea that GIs aim at ensuring the quality and authenticity of the product and, consequently, at protecting the reputation of the product and the GI against improper use by third parties.5 The classification of GIs as IPRs was and still is not undisputed. In the beginnings of the international protection for IPRs, GIs were not recognized as sui-generis rights but indirectly protected by provisions against false or deceptive indications of source. This approach was adopted by the Paris Convention of 18836 and by the Madrid Agreement for the Repression of False and Deceptive Indications of Source on Goods signed in 1891. The protection against deception of the relevant public provided under these agreements did not depend on any qualifying elements of the indications of source, or the products covered by those indications. It was only the Lisbon Agreement for the Protection of Appellations of Origin of 1958 that established an international protection for a category of GIs featuring elements of IPRs. According to Article 1(2) of the Lisbon Agreement this protection is based on a registration with the International Bureau of Intellectual Property in Geneva. Pursuant to Article 3 the protection extends to any usurpation or imitation of the appellation of origin, even if the true origin of the product is indicated or the appellation is used in translation or accompanied by terms such as “kind”, “type”, “make” or “imitation”. Whether the usurpation or imitation results in a deception of consumers is without relevance. The bilateral agreements on the protection of indications of source, appellations of origin and other geographical designations concluded between European countries as from the 1960s provided international protection according to the law of the country of origin for geographical indications of source or GIs of the contracting states included in lists of product names as annexes to the agreements. These lists were the result of the bilateral negotiations. The agreements were able to overcome the differences between the national GI systems, in particular the IPR-related schemes of protection which existed in Italy and France, and schemes ensuring protection according to unfair competition law, such as in Germany. This bilateral 3

4 5

6

Regulation (EEC) No 2081/92 of 14 July 1992 on the protection for geographical indications and designations of origin for agricultural products and foodstuffs, OJ L 208/1. CJEU C-108/01 of 20 May 2003, paragraph 64 – Prosciutto di Parma. CJEU C-108/01, ibid.; CJEU C-478/07, paragraphs 110 to 113 – Bud (relating to Regulation (EC) No 510/2006); CJEU C-56/16, paragraph 80 – Port Charlotte (relating to Regulation (EC) No 1234/2007). See Article 10 of the Paris Convention: “(1) The provisions of the preceding Article [Article 9: Seizure on importation, or other situations, of goods unlawfully bearing a mark or trade name] shall apply in cases of direct or indirect use of a false indication of the source of the goods or the identity of the producer, manufacturer, or merchant.”

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

GIs – Beyond Transition and Coherence

409

protection does not qualify as an IPR protection even though the CJEU, in its landmark judgment in the Turrones case,7 has recognized the character of this protection for the Spanish GIs “Turrón de Alicante” and “Turrón de Jijona” according to the bilateral agreement between France and Spain as falling within the scope of exceptions from the principle of free movement of goods which are justified on grounds of the protection of industrial and commercial property within the meaning of Article 36 EEC Treaty. A breakthrough for the international protection of GIs as IPRs was achieved by the TRIPs Agreement. GIs are covered in Part II of the Agreement as a category of IPRs to which Articles 22 to 24 apply. Nonetheless, the TRIPs Agreement has not overcome the different systemic approaches governing GI protection and has not introduced an obligation on Members to provide a sui generis protection for GIs. The gap between the countries of the Old and the New World with regard to GI protection has even deepened, and the issue of international GI protection has achieved a political dimension. Thus, the further strengthening of international GI protection is not a matter that can be realized within the framework of a TRIPs revision. Rather it has become a topic for Free Trade Agreements (FTAs). Similar to the bilateral agreements these FTAs afford international GI protection between parties operating with different systems for GI protection, mainly contracting parties that provide sui generis protection and those that rely on certification mark protection. The European Union is currently pushing the GI chapters of these FTAs as instruments to extend the protection of European GIs registered under the relevant EU Regulations to third countries, such as Canada, Japan, Singapore and Mexico. Finally, the revised Lisbon Agreement8 will establish a scheme of international GI protection that extends the existing international registration system for appellations of origin to geographical indications within the meaning of TRIPs. Article 2(1)(ii) of the Geneva Act defines the geographical indication as: any indication protected in the Contracting Party of Origin consisting of or containing the name of a geographical area, or another indication known as referring to such area, which identifies a good as originating in that geographical area, where a given quality, reputation or other characteristic of the good is essentially attributable to its geographical origin.

Thus, reputation-based GIs are admitted for registration under the revised Lisbon system. Since the Lisbon system covers not only agricultural products and foodstuffs but also non-agricultural products such as handicrafts, this revision has fostered the European Commission’s efforts to introduce a Union-wide system of GI protection

7 8

ECJ C-3/91 of 10 November 1992, paragraphs 37 and 38 – Exportur SA ./. LOR SA. The Geneva Act of the Lisbon Agreement was adopted on 20 May 2015 and will enter into force three months after five eligible parties have deposited their instruments of ratification or accession with WIPO.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

410

Roland Knaak

for non-agricultural products complementary to the existing scheme for agricultural products and foodstuffs under Regulation No 1151/2012.

34.2 characteristics of geographical indications While there is a clear international trend to accept GIs as IPRs it is not evident what makes GIs belong to the category of IPRs. Even the considerations of the CJEU in its case law on why protected designations of origin (PDOs) and protected geographical indications (PGIs) are industrial and commercial property rights are focused on the justification of this protection as IPRs. They do not explain which elements of GIs make them IPRs and what is the core element that turns a GI into an IPR. Unlike patents, industrial designs, copyrights and trademarks, which derive from conditions featuring a “novelty”9 effect, GIs are IPRs based on existing elements and conditions. They require traditions, frequently developed from natural resources, or a tradition-based reputation linked with a specific geographical region. The characteristics making up a GI cannot be produced by an individual person. They are the results of long-term processes in which many local producers are involved. While patents, industrial designs, copyrights and trademarks protect private interests of individual persons, GIs are incentives for all producers of the protected product originating in the relevant geographical area to maintain the qualities of their product. This function of the sui generis GI protection under EU law is emphasized by the EU legislator in Article 4 of Regulation No 1151/2012.10 Any producer or operator with his or her seat in the relevant geographical region who complies with the rules of the quality scheme for the protected product is entitled to join the controlling system under the EU Regulation. The right to use a GI registered pursuant to Regulation No 1151/2012 is granted to all operators in the geographical area which have entered into the system for verification of compliance with the product specification. In order to identify the IPR-related element of GIs it is helpful to look at other similar rights to a designation, namely trademarks and traditional specialities guaranteed (TSGs). Trademarks have been recognized since the beginnings of international IP law as IPRs and as such they are embedded into the protection schemes on the national and supranational levels. TSGs are likewise names or designations

9

10

That sounds strange for trademarks, but is the consequence of the earlier rights structure in trademark law. Article 4 of Regulation No 1151/2012 stipulates as the objective of the scheme for protected designations of origin (PDO) and protected geographical indications (PGI) “to help producers of products linked to a geographical area by (a) securing fair returns for the qualities of their products; (b) ensuring uniform protection of the names as an intellectual property right in the territory of the Union; (c) providing clear information on the value-adding attributes of the products to consumers”.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

GIs – Beyond Transition and Coherence

411

protected in the European Union under a specific regime.11 Nevertheless they do not belong to the category of IPRs. They are not subject of any international treaty on the protection of IPRs nor are they classified by the European legislator as IPRs. Contrary to the scheme for PDOs and PGIs which, according to Article 4 of Regulation No 1151/2012, aims at ensuring uniform protection of the names as IPRs, Article 17 of Regulation No 1151/2012 on the objective of the scheme for TSGs does not refer to a protection of those names as IPRs. The legal objective is defined as safeguarding traditional methods of production by helping producers of traditional products in marketing and communicating the value-adding attributes of their traditional products to consumers. This different approach to GIs on the one hand and TSGs on the other leads to the questions of what makes the difference between those names and what is the common IPR denominator of trademarks and GIs. According to established principles, trademarks are signs capable of distinguishing the goods or services of one undertaking from those of others. After registration, they confer an exclusive right to the owner to prevent third parties from using in the course of trade identical or similar signs for identical or similar goods where such use results in a likelihood of confusion. This exclusive right ensures protection for trademarks in their function as indicators of the commercial origin of the products. Contrary to trademarks, GIs are indications or names whose registration confers a right to all producers of the protected products where these products originate in a defined geographical area and possess certain characteristics which are due to this geographical origin. This exclusive right provides protection for GIs in their function as indicators of the geographical origin of products with specific characteristics linked to their geographical origin. The exclusive right, for both trademarks and GIs, is based on a link. In case of trademarks it is the link between the product and its commercial origin which is under the control of the trademark proprietor after the trademark has been registered. In case of GIs it is the link between the product and its geographical origin which affects certain characteristics of the product and is examined by the competent authorities in the application procedure before the GI is registered. In both cases the link is related to a person or group of persons. The link as a condition of a GI delimits the legitimate users. The exclusive rights deriving from a GI are conferred to these persons. Contrary to that, TSGs do not communicate such link. They are names for a product defined by its main features and produced according to a specific production method which has a traditional character. The group of legitimate users is not restricted. According to Article 23(1) of EU Regulation No 1151/2012 any producer which complies with the standards of the specific production 11

The scheme for protection of traditional specialities guaranteed in the European Union was established by EEC Regulation No 2082/92 of 14 July 1992 on certificates of specific character for agricultural products and foodstuffs and is currently subject of EU Regulation No 1151/2012.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

412

Roland Knaak

method is entitled to use the TSG. Even operators outside the EU may produce the specific product protected as a TSG, and may label it under the protected name and with the Union symbol.12 There is a relation between the TSG product and its tradition, but no link between the protected product and its tradition attributed to a specific group of persons who have established the tradition and are protected as the legitimate users. Even though the transitions between TSGs and GIs – in particular GIs with a reputation built on tradition – are fluent, TSGs lack a link related to persons who are entitled to their use as a feature turning them into IPRs.13 GIs have this link as the constitutive IPR element resulting in the granting of exclusive rights. Their characteristics qualify them, however, as specific IPRs which set them apart from all other IPRs.

34.3 geographical indications in transition As for the other IPRs, TRIPs has established minimum rules for the international protection of GIs. This minimum standard of protection, however, has been achieved on the basis of interests pursued by the TRIPs Members which widely contrast with the interests in the protection of the other IPRs. This contrast reflects the different interests of the Old and the New World. Within the Old World, including many developing countries, GIs are regarded as instruments of maintaining traditions and cultural heritage. This background resulted in the emergence of protective schemes that aim at transforming these assets into IPRs. The New World lacks these traditions and incentives. It has neither developed a factual framework for GIs, nor any legal scheme for their adequate protection. Protection for GIs is, in the countries of the New World, a matter for general principles of unfair competition law derived from the obligations under the Paris Convention, and for trademark law, mostly through the registration of certification marks. TRIPs has not overcome the varying systemic approaches to GI protection. The GI chapter in the TRIPs agreement is a compromise between the Old and the New World. The TRIPs obligations regarding protection for GIs are, according to Article 22(2), limited to principles against unfair competition including protection against any use of GIs which misleads the public as to the geographical origin of the product. Apart from these general rules, additional obligations under Article 23 apply to GIs for wines and spirits which have to be protected by the Members against any use for a wine or spirit not originating in the place indicated by the GI in question, regardless of whether this use results in a deception of the relevant public. This is an absolute IPRrelated protection independent of consumer confusion. The catalogue of exceptions from GI protection provided in Article 24 particularly reflects the interests of the New 12 13

Article 23(3) subparagraph 2 of Regulation No 1151/2012. See for example the specification of the GTS Mozzarella registered under EC Regulation No 2082/92 and published in OJ C 246 of 24 August 1996, p. 9.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

GIs – Beyond Transition and Coherence

413

World in the protection of trademarks14 and the continued existence of designations identical with GIs from the Old World and used as generic terms.15 Thus, in the process of transition of international IPR protection into a more balanced scheme of protection, the role of GIs should be separated from the other IPRs. With respect to GIs, TRIPS has not established a standard of protection that calls for more balancing of the rights conferred by GIs against the interests of producers or other market actors from territories outside the countries of origin for their free use. The GI protection in TRIPs is still underdeveloped. That is especially evident from the restrictions of GI protection to misleading use or acts of unfair competition under Article 22(2) and the right to continued use of GIs as generic terms under the conditions of Article 24(6). This unsatisfying result is confirmed by the commitment of the TRIPs Members in Article 24(1) of TRIPs to agree on entering into negotiations aimed at increasing the protection of individual GIs under Article 23. Twenty-five years later, the international protection of GIs is, instead of increasing on the multilateral TRIPs level, developing via FTAs and the revised Lisbon system towards a stronger sui generis protection covering more and more countries. In the FTAs between the EU and third countries outside the EU the GI chapters are the outcome of deals between the contracting parties on trade issues and market access. In principle, they extend the sui generis protection of GIs registered under the EU scheme to the contracting states and provide protection for GIs of these countries in the EU. According to some of these FTAs, GIs from the EU enjoy automatic protection in the contracting states without any need to pass through an application procedure in the contracting state.16 In the case of the CETA Agreement between the EU and Canada, Annex A featuring the GIs protected in Canada is completed by an additional Annex B containing a list of GIs that face specific limitations17 of this protection to keep these GIs free for a generic use. The acceptance of these exceptions by the European Commission in the negotiations with Canada is in particular disclosing the trade deal approach of the GI chapters in the FTAs. Transition of international GI protection requires more than trade deals. The Geneva Act of the Lisbon Agreement is a starting point for this transition in the GI 14

15

16

17

Pursuant to Article 24 paragraph 5 measures adopted to implement Articles 22 and 23 “shall not prejudice the validity of the registration of a trademark, or the right to use a trademark, on the basis that such a trademark is identical with, or similar to, a GI, where the trademark has been applied for or registered in good faith . . . either (a) before the date of application of these provisions in that Member, or (b) before the GI is protected in its country of origin”. Article 24 paragraph 6 provides that “nothing in this Section on GIs shall require a Member to apply its provisions in respect of a GI of any other Member with respect to goods or services for which the relevant indication is identical with the term customary in common language as the common name for such goods or services in the territory of that Member”. The Treaty between the EU and Singapore signed in October 2018 does not offer this automatic protection. GIs from the EU will be subject to a registration procedure in Singapore which will come into force upon ratification of the agreement by Singapore. For some GIs these exceptions cover a generic use in the official Canadian languages French and English.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

414

Roland Knaak

field and draws attention to the question of how to structure a convincing sui generis scheme of protection for GIs of non-agricultural products. The answer to this question is currently the big challenge in the transition process with respect to GIs.

34.4 geographical indications and coherence Coherence between the protection systems for IPRs is an overall objective widely recognized, but less implemented, by the EU legislator as well as the CJEU. The overlaps of IPRs and the accumulation of protection resulting therefrom has been uncritically accepted for a long time. Only in recent years can a trend be noted of greater reflection on these effects of IPR protection. With respect to GI protection in the EU, however, coherence does not appear to be an approach for a further development of the GI scheme. Due to the collective interests protected by GIs the parameters of GI protection are too distant from the criteria for protection of other IPRs, in particular trademarks. This view is confirmed by a look at the requirements of protection, the content of protection and the limitations of protection. GIs protect geographical names used for products from the relevant geographical area. The subject matter of GI protection is a geographical name or name with a geographical connotation. The protection requires a link between the geographical area indicated by the name and the product that originates in that area. Individual trademarks can, in principle, not protect geographical names because these names are descriptive18 and lack distinctiveness. They are signs or indications in the public domain that, according to established principles in trademark law, are excluded from protection unless they have acquired secondary meaning.19 GIs are protected against any use for products that do not originate in the relevant geographical area or comply with other product specifications.20 This protection extends to designations which are similar to or evoke the GI. Trademarks the use of which would contravene the protection of an existing registered GI are excluded from registration and protection. Their registration can be cancelled.21 In the conflict between an existing trademark and an emerging GI which is similar to the trademark and for which registration has been applied for covering products that are identical with the products of the trademark, the prior trademark confers, in principle, no protection against the GI. The exclusive right of the trademark holder

18

19 20 21

According to Article 74(2) of EU Regulation No 2017/1001 on the EU trade mark and Article 29 (3) of Trade Mark Directive No 2015/2436, this rule does not apply to collective marks while geographical certification marks are excluded from protection as EU trademarks (Article 83(1) of EUTMR) but permitted under the TMD on an optional basis as national trademarks (Article 28(4) of TMD). For EUTMs see Article 7 paragraphs 1(c) and 3 of EU Regulation No 2017/1001. Article 13(1)(a) and (b) of EU Regulation No 1151/2012. Article 14(1) of EU Regulation No 1151/2012.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

GIs – Beyond Transition and Coherence

415

is limited by the registered GI.22 GI and prior trademark coexist. Due to the collective interests in protecting a GI the principle of priority does not apply to these conflicts. Only in cases where a trademark enjoys reputation and renown and the GI would be liable to mislead consumers as to the true identity of the GI products, can the registration of a GI be refused.23 The contrasting parameters of protection exist with regard to the limitations of protection as well. Contrary to trademarks, GIs do not face a limitation for a descriptive use by competitors which do not belong to the group of authorized GI producers. The collective interests of the GI producers prevail. The descriptive use of a GI by competitors is per se inadmissible in order to prevent a GI becoming generic. This rule has been emphasized by the CJEU in the Champagner Sorbet judgment.24 According to the Court the purpose of protecting PDOs such as Champagne to assure consumers that products bearing such a PDO have certain specific characteristics and offer a guarantee of quality due to their geographical provenance would not be achieved if it were possible for a PDO to become a generic term. Therefore, any generic use of a PDO contravenes its protection. In the Champagner Sorbet case the CJEU concluded: If the fact that the name of a foodstuff, such as Champagner Sorbet, is commonly used by the relevant public to refer to that foodstuff, would be relevant for determining whether that use of the PDO as part of that name is unfair or not, that view would be tantamount to accepting that a PDO may be used generically and would be at odds with the GI protection.

Meanwhile, this very clear statement has been implemented by the German Supreme Court in the main proceedings.25

34.5 concluding remark GIs are a subject matter in the field of IP law distinct from all other IPRs. They do not fit into the general rules of IP protection and the themes of transition and coherence in IP law. Their specific characteristics as IPRs protecting collective interests focussed on public goods rather than private interests in exploiting exclusive rights require a different approach in developing GI protection further. Thus, Annette took a wise decision to exclude GIs from her former research topics on transition and coherence. Now she is supporting new research projects on GIs that aim at exploring in depth and on an empirical basis the sui generis scheme of GI protection in the EU. We are looking forward to the results of this research.

22 23 24 25

Article 14(1)(b) of EU Regulation No 2017/1001 and Article 14(2) of EU Regulation No 1151/2012. Article 6(4) of EU Regulation No 1151/2012. CJEU C-393/16 of 20 December 2017, paragraphs 47 and 48. Bundesgerichtshof, case I ZR 268/14, judgment of 19 July 2018, paragraph 30 – Champagner Sorbet II.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.045

35 Presence or Absence of Coherence in Trade Identity Protection in the European Union Alexander von Mühlendahl1

Annette Kur has dominated like few others the academic approach not only to trade mark law but also to design law and related matters of protection of intellectual property rights, branching broadly into private international law and international procedural law, both as a prolific author and as teacher, far beyond her German origin. The comparative law approach, which was and is the hallmark of the Munich Max Planck Institute for Intellectual Property and Competition Law, now named Institute for Innovation and Competition – her intellectual “Heimat”, has always moderated the search for the best answer. When recognizing a variety of possible solutions, we should be modest in propagating our own solutions. As an example for this I should like to refer to the famous Study on the Overall Functioning of the European Trade Mark System commissioned by the European Commission and published by the Max Planck Institute in 2013 – Annette Kur was one of the principal authors of this Study, which is an example of her skill in blending profound academic analysis and expertise with wisdom in proposing legislative solutions. My contribution to this Festschrift is meant as a small tribute to what Annette Kur represents today at the summit of intellectual property law.

35.1 introduction “Trade identity”, a term originating in the United States of America, embraces all “signs” used in trade that may be protected as trade marks, trade names, names of publications, or geographical indications. “Coherence” means that the provisions governing the various rights, despite their respective differences, lead to reasonable or appropriate solutions within the relevant IP right as well as in relation to the

1

Dr. iur., J.D., LL.M.; Ministerialrat a.D., ret. Vice-President of the EUIPO. Visiting Professor, Queen Mary University of London. Attorney-at-Law, Munich, Germany.

Downloaded from https://www.cambridge.org/core. 416 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

Coherence in Trade Identity Protection in the EU

417

respective other IP rights. Some of these issues are the subject of this chapter, from the perspective of European Union (EU) law.

35.2 protection of trade identity rights at the level of eu law The creation of EU law in the field of trade mark protection started with the First Trade Mark Directive, which was adopted in December 1988.2 This Directive was “codified” without any substantive amendment in 2008. As part of the so-called “Trade Mark Package”, a new Directive was adopted in 2015, replacing the “First Directive”.3 The implementation period ran until January 2019, for the introduction of an administrative cancellation procedure until 2023. The Directive was negotiated in parallel with the Community Trade Mark Regulation, the ambitious project for the creation of a true EU-wide trade mark law. The Regulation did not come to fruition until December 1993.4 This Regulation has been amended repeatedly since 1993 and was also “codified” in 2008.5 As part of the “Trade Mark Package”, the Regulation was again amended with effect from 16 March 2016 and 1 October 2017.6 The previous “Community trade mark” is now called the “European Union trade mark”, abbreviated as “EU trade mark”, and the Office for Harmonization in the Internal Market (OHIM) is now called the European Union Intellectual Property Office (EUIPO). EU legislation for sui generis protection of geographical indications (GIs) as separate IP rights began in 1992 with the adoption of the Regulation on the protection of GIs for agricultural products and foodstuffs.7 The Regulation was amended repeatedly and codified in 2006 without significant changes. Finally, in 2012, the Regulations on the protection of GIs and the protection of food specialties were combined in a single Regulation.8 The provisions distinguish between 2

3

4

5

6

7

8

First Council Directive 89/104/EEC of 21 December 1988 to approximate the laws of the Member States relating to trade marks, OJEU 1989 L 40 p. 1, hereafter abbreviated as TMD. Directive (EU) 2015/2436 of the European Parliament and of the Council of 16 December 2015 to approximate the laws of the Member States relating to trade marks (Recast), OJEU 2015 L 336 p. 1. Council Regulation (EC) of 20 December 1993 on the Community trade mark, OJEU 1994 L 11 p. 1 Council Regulation (EC) No 207/2009 of 26 February 2009 on the Community trade mark (codified version), OJEU 2009 L 78 p. 1. Regulation (EU) 2017/1001 of the European Parliament and of the Council of 14 June 2017 on the European Union trade mark (codification), OJEU 2017 L 154 p. 1, hereafter abbreviated as EUTMR. Council Regulation (EEC) No 2081/92 of 14 July 1992 on the protection of geographical indications and designations of origin for agricultural products and foodstuffs, OJEU 1992 L 208 p. 1. Regulation (EU) No 1151/2012 of the European Parliament and of the Council of 21 November 2012 on quality schemes for agricultural products and foodstuffs, OJEU 2012 L 343 p. 1.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

418

Alexander von Mühlendahl

“protected designations of origin” (PDOs) und “protected geographical indications” (PGIs). The protection of wine designations in general and of wine GIs in particular was eventually aligned with the approach to the protection of GIs in the agricultural products and foodstuffs Regulation and incorporated in the rules governing the CMO for all agricultural products.9 Aromatised wine products now also have a comprehensive regime for the protection of GIs.10 However, the scheme is limited to “geographical indications” without the additional category of “designations of origin”. Similarly, the protection of geographical indications for spirit drinks has since been adapted to the model of the agricultural products and foodstuffs Regulation in general.11 For spirits too, the protection is limited to “geographical indications” without the additional category of “designations of origin”. The initial broadly based approach of the Commission for harmonizing the law of unfair competition was eventually reduced to Directives on misleading and comparative advertising and consumer-protection based unfair commercial practices. The Directive on misleading advertising, initially proposed in March 1978, was finally adopted in September 1984. The Directive was extended to include the regulation of comparative advertising in 1997 and codified in 2006.12 The Directive on Unfair Commercial Practices13 covers conduct that runs counter to the interests of consumers. However, in preventing behavior that may mislead the consumer about the commercial or geographical origin of goods or services, the interests of competitors and consumers coincide. There are two rules in the Directive that are relevant in the present context: Article 6(1)(b) and Article 6(2) (a), the first prohibiting commercial practices considered to be misleading if the practice “misleads or deceives the average consumer in relation to . . . the essential 9

10

11

12

13

Regulation (EU) No 1308/2013 of the European Parliament and of the Council of 17 December 2013 establishing a common organisation of the markets in agricultural products and repealing Council Regulations (EEC) No 922/72, (EEC) No 234/79, (EC) No 1037/2001, and (EC) No 1234/2007, OJEU 2013 L 347 p. 671, as amended. Regulation No 251/2014 of the European Parliament and of the Council of 26 February 2014 on the definition, description, presentation, labelling and the protection of geographical indications of aromatised wine products and repealing Council Regulation (EEC) No 1601/91, OJEU 2012 L 84 p. 14. Regulation (EC) No 110/2008 was recently replaced by Regulation (EU) 2019/787 of the European Parliament and of the Council of 17 April 2019 on the definition, description, presentation and labelling of spirit drinks, the use of the names of spirit drinks in the presentation and labelling of other foodstuffs, the protection of geographical indications for spirit drinks, the use of ethyl alcohol and distillates of agricultural origin in alcoholic beverages, and repealing Regulation (EC) No 110/2008, OJEU 2019 L 130 p. 1. Directive 2006/114/EC of the European Parliament and of the Council of 12 December 2006 concerning misleading and comparative advertising (codified version), OJEU 2006 L 376 p. 21. Directive 2005/29/EC of the European Parliament and of the Council of 11 May 2005 concerning unfair business-to-consumer commercial practices in the internal market and amending Council Directive 84/450/EEC, Directives 97/7/EC, 98/27/EC and 2002/65/EC of the European Parliament and of the Council and Regulation (EC) No 2006/2004 of the European Parliament and of the Council (“Unfair Commercial Practices Directive”), OJEU 2005 L 149 p. 22.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

Coherence in Trade Identity Protection in the EU

419

characteristics of the product such as . . . geographical or commercial origin”; and in the second case if, taking into account all factual circumstances, the commercial practice is capable of inducing an average consumer to make a business decision which he otherwise would not have taken, and which includes any form of marketing of a product, including comparative advertising, which “creates a likelihood of confusion with any products, trade marks, trade names or other distinguishing marks of a competitor”.

35.3 subject matter of protection – gaps and absences A “complete” and coherent protection system requires that all types of intellectual property rights exist (and, moreover, are coherently protected). From this point of view, EU law has significant gaps because it does not provide protection for all trade identity rights and because the protection is inherently not consistent.

35.3.1 Trade Mark Law Protection by Registration or Use EU trade mark law, like the Trade Mark Directive in its original version and also in its revised version of 2015, merely regulates trade mark protection through registration. The protection for trade marks on the basis of use, which in principle is available on an equal footing with register protection, or at least for “notorious” or well-known marks within the meaning of Article 6 bis of the Paris Convention, has remained unregulated. To prepare the proposals for the so-called “Trade Mark Package”, the Commission had commissioned an investigation of the European trade mark system. This study was carried out by the then Max Planck Institute for Intellectual Property and Competition Law.14 The Study included the proposal of Union-wide protection for unregistered notoriously known marks.15 This proposal was not adopted, leaving the system with a significant gap, which can only partially be closed by national legislation. The absence of EU regulation also means that there are absolutely no common requirements for the protection of trade marks acquired through use. The resulting internal inconsistencies between use-based and register-based rights should be addressed in future attempts at improving Union-wide protection for trade marks.

14

15

Study on the Overall Functioning of the European Trade Mark System (2011) (Max Planck Study), available online at the website of the Max Planck Institute. Max Planck Study, para. 2.140 bis 2.144, pp. 94–95.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

420

Alexander von Mühlendahl

Trade Mark Categories Trade marks in EU law include individual marks (for goods and services) and collective marks. Since October 2017 the EU Trade Mark Regulation has also included the category of certification marks.16 The 2015 Trade Mark Directive leaves it to the Member States to provide (maintain or introduce) the category of certification or guarantee mark in national law.17 However, as this is optional, an imbalance will remain unless all Member States avail themselves of this option.

35.3.2 Trade Names Union law does not contain any rules on the protection of trade names. This is true even though the Paris Convention, in its Article 8, requires protection of (foreign) trade names, which in any case cannot be made dependent on a registration of the name. In that regard too, the undertakings operating in the Union must rely on nonharmonized national laws. The unfair competition law protection against likelihood of confusion cannot be regarded as a true substitute.

35.3.3 Geographical Indications The EU legislature has provided extensive sui generis protection for GIs used for agricultural products (including spirits) and foodstuffs. Conditions of protection are recognition at national level and subsequently at EU level and entry of the PDO or PGI in a register. Despite broad agreement on protection requirements, scope of protection and limits in the respective applicable Regulations, there are notable differences in the procedure for the acquisition of protection and even in the definition of the exclusive right or the rights conferred. For example, EU law only provides protection for “names”, but not for any other indications or signs used to identify the geographical origin of goods. There is no EU protection for GIs actually used, but not registered in the EU system, even if they should comply with the requirements for protection under EU law. There is no protection at Union level for GIs which do not fulfill the conditions for protection as “DO” or “GI” (often referred to as “simple” indications of origin). There is no EU GI protection for indications of origin for handicraft or industrial products, although there can be no doubt about their economic importance. There is no EU protection for GIs used for services. For GIs which fall within the mentioned gaps of protection, the protection according to the national law of the Member States remains possible. However, this is not unconditional because the Court of Justice of the European Union (CJEU) has taken the view that GIs which are in themselves 16 17

Article 83 et seq. EUTMR. Article 28 TMD.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

Coherence in Trade Identity Protection in the EU

421

capable of protection under EU law, but which are not actually recognized under the system, can no longer be protected under national law.18 35.3.4 Unfair Competition EU law provides for adequate protection against misleading practices. Nevertheless, protection remains by way of Directives, so that protection must be sought under the national laws of the Member States, that is, under the provisions of the Directives transposed into national law.19

35.4 “internal” conflicts between trade identity rights of the same category 35.4.1 Trade Mark Law In EU trade mark law and also in the harmonized national trade mark laws, conflicts between earlier and later marks are always resolved according to the principle of priority: the earlier right displaces or prevails against the later right. That is true irrespective of whether the earlier right is a national trade mark or an EU trade mark. In the former case, the conflict is governed by national trade mark law under the Trade Mark Directive, in the second case by the EU Trade Mark Regulation. The principle of priority is weakened by a number of specific provisions, including laches and acquiescence, and the recognition of so-called “intervening rights”. Overall, these rules appear appropriate and balanced. 35.4.2 Geographical Indications A coincidence of GIs may arise if the same designation occurs several times. Examples are the many homonymous names of cities or areas in the same country, and mountains, rivers, etc. that span several countries (for example the Alps, Rhine, and Moselle). 18

19

CJEU judgment of 18.11.2003, C-216/01 – Budejovicky Budvar, n.p. v. Rudolf Ammersin GmbH (BUD). Directives leave transposition to Member States which may choose solutions which may seem to be coherent in their systems, but which do not “match” what other Member States do. “Minimum standard” Directives provide a common floor, but still leave differences when transposed. Thus, adopting directly applicable Regulations is clearly preferable, as was done, e.g., in 2011 in the field of food labelling, by Regulation (EU) No 1169/2011 of the European Parliament and of the Council of 25 October 2011 on the provision of food information to consumers, amending Regulations (EC) No 1924/2006 and (EC) No 1925/2006 of the European Parliament and of the Council, and repealing Commission Directive 87/250/ EEC, Council Directive 90/496/EEC, Commission Directive 1999/10/EC, Directive 2000/13/ EC of the European Parliament and of the Council, Commission Directives 2002/67/EC and 2008/5/EC and Commission Regulation (EC) No 608/2004, OJEU 2011 No L 304 p. 18.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

422

Alexander von Mühlendahl

As far as there are Union rules, namely in the field of agricultural products and foodstuffs, products of viticulture and spirits, the Regulations also always contain provisions allowing for the co-existence of homonymous names, provided they are properly differentiated.20

35.4.3 Unfair Competition Law Since unfair competition law does not give any “rights” but only, in the present context, means to prevent confusion and deception, there can be no coincidence of rights in the sense outlined.

35.5 “external” conflicts between trade identity rights of different categories 35.5.1 Trade Mark Law Earlier Trade Identity Rights vs. Later Trade Marks Trade identity rights other than trade marks are available and protected under the laws of the Member States. Under the EU Trade Mark Regulation, such earlier rights can be asserted against a later EU trade mark in opposition and cancellation proceedings if they confer on their proprietor the right to prohibit the use of a later mark and the scope of protection is of more than mere local significance.21 Under the Directive Member States have the option of providing for such opposition and cancellation rights.22 As regards enforcing such rights in infringement actions against the use of a (later) EU trade mark, Articles 137 and 138 EUTMR provide that such infringement rights remain available. Both EUTMR and the Directive provide that earlier EU GIs may be enforced in opposition or cancellation proceedings against later trade marks.23 The decisive factor is whether the use of the later mark could be prohibited. The same principle applies for earlier national GIs. These GIs also constitute “absolute” grounds for refusal.24 These provisions mirror the conflict rules in the GI Regulations on agricultural products and foodstuffs.25 This rule follows the principle of priority and appears appropriate. Under these provisions it is assumed that GIs can actually be infringed by the use of trade marks. Infringements rights based on GIs are

20 21 22 23 24 25

E.g., Article 6(3) Regulation 1151/2012. Article 8(4) EUTMR. Article 5(4) TMD. Article 8(6) EUTMR, Article 5(3)(c) TMD. Article 7(1)(j) EUTMR. Article 14 Regulation 1151/2012.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

Coherence in Trade Identity Protection in the EU

423

governed by the respective provisions in EU or national law providing for GI protection. Earlier Trade Marks vs. Later Trade Identity Rights In a coherent system one would expect that trade mark rights prevail over later, other trade identity rights, under the principle of priority. However, this is not the case under EU trade mark law. The reason for this is that the exclusive rights of a trade mark proprietor set out in Article 9 EUTMR and Article 10 of the 2015 Directive only apply if the contested sign is used “in relation to goods or services”. According to the case law of the CJEU, this is not the case when use is made of a trade name without being at the same time use in relation to goods or services.26 This view is further reinforced by the fact that Article 10(6) of the 2015 Directive allows Member States to provide that trade mark rights may be enforced against the use of a mark for purposes other than the distinction of goods or services. The CJEU has understood this provision to mean precisely the prohibition of the use of a company name that coincides with an earlier trade mark.27 Thus, the Court has overturned the principle, which has been consistently accepted by the Member States, that a trade mark may be enforced against a trade name or company name if there is a likelihood of confusion, just as, conversely, the holder of a trade name has a claim against the use of a later trade mark. At the time of the last amendment of the EUTMR and the Directive, the opportunity was missed to provide for such coherence. While the amended Regulation and the 2015 Directive recognize that a trade mark may give rights vis-à-vis a later trade or company name, the condition remains that the use must be made “in relation to goods or services”. The proposal made in the Max Planck Study to include in the Trade Mark Regulation a provision corresponding to Article 10(6) of the 2015 Directive in order to close this gap28 was not included in the legislative proposal. Earlier Trade Marks vs. Later GIs If an earlier trade mark coincides with a more recent GI, it would be expected that the earlier right prevails under Article 9 of the EU Trade Mark Regulation and Article 10 of the 2015 Directive, subject to the fair use exception in the Regulation and the Directive.29 However, this is not what EU law provides. The conflict rules are not found in the Regulation or the Directive, but in the EU Regulations governing the protection of GIs. All these Regulations contain rules for the coincidence of PDOs and PGIs and trade marks, interestingly regardless of whether the trade marks are protected on the 26 27 28 29

CJEU judgment 11.9.2007, C-17/06 – Celine SARL v. Celine. CJEU judgment 21.11.2002, C-23/01 – Robelco NV v. Robeco Groep BV. Max Planck Study, paras. 2.221, 2.222. at p. 113, and para. 2.229 at p. 116. Article 14 EUTMR, Article 14 TMD.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

424

Alexander von Mühlendahl

basis of use or on the basis of registration. An obvious gap exists because the Regulations do not specifically include other trade identity rights, such as trade names. Earlier trade marks do not preclude the protection by registration of a DO or GI, unless the earlier mark is a mark having reputation. Thus, Article 6(4) of Regulation 1151/2012 reads as follows: 4. A name proposed for registration as a designation of origin or geographical indication shall not be registered where, in the light of a trade mark’s reputation and renown and the length of time it has been used, registration of the name proposed as the designation of origin or geographical indication would be liable to mislead the consumer as to the true identity of the product.

Comparable provisions are contained in the other Regulations. While the criteria “reputation”, “renown” and “length of use” are indeed appropriate criteria, the criterion “misleading about the true identity” appears to have been included in the Regulation without knowledge of trade mark conflict criteria. The basic idea, however, is that an interference with a reputed trade mark will exclude protection at EU level. The existence of earlier rights is also a possible ground for opposition to the protection granted under Article 10(1)(c) of Regulation 1151/2012. A cancellation of a registered PDO or PGI due to its conflict with an earlier reputed mark is not provided for. Earlier “normal” trade marks, namely those which do not fall into the category of “reputed marks”, are also recognized, but not as precluding protection of DOs or GIs, but as providing a right to continue use – thus, the outcome is coexistence.30

35.6 concurrent application of different laws Coherence also includes the issue of whether, and under what conditions, protection can be claimed in parallel under different protection rules. This question arises both with regard to the parallel protection of marks under EU and national law and with regard to the parallel application of rules of different categories.

35.6.1. Trade Mark Law Double Protection – EU Trade Mark, National Mark The trade mark law of the EU is based, among others, on the principle of coexistence: EU trade marks and national trade marks coexist. This concerns not only the continued existence of national trade mark laws, but also the admission of double protection. There is a certain barrier to double protection in that parallel or 30

For details, see Article 14 Regulation 1151/2012.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

Coherence in Trade Identity Protection in the EU

425

successive infringement claims arising from identical or similar EU and national trade marks are subject to certain limits.31 No “Priority” or “Preemption” over Competition Law EU trade mark law does not recognize the principle of “primacy” of trade mark law, according to which marks which enjoy protection as marks can no longer be protected under other legislation, such as unfair competition law.32 It is expressly stipulated that the holder of an EU trade mark may also apply for any existing claims for infringement of unfair competition law or civil liability.33 This does not answer the question as to how the relationship between trade mark protection against likelihood of confusion and unfair competition law protection against the use of misleading indications is designed in detail. In the present context, when applying harmonized or non-harmonized unfair competition law of the Member States, the question is whether a claimant may invoke, in addition to the protection of the IP right in question, also the protection available under unfair competition law. While trade mark protection is designed to protect individual IP rights and their assertion depends on the proprietor, unfair competition law prohibiting a likelihood of confusion seeks the protection of consumers, and its application thus cannot depend on the decision of an individual proprietor. It is therefore not surprising that in the case law of the Member States, as far as a conflict has been recognized, it is accepted that the unfair competition law claims arising from the harmonized EU law are on an equal footing with the special statutory claims for trade mark infringement. However, the parallel application of unfair competition law and trade mark law is not unproblematic: unfair competition law does not recognize the principle of priority, and the limits for enforcing trade marks, such as acquiescence or exhaustion, are not elements of unfair competition law. Since unfair competition law cannot totally override the evaluations of the trade mark law, an accommodation must be sought leading to a “peaceful” coexistence of these areas of law. In order to achieve this, a “supplementation” of unfair competition law is required by 31 32

33

Article 136 EUTMR. German case law, until quite recently, considered trade mark law “lex specialis” and refused to apply unfair competition law when trade mark law would not provide a remedy; the leading case is the decision of the German Federal Court of Justice of 30.4.1988, GRUR 1999, 161 – MAC Dog. The Court reversed its position only after the adoption of the Unfair Commercial Practices Directive in its judgment of 15.8.2013, GRUR 2013, 1161 – Hard Rock Cafe. While the parallel assertion by the trade mark proprietor seems to be accepted, the CJEU considered that the limits of exclusive trade mark rights are exhaustively provided for in the TMD and the EUTMR and would exclude that a defendant can invoke acts by the trade mark proprietor amounting to unfair competition as a defense to an infringement action: CJEU judgment of 19.3.2013, C-661/11, GRUR Int. 2013, 1036 – “Martin Y Paz”. Annette Kur has presented a convincing counter-argument, specifically to the CJEU judgment, in her article Vorrangtheorie à la Luxemburg? Zu den Auswirkungen der Funktionsrechtsprechung und der Entscheidung Martin Y Paz/Gauquie,in C. Alexander et al. (eds.), Festschrift fu¨r Helmut Ko¨hler zum 70. Geburtstag (C. H. Beck, 2014), p. 383.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

426

Alexander von Mühlendahl

“importing” the trade mark law barriers into competition law when the issue of protecting trade identity rights is involved.34

35.6.2 Geographical Indications Double Protection – EU and National GIs Unlike under trade mark law, there is no longer any protection under national law for DOs and GIs protected under EU law: the national protection which existed until the granting of EU protection is completely “lost”.35 In addition, as stated earlier, the CJEU ruled under the agricultural products and foodstuffs Regulation that GIs which meet the requirements of protection under EU law but which are not protected, for example because no application for protection has been made, also do not enjoy protection under national law – the EU law has “preempted” national protection.36 Still, protection as a “simple” GI may be claimed,37 and it is also assumed that the prohibitions against the use of misleading indications continue to apply. Double Protection – GIs and Trade Marks There is no provision in EU trade mark law or GI law which would exclude the protection of geographical indications as a trade mark – EU or national mark. However, inherent grounds preclude trade mark protection insofar as protection for the GI as such should be sought. GIs are by definition descriptive of goods from a specified place or region and therefore also lack distinctive character. However, the protection of geographical indications as collective marks is possible both in EU trade mark law and in national legal systems. Thus, if the protection of an EU-protected DO or GI as an EU or national mark were to be sought – be it as inherently protectable due to the addition of graphic elements, be it as a collective mark – the question would arise as to whether double protection should be excluded for systematic reasons. Case law is not yet available. In practice, however, there are very many registrations of collective marks in the EU Register consisting of PDOs or PGIs as such.

35.7 results The legal situation in the EU, with competences regarding the protection of intellectual property shared between the Union and its Member States, presents 34

35 36 37

This was successfully done by the German Federal Court of Justice in the Hard Rock Cafe decision, supra n. 32, with regard to acquiescence. Article 9 Regulation 1151/2012. See above at n. 18. See CJEU judgment of 7.11.2000, C-312/98 – Schutzverband gegen Unwesen in der Wirtschaft e.V. v. Warsteiner Brauerei Hans Cramer GmbH & Co. KG (Warsteiner).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

Coherence in Trade Identity Protection in the EU

427

an example of a complex system of mutually exclusive or overlapping regulations at EU and at Member State level. In the field of protection of trade identity rights, such as trade marks, trade names, and geographical indications, the existing “patchwork” presents difficulties and problems, such as double protection at the EU and Member State levels, which are largely unknown to other countries. The chosen rules are often incoherent and appear quite arbitrary. An overall concept of protection of trade identity rights under the relevant legislation is not apparent. However, as such coherence is often lacking at national level as well, we should not be too surprised.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.046

36 Virtue Ethics and Private Law – A Sketch Marcus Norrgård1

36.1 introduction I came to know Annette in the late 1990s when I was a young doctoral student at the University of Helsinki. Annette gave at the Hanken School of Economics in Helsinki a course on free movement in the EU. Finland joined the EU in 1995 and during my law studies at the University of Helsinki the role of EU law was almost negligible. Annette’s course was very useful and gave a solid basis for further studies. Since then our paths have crossed several times. I spent, together with my wife Petra, the spring of 2001 at the Max Planck Institute in Munich as a Stipendiat. Petra was a research assistant working on Annette’s UDRP project. A third pivotal moment was when I participated in Annette’s research project on international jurisdiction. For me, Annette has been the person who has opened the doors to continental Europe. I am very thankful for her support. I have tried not to close them. In this short piece I would like to look into transition and coherence from the point of view of virtue ethics. I believe that we need to transition from law dealing only with negatives (i.e. ‘though shalt not’) to positives (i.e. ‘do like this’). Further, I think that fragmentation in private law should be looked into. Cross-linking different areas of private law helps against the ever-increasing fragmentation of private law. A virtue ethical approach may help in de-fragmenting private law to a certain extent.

36.2 virtue ethics There are, simply put, three branches of ethical theories. Consequentialism looks at the consequences of decisions to find out whether a decision is ethical or not, 1

Professor of Law, University of Helsinki. The author and Professor Petra Sund-Norrgård are running a joint project on Virtue Ethics in Private Law, which covers not only intellectual property law, but also, for example, contract, labor and family law.

Downloaded from https://www.cambridge.org/core. 428 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.047

Virtue Ethics and Private Law – A Sketch

429

whereas deontology focuses on rules and duties (e.g. Kant). Consequentialism is common in intellectual property law especially in the form of law and economics. This short piece is not about consequentialism or deontology, but about the third branch of ethical theories, virtue ethics, which has been the focus of legal analysis to a much lesser degree. Virtue ethics has its roots especially in Aristotle’s thinking, but also in other later schools of thought, and it re-emerged (at least in the Anglo-American world) only in the 1950s.2 It looks into the character of people, trying to ascertain how a ‘virtuous’ person would act. Virtue ethics focuses more on the character of the actors than on their acts, thus being agent centered instead of act centered. Zagzebski is perhaps the philosopher who has taken this focus on the actor the furthest. She calls her theory an ‘exemplarist’ moral theory.3 Her moral theory builds upon the idea of an ‘exemplar’, a supremely admirable person who people want to emulate. She identifies the hero (dominated by the virtue courage), the sage (dominated by the virtue wisdom), and the saint (dominated by the virtue charity) as typical exemplars. There are many different taxonomies of virtues, but, for example, in positive psychology Peterson and Seligman have distinguished between six different main classes of character traits (i.e. virtues): (1) wisdom and knowledge; (2) courage; (3) humanity (which includes things such as love, friendship and compassion); (4) justice; (5) temperance; and (6) transcendence (which includes, for example, gratitude, hope and humor).4 It is not, however, necessary or perhaps even possible to enumerate all virtues. Sandler argues for a number of environmental virtues, such as virtues of sustainability, which shows that virtues need not only relate to the relation between people.5 According to classical virtue ethics (Aristotle), the aim of a virtuous life is ‘eudaimonia’, which translates as happiness, excellence or human flourishing. In modern virtue ethics this purpose (telos) of human existence has not always been accepted. Here, for the sake of simplicity, it is, however, assumed that the telos is eudaimonia in some sense of the word. It is also assumed that being virtuous is a practice much like playing an instrument; the more practice, the better the actor becomes.6 Practicing virtues always takes place in a community, which can be different for different actors and for different practices.7 A businessperson practicing a certain set of virtues that are relevant for doing business would do so in the

2 3 4

5

6

7

G. E. M. Anscombe, Modern Moral Philosophy, Philosophy 33, No. 124 (1958). Linda Zagzebski, Exemplarist Moral Theory (Oxford University Press, 2017). Christopher Peterson and Martin E. P. Seligman, Character Strengths and Virtues: A Handbook and Classification (Oxford University Press, 2004). Ronald L. Sandler, Character and Environment: A Virtue-Oriented Approach to Environmental Ethics (Columbia University Press, 2007). Alasdair MacIntyre, After Virtue: A Study in Moral Theory (University of Notre Dame Press, 2007). Ibid.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.047

430

Marcus Norrgård

community of business people. The set of virtues relevant for doing business might not be at all relevant in another community. In a situation of an ethical dilemma, the primary question in virtue ethics is not what the consequences of the decision are, or whether, in itself, it is considered right or wrong to act in a certain way. It is not about making the ‘right’ decision out of mere duty or because a certain rule prescribes a certain type of conduct. Instead, the question to be answered revolves around virtuous behavior and the character of the actor. One common way is to ask what the virtuous person would do in the context at hand: an action is right if, and only if, it is what a virtuous actor would characteristically do in the circumstances.8

36.3 virtue jurisprudence Legal analysis based on virtue ethics has been called ‘virtue jurisprudence’.9 Apart from purely legal theoretical analysis,10 virtue jurisprudence has found its way into public law and, to a lesser degree, into private law. What is striking is that mainly American law has been the focus of research. One strand of enquiry has been a virtue-based theory of judging, namely a study on the virtues a judge should hold, such as judicial impartiality and judicial integrity.11 Michelon has especially focused on a judge’s practical wisdom (phronesis, prudentia) in his analysis.12 Ralli has studied the intellectual virtues of the judge from a purer Aristotelian point of view.13 Van Domselaar has introduced the neoAristotelian concept of civic friendship into the discussion on adjudication as a response to Solum.14 Van Domselaar has also discussed ‘judicial perception’ as an ethical skill for judges.15

8 9

10

11

12

13

14

15

Rosalind Hursthouse, On Virtue Ethics (Oxford University Press, 1999). Lawrence B. Solum, Virtue Jurisprudence: A Virtue-Centered Theory of Judging, 34 (1/2) Metaphilosophy 178–213 (2003). See, for example, Amalia Amaya, Exemplarism and Judicial Virtue, 25(3) Law and Literature (2013). Available at SSRN: https://ssrn.com/abstract=2489879 or http://dx.doi.org/10.2139/ssrn .2489879. See also Sandrine Berges, Education and Paternalism: Plato on Virtue and the Law, in: Amalia Amaya and Ho Hock Lai (eds.), Law, Virtue and Justice (Hart Publishing, 2013). See also Linghao Wao and Lawrence B. Solum, Confucian Virtue Jurisprudence, in: Amaya and Lai (eds.) (ibid). Solum (supra n. 9); Søren Stig Andersen, Justice and Other Juridical Virtues, 34(3/134) Retfærd: Nordisk Juridisk Tidsskrift 85–98 (2011). Claudio Michelon, Practical Wisdom in Legal Decision-Making, in: Amaya and Lai (eds.) (supra n. 10); Wao and Solum (supra n. 10). Tommi Ralli, Intellectual Excellences of the Judge, in: Liesbeth Huppes-Cluysenaer and Nuno M. M. S. Coelho (eds.), Aristotle and the Philosophy of Law: Theory, Practice and Justice (Springer, 2013). Iris van Domselaar, Moral Quality in Adjudication: On Judicial Virtues and Civic Friendship, 1 Netherlands Journal of Legal Philosophy 24–46 (2015). Iris van Domselaar, The Perceptive Judge, 9(1) Jurisprudence 71–87 (2018).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.047

Virtue Ethics and Private Law – A Sketch

431

Virtue ethics has also been used in the field of legal ethics in trying to ascertain how a ‘good’ lawyer should behave.16 Kaptein has especially focused on client confidentiality.17 Virtue ethics has also found a place in American constitutional theory.18 Solum is even talking about an ‘aretaic turn’ (i.e. a virtue-ethical turn) in constitutional law.19 A theory of virtue legislation or aretaic legislation has been studied, where such legislation would take human flourishing as the proper end of the law.20 Clark looks at public virtues and law-making.21 Criminal law has been the subject of virtue theoretical analysis,22 and environmental law23 and public international law24 have been studied as well. Private law, however, has to a much lesser degree been the focus of virtue jurisprudence. There is some research on contract,25 property26 and tort law.27 In

16

17

18

19

20

21

22

23

24

25

26

27

See, e.g., Michael S. McGinnis, Virtue Ethics, Earnestness, and the Deciding Lawyer: Human Flourishing in a Legal Community, 87 North Dakota Law Review 19 (2011). Hendrik Kaptein, Virtues of Truthfulness in Forbearing Wrongs: Client Confidentiality Qualified by Legal Symmetry of Past and Future Harm, in: Amaya and Lai (eds.) (supra n. 10). L. J. Strang, Originalism and the Aristotelian Tradition: Virtue’s Home in Originalism, 80 Fordham Law Review 1997 (2012). Lawrence B. Solum, The Aretaic Turn in Constitutional Theory, 70 Brook. L. Rev. 475 (2004– 2005). Donald J. Kochan, The Mask of Virtue: Theories of Aretaic Legislation in a Public Choice Perspective, 58 St. Louis U. L. J. 295 (2014). Lawrence B. Solum, Virtue as the End of Law, 9 Jurisprudence 6–18 (2018). Available at SSRN: https://ssrn.com/abstract=3165570. Sherman J. Clark, Neoclassical Public Virtues: Towards an Aretaic Theory of Law-Making (and Law Teaching), in: Amaya and Lai (eds.) (supra n. 10). Katrina L. Sifferd, Virtue Ethics and Criminal Punishment, in: Alberto Masala and Jonathan Webber (eds.), From Personality to Virtue: Essays on the Philosophy of Character (Oxford University Press, 2016), 35–61; Jonathan Jacobs, Character, Punishment, and the Liberal Order, in: Masala and Webber (eds.) (ibid.), 9–34; Kyron Huigens, Motivating Intentions, Reciprocal Specification of Ends and the Assessment of Responsibility, in: Amaya and Lai (eds.) (supra n. 10), 155–168. See also Ekow N. Yankah, Liberal Virtue, in: Amaya and Lai (eds.) (ibid.), 169–194; and Ekow N. Yankah, Legal Vices and Civic Virtue: Vice Crimes, Republicanism and the Corruption of Lawfulness, in: Huppes-Cluysenaer and Coelho (eds.) (supra n. 13), 203–222. Anonymous, Uncommon Goods: On Environmental Virtues and Voluntary Carbon Offsets, 123 Harvard Law Review 2065 (2010). Jamie Gaskarth, The Virtues in International Society, 18(3) European Journal of International Relations 431–453 (2011). Avery Katz, Virtue Ethics and Efficient Breach, 45 Suffolk U. L. Rev. 777 (2012). Chapin F. Cimino, Virtue and Contract Law, 88 Or. L. Rev. 703 (2010). Gregory S. Alexander, The Social-Obligation Norm in American Property Law, 94 Cornell L. Rev. 745 (2009). David Lametti, The Objects of Virtue, in: Gregory S. Alexander and Eduardo M. Peñalver (eds.), Property and Community (Oxford University Press, 2010), 1–37. Available at SSRN: https://ssrn.com/abstract=1758859. Eduardo M. Peñalver, Land Virtues, 94 Cornell L. Rev. 821(2009). Eric R. Claeys, Virtue and Rights in American Property Law, 94 Cornell L. Rev. 889, 917 (2009). Heidi Li Feldman, Prudence, Benevolence, and Negligence: Virtue Ethics and Tort Law, 74 Chi.-Kent L. Rev. 1431, 1435 (2000).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.047

432

Marcus Norrgård

corporate law, for example, Colombo has applied virtue ethics.28 In copyright law, Lametti has proposed that copyright holders should act ethically, specifically virtue ethically, in their use of so-called digital locks in Canadian copyright law.29 His argument is based on the fact that he sees the Canadian Copyright Bill C-32 as ‘seriously flawed’ and only reasonable copyright holders might ‘save’ it. Opderbeck has studied intellectual property in biotechnology from a virtue perspective, with focus on virtues in open science and open source production.30 He identifies beneficence, fidelity to trust, justice, and practical wisdom as foundational virtues in that setting. Generally, it can be said that virtue ethics has been gaining traction, in private law also, including intellectual property law. One obvious obstacle is how to interlink virtue ethics, with its focus on excellence in human behavior, and legal norms, which at their core lay down ‘mere’ minimum standards of legal conduct or otherwise focus on how ‘non-excellent’ a person is allowed to be. Whereas virtue ethics focuses quite clearly on positive human qualities (such as generosity and honesty), law focuses generally on the negative aspects of human behavior. Ethical principles and legal norms seem thus to work on different levels. There are however apparent connections between virtue ethics and private law, which I will briefly touch upon below. At least three levels of questions can be identified. The first level is the substantive conflict-solving rule itself, for example whether there is an infringement or not. The second level is the question of how the actors are allowed to go about solving their conflict. It is clear that virtue ethics does play a part in the ‘how’; a rightsholder (and also an alleged infringer) can be more or less aggressive, rely on ‘dirty tactics’ in trying to ‘win’ the case, etc. The third level is the evidentiary level, where the character of a witness or a party is praised or comes under attack. Neither the ‘how’ nor the evidentiary side of things is here of primary interest. Rather, the fundamental question is whether virtue ethics can play a part in resolving a private law conflict, for example in intellectual property law or unfair competition law. Not all parts of private law can accommodate virtue ethics to the same degree. Private law rules which need to be very clear-cut ex ante and that do not take into account the ethical dimension of the actor’s conduct have less room for virtue ethics. Statutes of limitation would be a prime example of such rules: after a given 28

29

30

Ronald J. Colombo, Toward a Nexus of Virtue, Hofstra University Legal Studies Research Paper No. 11-17 (2011). Available at SSRN: https://ssrn.com/abstract=1908889 or http://dx.doi .org/10.2139/ssrn.1908889. David Lametti, How Virtue Ethics Might Help Erase C-32’s Conceptual Incoherence, in: M. Geist (ed.), From “Radical Extremism” to “Balanced Copyright”: Canadian Copyright and the Digital Agenda (Irwin Law, 2010), 309–340. Available at SSRN: https://ssrn.com/abstract=1758883. David W. Opderbeck, A Virtue-Centered Approach to the Biotechnology Commons (or, the Virtuous Penguin), 59 Me. L. Rev. 315 (2007).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.047

Virtue Ethics and Private Law – A Sketch

433

time limit, the obligation ceases to exist. However, the more the rule in question allows the context of the individual situation to be taken into account, including the ethical dimension, the greater the room for ethical analysis, including virtue ethics. For example, Article 14(1)(c) of the EU Trademark Regulation (EUTMR, 2017/ 1001) provides the right to use a trademark in order to indicate the intended purpose of a product or service, in particular as accessories or spare parts. However, this use needs to be in accordance with ‘honest practices in industrial and commercial matters’ (Article 14(2) of EUTMR). Traditionally this has been analyzed by asking what such honest practices might be or whether a certain activity in a certain setting is to be seen as an ‘honest practice’. A virtue ethical approach would instead focus on the character traits of the market participants and ask what an exemplary (or admirable) business would do in the context at hand. Another example is unfair competition law. The foundation of Finnish unfair competition law lies in § 1 of the Unfair Business Practices Act (1061/1978), which states that ‘good business practice may not be violated nor may practices that are otherwise unfair to other entrepreneurs be used in business’. The content of ‘good business practice’ and ‘otherwise unfair’ are closely tied to the conduct of the market participants. Although different types of conduct, such as slavish imitation and misuse of goodwill, have over time been identified as activities typically being within the scope of § 1 and where, at least in principle, rather clear-cut rules have been developed, § 1 still allows ample room for taking into account the context of the case, including the degree to which the market participants have conducted themselves in an ethical manner. Prima facie, without any knowledge of the specifics of a case, I think that it is not that controversial to argue in a case of unfair business practices that an honest and moderate market participant who displays benevolence and justice towards the other party would generally have the upper hand against an opposing party who has acted dishonestly and immoderately, and has displayed malevolence.

36.4 virtuous businessperson The bridge between virtue ethics and legal analysis lies, I would argue, at least in part in exemplars. As already highlighted above, exemplars such as the sage, the hero or the saint are admirable persons who other people want to imitate. In legal analysis, the exemplars necessarily work in a slightly different fashion than in ethics. Law, at least traditionally, seeks to find a proper balance between competing interests or persons. Law decides where to draw the line, whereas virtue ethics is about showing how to strive for excellence or human flourishing. Nevertheless, by modifying exemplars into something that is not as ‘excellent’ creates something that we already recognize in private law. In tort law, the good family man (bonus pater familias) is (still) used as a standard when ascertaining culpability, the question being how the bonus pater familias would have acted in the Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.047

434

Marcus Norrgård

situation at hand. Patent law recognizes the ‘person skilled in the art’, trademark law the ‘average consumer’ and design law the ‘informed user’. I call this type of legal standard referring to a (fictitious) person an ‘archetypal’ legal standard or technique. Legislating using archetypes is thus not unheard of in private law, although they are not exemplars in the virtue ethical sense. I would argue that it is possible to go a step further towards the virtue ethical approach and analyze legal problems applying exemplar thinking. This would however require that the fictitious person is introduced in legal interpretation without the express support of statute law. Instead of focusing on the practices, namely conduct or acts, a virtue ethical legal analysis would focus on the person engaging in the commercial practices, and the question would be what a certain type of person would have done in a certain situation. This person can be more or less virtuous, but if the question relates to ‘honest’ commercial practices, the person necessarily needs to possess some character traits that generally would be viewed as ‘good’ (or admirable or worthy of emulation/imitation). Businesspersons are normally seen as (more or less) rational actors aiming at profit maximization. Profit maximization, at least in the sense of maximizing the actor’s own profit, is not, however, generally viewed as a virtue. There are philosophies that emphasize self-interest, but in traditional virtue ethical thinking greed, for example, has been seen as a vice. Focus on profit maximization (and self-interest in general) does not seem to leave much room for virtues such as generosity (for example, in the form of giving money to someone without having a legal obligation to do so), compassion (in the sense of being aware of someone’s suffering and wishing to relieve that suffering) or temperance (i.e. moderation, for example asking for less than the law gives the right to) unless these virtues help increase profit. In other words, profit maximization becomes the primary goal and everything else becomes secondary to that. A traditional way of reining in the drive to maximize profit has been legislation and self-regulation. If, however, the law accepts that companies are allowed to maximize profits and the only opposing force is mandatory legislation, then law becomes only a tool that draws the line between self-interested parties or sets a bare minimum standard of conduct. This is, basically, the situation today. There are, however, businesses known for their virtuous activities. The Religious Society of Friends, colloquially known as the Quakers, introduced fixed prices (and refused to haggle) as early as the seventeenth century, because they felt – for religious moral reasons – that everyone was equal and therefore the same price should apply to everyone. Fixed prices were also a way to combat greed and dishonesty. Thanks to fixed prices parents let their children go to Quaker shops without having to fear the children being ripped off. The honesty of the Quakers led to success in business, although this was not their primary motivation. Businesses in general cannot, however, be relied upon to change their motivation without external help, for example in the form of the law. In the best of worlds, the Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.047

Virtue Ethics and Private Law – A Sketch

435

law could create incentives for positive behavior by promoting good behavior instead of, as now, only disincentivizing negative behavior. Instead of saying what cannot be done, the law could tell us how to achieve excellence in the field of ethical behavior. This kind of utopian idea can seem quite delusional for anyone dealing with legal analysis and not with business ethics. I think, however, we all intuitively know of businesses that are able to balance making a profit with other important values (i.e. virtues), such as honesty, moderation and compassion. Some of these businesses have realized that finding a proper balance is in the long term better for profits, but others emphasize different values because it is the right thing to do regardless of the effect on the bottom line. What if law could take the genuinely good businesses and make them into exemplars and say, ‘this is the way you should act if you want to conduct yourself in a truly honest way’? I think it is possible. Especially in legal contexts that ask for honesty and loyalty or require taking into account weaker parties or third-party interests, a virtue ethical approach would be a viable option. Such an approach would shift the focus from the acts to the character of the actors, which in itself may create impulses for new ideas. Asking what the ‘virtuous businessperson’ would do could potentially allow for a broader picture to be taken into account by the decision maker with ensuing richer argumentation. The next step in our research project is to operationalize the virtue ethical approach. That, however, I must leave for another article.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.047

37 Closing the Gap: How EU Law Constrains National Rules Against Imitation Martin Husovec1

37.1 introduction If imitation was simply bad, society could just proclaim it unfair. But it is not. Imitation is a lifeblood of economy and culture.2 It fuels progress because it facilitates diffusion of new ideas. As much as innovators push the boundary of our knowledge, imitators help to disseminate it. Intellectual property rights are a tool for achieving balance between innovators and imitators. They try to motivate innovators to come up with new ideas and commercialize them, while keeping imitators at bay for a while. This artificial lead time allows them to recoup R&D investments so that they are economically motivated to innovate again. Intellectual property rights are designed by legislators in a highly political process. Although contours of these choices are still left to the courts to determine, in particular as new technological uses develop, the most fundamental trade-offs are undoubtedly made directly by the legislator. Next to these exclusive rights, many countries regulate imitation also by means of judge-made doctrines against misappropriation.3 These are based on national tort law, or unfair competition law, and equally constrain imitation. It is often said that unlike IP rights which regulate immaterial objects, this supplementary protection against imitation relies on case-bycase interventions against conduct. Very often, they are anchored in the framework 1

2

3

Assistant Professor, Department of Law, London School of Economics and Political Science (LSE). The author would like to thank Graeme Dinwoodie, Lionel Bently and Ansgar Ohly for their careful comments on the earlier version of this article. All omissions and mistakes are the author’s own. Cf. the US Supreme Court decision in Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (‘From their inception, the federal patent laws have embodied a careful balance between the need to promote innovation and the recognition that imitation and refinement through imitation are both necessary to invention itself and the very lifeblood of a competitive economy.’). In some countries, such as Germany, these doctrines were later codified to some extent in the law.

Downloaded from https://www.cambridge.org/core. 436 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

How EU Law Constrains Rules Against Imitation

437

of fairness of commercial practices. These doctrines differ considerably across the EU.4 Some countries allow judges to supplement the judgment of the legislator by means of ad hoc fairness considerations. Others find the idea of doing so irresponsible and consider either the legislator to be the only authority that should make such choices, or the very idea of abstract fairness to be inherently incapable of any predictable adjudication.5 The goal of this short essay is not to discuss which regulatory approach to imitation is superior. Annette Kur dedicated some of her finest work to the study of this issue.6 Instead, my goal is to highlight the European dimension of the debate which appears neglected. This dimension points towards a much more constrained regulatory environment for national policy makers. Annette Kur has been one of the few scholars systematically reminding us of this important issue for over a decade;7 her work in the area underscores a consistent feature of her scholarship – a rigorous study of fundamentals that matter. The cornerstone of the European project has always been its mechanism of free movement, embodying the idea that unnecessary barriers to cross-border trade should be avoided in order to build a brighter future. In the early days, intellectual property rights often played a role of justification of local idiosyncrasies erecting barriers to intra-EU trade. This is also why so much effort was spent on approximating many of the aspects of intellectual property rights in the past two decades. At the same time, local doctrines of misappropriation grounded in unfair competition and tort laws were left largely non-harmonized. However, some of them were incorporated (in form or substance) into newly formalized Union legislation on IP rights. Therefore today, the Union’s constraints on the national rules against imitation are twofold: (1) constraints imposed by formalized IP protection which incorporated national misappropriation doctrines and therefore pre-empts them;

4

5

6

7

Rogier de Vrey, Towards a European Unfair Competition Law: A Clash Between Legal Families (Brill, 2005); Ansgar Ohly, The Freedom of Imitation and Its Limits – A European Perspective (2009) 41 IIC, pp. 506–524. L’Oréal v. Bellure [2007] EWCA Civ 968 at paras. 139, 140 per Jacob LJ (‘What one man calls “unfair” another calls “fair” ... I think there are real difficulties in formulating a clear and rational line between that which is fair and that which is not, once one goes outside the requirement of no deception.’). See Annette Kur, What to Protect, and How? Unfair Competition, Intellectual Property, or Protection Sui Generis, in Nari Lee et al. (eds.), Intellectual Property, Unfair Competition and Publicity: Convergences and Development (Edward Elgar, 2014), pp. 11–32. Annette Kur, (No) Freedom to Copy? Protection of Technical Features under Unfair Competition Law, in Wolrad Prinz zu Waldeck und Pyrmont et al. (eds.), Patents and Technological Progress in a Globalized World (Springer, 2009), p. 521 ff; Annette Kur, Nachahmungsschutz und Freiheit des Warenverkehrs – der wettbewerbsrechtliche Leistungsschutz aus der Perspektive des Gemeinschaftsrechts, in Hans-Jürgen Ahrens et al. (eds.), Festschrift fu¨r Eike Ullmann (Juris, 2006), pp. 717–735.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

438

Martin Husovec

and (2) constraints imposed beyond these situations by the back-stop mechanism of free movement.

37.2 constraining by legislation: secondary eu law pre-emption The EU legislator can choose from different approaches when approximating existing national rights. It can harmonize without leaving any room to deviate to the Member States (full harmonization) or harmonize but leave some room for stricter or more detailed national provisions (minimum harmonization). In both situations, the secondary EU law consequently pre-empts national rules. In the former case, it pre-empts any deviation; in the latter case, deviation which provides less protection or contradicts this minimum. Once pre-emption applies, the principle of primacy requires that Union law takes precedence in case of conflict.8 Such conflicting national rules then have to be repealed by the Member States due to the principle of loyal cooperation,9 and potentially set aside by courts due to the doctrine of direct effect.10 To clarify which EU rules pre-empt which national rules, it is necessary to conduct a two-prong analysis: (1) it has to be determined whether national and EU rules overlap in scope (scope of applicability); and (2) if they do, it has to be examined whether the EU legislator intended to exclude national rules such as those at stake.11 As explained by Arena, in practice, the Court of Justice of the European Union (CJEU) can find pre-emption because of: (A) an actual conflict due to wording and/or frustration of goals of harmonization (conflict pre-emption); or alternatively, (B) an intention of the EU legislator to displace any national regulation in the area (field pre-emption).12 37.2.1 Conflict Pre-emption Today’s trademark law doctrines bringing referential uses within the scope of potential infringement, or anti-dilution protection used to be part of the unfair competition area of the toolkit in some countries.13 Since secondary EU law, as interpreted by the CJEU, legislated protection against dilution and a broad set of 8 9 10 11

12 13

Flaminio Costa v. ENEL, C-6/64. Commission v. Germany, C-74/86, para. 11. Simmenthal, C-106/77, para. 17. Amedeo Arena, Exercise of EU Competences and Pre-emption of Member States’ Powers in the Internal and the External Sphere: Towards ‘Grand Unification’? (2016) 35(1) Yearbook of European Law, pp. 28–105. Ibid., Section B. See Graeme B. Dinwoodie, Dilution as Unfair Competition: European Echoes, in Rochelle C. Dreyfuss and Jane C. Ginsburg (eds.), Intellectual Property at the Edge: The Contested Contours of IP (Cambridge University Press, 2014), pp. 81–102; Martin

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

How EU Law Constrains Rules Against Imitation

439

uses of signs directly into trademark law, the underlying causes of action in the national law, whatever their origin and labels, automatically became substance regulated by EU law too. Due to the principle of primacy, the non-harmonized body of unfair competition law now has to yield to the harmonized EU trademark law as far as they overlap. As a consequence, Directive 2015/2436 pre-empts national unfair competition or tort law offering more rights for the same reasons.14 For instance, the CJEU’s keyword advertising case law cannot be circumvented by strategic use of national unfair competition law to provide more protection.15 The harmonization, after all, affects the entire legal order,16 not just pieces of national law that are identically titled. Pre-emption can also be less straightforward to apply. For instance, harmonization of exclusive rights in copyright law by means of Directive 2001/29 only touches upon some of the economic rights. Therefore, in theory, the national parliaments can legislate additional rights for the benefit of authors. For instance, Czechia, Germany and Slovakia all recognize innominate exclusive rights – further unspecified uses of protected subject matter which can be potentially developed by courts.17 Such innominate rights can in essence constitute protection against misappropriation. Similar rights could be furnished by courts by means of unfair competition law in some countries.18 Following the logic of incomplete EU harmonization, the tentative conclusion would appear to be that since the secondary EU copyright law does not harmonize all exclusive rights, it should not limit such doctrines (whether rooted in innominate exclusive rights or unfair competition law) either. The CJEU, however, also protects against conflicts with harmonization objectives.19 In Svensson, the Court held that Swedish law may not extend national

14

15

16 17

18

19

Husovec, Trademark Use Doctrine in the European Union and Japan (2017) 21 Marq. Intell. Prop. L. Rev. pp. 1–30. Dinwoodie (ibid.), p. 92 (noting that the scope of EU intrusion into unfair competition via dilution has turned out to be greater than would have been evident on the face of Article 5(2)). Opposite view on pre-emption in passing-off context pointing to Recital 40 of Directive 2015/ 2436: see de Vrey (supra n. 4), p. 290. See for the criticism of the Slovak and Czech practice: Martin Husovec, Koncept parazitujúcej nekalej sút’aže, in Marian Kropaj (ed.), Sub Specie Aeternitatis: Zˇivot a dielo profesora Ja´na Sˇvidrona (Faber, 2015). Pfeiffer and Others, C-397/01, para. 115. See Section 12(5) of the Czech Copyright Act; Section 15(1) of the German Copyright Act; Section 19 of the Slovak Copyright Act. The infringement of such innominate exclusive rights and of unfair competition protection against misappropriation might differ in their consequences. First, the extent to which they are protected as a form of ‘property’ by Article 17(2) of the EU Charter might differ. Second, the extent to which other pieces of EU law apply to the protection against misappropriation might differ (e.g. the Enforcement Directive). Third, while exclusive rights might create a presumption of harm, unfair competition law traditionally requires proof of harm every time the plaintiff asserts it. It is worth drawing a parallel here with the US law, which has extensive experience with the doctrine of pre-emption in conflicts between federal intellectual property laws and national laws, including unfair competition law. See Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

440

Martin Husovec

protection against hyperlinking that is rendered permissible given the scope of communication to the public right. Moreover, the CJEU talks about ‘precluding a Member State from giving wider protection to copyright holders’.20 This is a much stronger statement than for instance Football Dataco’s wording of pre-emption, which speaks of ‘[precluding] national legislation which grants databases, as defined in Article 1(2) of the directive, copyright protection under conditions which are different to those set out in [the Directive]’.21 The Svensson holding would seem to equally affect any law willing to protect for the same reason. The CJEU argues that additional national protection would undermine the effectiveness of the common concepts.22 If the effectiveness of the Union concepts can be undermined by national deviations, then those concepts of secondary law will pre-empt such deviations. Even though the Football Dataco wording of pre-emption is much narrower, due to the Svensson arguments, its effects could be more horizontal too. Pre-emption extends also to optional elements. They are binding once the national legislator has decided to exercise the choice in that particular area. One illustration of this is the concept of dilution which used to be optional for national trademarks under Directive 89/104. However, as clarified by the CJEU, the contours of protection do not allow for deviation once a Member State has decided to exercise the choice of having such protection.23 In other words, despite optionality, the national legislator has only two choices: provide no extra protection, or provide it in the scope envisaged by the EU law. And because of the arguments noted in the previous paragraph, it is immaterial if the legislator would exercise such choices by means of trademark law, unfair competition, or other field of law.

37.2.2 Field Pre-emption Interpreting the EU legislator’s intentions can be complicated, especially when trying to determine if the EU law is meant to displace any national laws on the topic.24 For intellectual property rights, the explicit goals of rules are very rarely about providing an undisturbed positive space for imitation. On the contrary, they usually concern the narrowly defined reasons why imitation needs to be restricted in a particular case. For imitation, the answer is often only partial (e.g. ‘free as far as

20 21 22

23 24

U.S. 141 (1989) (holding a state anti-plug molding law partially pre-empted by US federal patent law); Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225 (1964); Compco Corp. v. Day-Brite Lighting, Inc., 376 U.S. 234 (1964) (both holding state unfair competition laws pre-empted by US federal patent law; state laws protected subject matter which was found unprotected by federal patent laws). Svensson, C-466/12, ratio decidendi. Football Dataco, C-604/10, ratio decidendi (emphasis mine). Svensson, C-466/12, para. 35. Contrast this with the opposite holding regarding additional rights in C More Entertainment, C-279/13, paras. 35–37. Adidas-Salomon, C-408/01, para. 18. See Arena (supra n. 11), Section D.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

How EU Law Constrains Rules Against Imitation

441

copyright law is concerned’). This is understandable given that the rules are predominantly driven by opposition to free-riding in a specific sector (vertical view) and any limiting principles are often seen as a form of balancing within the right’s design. Unless the legislator understands the IP rule at stake as the only one addressing the issue, its outlook is not the one of aggregate balance between innovation and imitation (horizontal view). This set-up can have important consequences for the notion of field pre-emption under secondary EU law. Consider the example of Lego bricks. The CJEU in its seminal Lego decision held that the shape of a toy brick could not be registered as a trademark due to its purely technical character. The Court emphasized the need to keep product shapes open to competitors on the market.25 One could argue that EU trademark law should impact the rules of national unfair competition law that circumvent this lack of registrability. The problem with this argument is that pre-emption would likely fail at the first stage – that of field applicability. Once a trademark is denied protection through a registration system, its regulation is outside EU trademark law, which deals only with registered signs. Any claims for extending the ‘trademark’ principles to protection of non-registered signs therefore need to be made on the level of consistency or policy arguments within the national law.26 To be sure, the EU law does not compel such arguments. We can contrast this with the recent Levola Hengelo case,27 which concerned copyrightability of a specific taste of cheese. In this case, the CJEU held that such subject matter does not constitute a ‘work’ because it does not constitute expression and thus is not eligible for copyright protection. Would a national unfair competition law providing some level of protection to taste conflict with this injunction of EU copyright law? Unlike in Lego, the scope of applicability would not be an issue. Protection would not require any formalities in either case and has the same underlying rationale. In the second step, the CJEU would have to determine whether such national protection is meant to be displaced by EU law or leads to an actual conflict. Because copyright law usually does not make an exhaustive statement about what can be imitated (Premier League),28 the likely outcome would be that no Union field pre-emption applies. The CJEU could still test whether unfair competition protection of taste would not endanger goals of copyright law (see above). But with a broadly defined subject matter, a finding of a conflict preemption seems unlikely too.

25 26

27 28

Lego, C-48/09 P, para. 60. National pre-emption by IP laws exists in some countries – see de Vrey (supra n. 4), p. 290. The idea behind such domestic consistency arguments is often that the law strives to achieve internal coherence – a rational legal order. Without justification, it should not treat more favorably situations in which the legislator previously did not feel a need to intervene by explicitly protecting the interests at stake. Levola Hengelo, C-310/17. Premier League, C-403/08, C-429/08, paras. 99–100.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

442

Martin Husovec

If the notion of ‘matter treated’ is read differently, however, such findings cannot be ruled out. According to CJEU case law, ‘the common rules may pre-empt a field that is broader than the one they actually regulate’.29 For instance, on the level of industry-specific exceptions, arguments about imitation can be made more persuasively. A good example could be the decompilation exception in the Software Directive (Directive 2009/24).30 Article 6 lays down rules concerning when a computer program can be translated from a machine code to a source code in order to build a competing product. Although the Directive deals with copyright law only, and mentions that it does not impact rules of unfair competition, it is unlikely that the CJEU would simply accept national misappropriation doctrines blocking the effects of Article 6.31 The Court would most certainly interrogate the reasons for such limitations on imitation. If the reasons behind the national rules and copyright law would align in essence, the national rules limiting imitation are likely to be preempted by the Directive. The Court could therefore consider Article 6 an exhaustive statement of the law on the matter which displaces any other rules on the national level; an argument that is reinforced by the prohibition of contractual deviations.32 To conclude, secondary EU law imposes an important constraint on national misappropriation doctrines. The extent of this depends on the goals behind the EU acts, in particular to what extent they are interpreted as exhaustive statements of the law and to what extent their goal is also to permit certain acts (horizontal view) and not only prohibit them from some limited angle (vertical view). In Lego, for instance, the Court refers to potential protection under national unfair competition laws after it refuses to protect a sign by a system of registered protection for its technical character – thus signaling its vertical view.33 It therefore requires a case-bycase assessment to determine whether a particular rule of EU law aims to safeguard the protection zone only, or also the corresponding imitation space. 29 30 31

32

33

Arena (supra n. 11), Section D(i). Directive 2009/24/EC. By the same token, if Article 6 is an exhaustive statement, in absence of contradictory signals in other pieces of EU legislation, it is likely to be viewed as lex specialis to other rules, such as Recital 16 of the Trade Secrets Directive, which would otherwise allow contractual arrangements. Compare interpretation of Article 9 of Directive 2009/24/EC (extending this effect to Article 6) in connection with Article 5(3), an exception for lawful acquirer, in SAS Institute Inc. v. World Programming Ltd [2010] EWHC 1829 (Ch), para. 312 (upholding contract-proof nature). In the subsequent litigation, the High Court, in the context of the fraud claims based on the US law, came to the following conclusions: ‘[Article 5(3) is an exception for, on the one hand, rights in the intellectual property family (“provisions such as those concerning patent rights . . . semiconductor products” and also “unfair competition” which apparently equates to passing off ) and, on the other hand, rights in contract (save to the extent that they conflict with Art. 5). This indicates clearly that the Directive does impact on other areas of law and that national laws outside the specific categories in Art. 9(1) must be read and applied to give indirect effect to the Software Directive, save insofar as they fall within the limited exceptions.’ (SAS Institute Inc. v. World Programming Ltd [2018] EWHC 3452 (Comm.), para. 180). Lego, C-48/09 P, para. 61.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

How EU Law Constrains Rules Against Imitation

443

Even when pre-emption by the EU secondary law does not apply, the CJEU can resort to a second level of scrutiny – that structured by concerns of free movement. Unlike pre-emption analysis, where protection against misappropriation in various IP laws is the predominant paradigm, free movement analysis is animated by the goal of unrestrained trade and thus freedom to imitate, against which national deviations have to justify their existence and proportionate design.

37.3 constraining intended gaps: free movement Incomplete harmonization through EU IP law assures that the Member States are free to experiment in their national law by providing less room to imitators and more rights to innovators. However, even if harmonization is incomplete and provides only a set of minimum or partial requirements in the area of intellectual property, national regulation internalizing them still has to comply with the ‘ceiling’ of obligations presented by the provisions of free movement, fundamental rights, and other provisions of EU primary law.34 In particular, any limitation on free imitation of goods stemming from additional national misappropriation doctrines is likely to interfere with free movement of goods (Article 34 TFEU) and, possibly, services (Article 56 TFEU). Such interference then has to be justified by a legitimate goal, and be necessary and proportionate.35 Among the public interest justifications, the CJEU recognizes two applicable sets of interests: ‘protection of industrial and commercial property’ and ‘protection of consumers and fairness in commercial transactions’. Deepening legislative harmonization creates tougher scrutiny by the free movement rules. As explained by Advocate General Colomer in Budˇejovický Budvar, ‘[r]eliance on the exceptions under [free movement] may no longer be justified, however, if a Community provision comes to safeguard the same interests as the national provision, once harmonisation is complete’.36 It might simply be more difficult to justify deviations and fragmentation since the EU legislator considered safeguarding particular interests on the EU level. In Clinique, for instance, the CJEU rejected higher protection of consumers against confusion, despite the fact that the EU law at stake was based on minimal harmonization. The Court viewed the national deviation as a disproportionate limitation on the free movement of goods. In a similar fashion, one can question, for instance, additional protection against imitation of non-registered designs after expiration of their pan-European protection through the principle of free movement even for intended gaps in harmonization.37 Although such intended gaps have to be respected as a matter of 34 35

36 37

Arena (supra n. 11), Section D(ii). See Articles 36 and 52 of the TFEU. Moreover, it would have to satisfy the so called ‘Keck criteria’ – see Keck and Mithouard, C-267/91, para. 16. Opinion of the Advocate General, Budˇejovický Budvar, C-478/07 (emphasis mine), para. 94. Posed by Kur 2006 (supra n. 7), p. 720.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

444

Martin Husovec

secondary EU law, increased protection of a subject matter by other Union instruments decreases likelihood of legitimacy of national rules, which always have to be defended against the principle of free movement stemming from primary EU law. To date, the CJEU has not had the chance to address a single case of pure imitation under unharmonized national laws from the perspective of free movement.38 The case that comes the closest – Beele (1982)39 – dealt with an imitation of a formerly patented product the protection of which had expired. According to the Dutch courts, the imitation was also misleading. The Dutch Supreme Court had doubts about issuing an injunction against such products which it presumed had been lawfully marketed in other Member States. The CJEU recognized the lack of common rules in the area and that an injunction would constitute an obstacle to free movement. After noting that the Dutch tortious doctrine against ‘precise imitation’ is not discriminatory,40 the Court reviewed its justification as follows:41 National case-law prohibiting the precise imitation of someone else’s product which is likely to cause confusion may indeed protect consumers and promote fair trading; these are general interests, which ... may justify the existence of obstacles to movement within the community resulting from disparities between national laws relating to marketing of products. That such a rule does meet mandatory requirements is moreover borne out by the fact that it accords with the principle underlying Article 10 bis of the Paris Convention ... which prohibits inter alia all acts of such a nature as to create confusion with the goods of a competitor, and by the fact that this rule is recognized in principle in the case-law of the most Member States.

The CJEU then emphasized that the injunction concerns a product which ‘for no compelling reasons [is] practically identical to the products imitated’, which ‘needlessly causes confusion’.42 The decision therefore rests on two important assumptions: (1) imitation was causing confusion; and (2) there were no common Community rules applicable to confusing products. As pointed out by Annette Kur, the extent of confusion would today be considered against several sources of EU law.43 Therefore it is possible that the CJEU would judge the presence of misrepresentation differently, and be actually engaging with the factual pattern. If the deception could be said to affect consumer interests, the Unfair Commercial Practices Directive could apply.44 Misleading imitations impairing consumer interests are likely to be caught by its provisions. In fact, the Directive even foresees one such situation on its blacklist of practices that are unfair 38

39 40 41 42 43 44

Andreas Wiebe, in Andreas Wiebe and Georg Kodek (eds.), Kommentar zum UWG: 2. Auflage (Manz, 2012), 4 Lfg, p. 6. Beele, C-6/81. Ibid., para. 7. Ibid., para. 9. Ibid., para. 11. Kur 2006 (supra n. 7), pp. 719–720. Directive 2005/29/EC – see Kur, ibid., p. 720.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

How EU Law Constrains Rules Against Imitation

445

per se.45 In a purely B2B setting, on the other hand, where consumer behavior is not at stake, the CJEU could draw insights from the potentially applicable Directive on Misleading and Comparative Advertising, and EU trademark law.46 Even if their provisions would not apply, they might provide a healthy benchmark for confusion tests in order to assess when the national considerations presented as misrepresentation are still necessary and proportionate. This point is reinforced by Dior, in which the CJEU extended limits of the trademark law into copyright law on the basis of free movement.47 In a purely B2B scenario with no negative impact on consumers, deviations might be harder to justify. In absence of any secondary EU law, the CJEU would question a need to limit deviations on the basis of ‘fairness in commercial transactions’ (if based on unfair competition law) or ‘industrial and commercial property’ (if based on specific IP legislation). However, situations without confusion are not explicitly envisaged by Article 10 bis of the Paris Convention. In Beele, the referring court stressed that the assessment of confusion was made not only from a technical but also from economic and commercial points of view.48 This suggests that the confusion could be understood very differently, and in fact incorporate considerations which other countries might view as pure misappropriation.49 The CJEU should therefore analyze national rules holistically and, in absence of applicable EU benchmarks, try to understand policy rationales – what interests are being safeguarded for what reason – rather than be misled by their labeling. If consumers are missing in the picture, the argument needs to be made why crossborder trade has to be hindered by national rules prohibiting particular acts of pure imitation given other IP tools that are already at the disposal of the innovator.50 This further raises the evidentiary burden compared to conventional arguments made by parties in a usual unfair competition dispute. If rigorously adopted, it could force such doctrines to be more circumscribed by private and general interests – which has a standalone value.

37.4 conclusions This chapter pointed out that there are two EU restraints on national rules against imitation: (1) pre-emption mechanisms imposed by secondary EU law; and (2) principles of free movement of goods and services imposed by primary EU law. 45 46

47 48 49 50

See Annex 1, number 13, prohibiting instances of ‘deliberately misleading’ products. Directive 2015/2436 (imposing only minimal standards); Directive 2006/114/EC (full harmonization). Dior, C-337/95, para. 58. Beele, C-6/81, para. 11. For instance: BGH GRUR 2006, 79 – Jeans I. For criticism, see Ohly (supra n. 4), p. 514 ff. For instance, why firms investing in computer programs should appropriate functionality beyond the limits of copyright and patent law. Unless SAS, C-406/10, triggers field preemption, this is a free movement issue.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

446

Martin Husovec

Together they are tightening the noose on national regulatory deviations from the EU IP landscape. With the mushrooming legislative activity in the EU against imitation and different forms of appropriation, the regulatory freedom of the Member States is arguably decreasing, and this is true also for the non-harmonized unfair competition law. If the issue receives more attention from scholars, eventually national judiciaries might realize this too.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.048

38 European Union Law and Slavish Imitation: An Update in Honour of Annette Kur Matthias Leistner1

38.1 introduction In 2006 Annette Kur’s seminal essay on unfair competition protection against imitations and its impact on the free movement of goods was published in the liber amicorum for Eike Ullmann.2 In essence, she argued that in the light of ongoing harmonization of European IP and unfair competition law, namely the introduction of unregistered Community design protection and other secondary law instruments, the old carte blanche for national protection against slavish imitations from the Beele judgment no longer applied – or rather that Beele had never been carte blanche for national law.3 Since then, the German Act on Unfair Competition4 was restructured in 2015, and German case law in the field has developed rapidly. In view of Kur’s 2006 paper, some of these case law developments have to be met with approval. Others have remained controversial (see 38.2 below). More importantly, European Union secondary law has further developed too. In particular, meanwhile, a number of Court of Justice of the European Union (CJEU) cases (in other areas) have dealt with the Unfair Commercial Practices Directive (UCPD), which arguably overlaps with national protection against unfair imitation, at least insofar as product marketing

1

2

3 4

Prof. Dr. iur., LL.M. (Cambridge); Chair of Private Law and Intellectual Property Law, with Information and IT-Law (GRUR-Chair), Ludwig Maximilian University (LMU), Munich. Annette Kur, Nachahmungsschutz und Freiheit des Warenverkehrs – der wettbewerbsrechtliche Leistungsschutz aus der Perspektive des Gemeinschaftsrechts, in: Ahrens/Bornkamm/ Kunz-Hallstein (eds.), Festschrift fu¨r Eike Ullmann (C. H. Beck, 2006), p. 717. Ibid., p. 721 et seq. Gesetz gegen den unlauteren Wettbewerb of 17 February 2016, BGBl. I 233; provisions without reference in the following text are provisions of this Act.

Downloaded from https://www.cambridge.org/core. 447 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

448

Matthias Leistner

which creates consumer confusion is concerned.5 This requires an ‘update’ of Kur’s analysis in regard to the relevance of European secondary law (see 38.3 below).6

38.2 slavish imitation in german unfair competition law since 2006 38.2.1 Legislative Activities since 2006 Since its latest major revision in 2015, the German Unfair Competition Act covers unfair imitations in section 4(3), which, according to its heading and to the explanatory memorandum,7 specifically protects competitors’ interests.8 Of the different case groups of the provision, only section 4(3)(a) is directly relevant to our topic. It addresses imitations of competitors’ products or services which unnecessarily mislead customers about their origin. In addition, section 5(2), which regulates misleading commercial practices in the interest of consumers, provides for a prohibition of any marketing of products or services which might lead to confusion with a product or service of a competitor. The prevailing opinion in German doctrine distinguishes between section 5(2), which more or less directly corresponds to Article 6(2)(a) UCPD and which consequently implements fully harmonized European rules into German law, and section 4(3)(a), which is regarded as being outside the scope of the Directive due to Recital 6 sentence 3 UCPD. This Recital exempts ‘commercial practices which harm only competitors’ economic interests or which relate to a transaction between traders’. However, as will be shown below,9 in the light of the CJEU’s judgments on this issue, the opinion that Recital 6 sentence 3 covers section 4(3)(a) is hardly tenable, since this provision of German law clearly not only concerns the interests of competitors, but also at the same time the interests of consumers. Thus, both provisions on misleading practices in connection with the imitation of goods or services, namely sections 4(3)(a) and 5(2), fall into the scope of the Directive, while only section 4(3)(b) on taking undue advantage of or being 5 6

7

8

9

Cf. already clear-sighted Kur, supra n. 2, p. 719 et seq. Given the limitations of the possible scope of this contribution, the arguments will mainly focus on the issue of unfair competition protection against confusing slavish imitations of technical product features, which are or were susceptible of patent protection, but have not been registered or for which patent protection has expired. Begru¨ndung Regierungsentwurf Gesetz gegen den unlauteren Wettbewerb 2008, BT-Drs, 16/10145, p. 17; cf. also Beschlussempfehlung und Bericht des Ausschusses fu¨r Recht und Verbraucherschutz, BT-Drs. 18/6571, p. 15. To be sure, these aspects were already allegedly separated in the previous Act on Unfair Competition of 2004 where the legislator was also of the opinion that Art. 6(2)(a) UCPD was implemented by s. 5(2) notwithstanding s. 4(9) (as it then was), which latter allegedly only concerned the protection of competitors – see e.g. Federal Court of Justice (BGH) GRUR 2010, 80 para. 17 – LIKEaBIKE; Köhler, GRUR 2009, 445 (447 et seq.). See 38.3.2 below.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

EU Law and Slavish Imitation

449

detrimental to another competitor’s products or services is outside the realm of the UCPD’s full European harmonization of unfair commercial practices (see further 38.3.2 below). 38.2.2 Selected Case Law since 2006 Since 2006, the Federal Court of Justice has structured and specified unfair competition protection against slavish imitations (under today’s section 4(3)) in a clear way, which in fact takes into account some of Kur’s concerns. At the same time, some of the developments in case law, which she had already hinted at, have meanwhile been completed by further judgments of the Court. a. Some favourable developments Namely, the case group of imitation of modular products,10 which had already been substantially limited by the time of Kur’s paper, has meanwhile been abandoned altogether in the Segmentstruktur judgment.11 Also, the case group of short seasonal protection for certain fashion designs, which initially filled the gap left by the unavailability of unregistered design protection in German law, but which had overlapped with European (and German) design protection after the introduction of an unregistered Community design right and which had become untenable, given the differentiated conditions and limitations of these specific IP rights, was expressly given up by the Federal Court of Justice in the same judgment. Consequently, today only cases of genuine unnecessary causation of consumer confusion as to the origin of the goods or cases of undue image misappropriation or damage to another competitor’s reputation fall under section 4(3)(a) and (b). However, even in these cases there might be contradictions to certain conclusive conditions and limitations of the existing IP rights system.12 This problem is aggravated by the fact that the German courts still tend to assume consumer confusion in the cases of identical or almost identical imitations from the very fact of the unnecessary identical or almost identical (slavish) imitation as such. In fact, the Federal Court of Justice’s judgments in the cases Exzenterzähne13 and Bodendübel14 are cases in point. These cases concerned technical products for which patent protection had lapsed. 10

11 12 13

14

This concerned the imitation of modular products, where the amortization of the investment is particularly dependent on the sales of compatible supplementals, such as in the case of LEGO bricks and comparable modular systems. BGH, GRUR 2017, 79 – Segmentstruktur. Cf. Ohly, IIC 2010, 506 (513). BGH, GRUR 2015, 909 – Exzenterzähne. This concerned the sales of imitations of certain dowels for the fixing in position of electrical cables, which were characterized by eccenter teeth at the edges which helped to fix the dowels in the walls. BGH, GRUR 2017, 734 – Bodendübel. This concerned the sales of imitations of ground dowels.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

450

Matthias Leistner

b. Protection of Formerly Patent-Protected Elements against Imitation: Exzenterzähne and Bodendübel In Exzenterzähne, the Court restated the cumulative and independent application of section 4(3)(a) and specific IP rights and justified this by pointing out that section 4(3) had its own objective and conditions, namely likelihood of consumer confusion regarding the origin of the product in the case of lit. (a). Hence, according to the Court, even (formerly) patent-protected elements, which essentially perform a technical function, could give a product so-called ‘competitive individuality’, which is a necessary precondition of protection against imitations under German law. Only if an element is genuinely indispensable for achieving a technical result, namely if there is no viable alternative for performing the same technical function, will it be excluded from possible protection against imitations. Also, as the main additional condition of section 4(3)(a) is consumer confusion, section 4(3)(a) will not apply either (at least in theory) if the danger of customer confusion can be avoided by adequate labelling. Against this background, the judgments in Exzenterzähne and Bodendübel concerned cases where, because of the characteristics of the products and markets in question, effective labelling was not sufficient to exclude the likelihood of customer confusion in the relevant markets.15 For this reason, the Court decided in Exzenterzähne that even if patent law protection for certain technical elements had expired, competitors were not allowed to imitate these elements in an identical or almost identical form (slavish imitation), if this led to consumer confusion, as long as there were viable alternatives for reaching the same overall technical result. Indeed, in theory, this approach allows a consistent solution to avoid contradictions in the negative overlap of expired patent protection and possible unfair competition law protection.16 At a closer look, however, it becomes clear that the two different objectives of regulation of competition in these two areas of law contradict each other in these situations. While unfair competition law regulates identical or slavish imitation in order to protect consumers and competitors against the effects of market confusion, the fundamental ratio of patent law goes in another direction: in fact, patent law grants limited protection in order to incentivize the disclosure of certain technological solutions and to make them accessible for dynamic competition after the expiry of the protection term in the interest of competitors.17 Also, the interest of consumers with regard to market differentiation 15

16

17

In fact, the products in question (see supra nn. 13 and 14) were so small and so much characterized by their technical function that an effective labelling would not have been possible. See on what we call negative or a posteriori overlaps Derclaye and Leistner, Intellectual Property Overlaps: A European Perspective (Hart Publishing, 2011), p. 3 and on the appropriate solution in German law, p. 275 et seq. Cf. e.g. Kraßer and Ann, Patentrecht (C. H. Beck, 2016), § 3 para. 11; Leistner, GRUR 2018, 697 (698).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

EU Law and Slavish Imitation

451

and lowering of prices after that limited protection term has to be considered. As a result, the balancing of these different regulation objectives and interests might even require the acceptance of a certain degree of consumer confusion in the market for a limited period of time, as a necessary ancillary effect of the development of price competition after the expiry of the protection term. This is even more evident in product or service markets where the permission of imitations allows customers to ultimately develop the necessary knowledge to distinguish between the original and the copy within a reasonable transitional period. By contrast, the application of unfair competition protection effectively cements the monopolistic market position of the former patent holder and thus delays possible competition by imitation for a potentially unlimited period of time, given that the former patent holder can effectively use his ‘springboard’ to successfully perpetuate his market position.18 On a more practical note, at any rate, the consistent application of the Federal Court of Justice’s ratio would require a very disciplined application of the condition of avoidable consumer confusion by the lower courts, which should asses in every single case, depending on the specific market situation, whether a real danger of customer confusion in the relevant markets exists, instead of deriving consumer confusion from the very fact of identical or almost identical imitation of certain essential characteristics of the product in question (which in such cases will indeed often be the very (formerly) patented, innovative technical elements).19 In fact, the follow-up implementation of the Exzenterzähne doctrine in the lower courts has not always lived up to the necessary strict standard in that regard, which illustrates the potential danger that the Federal Court of Justice’s ratio tends to be misused to unjustifiably perpetuate expired patent law protection by means of unfair competition law.20 In fact, it seems the Federal Court of Justice has meanwhile also realized that danger: in a more recent judgment the Court has at least considerably lowered the bar for sufficient measures avoiding the likelihood of customer confusion, if the characteristics and use of the product do not allow for particularly effective measures in order to distinguish the ‘imitated’ product.21 However, although this is a favourable adjustment, it will have to be awaited whether this new development will effectively discipline the practice of the lower courts, which (also historically) have often tended to grant unfair competition law protection against slavish imitation following a mere hunch that such identical or almost identical imitation is unfair, unless it is technologically indispensable. What is more, the fundamental conflict between patent law’s objective to make new technology accessible for competition

18 19 20 21

Cf. Leistner, UWG – Großkommentar (De Gruyter, 2013), § 4(9) paras. 103 and 197. Leistner, GRUR 2018, 697 (698 et seq.). Cf. comprehensively Leistner, GRUR 2018, 697 (700 et seq.) with further references. BGH, GRUR 2017, 1135 – Leuchtballon; see case annotation: Ullmann, jurisPR-WettbR (juris GmbH), 11/2017, Anm. 3.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

452

Matthias Leistner

by imitation and follow-up innovation22 after the expiry of the protection term, and possible unfair competition protection against slavish imitation which can lead to a certain degree of customer confusion, remains unsolved by this more practical and only nuanced correction of the doctrine by the Federal Supreme Court. Even from the perspective of German law, this raises the question whether technical elements for which patent protection has expired should not be excluded even more strictly from unfair competition protection in order to avoid the otherwise impending tensions with the objectives of patent law. While the situation in German law in that regard has been discussed in a differentiated way in another paper by this author23 and also by many other German authors,24 it is also an intriguing task to take a closer look at the legal situation in regard to this issue in EU law. To honour Kur’s essay, this shall be the main focus of the following ‘update’ of Kur’s analysis from 2006.

38.3 european union law, slavish imitation of formerly patent-protected product elements in german law and the freedom of imitation 38.3.1 Increased Importance of the UCPD Although this is disputed in parts of German legal doctrine25 in regard to section 4(3) (a), the most striking development in European law since 2006 has been that the conflict between unfair competition law regulation of identical or slavish imitations and well-considered limitations of the IP rights system has meanwhile reached the heart of European secondary law on unfair competition.26 This is mainly because of the general provision on misleading commercial practices by way of marketing a product which creates confusion with a competitor’s products, trade names or other distinguishing marks in Article 6(2)(a) UCPD.

22

23

24

25

26

Cf. particularly problematic in that specific regard Frankfurt Regional Court (LG), Judgment of 21.3.2018, 2-06 O 98/18, p. 17 (= GRUR-RS 2018, 9284) – Exzenterzähne III: even certain changes of the shape of the product would be perceived by customers as mere improvements of quality by the original manufacturer, if they were advantageous for the technical use of the product. Leistner, GRUR 2018, 697 (on the basis of a legal opinion commissioned by one of the parties in the Exzenterzähne case). See the references in Leistner, GRUR 2018, 697; most clearly: Ullmann, jurisPR-WettbR (juris GmbH), 9/2015, Anm. 3. See for the opinion, which is also followed here, Sack, WRP 2017, 650; for a brief overview of both sides of the discussion cf. Ohly and Sosnitza, UWG – Großkommentar (De Gruyter, 2016), § 4(3) para. 3/10. In the absence of any relevant case law on the matter, Kur had to remain undecided in regard to this point, which she already considered in her 2006 essay, when the UCPD was not even implemented into the national laws of the Member States; supra n. 2, p. 719 et seq.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

EU Law and Slavish Imitation

453

In fact, while hitherto the importance of European secondary law was arguably comparatively limited in the field (although already clearly present because of Article 4(g) Misleading and Comparative Advertising Directive),27 from today’s viewpoint the possible tension or even conflict between the differentiated IP rights system and unfair competition law protection for certain formerly IP-protected elements on the ground of misleading marketing of copycat products by competitors has meanwhile been ‘internalized’ in European secondary law. Consequently, as far as European secondary law harmonization goes, the old Beele case law has meanwhile to be replaced by an interpretation of the respective Directive provisions, which is in line with European primary law.

38.3.2 Section 4(3)(a) Not Exempted From the Directive’s Scope As for the question of applicability of Article 6(2)(a) UCPD in regard to provisions of national law on unfair imitation, Recital 6 sentence 3 of the Directive sets the course. Hence, it is decisive whether such national provisions only protect competitors’ interests. In the Mediaprint judgment the Court emphasized the broad scope of the Directive and clarified that the exemption from the fully harmonized scope of the Directive only applies if a national provision exclusively relates to unfair commercial practices which harm ‘only’ competitors; any additional objectives, not covered by Recitals 6 or 9, and in particular any objectives which relate to the protection of consumers, will trigger the applicability of the UCPD.28 Since then, in Wamo29 and RLvS Verlagsgesellschaft,30 the Court has ruled that under the division of jurisdiction between the EU courts and the Member States’ courts it is for the national court to interpret the national provisions at issue in regard to the question whether they (also) pursue objectives relating to consumer protection.Thus, legal doctrine in Germany and the German legislator seem to feel that there is a certain leeway for what might be called a subjective ‘definition’ of the objective of certain unfair competition rules by way of explanatory memoranda and headlines relating to older case law.31 In effect, with regard to product imitations in German law, this amounts to splitting up the assessment of one and the same commercial practice into a consumer protection related part (regulated by section 5(2)) and a competitor 27

28

29 30 31

See further Ohly, IIC 2010, 506 (520 et seq.); cf. also Art. 4(h), Misleading and Comparative Advertising Directive. ECJ, C-540/08, ECLI:EU:C:2010:660 para. 27 et seq. – Mediaprint. This of course could not yet be taken into account in BGH, GRUR 2010, 80 para. 17 – LIKEaBIKE, where the Court argued that s. 4(9) (as it then was) was outside the scope of the UCPD because it ‘predominantly’ protected the competitors’ individual effort; however, according to Mediaprint, predominance of the objective to protect competitors is no longer sufficient to profit from the exemption in Recital 6 UCPD. ECJ, C-288/10, ECLI:EU:C:2011:443 para. 22 et seq. – Wamo. ECJ, C-391/12, ECLI:EU:C:2013:669 para. 35 – RLvS Verlagsgesellschaft. Cf. part of the references in supra nn. 7, 8 and 25.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

454

Matthias Leistner

related part (regulated by section 4(3)(a)). For German unfair competition law, which has traditionally been characterized by a unitary approach, combining the objectives of protecting consumers, competitors and the general public, this seems a rather artificial split.32 More importantly, from the viewpoint of European law this approach seems not entirely convincing either for a couple of reasons. First, in Wamo, the national court had given no further information whatsoever on the legal context of the prohibition in question and actually, according to the national court, it was even doubtful whether the provision contributed to the benefit of the consumers at all. This uncertainty seems to have been at least part of the reason why the Court – in such a case of doubt and lacking the necessary information – applied a light touch and left the assessment to the national court.33 Under that perspective, of the German prohibitions on the marketing of product imitations, only section 4(3)(b) on unduly taking advantage of the repute of a competitor’s product (without misleading consumers) would be a case comparable to Wamo, because it is indeed doubtful whether this prohibition serves consumers’ interests at all. By contrast, there can be no doubt that the prohibition of the marketing of product imitations which can lead to confusion of customers in regard to the origin of the product, always also serves consumers’ interests; this was even impliedly acknowledged by the BGH in its pre-Mediaprint LIKEaBike judgment.34 This objective fact can hardly be ‘defined away’ by a subjective interpretation of national law (or any guidance in that regard by the national legislator). Moreover, such interpretation, even if it were still in the jurisdiction of the national courts and methodologically tenable, would violate the principle of sincere cooperation enshrined in Article 4(3) TEU. Admittedly, in many countries there is a sharp systematic distinction between consumer protection on the one hand, and the tort of passing off on the other, and the UCPD only intended to harmonize the consumer protection side. Therefore, the respective distinction is certainly not purely artificial. However, where it conflicts with the unitary market objective of the UCPD, from this author’s viewpoint, the objectives of European law should take precedence. Consequently, in cases of genuine overlap, where the protection of competitors is granted for reasons which – on the corresponding demand side of the market – directly mirror the UCPD’s scope and consumer protection objectives, the European harmonization should be regarded as conclusive and consequently also affects the tort of passing off.35 What is more, secondly, in Beele the ECJ already decided clearly that product imitation which leads to consumer confusion indeed is prohibited in the consumers’ interest.36 This is because in Beele, actually, the Court would have had the choice 32 33 34 35 36

Which reminds this author of Magritte’s painting Ceci n’est pas une pipe. Cf. ECJ, C-288/10, ECLI:EU:C:2011:443 para. 20 et seq. – Wamo. Cf. supra n. 28. Cf. Gibson and Johnson (2015) 131 LQR 474. ECJ, Case 6/81, ECLI:EU:C:1982:72 paras. 7 and 9 – Beele.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

EU Law and Slavish Imitation

455

whether to justify the Dutch provision at issue on the grounds of Article 30 TEC (today’s Article 36 TFEU), namely the protection of industrial and commercial property, or on the grounds of the Cassis case law on mandatory requirements relating to the protection of consumers and fairness in commercial transactions, and expressly chose the second alternative.37 Thus, Beele had already established an autonomous interpretation in EU law, which unambiguously clarifies that a national provision, such as section 4(3)(a), serves the objective of consumer protection. Against this background, from this author’s viewpoint, there can be no doubt that section 4(3)(a) falls into the scope of the UCPD and thus implements fully harmonized European law. The consequences of this for section 4(3)(a) are lopsided, however.38 Since section 5(2) implements the protection standard of the Directive anyway, applying a more liberal standard under section 4(3)(a) would be permissible from the EU law viewpoint.39 By contrast, applying a stricter standard than the Directive’s standard is not possible in the fully harmonized field of application of the UCPD since this could not be balanced out by section 5(2). Hence, from the perspective of EU law, the assessment of the German law’s current status as regards slavish imitation under section 4(3)(a) hinges mainly on the interpretation of Article 6(2)(a) UCPD. This interpretation has to be carried out in line with EU primary law and in the overall context of EU secondary law.

38.3.3 Slavish Imitations of Formerly Patent-Protected Elements under Article 6(2)(a) UCPD a. Requirement of Commercial Relevance A first and generalizable consequence of this approach is that in the context of section 4(3)(a) the relevance requirement of Article 6(2) UCPD has to be applied consistently.40 Hence, even in the case of slavish imitation, not only has it to be specifically shown in each case that the imitation leads to avoidable confusion of customers, but also that this causes or is likely to cause the average consumer to take a transactional decision that he or she would not have taken otherwise. While the 37

38 39

40

Ibid. In fact, even in ECJ, C-238/89, ECLI:EU:C:1990:473 – Pall/Dahlhausen, where the Court indeed also considered the competitors’ interests (see para. 20 et seq.) in a case which was at least close to misleading commercial practices in product marketing. The main aspect was also protection of consumers against being relevantly misled (see para. 16 et seq.). See also Sack, WRP 2017, 650 (653 et seq.). Also, of course the UCPD is only binding for commercial practices vis-à-vis private consumers, whereas commercial practices targeted to commercial customers are outside its scope – see Recital 6 sentence 3. However, since s. 4(3) does not distinguish between private and commercial customers, and since no other will of the legislator is discernible, a unitary interpretation in line with EU law should apply in German law. See Sack, WRP 2017, 650 (653).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

456

Matthias Leistner

Federal Court of Justice’s case law is generally in line with this requirement, some of the case law of the lower courts – and in particular the application of the Federal Court of Justice’s doctrine in Exzenterzähne by the follow-up case law of the lower courts – is not.41

b. Product Elements which Perform a Technical Function Secondly, the crucial question on the interpretation of Article 6(2)(a) UCPD is whether, in the context of the Philips/Remington judgment on Article 4(1)(e)(ii) Trademark Directive42 and the provision of Article 7(1) Design Directive, the imitation of product elements which essentially perform a technical function should in principle be excluded from the protection, granted to consumers by Article 6(2) (a) UCPD.43 Indeed, in Philips/Remington the public interest in free use (and choice) by all competitors of all product features which perform a technical function was regarded as a legitimate aim, which on balance weighed heavier than the interest to register such elements for trade mark protection, even if there were technical alternatives for competitors.44 Also, the court expressly explained that this should prevent the establishment or perpetuation of exclusive rights relating to technical solutions on the basis of trade mark law. Nonetheless, on the basis of a mere grammatical and contextual interpretation it seems doubtful whether the provisions in the Trade Mark and Design Directives can indeed be generalized in order to exclude any prohibition under Article 6(2)(a) UCPD of misleading practices based on the imitation of technical elements of a product. This is, first, because a comparable exclusion is obviously missing in the text of the UCPD and thus on the mere basis of contextual interpretation, an e contrario argument seems at least as legitimate as any attempt to broaden the Trademark and Design Directives’ exclusions by way of an analogy. Indeed, as Article 6(2)(a) UCPD expressly and specifically requires commercially relevant confusion of consumers, also from the viewpoint of teleological interpretation a specific distinctive element is present in the specifically consumer protection oriented Directive’s provision, which, characteristically, requires that the average consumer is likely to be caused to take a distorted transactional decision that he or she would not have taken otherwise.

41 42 43

44

See Leistner, GRUR 2018, 697 (702 et seq.) with further references. ECJ, C-299/99, ECLI:EU:C:2002:377 para. 79 et seq. – Philips/Remington. Cf. on German law Ullmann, jurisPR-WettbR (juris GmbH), 9/2015, Anm. 3; Sack, GRUR 2015, 442 (446). See further on the express ‘freedom of competition’ rationale underlying this judgment Annette Kur, Too Pretty to Protect? Trade Mark Law and the Enigma of Aesthetic Functionality, in: Drexl/Hilty/Boy et al. (eds.), Technology and Competition – Contributions in Honour of Hanns Ullrich (Larcier, 2009), p. 139 (147).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

EU Law and Slavish Imitation

457

Hence, to be sure, Article 6(2)(a) UCPD requires actual confusion of the consumers in the specific market45 as well as actual commercial relevance in the specific sales situation which is typical for the imitated product. However, if these requirements are actually met in a given case, imitation of technical elements which creates confusion with products of a competitor and thus effectively and relevantly misleads consumers is not per se excluded from the prohibition under Article 6(2)(a) UCPD. If the resulting confusion could have been avoided by the competitor by way of labelling or other reasonable measures in order to distinguish its products from the originals, Article 6(2)(a) UCPD straightforwardly applies and the marketing of the copycat products is prohibited under the Directive.

c. Cases of Unavoidable Consumer Confusion Against this background, the most intriguing problem, which also concerned the Federal Court of Justice in its Exzenterzähne judgment, is how to assess situations where, with regard to the ‘legitimate’ application by a competitor of technical solutions which were eligible for patent protection but never registered by the original manufacturer or for which patent protection has expired, the resulting confusion in the markets cannot be avoided.46 In fact, if the conditions of consumer confusion and commercial relevance are applied with the necessary strictness by the courts, such situations should be very rare, since typically the average consumers will be able to distinguish between the original and the copycat product if appropriate information measures are taken.47 However, in rare cases where such measures are actually not possible or not sufficient to avoid consumer confusion, inevitably the public interest in free access to the technical solutions (which is also, typically, in the long-term interest of consumers48) and the short-term interest to protect consumers against any inevitably misleading marketing of products in the transitional phase – when the repute of the original product is still relevant for consumer expectation – clash with each other. In result, the short-term interest of consumers in being protected against any misleading marketing of products (which stands behind Article 6(2)(a) UCPD) flies in the 45

46 47

48

Here, in regard to private consumers, naturally, the strict standard of the European average consumer applies. When commercial customers are concerned, in such cases – because of the mandatory character of European primary law – the even stricter standard of a ‘prudent economic operator’ (see ECJ, C-238/89, ECLI:EU:C:1990:473, para. 21 – Pall/Dahlhausen) has to be applied. See on the interpretation of s. 4(3)(a) in line with the Directive (also in regard to commercial customers) supra n. 39. See also Sack, WRP 2017, 650 (655). This is even clearer in the case of ‘prudent economic operators’, which according to the judgment in Pall/Dahlhausen (supra n. 45) might even be required to consult additional sources of information if they are in doubt about certain facts concerning a marketed product. This is because once clarity has been re-established in the markets, consumers can profit from their larger range of choice of different products. Cf. at n. 18 supra.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

458

Matthias Leistner

face of their own long-term interest to profit from a larger choice of different products, applying the same and formerly patented technical solution, and also of the public interest in the freedom of the technological state of the art after expiry of patent protection. Sure enough, as has been said, if confusion can be avoided, Article 6(2)(a) UCPD requires the copycat to apply the necessary measures to do so, even if only unprotected elements were slavishly imitated by it. However, if this is not possible, the interest behind Article 6(2)(a) UCPD has to be balanced against the countervailing public and long-term interests which are behind the specific conditions and limitations – in particular the limited protection term- of patent law.49 d. A Solution Based on an Interpretation of the Directive in line with EU Primary Law As for the specifics of this balancing of interests, the text of the UCPD is silent.50 From this author’s viewpoint, the solution must therefore lie in an interpretation of the Directive in line with primary law and in particular with the objective of the fundamental freedoms. In that regard, first, we should recall that exclusive IP protection in EU law, insofar as it potentially clashes with the fundamental freedoms, is always only guaranteed to the extent of the specific subject matter of the IP right in question.51 In particular, according to some of the more recent judgments in the copyright field, in regard to the exclusive use rights this shall give the right holder an adequate chance to get appropriate remuneration for the marketing of products based on his or her intellectual effort, while further remuneration, achieved by controlling markets beyond the direct reach of the exclusive rights, is not justifiable.52 Hence, it is clear indeed – as the Court also ruled in Philips Remington53 – that from the viewpoint of EU law there is a legitimate public interest in free access to technical solutions, when the conditions of patent law are not met or when exclusive patent protection has expired. Secondly, and here the circle closes to Kur’s seminal 2006 essay, in Beele the Court already explained that protection against identical or slavish imitations which 49

50

51

52

53

Ohly, IIC 2010, 506 (513); Sack, WRP 2017, 650 (655) with further references, who wants to solve the problem along the lines of German law principles, i.e. by applying ‘inherent limitations’ to the prohibition of misleading practices, which shall be specified in such cases on the basis of a balancing of interests. See on the problems in that regard (also in the light of the Court’s case law on Art. 6 UCPD) Sack, WRP 2017, 650 (655). See very early for patent law e.g. ECJ, Case 15/74, ECLI:EU:C:1974:114 – Centrafarm/Sterling Drug. Cf. ECJ, C-403/08, ECLI:EU:C:2011:631 para. 106 et seq. – Murphy; ECJ, C-128/11, ECLI:EU: C:2012:407 para. 62 et seq. – UsedSoft. In the context of the problem at hand, i.e. the interpretation of Art. 6(2)(a) UCPD, it has to be emphasized that both judgments also concerned the interpretation of secondary EU copyright law. See supra n. 42.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

EU Law and Slavish Imitation

459

can lead to consumer confusion is inherently limited if such imitations are ‘necessary’ from a technical or ‘from the economic and commercial point of view’.54 This indeed helps to solve the issue at hand. If the identical or almost identical imitation of certain product elements is indispensable to achieve a certain technical result, and if patent protection for these elements is not present or has expired, the imitation is necessary from a technical point of view and, in absence of possible measures to avoid consumer confusion, has to be accepted in the interest of free access to the technological state of the art. Insofar, the balancing of interest in EU law neatly reflects a common attitude in the overwhelming majority of Member States.55 However, and this is the crucial difference to the Federal Court of Justice’s Exzenterzähne judgment, even when there are technical alternatives to achieve the same result, according to an interpretation of Article 6(2)(a) UCPD in line with primary EU law, slavish imitation which unavoidably leads to consumer confusion is justified if the implementation of the very technical solution of the original manufacturer was necessary for economic and commercial reasons, and in particular if it was necessary to newly enter a market which was hitherto dominated by the original manufacturer because of formerly existing patent law protection with a product, designed in accordance with the principle of free competition, namely without being hampered by unfair competition law perpetuation of former patent protection. This balanced result, which for German law already reflects the prevailing opinion in literature,56 is less obvious and not easy to construe when the interpretation of Article 6(2)(a) UCPD is concerned.57 Nonetheless, this is necessary since the conflict of interests, which Kur analyzed in a more general context and mainly with regard to the free movement of goods and national law, is meanwhile enshrined in European secondary unfair competition law itself. In this new context it has lost nothing of its original relevance, however. Hence, notwithstanding the developments of the last ten years, it has been Kur’s great essay of 2006 which has guided the way to a balanced result under today’s legal conditions in the framework of European secondary unfair competition law.

54 55 56 57

ECJ, Case 6/81, ECLI:EU:C:1982:72 para. 11 – Beele. Cf. Sack, WRP 2017, 650 (655) with further references. See the references in Sack, WRP 2017, 650 (655); Leistner, GRUR 2018, 697 (703 et seq.). See also Sack, ibid.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.049

39 The German Misappropriation Origins of Trademark Antidilution Doctrine: A Translation of the 1924 Odol Opinion of the Elberfeld Landgericht Barton Beebe1

Frank Schechter introduced the concept of trademark dilution to American readers in his 1927 Harvard Law Review article “The Rational Basis of Trademark Protection.”2 He concluded the article by noting that his formulation of the concept “is fortified by the doctrine that has developed within recent years in German law on this same point.”3 In waiting until the end of his article to refer to the German case law, Schechter effectively implied that dilution was a home-grown American idea – indeed, that it was Schechter’s own, original idea – and that the German courts just happened to have come up with a similar notion. In fact, Schechter appears to have taken the concept of trademark dilution and much of the rhetoric he used in describing it – including “selling power” and the very term “dilution” – from a single German trial court opinion, specifically, the opinion of the Elberfeld Landgericht in the 1924 Odol case.4 The enormous significance of the Odol opinion for American and global trademark law still remains unappreciated. In honor of Annette Kur’s path-breaking and decades-long work in forging mutual understanding between the European and American trademark law systems, presented here is the first full translation into English of the Odol opinion, trademark dilution’s “ur-case.”5 1

2 3 4

5

John M. Desmarais Professor of Intellectual Property Law, NYU School of Law. Many thanks to Julie Hagedorn for careful and comprehensive help with the translation. Frank I. Schechter, The Rational Basis of Trademark Protection, 40 Harv. L. Rev. 813 (1927). Id. at 831. LG Elberfeld, 14 September 1924, 204 – Odol. For a fuller discussion of Schechter’s uses of the Odol opinion, see Barton Beebe, The Suppressed Misappropriation Origins of Trademark Antidilution Law: The Landgericht Elberfeld’s Odol Opinion and Frank Schechter’s ‘The Rational Basis of Trademark Protection’, in Intellectual Property at the Edge: The Contested Contours of IP 59 (Rochelle Cooper Dreyfuss and Jane C. Ginsburg eds., Cambridge University Press, 2014). Dr. Ludwig Wertheimer produced a partial English translation of the case in Dr. Ludwig Wertheimer, Broadened Protection of Names and Trade-Marks Under the German Law, 20 T. M. Bull. (N. S.) 75, 76–77 (1925). Schechter quoted at length from Dr. Wertheimer’s

Downloaded from https://www.cambridge.org/core. 460 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.050

A Translation of the Odol Opinion

461

As a full reading of the Elberfeld Landgericht’s Odol opinion will confirm, trademark dilution, at least in its origins, was a misappropriation law concept rather than a trademark law concept. The Odol court did not speak of anything like “search costs”6 or “imagination cost,”7 or of “blurring”8 or “tarnishment.”9 Instead, with reference to the defendant’s conduct, the essence of the court’s holding was that “[i]t is contrary to good morals to appropriate the fruits of another’s labor in this manner, knowing that the other will or could be harmed by doing so.”10 Indeed, in the final paragraph of the opinion, the Odol court explicitly stated that having resolved the matter before it under German misappropriation law, “it is not necessary to examine whether the complainant’s claims are also justified on the basis of trademark law.”11 In short, Odol was a misappropriation case in which the misappropriated property at issue happened to be a trademark. But Schechter sought in various ways to conceal the fact that antidilution protection was a form of misappropriation law. For example, though he provided a lengthy block quotation from the Odol opinion in the conclusion of his “Rational Basis” article,12 there is one sentence from the opinion that he appears deliberately to have replaced with an internal ellipsis: the sentence quoted above stating the essence of the court’s holding that the defendant had misappropriated the fruits of another’s labor. Why would Schechter have sought to obscure antidilution protection’s true genealogy? As explained in detail elsewhere,13 Schechter was writing at the time to a legal community increasingly under the influence of American Legal Realism, particularly at his intellectual home, Columbia Law School. Exponents of legal realism were openly contemptuous of the Supreme Court’s holding in the 1918 misappropriation case of International News Service v. Associated Press,14 in which the Court found that the petitioner was improperly “endeavoring to reap where it has not

6

7

8

9

10 11 12 13 14

translation in his Rational Basis article. See Schechter, supra note 2, at 832. This translation has been made by myself and Julie Hagedorn. William Landes and Richard Posner, The Economic Structure of Intellectual Property Law 207 (Harvard University Press, 2003) (describing trademark dilution by blurring as causing “an increase in consumer search costs”). See Ty, Inc. v. Perryman, 306 F.3d 509, 511 (7th Cir. 2002) (describing trademark dilution by blurring as causing a consumer to “incur as it were a higher imagination cost”). See 15 U.S.C. § 1125(c)(2)(B) (defining “dilution by blurring” as “association arising from the similarity between a mark or trade name and a famous mark that impairs the distinctiveness of the famous mark”). See 15 U.S.C. § 1125(c)(2)(C) (defining “dilution by tarnishment” as “association arising from the similarity between a mark or trade name and a famous mark that harms the reputation of the famous mark”). See Odol, infra paragraph 6. See Odol, infra paragraph 8. See Schechter, supra note 2, at 832. See Beebe, supra note 4. 248 U.S. 215 (1918).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.050

462

Barton Beebe

sown.”15 In this context, Schechter likely sought to downplay the obvious similarities between the reasoning in Odol and language in Justice Pitney’s opinion for the majority in International News Service. More generally, Schechter likely sought to obscure the degree to which the antidilution doctrine he was promulgating in his “Rational Basis” article was essentially formalist in nature.16 This may help to explain why in “Rational Basis” Schechter offered so many different formulations of the concept of trademark dilution, some of them not always consistent with each other.17 But Schechter’s contemporaries saw through his ruse. Felix Cohen’s 1935 article “Transcendental Nonsense and the Functional Approach” remains one of the most important and influential works of American Legal Realism.18 In it, Cohen levelled a scathing attack on “Rational Basis,” accusing it of “economic prejudice masquerading in the cloak of legal logic”19 and repeatedly using the phrase “vicious circle” to describe Schechter’s purportedly circular reasoning.20 Other scholars, even those sympathetic to it, similarly recognized the basic formalism of Schechter’s version of antidilution protection.21 But Schechter’s efforts at disguise did eventually succeed in at least one respect. The many different definitions of trademark dilution that appear in “Rational Basis” sufficiently confused the concept that it would become, at least for subsequent generations of trademark lawyers, a great mystery as to what exactly trademark dilution is. Further complicating matters was that Elberfeld Landgericht’s opinion in Odol formulated the concept of trademark dilution both in terms of the tort of misappropriation and a potential violation of unfair competition law. Into the void of meaning could flow incoherent concepts like trademark blurring and trademark tarnishment and the inane notion that antidilution protection ultimately seeks to minimize consumers’ “imagination costs.”22 Perhaps a return to the clear and straightforward discussion of the concept originally offered by the Odol court will help to clarify that trademark antidilution law is essentially misappropriation law – 15 16

17 18

19 20 21

22

Id. at 239. See Beebe, supra note 4, at 75 (discussing “the essentially formalist spirit of Schechter’s proposed test for dilution – which would simply ask (1) does the plaintiff’s mark qualify for antidilution protection by virtue of its inherent distinctiveness and fame, and (2) is the defendant’s mark identical or closely similar to the plaintiff’s mark?”). But see Robert Bone, Schechter’s Ideas in Historical Context and Dilution’s Rocky Road, 24 Santa Clara Computer & High Tech. L.J. 469, 489 (2008) (describing Schechter’s Rational Basis article as a “quintessential legal realist project”). See Beebe, supra note 4, at 77–79. See Felix Cohen, Transcendental Nonsense and the Functional Approach, 35 Colum. L. Rev. 809 (1935). Id. at 817. Id. at 814–815. See, e.g., John Wolff, Non-Competing Goods in Trademark Law, 37 Colum. L. Rev. 582, 602 (1937) (comparing antidilution protection to misappropriation protection in International News Service). See also Beebe, supra note 4, at 75–76 (discussing Wolff’s criticisms of Schechter’s concept of dilution). See Ty, Inc. v. Perryman, 306 F.3d 509, 511 (7th Cir. 2002).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.050

A Translation of the Odol Opinion

463

albeit one that, as Graeme Dinwoodie has emphasized, is best understood within the broader context of unfair competition law23 – and perhaps this will help clarify for American courts exactly why so many of them, “all realists now,”24 are reluctant to grant antidilution protection. Odol May Not Be Permitted To Be Used Even for Goods Entirely Dissimilar to Mouthwash.

Odol darf auch für gänzlich verschiedene Waren wie Mundwasser nicht verwendet werden.

Decision of the District Court of Elberfeld, 14 September 1924, 13.O.89/24. Facts

Entscheidung des Landgerichts Elberfeld vom 14. Sept. 1924 13. O. 89/24. Tatbestand

1 The complainant filed suit requesting the court to order the respondent to consent to the cancellation of the wordmark registered at the patent office under the file number M 35877/9b Wz in connection with the respondent’s business operation, a steel goods factory, and its goods, which include cutlery and cutting tools, to order the respondent to compensate the complainant for all damages it has incurred or will incur from the respondent’s use of the mark “Odol,” and finally to prohibit the respondent under threat of penalty from affixing the mark “Odol” to its goods and their packaging or wrapping, from placing goods thus labeled on the market, and from affixing the mark on its announcements, price lists, business letters, remittances, invoices, and the like.

1 Klägerin hat Klage erhoben mit dem Antrage, die Beklagte zu verurteilen, darin einzuwilligen, daβ das für den Geschäftsbetrieb, Stahlwarenfabrik der Beklagten und für ihre Waren Messerachmiedewaren in die Warenzeichenrolle des Patentamtes unter dem Aktenzeichen M 35877/9b Wz eingetragene Wortzeichen geloscht ¨ werde, daβ Beklagte der Klägerin allen Schaden zu ersetzen habe, der der Klägerin durch Verwendung des Zeichens „Odol“ entstanden sei und entstehen werde, endlich der Beklagten bei Vermeidung von Strafe zu untersagen, ihre Waren oder deren Verpackung oder Umhüllung mit dem Warenzeichen „Odol“ zu versehen und die so bezeichneten Waren in Verkehr zu setzen, sowie an Ankündigungen, Preislisten, Geschäftsbriefen, Empfehlungen, Rechnungen oder dergl. dieses Zeichen anzubringen.

2 The complainant asserted that the wordmark “Odol” was registered to it on 5 March 1895 on the trademark register and has been used by the complainant since that time as a mark for its mouthwash and toothpaste. On 22 March 1923, the respondent

2 Sie hat geltend gemacht, daβ für sie das Wortzeichen „Odol“ am 5. März 1895 in die Warenzeichenrolle eingetragen worden sei und von ihr Seit dieser Zeit als Zeichen für ihr Mundwasser und für ihre Zahnpasta benutzt werde. Am 22. März 1923 habe die

23

24

Graeme B. Dinwoodie, Dilution as Unfair Competition: European Echoes, in Intellectual Property at the Edge, supra note 4, 81, 101 (“[T]he European experience post-L’Oreal suggests that that characterization of the cause of action [as misappropriation] need not result in formalist, unlimited protection of property. Instead, if understood as part of unfair competition law, there is hope that it can escape the formalism that (may have) motivated Schechter and provide a measured form of relief reflecting a common normative impulse without imperiling free competition.”). Joseph William Singer, Legal Realism Now, 76 Cal. L. Rev. 465, 467 (1988).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.050

464

Barton Beebe

had filed for registration of the wordmark “Odol” for its business operation, a steel goods factory, and for its goods, which include cutlery and cutting tools and skin care and nail care articles. The registration was granted, except with respect to skin care and nail care articles. However, the complainant requested that the remaining registration also be cancelled. The respondent requested that the complaint be dismissed. Documents relating to the protection of the complainant’s wordmark are attached. The written submissions are referenced.

Beklagte das Wortzeichen „Odol“ für ihren Geschäftsbetrieb und ihre Waren, Stahlwarenfabrik und Messerschmiedewaren, Haut- und Nagelpflegegeräte zur Eintragung angemeldet. Die Eintragung sei erfolgt, abgesehen von der für Haut- und Nagelpflegeartikel verlangten, aber auch im übrigen müsse die Loschung ¨ erfolgen. Beklagte hat Abweisung der Klage beantragt. Die den Zeichenschutz der Klägerin betr. Akten sind beigefügt worden. Auf die Schriftsätze wird verwiesen.

Grounds for the Decision

Entscheidungsgründe

3 The court considers the complaint to be well-grounded in § 826 of the Civil Code.

3 Das Gericht erachtet die Klage für jedenfalls aus § 826 BGB [Bürgerliches Gesetzbuch] begründet.

4 After the complainant proved that the word “Odol” had long since been registered to it as a trademark, the respondent no longer contested this fact. It is furthermore known to the court that the complainant firm and the mouthwash it produces are referred to by the name “Odol.” It can therefore well be said that the word “Odol” developed into a catchword that has come to distinguish the complainant’s goods and has acquired an advertising strength that goes beyond the otherwise typical function of trademarks, such that when anyone reads or hears the word “Odol,” they think of the complainant’s mouthwash. These facts are of the utmost importance for the complainant firm: their goods have a good reputation, and any goods bearing the name “Odol” can presume to have the reputation among the public of being of high quality.

4 Nachdem die Klägerin nachgewiesen hat, daβ für sie seit langer Zeit das Wort „Odol“ als Warenzeichen eingetragen ist, hat die Beklagte diese Tatsache nicht mehr bestritten. Es ist ferner gerichtskundig, daβ die klägerische Firma und das von ihr hergestellte Mundwasser unter dem Namen „Odol“, Werke bezeichnet werden. Man kann daher wohl sagen, daβ das Wort „Odol“ ein Schlagwort geworden ist, das für die Waren der Klägerin kennzeichnend geworden ist und eine über die sonstige Bedeutung von Warenzeichen hinausgehende Werbekraft erlangt hat, so daβ jedermann, wenn er das Wort Odol liest oder hort, ¨ an das klägerische Mundwasser denkt. Diese Tatsachen sind für die klägerische Firma von groβter ¨ Bedeutung; ihre Ware ist als gut bekannt und eine mit dem Namen „Odol“ bezeichnete Ware hat beim Publikum die Vermutung für sich, daβ sie von guter Beschaffenheit sei.

5 As a result, the complainant has the greatest interest in ensuring that its mark not become diluted; it would lose in advertising strength if everyone were to use it to designate their goods. The trademark will become devalued, if it is used for other, wholly different goods.

5 Die Klägerin hat infolgedessen das groβte ¨ Interesse daran, daβ ihr Zeichen nicht verwässert wird; es würde an Werbekraft einbüβen, wenn jedermann es zur Bezeichnung seiner Waren verwenden würde. Das Warenzeichen wird für die Waren der Klägerin entwertet, wenn es für andere gänzlich verschiedene Waren verwendet wird.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.050

A Translation of the Odol Opinion

465

6 The respondent registered the mark for its steel goods with the obvious intention to profit from the strength of the mark for the distribution of its goods. There are, of course, innumerable fine-sounding words that the respondent could have used to designate its goods; in choosing precisely the word “Odol,” it is clear that the respondent did so because the mark has acquired an especially good reputation as a result of the efforts of the complainant. The respondent furthermore must have recognized, and surely did, that there was at least a possibility that its actions would cause damage to the complainant. Knowing this, the respondent had its mark registered nonetheless; that is to say, it behaved at least with reckless intent. It is contrary to good morals to appropriate the fruits of another’s labor in this manner, knowing that the other will or could be harmed by doing so. Also the respondent must recognize that when anyone calls his goods “Odol,” the word “Odol” loses value for the complainant.

6 Die Beklagte hat sich das Warenzeichen für ihre Stahlwaren in der offenen Absicht eintragen lassen, aus dessen Schlagkraft für die Verbreitung ihrer Waren Nutzen zu ziehen. Es gibt selbstverständlich unzählige wohlklingende Worte, die die Beklagte zur Bezeichnung ihrer Waren hätte verwenden konnen; ¨ wenn sie nun gerade das Wort „Odol“ wählte, so erhellt, daβ sie es deshalb tat, weil dieses Zeichen durch die Bemühungen der Klägerin einen besonders guten Klang bekommen hat. Sie muβte sich ferner sagen, und hat es sich auch zweifellos gesagt, es bestehe mindestens die Moglichkeit, ¨ daβ die Klägerin durch ihr Vorgehen geschädigt werde. In diesem Bewuβtsein hat die Beklagte ihr Zeichen eintragen lassen, also mindestens mit Eventualdolus gehandelt. Es verstoβt ¨ gegen die guten Sitten, sich die Früchte fremder Arbeit in dieser Weise anzueignen in dem Bewuβtsein, daβ der andere dadurch geschädigt werde bezw. geschädigt werden konne. ¨ Auch die Beklagte muβte sich sagen, daβ, wenn jedermann seine Ware „Odol“ nennt, das Wort „Odol“ für die Klägerin an Wert verliert.

7 To be sure, the parties cannot be said to stand in competition with one another due to the completely different nature of the goods they sell. This, however, is beside the point. The respondent advertised its products by means of a word with strong appeal that had acquired this appeal precisely through the years-long and extensive activity of the complainant. Paragraph § 1 of the Law against Unfair Competition can also serve to support the claim, because this legal provision does not require that the company acting contrary to good morals stand in competition with the company seeking an injunction and damages. Instead, it is enough that the company against which claims have been asserted intervenes in a manner that hinders competition between two competitors (see Rosenthal Note 63 at § 1). Here, however, the complainant’s ability to compete, specifically its ability to compete

7 Freilich stehen die Parteien wegen der ganzlichen Verschiedenheit der von ihnen vertriebenen Waren nicht im Wettbewerb miteinander. Dies tut aber nichts zur Sache. Die Beklagte hat für ihre Ware Reklame gemacht, indem sie sich eines zugkräftigen Wortes bediente, das gerade durch die langjährige und ausgedehnte Tätigkeit der Klägerin allein seine Zugkraft erlangt hatte. Auch der § 1 des Gesetzes gegen den unlauteren Wettbewerb kann den Klageantrag stützen, weil diese Gesetzesbestimmung nicht voraussetzt, daβ der gegen die guten Sitten Verstoβende im Wettbewerb gerade mit dem ihn auf Unterlassung und Schadenersatz in Anspruch nehmenden steht, vielmehr genügt es, daβ der in Anspruch genommene in den Wettbewerb zweier Konkurrenten hindernd eingreift. (s. Rosenthal Note 63 zu § 1). Hier aber wird die Wettbewerbfähigkeit

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.050

466

Barton Beebe

with other mouthwash manufacturers, would be damaged if the significance of the mark were to be diminished.

der Klägerin, ihre Fähigkeit mit anderen Mundwasserfabrikanten zu konkurrieren, geschädigt, wenn die Bedeutung ihres Zeichens geschmälert würde.

8 The court finds itself in agreement with the decision published in the Juristisches Wochenblatt, Volume 24, Double Issue 9/10, page 722, of the District Court of Chemnitz from 11 April 1923, which further pursued and refined the legal principles set out by the Reichsgericht (see R. G. 66, 239). Seeing as the reference to § 826 of the Civil Code and the Law on Unfair Competition is sufficient, it is not necessary to examine whether the complainant’s claims are also justified on the basis of trademark law; instead, the matter was resolved as above.

8 Das Gericht befindet sich bei einer Entscheidung in Uebereinstimmung mit der in der JW. 24 Doppelheft 9/10 Seite 722, abgedruckten Entscheidung des Landgerichts Chemnitz vom 11. April 23, die die reichsgerichtlichen Grundsätze (s. R. G. 66, 239) weiter verfolgt und vervollständigt hat. Es braucht, da die Bezugnahme auf § 826 BGB und das Gesetz über den unlauteren Wettbewerb genügt, nicht erst untersueht zu werden, ob auch auf Grand des Warenzeichengesetzes die Anträge der Klägerin gerechtfertigt, sind; vielmehr war wie geschehen zu erkennen.

Contributed by R. A. [Attorney] Dr. Rosenthal, Hamburg.

Mitgeteilt von R. A. Dr. Rosenthal, Hamburg.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.050

40 The Relationship Between the Unfair Competition Regime and IP Law Gustavo Ghidini1 and Giovanni Cavani2

Jeder Rechtsgegensatz ist ein Weltanschauungsgegensatz H. Kelsen

40.1 unfair competition: an old international regulation ... As known, whereas the EU enacted a European antitrust regime a long time ago, there is no EU regulation of unfair competition (UC)3. Specific statutes only exist in some Member States, such as Germany, Spain and Italy. In others, like the UK, common law principles apply. In France, the homeland of the legal category of unfair competition (and of the term itself: concurrence déloyale), the Courts have

1 2 3

Professor Emeritus, University of Milan; LUISS University, Rome. Professor of IP and Competition Law, University of Modena-Reggio Emilia. On the topic see, ex multis, R. Hilty and F. Henning-Bodewig, Law Against Unfair Competition: Towards a New Paradigm in Europe?, Springer 2007; F. HenningBodewig, Unfair Competition Law: European Union and Member States, Kluwer Law, 2006. Similarly, in the US there is no federal statutory law of general unfair competition. Sec. 43(a) of the Lanham Act in fact is not a general federal rule preventing any act of UC; it is limited to a prohibition against some form of infringement of unregistered marks, trade names, or trade dress, false advertising and product disparagement. On the development of the interpretation of Sec. 43(a), see J. T. McCarthy, Lanham Act § 43(A): The Sleeping Giant is Now Wide Awake, (1996) 59(2) Law and Contemporary Problems 45. At page 51 the author affirms that: ‘Section 43(a) can never be a federal “codification” of the overall law of “unfair competition” because of its inherently limited wording. Advocates of expanded federal law dreamed that section 44 could accomplish this goal because of its expansive language of “protection against unfair competition.” But even the most expansive language of section 43(a) is limited to a prohibition against some form of false designation or misleading representation. Such a prohibition is quite a bit narrower in scope than the whole genus of “unfair competition” law.’.

Downloaded from https://www.cambridge.org/core. 467 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

468

Gustavo Ghidini and Giovanni Cavani

applied the general norm on damaging tortious conducts4 since the midnineteenth century. The only provision which has a transnational, hence also ‘European’ scope, is embodied in Articles 10bis and 10ter of the Paris Convention (PC). These rules lay down a set of guiding principles whose vagueness allowed national legal orders to interpret and apply the PC provisions, in particular those of Article 10bis as amended by the 1925 Hague Revision,5 in ways that often came to reflect sharply divergent jurispolitical inspirations (as we will stress below, comparing Continental approaches with Anglo-Saxon ones).6

40.2 ... and business oriented Article10bis PC establishes a general criterion of competitive ‘unfairness’ of explicit professional imprint: the “usages honnêtes dans le commerce et l’industrie” (Art. 10bis (2)). This criterion openly diverges from the general concept of ‘unjustice’, that in Continental civil law has traditionally defined tortious conducts in harmony with the civic values of the social community as a whole (see e.g, the reference to ‘general’ good customs, gute Sitten, in § 826 BGB, Art. 1240 French civil code, Art. 1151 Italian civil code). The ‘professional’ inspiration – and textual reference to business standards7 – is further revealed by the sole mention of professional (i.e. industrial and trade) associations legitimate to enforce UC law (Art.10ter (2)). This normative background, and its jurispolitical rationale, fostered by the major industrial and trade associations and Chambers of Commerce, obviously influenced by the position of the leading represented firms, well explains the extensive (albeit not unanimous) success among judges and scholars (especially those intensely engaged in legal practice, more in Italian and Belgian experience than in other Continental countries) of a quite debatable line of interpretation of the scope and application of UC provisions.

4

5

6

7

Article 1240, formerly 1382, of the Civil Code – offspring of the Roman law rule of neminem laedere: lex Aquilia, anno 286 a. C. Which introduced in Art. 10 bis – originally inserted in the PC by the Brussels Revision of 1900 and containing a general provision concerning the repression of UC – a definition and examples of acts of UC. The Revision Conference of Lisbon of 1958 introduced paragraph 3 no. 3 on fake and misleading communications. This is also why so far no attempt has successfully established a modern, well-structured EU regulation of unfair competition. We hope that possible future attempts will follow the blueprint of the 2004 ‘breakthrough’ German Act Against Unfair Competition as revised in 2015 (see 40.9 below). A reference absent in § 1 of the German Act Against Unfair Competition of 1909, for which § 826 BGB served as a blueprint. Both provisions refer to the boni mores (gute Sitten). On this see further 40.9 below.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

The Unfair Competition Regime and IP Law

469

40.3 consequent over-protectionist interpretations/applications It is a line that in substance has led, and still often leads – at different degrees of influence from country to country – to the protection of goodwill as “quasi-property”, providing protection to commercial activities and innovations even in absence of IP protection or passing-off. In particular, it often occurs (especially in Continental Europe, as we said) that in legal practice and court decisions the evocation of acts of “unfair competition” works as a substitute, so to speak, for absent or no longer applicable or tout court inapplicable IP protection. This frequently occurs in cases dealing with imitations of competitors’ products (at times, also of commercial methods or advertising strategies), or imitations of functional or aesthetic features not covered or no longer covered by patents, utility models or design rights. In substance, when this happens, the recourse to UC functions as a tool for circumventing either the expiry or the absence of IPR protection (whether de facto, or by law: think, for example, of commercial methods, declared unpatentable by Article 52(2)(c) EPC). Further, that evocation is often used to bypass PC’s own specific norm on product imitation, Article 10bis (3)(No. 1), which prohibits (not every act of imitation but) only all acts of imitation (if and when) producing confusing effect. The bypass is achieved by relying on the general criterion of fairness in competition (“honest commercial uses”, Article 10bis (2)), of which the former provision features as a particular example, and not as an independent rule.8 This allows prohibiting even non-confusing imitations (‘lookalikes’) of the product’s ‘dress’, even in the absence of any likelihood of confusion between the rival products. As we will further note in 40.8 below, the UK adopts a more pro-competitive approach according to which, in a nutshell, there is ‘no misappropriation without misrepresentation’. Less recently, until the implementation of Directive 2006/114/EC, which restricts the freedom of comparative advertising, that evocation was also used in some countries to prevent even the truthful criticism of competitors’ products, thus allowing bypass of the explicit limitation to ‘false allegation’ set by Article 10bis (3)(No. 2) PC.9

8

9

Note the expression ‘Notamment devront être interdits. . .’ (‘In particular shall be prohibited. . .’): incipit of Art. 10bis (3); followed by the exemplification of more frequently recurring ‘dishonest’ practices, such as, as stated, passing-off, disparagement of competitors, misleading consumer information. It might be of some interest to recall that even after the adoption of the above-mentioned Directive the CJEU in L’Oreal v. Bellure (C-487/07) interpreted Art. 4(g) (ex-Art. 3a(h)) of the Directive in the sense that it forbids a comparative advertisement informing consumers that a product is similar or identical to another, better known, product, notwithstanding the fact that the former is not covered by a patent or other IP right, and the imitation is non-confusing. With the surprising result that whilst manufacturing and selling that product is perfectly lawful, it

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

470

Gustavo Ghidini and Giovanni Cavani

40.4 the ‘moral’ justification A specific ‘moral’ justification, which drew on the notion of ‘honest practices in industrial or commercial matters’ for defining ‘fairness’ in competitive relations, was traditionally offered in doctrinal and jurisprudential dicta under the theory of ‘misappropriation’ of the achievements (and related goodwill) obtained by other firms through labour and investments – ‘toils and expenses’, a phrase first adopted, as far as we know, by the Greek poet Pindar to describe the strong engagement of the participants in the Olympian games. Under this theory, at times tautologically identified with the principle of ‘unjust enrichment’,10 a trader is regarded as competing fairly only if it does so by virtue of ‘its own’ performance (‘Leistungswettbewerb’ or ‘competition on [one’s own] merits’).11 In some countries, the courts rely on the same theory when proscribing ‘parasitic competition’.12 In other words, this theory, however labelled, assumes that only those who have the capacity (and thus ‘the merit’) to innovate – instead of imitating – can legitimately enter the market and operate on it. This theory, also supported by scholars like Rudolf Callmann,13 originated in the US, with the Supreme Court decision of 1918 in International News Service v. Associated Press,14where Justice Pitney, in delivering the majority opinion, ruled that she/he who has invested to create something has a (quasi) property on the result of her/ his investment. But, as Justices Brandeis and Holmes argued in their dissenting opinion, ‘the fact that a product of the mind has cost its producer money and labor, and has a value for which others are willing to pay, is not sufficient to ensure to it this legal attribute of

10

11

12

13 14

would not be lawful to say so! See V. Di Cataldo, The Trade Mark with a Reputation in EU Law – Some Remarks on the Negative Condition “Without Due Cause”, [2011] IIC 833–845. See e.g. R. Callmann, He Who Reaps Where He Has Not Sown: Unjust Enrichment in the Law of Unfair Competition, (1942) 55 Harv. L. Rev. 595. For a harsh criticism of the ‘concept’ of competition on the merits, see also OECD, Competition on the Merits: Policy Round Table (2005); see further below, in this section. See, significantly and exemplary, in Italian jurisprudence: Milan Court May 12, 2016 in Giurisprudenza Annotata di Diritto Industriale (G.A.D.I.) 2016, 938 ff and February 17, 2016, 682 ff. That general ‘moral’ justification (the adjective ‘parasitic’ itself evokes a reprehensible and disqualifying behaviour) has often, wholly inappropriately, been anchored in the biblical reference to ‘reap where one has not sown’. But in truth that biblical reference is a gross ‘misappropriation’ of the religious texts, whose correct interpretation goes in the opposite direction: that of evoking God’s generous promise to Israel ‘to give you a land with large and beautiful cities which you did not build, houses full of all good things, which you did not fill, hewn-out wells which you did not dig, vineyards and olive trees which you did not plant’ (Deuteronomy, 6:10–11; almost identical words in Joshua, 24:13). Thus, the parable of the Sower (see, in the New Testament, John, 4:38) does not disqualify at all the conduct of the Reaper! In a respectful transposition to a market economy it would rather support a general ‘open access’ approach. See note 10, supra. 248 U.S. 215 (1918).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

The Unfair Competition Regime and IP Law

471

property’. Subsequently, however, the theory has repeatedly been dismissed by the Supreme Court, in the famous cases Sears, Roebuck & Co. v. Stiffel Co.,15 Compco Corp. v. Day-Brite Lighting, Inc.16 and, more recently, in Bonito Boats, Inc. v. Thunder Craft Boats, Inc.17 and Dastar Corp. v. Twentieth Century Fox Film Corp.18 And rightly so.

40.5 systemic inconsistencies and contradictions First, leaving aside for the moment the impact that it has on competition and ‘coherence’ with law and economics arguments on competitive dynamics, the interpretative line stemming from this theory contrasts with basic legal-logical and systemic principles. In particular: (a) on a ‘horizontal’ plane, the theory violates the principle of noncontradiction (i.e. coherence) between distinct areas of law concerning the same substantive phenomena. It is the principle that provides the guideline for addressing the general problem of ‘overlaps’, to which Annette Kur dedicated poignant reflections. Here, the contradiction is indeed all too evident, as the rules on UC are invoked effectively to extend the statutory term of exclusive protection under a design registration or a patent, or to ex novo institute an exclusionary (erga omnes) protection compensating for, even without time limits, the absence of IPRs. Moreover, in the cases last described, an internal contradiction emerges within the area of UC. In fact, by relying on the general clause of ‘usages honnêtes’, the interpretation here opposed circumvents also the principle provided by Article 10bis (3)(No. 1), according to which the imitation of a competitor’s products is forbidden (not per se, but) if it causes confusion with the latter (had PC allowed the extension of the prohibition to imitate independently from confusion, it would have, as minimum, put a ‘e.g. confusingly’). The lack of coherence also extends: (b) to the ‘vertical’ plane, i.e. the interpretation of ordinary norms in contrast with overarching legal principles. We refer, in particular – (i) to the principle of the numerus clausus of IP rights which is itself an expression of the exceptional character of statutory exclusionary limits (which of course the legislator – but not the interpreter! – can extend: but in any case not so largely to transform the exception in the rule) to the general freedom to imitate, as long as the 15 16 17 18

376 U.S. 225 (1964). 376 U.S. 234 (1964). 489 U.S. 141 (1989). 539 U.S. 23 (2003), on which see further below, at 40.5.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

472

Gustavo Ghidini and Giovanni Cavani

imitation does not cause confusion among consumers. We will return to this contradiction, which intersects with the anticompetitive profiles of the criticized interpretative line;19 and (ii) to the systemic guideline according to which in a market economy focused on freedom of competition norms of ordinary rank should be interpreted in coherence with the positive constitutional principle of freedom to conduct a business (Art. 16 ChFREU). Now, this guideline requires that when provisions of ordinary rank allow different interpretations, the interpreter must privilege the solution which is more consistent, or at least less in contrast, with said overarching principle. In sum, the criticized approach leads to a restriction of the general freedom to imitate, which is nothing else than an instance of the general and overarching principle of freedom to compete, as well stated by Justice O’Connor in Bonito Boats:20‘From their inception, the federal patent laws have embodied a careful balance between the need to promote innovation and the recognition that imitation and refinement through imitation are both necessary to invention itself, and the very lifeblood of a competitive economy.’ Even later, the freedom to copy what is not protectable, or no longer protected by a patent or copyright, has been reaffirmed by the US Supreme Court in Dastar,21 and by no less than Antonin Scalia. As the great conservative – and libertarian – judge wrote: ‘The right to copy, and to copy without attribution, once a copyright has expired, like the right to make [an article whose patent has expired] including the right to make it in precisely the shape it carried when patented – passes to the public.’ (Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225, 230 (1964); see also Kellogg Co. v. National Biscuit Co., 305 U.S. 111, 121–122 (1938)). ‘In general, unless an intellectual property right such as a patent or copyright protects an item, it will be subject to copying.’ (TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23, 29 (2001)). The rights of a patentee or copyright holder are part of a ‘carefully crafted bargain’ (Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 150–151 (1989)) under which, once the patent or copyright monopoly has expired, the public may use the invention or work at will and without attribution. In other words, the criticized interpretative approach would mean overturning the fundamental principle according to which ‘the temporal limitation of patents and copyrights augments the public domain because once the patent or copyright expires, the patented or copyrighted work cannot be reappropriated; it is forever a part of the public domain’.22 19 20 21 22

See below, at 40.6. Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 146 (1989). Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23 (2003). R. Posner, Intellectual Property: The Law and Economics Approach, (2005) 19(2) J. of Econ. Perspectives 57–73, at 60 (emphasis added).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

The Unfair Competition Regime and IP Law

473

40.6 anti-competitive consequences Apart from being systemically incoherent with the overall legal framework of IP law (let alone the fake biblical evocation), the ‘misappropriation’ theory also runs contrary to solid arguments of law and economics in defence of competitive dynamics. In a liberal economic order, governed by the constitutional principle of freedom of enterprise and (hence) competition on equal legal terms, one cannot but question the soundness of adopting, as a legal standard of competitive fairness, the pseudo-moralistic assumption whereby the firm that opens new industrial or commercial paths should enjoy the right to bar competitors from following those paths and should vice versa open in their turn different, autonomous new paths. Why on earth should firms that are not capable of opening, or choose not to open, new tracks not have the right to enter or remain in a given market? What is more,23 expanding the prohibition to imitate beyond the boundary of counterfeiting and passing-off reduces, not enhances, the efficiency of dynamic competition. This, not only because such expansion would limit consumer choice by depriving consumers of a second or third best option. But also because the reduction of the number of follow-on competitors (also further discouraged by the vagueness, hence uncertainty of the lawfulness of even non-counterfeiting nor confusing follow-on initiatives they might be considering) brings in itself the risk of creating, to the benefit of the market leaders, rent-seeking positions – positions supported by a protection significantly superior (not subject to time limits as well as inclusive of non-patentable subject-matter and non-confusing competing products) to that stemming from the mere statutory enforcement of counterfeiting and passing-off. In sum, the criticized theory reduces both the incentive for market leaders to keep rolling up their sleeves to innovate further and the chances of followers to compete on quality and price (‘price’ is quite often the real, not-so-hidden reason for the criticized over-protectionist approach) to the benefit of consumers, as well as to develop follow-on innovation.

23

As concerns the modern debate on (non-counterfeiting nor confusing) imitation as a factor propulsive of dynamic competition, see for example C. S. Baradello and A. Salazzaro, The Role of Imitation in Global High-Tech Product Development, Symphonya: Emerging Issues in Management, No. 1, 2012, available at: http://symphonya.unicusano.it/article/view/2012.1 .05baradello.salazzaro; O. Shenkar, Copycats: How Smart Companies Use Imitation to Gain a Strategic Edge (Harvard Business School Publishing Corporation, 2010); I. Abel, From Technology Imitation to Marker Dominance: The Case of IPod, Competitiveness Review: An International Business Journal , 2008, 257–274; S. P. Schnaars, Managing Imitation Strategies: How Later Entrants Seize Markets from Pioneers, The Free Press, 1994;M. K. Bolton, Imitation versus Innovation: Lessons to be Learned from the Japanese, Organizational Dynamics, 1993, 30–45; T. Levitt, Innovative Imitation, Harvard Business Review, 1966, 63–70.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

474

Gustavo Ghidini and Giovanni Cavani

40.7 follows: the equivocal argument of protection of ‘investments’ Finally, it cannot be argued that the position supported here would discourage ‘investment’: it simply does not reward investment that leads to ‘non IPRs-deserving’ objective results. Indeed, on the basis of a systemically coherent interpretation of IP law, ‘investments’ per se cannot entitle the investor to an exclusionary protection. In this we disagree with the recent dictum of the Court of Justice of the European Union (CJEU) about a general (generic, indeed) ‘investment function’: a dictum, related to the discredited ‘if value then right’ doctrine,24 which has so far been limited to wellknown trademarks, but whose underlying ‘philosophy’ risks empoisoning the whole field of IP law. IP can perform a dynamic (and pro-competitive) mission only if it protects, within the statutorily set time (and other) limits, not investments as such, but specific intellectual achievements,25 statutorily qualified as worthy of such protection. If costly research does not lead to an innovative result, the patent cannot be granted, no matter how big the investment was. So it is for works that do not qualify as copyrightable, or trademarks that do not reach distinctive character – whatever the investment their creation required. The regime of UC must ‘respect’ that IP mission, to which those limits are functional.

40.8 the correct systemic framework Thus, the anti-competitive and consumer-unfriendly features of the criticized approach marry with its described intrinsic systemic contradiction/incoherence with the exceptional nature of IPRs in an overall legal order overarched by the principle of freedom of initiative and competition; a legal order where IPRs are and must remain islands in a sea where even small boats are free and encouraged to navigate (pirates excluded, of course). We then recommend abandoning the ambiguous slippery terrain of the reference to ‘morals’26 in UC law, because it only serves as a 24

25

26

See F. S. Cohen, Transcendental Nonsense and the Functional Approach, (1935) Columbia L. Rev. 809, 815. Drexl fully agreed, in a personal correspondence, and Loewenheim confirms (about copyright, but with a self-evident general reach): ‘Das Urheberrecht schützt die geistige Leistung und nicht die Investition’: U. Loewenheim, reviewing a book by M. Bisges, in Zeitschrift fu¨r Geistiges Eigentum, 2016, 568, 571. For those fond of ‘ethical’ approaches to the regulation of economic phenomena, the ‘misappropriation’ doctrine can be harmlessly (albeit superfluously) evoked only to assert an ‘ethic’ ground of the positive legal prohibitions of counterfeiting of patents, models, trademarks, and acts producing risk of confusion (passing-off ). No more: not, in particular, to provide a justification to the several recalled cases where, as said, the reference to UC is (mis)used to

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

The Unfair Competition Regime and IP Law

475

cloak to cover the over-protectionist interest of the market leaders, and the adoption of a concept of competitive fairness aimed at a goal of market efficiency based on the model of effective (workable) competition also between leaders and followers, which creates stimuli to innovation and enhances consumer welfare. This is more than just wishful thinking. Indeed, the Anglo-Saxon jurisprudence, as the first in Europe, moved beyond this sectorial business approach. In the US the misappropriation doctrine had little support from the courts either.27 The courts generally refused to extend proprietary protection beyond the IPRs established by law, outside whose confines only fraudulent operations on the market were held to be unlawful (‘no misappropriation without misrepresentation’). This line of thought is put in a nutshell by Jacob J (as he then was): ‘There is no tort of taking a man’s market or customer. Neither the market nor the customers are the plaintiff’s to own. There is no tort of making use of another’s goodwill as such.’28, 29

40.9 a ray of light: the new (2015) german law on unfair competition (uwg) Moreover, witnessing that ‘the times they are a-changin’, the approach supported here has recently found quite an important positive law support in the German Act Against Unfair Competition of 2004, as amended in 2015 – a reform scarcely commented on in the international doctrine. The new German Act – which incorporates the key concepts of the UCP Directive, and in ultimate systemic perspective ‘reaps the fruits’ of the doctrinal and policy seeds sowed by scholars like

27

28 29

bypass the limits of IP law, in part e.g. either to ‘strike’ imitations of functional or aesthetic features not covered or no longer covered by patents or registered models, or to enforce nonconfusing imitations (‘lookalikes’) of products’ ‘dress’ (see 40.3 above). See 40.2 above, and B. Beebe, The Suppressed Misappropriation Origins of Trademark Antidilution Law: The Landgericht Elberfeld’s Odol Opinion and Frank Schechter’s ‘The Rational Basis of Trademark Protection’’, in R. C. Dreyfuss and J. C. Ginsburg (eds.), Intellectual Property at the Edge: The Contested Contours of IP, Cambridge University Press, 2014, 59. Hodgkinson Corby Ltd and Another v. Wards Mobility Services Ltd, [1995] RSR 169, 174–175. This approach is possibly a by-product of the emergence, after World War II, of a new vision of capitalism inspired by the postulates of the ‘social market economy’ (soziale Marktwirtschaft). As concerns competition, this vision translated into two basic guidelines: (a) the adherence to the principle of freedom of competition (as acknowledged in its adversarial toughness); and (b) the acknowledgement that the limits of this freedom should be fixed where private business interests may collide with social welfare – meaning both collective consumers’ interests and the general interest in a system based on diffuse competition. In the words of the Italian Supreme Court (Cass. 1185/1997), the ‘guiding value’ for unfair competition is ‘the freedom of competition, which is both the object and extent of the protection in question: a value to defend – it must be added – against the risk that the individual freedom of enterprise can kill another’s freedom of competition’ (emphasis added).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

476

Gustavo Ghidini and Giovanni Cavani

W. Hefermehl30 – significantly modifies the law of 1909. First, in § 1, it defines the Act’s purpose as protecting competitors, consumers and other market stakeholders, as well as the public interest in undistorted competition. Then, in § 3, it defines as ‘unfair’ (note: instead of the former concept of gute Sitten the new Act postulates the concept of unfairness – Unlauterkeit) those practices that, while not complying with ‘professional diligence’, are apt to ‘materially distort the economic behaviour of consumers’.

40.10 conclusion: some specific implications of the anti-protectionist stance The repudiation of the ‘misappropriation approach’ has some precise consequences, whose common denominator is that UC must not serve as substitute or supplement to the absence of a statutory IP protection, nor bypass the limit of confusion expressly set by Article 10bis PC for imitations. (i) UC protection cannot in principle be invoked where a competitor reproduces technical features which have not been patented by the inventor or which are not protected by a patent any more. According to the criticized approach that still prevails in some jurisdictions, the unauthorized reproduction of technical features can amount to an ‘unfair misappropriation’. The paradoxical result is that since protection against unfair competition lasts of course in perpetuity, the unpatented components would be protected against reproduction imitation for longer than ones claimed in the patent. We think that such a reproduction is in principle lawful, since (a) unpatented technology cannot be the object of exclusive exploitation, and (b) in the case of imitation of internal parts no risk of confusion is normally conceivable. (ii) UC cannot be evoked against the imitation of product features which were protected by an expired patent or (the more frequent case) an expired registered design or model, nor against an unregistered design after expiry of the three-years term of protection allowed by the Community Design Regulation. (iii) Once, or even better before, a design right is expired it is up to the former owner, not the competitor, to add ‘distinctive’ and nonfunctional variations to that shape – however made well distinguishable by a specific trademark on the product and/or its packaging. Otherwise, the former registrant would de facto perpetually prolong 30

On the general principles of constitutional rank overarching the regulation on unfair competition, see W. Hefermehl, Wettbewerbsrecht, 8th ed., C. H. Beck, 2008 – in particular Part I, on the fundaments of the discipline.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

The Unfair Competition Regime and IP Law

477

its monopoly on the exact object of the expired registered design. Again the words of Justice Scalia in Dastar should be recalled: ‘the right to make [an article whose patent has expired,] including the right to make it in precisely the shape it carried when patented – passes to the public’.31 In turn, and of course, competitors of the former registrant should abstain from reproducing those ‘capricious’ variations: otherwise, they would act unfairly: by causing confusion.

31

See above, note 18; emphasis added.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.051

41 Comparative Advertising: Does Trade Mark Law Over- or Under-Protect the Average Consumer? A Couple of Recent Examples of Asian Jurisdictions Going their Own Way David Llewelyn1

In the leading judgment in the recent UK Supreme Court decision in WarnerLambert, a patent validity and infringement case that has provoked divergent views on what actually it decided, Lord Sumption cautioned that it is important to guard against the over-ready transfer of concepts from one legal system to another, in which the legal context may be different.2

In today’s world in which legislation and legal decisions from a multitude of jurisdictions are available almost instantaneously to researchers and practitioners all around the globe on the press of a few keys and clicks of a mouse, such words of caution are both welcome and apposite at a time when even countries formerly of like minds and background seem to have started to go their own way.

41.1 all change in uk trade mark law An important export by the United Kingdom to many countries in the past 30 years has been of various provisions in the Trade Marks Act 1994 (1994 UK Act). The 1994 UK Act is a piece of legislation that led the authors of the 13th edition of practitioners’ work Kerly’s Law of Trade Marks and Trade Names (Kerly), the first after the coming into force of the new Act, to start its “Introduction” chapter thus: Since the last edition of this work [in 1986, although there was a rather concise Supplement by Sir Robin Jacob, David Kitchin QC and James Mellor in 1994], registered trade mark law in the United Kingdom has undergone fundamental 1

2

Professor (Practice), School of Law, Singapore Management University; Professor of Intellectual Property Law, King’s College London; Managing Director, David Llewelyn & Co LLC (Singapore). Warner-Lambert Company LLC v. Generics (UK) Ltd t/a Mylan and another [2018] UKSC 56 at [85].

Downloaded from https://www.cambridge.org/core. 478 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

Comparative Advertising and Trade Marks in Asia

479

changes and to a large extent has started from scratch. The trade marks regime begun by the Trade Marks Act, 1875, which was itself founded upon earlier equitable principles, has been replaced exactly a century after the first edition of Kerly by a system much of the detail of which is dictated by developments at the European Community and international level.3

As is now well-known, it took the English judges some time to accept the extent of those fundamental changes and the fact that the law had “started from scratch”, although it is fair to say that they had done so wholeheartedly by the time of the Brexit referendum in 2016. Indeed, the “Introduction and Future Developments” chapter in the 16th edition of Kerly, published in 2018, commented in anticipation of Brexit: there has been substantial progress to resolving many of the moot points in the regime, although there remain areas where criticism is both justified and necessary: in those areas the UK approach will become the only directly relevant approach as the UK distances itself from some of the influences of continental European thinking.4

41.2 otiose, but still alive On the other hand, it is noteworthy that in the United Kingdom, the very country in which it has its fons et origo,5 section 10(6) of the Trade Marks Act 1994 – one of the provisions that have been transposed verbatim (or practically so) in a number of other jurisdictions including Hong Kong, India and Ireland – has been (to adopt the words of Kerly again) “rendered otiose by the CJEU’s decisions which exclude lawful comparative advertising from the scope of trade mark infringement”.6 Instead, as made clear inter alia in the O2 v. Hutchison7 and L’Oréal v. Bellure8 3

4

5

6

7 8

The international developments that dictated change were the Madrid Protocol and, somewhat late in the day, Article 6 bis of the Paris Convention 1883. Kerly’s Law of Trade Marks and Trade Names (16th ed., Sweet & Maxwell, 2017), Mellor et al., eds., at [1-018]. It can be safely assumed that even such a far-sighted commentator as Sir Robin Jacob, when stating in 2011 in the foreword to the 15th edition of Kerly that ‘[s]ettled law is akin to the system of fixed stars. Commercial men can steer by them . . . By way of contrast, the greater the ambit of uncertain law, of planets of unpredictable orbit, the more commercial men will find themselves at sea without clear guidance as to where they can steer and where there may be rocks’, did not have in mind the possibility of the UK leaving the European Union when he proceeded to bemoan the fact that in trade mark law “there are fewer fixed stars and more wandering planets than in most branches of the law”, at [v]. As noted by Lord Strathclyde on behalf of the Government in the Parliamentary debates on the Trade Marks Bill in 1994: “The government have been persuaded that there is no harm in comparative advertising . . . provided it makes fair use of a registered trade mark for the purpose of informing the public” (HL Official Report, Public Bill Committee on the Trade Marks Bill, Second Sitting, 18 January 1994, col. 42). Kerly’s Law of Trade Marks and Trade Names (15th ed., Sweet & Maxwell, 2011), Mellor et al., eds., at [15-059]. C-533/06 [2008] RPC 33. C-487/07 [2010] RPC 1.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

480

David Llewelyn

judgments of the Court of Justice of the European Union (CJEU), the legality of this kind of advertising is to be dealt with primarily by reference to the criteria contained in the Directive concerning misleading and comparative advertising;9 use which complies with the requirements of the Directive is not liable to affect any of the functions of the registered mark and does not infringe. Section 10(6) was a home-grown provision in the 1994 UK Act that, in the words of the 13th edition of Kerly, “reflects something of a sea change” in the attitude to comparative advertising in the UK, which had been curtailed under earlier legislation from 1938.10 However, after the subsequent intervention of the CJEU, Lord Justice Jacob (as he then was) was typically robust in his criticism: It is a pointless provision . . . It should be repealed as an unnecessary distraction in an already complicated branch of the law.11

Nevertheless, a provision in identical (or practically identical12) terms remains part of trade mark law in many jurisdictions, including Hong Kong. A recent decision of the Hong Kong Court of First Instance, PCCW-HKT Datacom Services Limited & Ors v. Hong Kong Broadband Network Limited,13 cut away the European encrustations and applied the provision as it stands in section 21 of the Trade Marks Ordinance,14 that is in terms identical to section 10(6) of the UK Act. This section of the Hong Kong Ordinance permits the use of competitors’ trade marks for identifying goods or services of the trade mark owners in accordance with honest practices in industrial or commercial matters. In addition, it sets out, in section 21(2), that the factors the court may consider when determining “honest practices” include whether:  the use takes unfair advantage of the trade mark;  the use is detrimental to the distinctive character or repute of the trade mark; or  the use is such as to deceive the public. 9

10

11

12

13 14

Directive 2006/114/EC. In the EU Trade Mark Regulation, the types of infringing use in Article 9(3) have been amended to include use of a sign in comparative advertising “in a manner contrary to [the] Directive”. Kerly’s Law of Trade Marks and Trade Names (13th ed., Sweet & Maxwell, 2000), Kitchin et al., eds., at [13-070]. O2 v. Hutchison 3G Ltd [2007] RPC 19, at [58]. The 15th ed. of Kerly (2011 – note 6, supra) went so far as to state, at [15-059], that if it had any effect, it “would be ultra vires”. Or, to put it in trade mark law language, the differences are so insignificant they may go unnoticed by the average lawyer: see LTJ Diffusion v. Sadas [2003] FSR 34, CJEU. For example, section 94 of the New Zealand Trade Marks Act 2002 provides a specific exception for use of a sign in comparative advertising, although that term is not defined. It also adopts verbatim most of the wording from section 10(6) of the 1994 UK Act. [2018] HKCFI 2037. Order No. 35 of 2000.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

Comparative Advertising and Trade Marks in Asia

481

It was noted by the judge,15 Mimmie Chan J, that the enumerated factors contained in section 21(2) do not include whether the use discredits or denigrates the registered trade mark or is misleading or creates confusion, unlike Article 3a of the European Union’s 1997 Misleading Advertising Directive16 that specifically deems such use misleading. The defendant, telephone and internet service provider Hong Kong Broadband Network, relied on section 21 to justify its use in advertising its mobile telephone services of five statements addressed to consumers using the services of the plaintiff’s group of companies. These statements used the plaintiff’s registered trade marks “HKT”, “PCCW”, “eye” and “電訊盈科” to draw an unfavourable comparison with the defendant’s services: exhorting the plaintiff’s customers to “say goodbye to bloated monthly fees” and describing the plaintiff in Cantonese as “大食”(translated in the judgment as, literally, big eater or gluttonous).17 The defendant relied on a quartet of English cases decided before the CJEU decisions referred to above: Barclays Bank PLC v. RBS Advanta,18 Vodafone v. Orange Personal Communications Services Ltd,19 Cable and Wireless PLC v. British Communications PLC20and British Airways PLC v. Ryanair Ltd.21 Unsurprisingly, the plaintiff argued that these cases should be ignored as it was clear from subsequent CJEU decisions including O2 v. Hutchison and L’Oréal v. Bellure that they had not given proper consideration to the second half of the proviso in section 10(6), namely whether “the use without due cause takes unfair advantage of, or is detrimental to, the distinctive character or repute of the trade mark”. However, the judge noted that the purpose of section 21 is to permit comparative advertising and held (surely correctly) that the provision should be interpreted on its face in the light of all the evidence and bearing in mind Article 17 of TRIPS, under which the legitimate interests of the owner of the trade mark and of third parties should be taken into account. Thus, although reference to decisions of courts in other jurisdictions with the same or similar legislation may offer a guide: the fact remains that the Directives of the European Community, and the decisions of the ECJ (and of the English court), and their interpretation of the Directives, have no binding effect in Hong Kong.22

15

16

17 18 19 20 21 22

PCCW-HKT Datacom Services Limited & Ors v. Hong Kong Broadband Network Limited [2018] HKCFI 2037, at [22]. Directive 97/55/EC of the European Parliament and of the Council of 6 October 1997 amending Directive 84/450/EEC concerning misleading advertising. [2018] HKCFI 2037 at [7]. [1996] RPC 307. [1997] FSR 34. [1998] FSR 383. [2001] FSR 541. PCCW-HKT Datacom Services Limited & Ors v. Hong Kong Broadband Network Limited [2018] HKCFI 2037, at [22].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

482

David Llewelyn

Notwithstanding this, the judge agreed with Laddie J in Barclays Bank v. RBS Advanta that these matters should be decided by considering their impact on the reasonable reader, rejecting any suggestion that such a reader in Hong Kong would be less aware than her counterpart in the UK of the wiles and wheezes of those promoting and advertising the merits of their goods and services. Indeed, she considered that: The average consumer in Hong Kong can be said to be just as savvy to advertising language, exaggeration and trade puffery as the average consumer in the UK – if not the more so, bearing in mind the colloquial and often colourful and exaggerated terms frequently and commonly used in the Chinese media in Hong Kong. Just by way of illustration, a reasonable consumer and reader of advertisements in Hong Kong would be impervious to straplines describing a residential unit “with unrivalled seaview” for sale, well knowing that in reality, that description would mean in most cases no more than a narrow view of water, seen through gaps in buildings fronting the unit in question. When real estate in Kowloon is advertised against a backdrop of the French Riviera, consumers in Hong Kong can only be skeptical and hardened against advertising puffery. To such extent, the average reader and consumer is well educated in the inherent exaggeration of advertising in Hong Kong, and will not be inclined to take advertising language seriously, without a healthy dose of skepticism.23

In the light of this, the judge (in terms redolent of the style of both the English judges in the above quartet of cases, Laddie and Jacob JJ) considered the meaning of the words and concluded: given the fact that the average consumer and reasonable reader of the Advertisement in Hong Kong is more than used to hyperbole and exaggeration, the words complained of by the Plaintiffs, namely “bloated” and “大食”, mean no more than “expensive”, expressed in sensational and colored [sic] language, as is usual in advertising. The average consumer of electronic communications service in Hong Kong is accustomed to price comparison and assertions of good, better or the best deals in monetary terms, claims made in a highly competitive and fast moving market. The average consumer reading the Advertisements will not expect to find, or look for, a serious comparison of details, nor an itemized analysis of the terms offered. The reasonable person realises that he/she is not reading a research paper. The attraction and focus of the Advertisements and straplines are simply the headline price for a general category of service. The message conveyed by the Advertisements is simply that a customer will get a cheaper price of $99 per month, by subscribing to the Defendant’s home telephone and fiber optic broadband service, instead of using the Plaintiffs’ same services at their expensive (“bloated/ 大食”) monthly fees. That is the “take home message”. I do not consider that the average reasonable reader will take the statements in the Advertisements seriously, to carry any broader meaning, that the Plaintiffs’ 23

Ibid., at [35].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

Comparative Advertising and Trade Marks in Asia

483

prices will “invariably” be cheaper than the Defendant’s, in all cases, irrespective of the location, term of the plan, and combination of the services sought. Nor do I consider that to the average reader hardened to advertising language in Hong Kong, “bloated” or “大食” will be considered to carry any derogatory or sinister meaning, of the Plaintiffs cheating their customers. The words simply poke fun at the Plaintiffs’ prices, and will be reasonably understood as such.24

On the plaintiffs’ argument that the use was such as to give the defendant an “unfair advantage”, she was unwilling to focus solely on the word “advantage”, having earlier noted that the very purpose of comparative advertising means that the other party’s goods with which the advertiser is comparing are never put in the most favourable light. This has the consequence that “unfair advantage” needs to be interpreted in the context of a provision that aims to permit such advertising, provided the legitimate interests of the trade mark proprietor are taken into account.25 As a result, she held: Even if it can be said that by referring to the Plaintiffs’ Marks, the use of the Marks in the Advertisements took “advantage” of the Marks, by riding on their coat-tails and reputation, despite the contrast made in the differences in the pricing characteristics of the services of the Plaintiffs and the Defendants (sic), the message conveyed in the Advertisements is substantially true, and not misleading: that the Plaintiffs’ services are more expensive than the Defendant’s. For that reason, any advantage arguably taken of the Marks in the comparative advertising is not “unfair”.26

This can of course be contrasted with the somewhat controversial view expressed by the CJEU in L’Oréal v. Bellure27, that “an advantage may be unfair even if it causes no harm to the distinctive character or repute of the mark or more generally to the interest of the trade mark proprietor”.28 Instead, it is more in line with the English Court of Appeal’s decision in Whirlpool v. Kenwood,29 that a mere advantage is insufficient for a finding of infringement – there must be some factor making unfair the advantage taken. In a recent English Court of Appeal decision, Argos Limited v. Argos Systems Inc,30 it was noted by Floyd LJ of the CJEU’s view: 24 25

26 27

28 29 30

Note 22, supra, at [41]–[42]. Ibid., at [32], referring with approval to the observations to like effect by Lewison J (as he then was) at first instance in O2 Holdings Ltd, O2 (UK) Ltd v. Hutchison 3G Ltd [2006] EWHC 534 at [145]: “The case of comparative advertising is not quite the same [as the normal detriment case under s 10(3) TMA 1994]. It is to be expected that in robust competition competitors will not be over-sensitive to their competitors’ marks.” Note 22, supra, at [57]. [2010] RPC 1. Another finding by the Hong Kong judge that stands in stark contrast to the findings of the CJEU is that at [61]: “A statement of truth cannot be detrimental to the reputation of the Marks, even if the Plaintiffs find it uncomfortable to be confronted with a substantial truth.” [2010] RPC 1, at [50]. [2010] RPC 2, at [136]. [2018] EWCA Civ 2211.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

484

David Llewelyn

“Such a general statement is comprehensible in the context of a case where there is, or is likely to be, a transfer of the image of the proprietor’s reputation to the defendant’s goods or services.”31 However, the learned judge, with whom Lord Kitchin and Sir Colin Rimer agreed, proceeded to reject the contention that all that was needed was an advantage: “So to hold would be to empty the word ‘unfair’ of any meaning.”32 This seems consistent with the CJEU’s views in L’Oréal v. Bellure, although is perhaps expressed in a clearer fashion. This judgment, which contrasts so noticeably with the more protectionist attitude of many European courts, may usefully be compared also with the treatment of comparative advertising in the trade marks legislation of another Asian jurisdiction, Singapore, that introduced in 2004 a specific fair use exception into its trade marks legislation, to replace section 27(6) that was identical to section 10(6) of the 1994 UK Act. This new “fair use” defence, in section 28(4) of the Trade Marks Act, states: (4) Notwithstanding section 27, a person who uses a registered trade mark does not infringe the trade mark if such use— (a) constitutes fair use in comparative commercial advertising or promotion; (b) is for a non-commercial purpose; or (c) is for the purpose of news reporting or news commentary.

Section 28(4) was considered by the Singapore High Court for the first time in Allergan, Inc v. Ferlandz Nutra Pte Ltd.33 In this case, the plaintiff, Allergan, alleged that the defendant, Ferlandz Nutra, had infringed registered mark “LATISSE” under section 27(1) (which is in terms identical to section 10(1) of the 1994 UK Act) by using it in a promotional brochure comparing Allergan’s Latisse and the latter’s Lassez eyelash-enhancing products. George Wei JC emphasised34 that the court no longer had to consider local and UK case law decided under section 27(6) as it is no longer relevant. At the same time, however, Singapore courts are not bound to follow US case law on fair use. Instead, in the judge’s view, it is best for local courts to develop their own approach to section 28(4); a provision that, unlike UK and EU law which adopts the Paris Convention wording of “honest practices in industrial and commercial matters”, introduces into Singapore trade mark law the concept of “fair use”. On the facts, the judge found that the defendant’s use of the “LATISSE Mark” was both misleading and likely to give rise to confusion in the minds of end-users as to whether there is a trade connection between the products: the court is entitled to have regard to both factors in reaching a decision on whether the use was fair. Section 28(4)(a) recognises that a balance must be struck between 31 32 33 34

[2018] EWCA Civ 2211 at [103]. Ibid., at [108]. [2016] 4 SLR 919. Ibid., at [147].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

Comparative Advertising and Trade Marks in Asia

485

the rights of the trade mark owner and the right of competitors to enter the market and to provide the public with alternative choices. Section 28(4)(a) permits a fair use of the trade mark for commercial comparative advertising or promotion. By doing so, the public interest in fair competition is safeguarded. Whether or not the inclusion of a false statement negates fair use must depend on the nature of the statement (eg, does it significantly undermine the basis of the comparison) and whether the defendant knew or ought to have known that it was false. In those cases where there is a dispute as to whether the statement is misleading, this is best resolved by asking whether the average consumer would find the statement misleading in a material manner. In coming to this decision I hasten to add that control of misleading advertising is a much wider subject than trade mark infringement and comparative advertising.35

Notwithstanding this, the judge found in the circumstances that the defendant’s use did not fall within the comparative advertising defence and was therefore an infringement. Both the Hong Kong and Singapore decisions highlight the extent to which both jurisdictions are putting clear water between themselves and the position under EU law. At this juncture in the development of the European Union, it is perhaps apposite to wonder whether the same may happen with the United Kingdom in the future!

35

Note 33, supra, at [150].

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.052

c

(Un-)fairness

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Conclusion

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

42 Transition and Coherence in Intellectual Property Law Graeme B. Dinwoodie1 and Ansgar Ohly2

42.1 introduction The nature and content of intellectual property law, which is heavily contingent on the state of technology and on social and market developments, have always been subject to ongoing transitions. How those transitions are effected and the shape they take is crucial to the ability of intellectual property (IP) to achieve its stated goals and to provide the necessary climate for investment in creativity, innovation and brand differentiation. The first half of this book explored those questions, as discussed more fully below in section 42.2. The challenges presented by transitions include a challenge to the coherence of the IP system, both within and across time and space, as well as within and across distinct IP regimes. Of course, this proposition glosses perhaps too quickly over a preliminary question: to what extent is coherence desirable? There are good reasons why different intellectual property rights take different forms. Would imposing an identical or similar rule on types of intellectual property designed to achieve different results for different reasons be anything other than surface coherence, masking substantive confusion? And the notion that different countries might appropriately adopt divergent rules is well-accepted as an important recognition both of national sovereignty and the adage that “one size does not fit all.” Yet, in both respects these sound justifications for difference can, if they foreclose proper assessment of the nature and reasons for difference, potentially undermine the purposes of all the regimes in question. This dilemma, which is explored more fully below in section 42.3, was addressed by authors in the second half of this book. 1

2

Global Professor of Intellectual Property Law, IIT Chicago-Kent College of Law; Visiting Professor of Law, University of Oxford. Prof. Dr. jur., LL.M. (Cantab.), Chair of Private Law, Intellectual Property and Unfair Competition at the Ludwig Maximilian University of Munich; Visiting Professor of Law, University of Oxford.

Downloaded from https://www.cambridge.org/core. 489 , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

490

Graeme B. Dinwoodie and Ansgar Ohly

42.2 transitions Transitions may occur through incremental evolution or they may represent sharp departures from the status quo. Which approach is chosen by those seeking to change the law, or which result ensues from a transition that occurs, turns on a number of variables, including the form taken by a transition.

42.2.1 Forms of Transition Sharp transitions can occur inter alia through a reversal of a lower court ruling, the explicit over-ruling by a superior court of prior binding decisions, or the repeal or amendment of an existing statutory provision. These represent clear forms of transition; indeed, some might regard “transition” as connoting a less disruptive change in law than some of these examples might suggest. Alternatively, a transition can be the result of a court applying existing law to a new set of facts, thus making new law or extending the scope of an applicable principle. Here, the fact that a change has been made may not be immediately apparent, as the transition is as much a change in the law in practice as it is a real change in the applicable legal rule. Indeed, one could regard a transition of this kind as an effort to preserve coherence and prevent disparate results occurring in analogous situations that differ only in the particular technology or social setting at issue.3 This form of transition is not unusual in intellectual property law where, because of its close connection to technological development, the law often appears in a state of constant flux, requiring continuing adaptation. The doctrinal content of copyright law in particular has frequently been a response to new technological developments.4 Indeed, this has been true from the very inception of copyright – the development of privileges in the Stationers Company in London, which are seen as a clear forerunner of copyright, was a response to the introduction of the printing press to England by William Caxton.5 And this remains true today, with transitions contemplated in chapters in this book to take account of the growth in the potential of artificial intelligence as well as the opportunities to make use of data.6 3

4

5

6

Cf. Goldman v. Breitbart News Network, LLC, 302 F. Supp. 3d 585, 586 (S.D.N.Y. 2018) (“That technology and terminology change means that, from time to time, questions of copyright law will not be altogether clear. In answering questions with previously uncontemplated technologies, however, the Court must not be distracted by new terms or new forms of content, but turn instead to familiar guiding principles of copyright.”). See Sony Corp. v. Universal City Studios, 464 U.S. 417, 430 (1984) (“From its beginning, the law of copyright has developed in response to significant changes in technology.”). See Harry Ransom, The First Copyright Statute 17–53 (1956) (discussing the role of The Stationers’ Company in giving “quasi-legal status to literary property” in England during the period from 1559 to 1601). See supra Chapters 12, 13, 14 and 16 (Gangjee, Khan, Luginbuhl and Drexl).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

491

As technology changes, incumbent stakeholders often contest the right to control exploitation of intellectual property in new technological settings.7 Depending upon the political economy of the relevant lawmaking unit, this contest might take the form of requests for specific legislative intervention tied directly to the technological development. For example, when technology developed to allow families to “sanitize” potentially offensive content in movies, the US Congress wrote a specific exemption allowing such technology notwithstanding that it potentially created a derivative work without the permission of the copyright owner.8 Legislatures pursuing this route, however, run the risk that technology quickly renders the provision irrelevant, unless courts are willing to read detailed language as reflective of a broader principle. And if the legislation is so specific as to reference particular technology, this will be impossible for courts to do.9 Alternatively, legislatures enacting generally applicable statutory reforms have often included provisions that purport on their face to apply to the entire range of technological options, including those not envisaged by the legislature at the time of enactment.10 Courts are peppered with arguments for technological neutrality, namely, that as technologies evolve the same basic principles should extend to all.11 Indeed, it is not unusual for fault lines in “new technology” cases to open up 7

8

9

10

11

See Ben Depoorter, Technology and Uncertainty: The Shaping Effect on Copyright Law, 157 U. Pa. L. Rev. 1831, 1847 (2009). See, e.g., 17 U.S.C. § 110(11) (immunizing: “the making imperceptible, by or at the direction of a member of a private household, of limited portions of audio or video content of a motion picture, during a performance in or transmitted to that household for private home viewing, from an authorized copy of the motion picture, or the creation or provision of a computer program or other technology that enables such making imperceptible and that is designed and marketed to be used, at the direction of a member of a private household, for such making imperceptible, if no fixed copy of the altered version of the motion picture is created by such computer program or other technology.”). See Monica Zhang, “Fair Compensation” in the Digital Age: Realigning the Audio Home Recording Act, 38 Hastings Comm. & Ent. L.J. 145, 147–48 (2016) (“the AHRA has not been able to keep up with the rapid pace of technological advancement due to Congress’s passage of narrow statutory language; proving that Congress’s view was shortsighted by a general consensus that the AHRA is irrelevant despite of its well-intentioned bargain. A robust copyright regime can only be as effective as it is adaptable to the unpredictable technological environment in which it exists because copyright is a reactive body of law.”). See, e.g., 17 U.S.C. § 102(a) (“Copyright protection subsists, in accordance with this title, in original works of authorship fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device”) (emphasis added); Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on Copyright and Related Rights in the Digital Single Market and Amending Directives 96/9/EC and 2001/29/EC, OJ L 130, 17.5.2019, 92, Recital 3 (“Relevant legislation needs to be future-proof so as not to restrict technological development.”). See Lyria Bennett Moses, Recurring Dilemmas: The Law’s Race to Keep Up with Technological Change, U. Ill. J.L. Tech. & Pol’y 239, 270 (2007) (“Technological neutrality is often seen as a positive feature of legislation, since being independent of any particular technology ensures the law will apply effectively and fairly in different technological contexts. However, the use of

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

492

Graeme B. Dinwoodie and Ansgar Ohly

along the question whether regulating the new technology should be left to the courts or Congress.12 Sometimes, this question is affected by the nature of the statutory scheme that the legislature has constructed; at least in common law regimes, the legislature might consciously leave matters for judicial development.13 And courts are able to develop existing principles by reference not only to prior case law but also to legislative policy choices in adjacent areas of law that signal the direction that the law should take.14

12

13

14

seemingly technology-neutral language will not always achieve this goal, at least in areas where technologies are evolving.”) Compare Am. Broad. Companies, Inc. v. Aereo, Inc., 573 U.S. 431, 449 (2014) (Breyer J) “(Aereo and many of its supporting amici argue that to apply the Transmit Clause to Aereo’s conduct will impose copyright liability on other technologies, including new technologies, that Congress could not possibly have wanted to reach. We agree that Congress, while intending the Transmit Clause to apply broadly to cable companies and their equivalents, did not intend to discourage or to control the emergence or use of different kinds of technologies”) with id. at 462 (Scalia J., dissenting) (“It is not the role of this Court to identify and plug loopholes. It is the role of good lawyers to identify and exploit them, and the role of Congress to eliminate them if it wishes. Congress can do that, I may add, in a much more targeted, better informed, and less disruptive fashion than the crude ‘looks-like-cable-TV’ solution the Court invents today.”). Similar debates can be found in trademark law. Compare Graeme B. Dinwoodie and Mark D. Janis, Lessons From the Trademark Use Debate, 92 Iowa L. Rev. 1703, 1710–11 (2007) (“Congress has, on occasion, explicitly prompted further expansion of rights, but on the whole, Congress has been comfortable in allowing courts to develop the basic contours of trademark protection consistent with its purpose, codifying (or implicitly endorsing by silence) those developments where necessary”) with Stacey L. Dogan and Mark A. Lemley, Grounding Trademark Law Through Trademark Use, 92 Iowa L. Rev. 1669, 1670 (2007) (“The question raised by the trademark use cases, as we view it, is whether courts should countenance a radical departure from that traditional model without specific instruction from Congress. We think they should not.”). See H.R. Rep. No. 1476, 94th Cong., 2d Sess. 66, reprinted in 1976 U.S. Code Cong. & Ad. News 5659, 5680 (“there is no disposition to freeze the [fair use] doctrine in the statute . . . courts must be free to adapt the doctrine to particular situations on a case-by-case basis”). Compare Graeme B. Dinwoodie, The Common Law and Trade Marks in an Age of Statutes, in Lionel Bently, Catherine Ng and Giuseppina D’Agostino (eds.), The Common Law of Intellectual Property: Essays in Honour of Professor David Vaver (Hart Publishing, 2010), 331 (arguing that “despite substantial legislative intervention, both Congress and the Supreme Court appear content that the development of trademark and unfair competition law in the United States remain heavily dependent on common law lawmaking by the courts”) with Michael Grynberg, Things Are Worse Than We Think: Trademark Defenses in A "Formalist" Age, 24 Berkeley Tech. L.J. 897, 903 (2009) (“The doctrinal realities of modern trademark law make reform efforts based on judicial action unlikely to succeed.”). See Erven Warnink BV v. J Townend & Sons (Hull) Ltd, [1979] AC 731, 743 (Lord Diplock) (“Where over a period of years there can be discerned a steady trend in legislation which reflects the view of successive Parliaments as to what the public interest demands in a particular field of law, development of the common law in that part of the same field which has been left to it ought to proceed upon a parallel rather than a diverging course”); Joined Cases C-236/08 to 238/08, Google France v. Louis Vuitton Malletier [2010] ETMR 30, para. 86 (shaping trade mark liability by reference to principles of transparency in the EU Directive on electronic commerce).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

493

42.2.2 Institutional Choices As implicitly suggested above, the form adopted by a particular transition may also vary by the institution that makes it happen. For example, courts are typically constrained from the kind of radical re-orientation of the law that legislatures might direct to cope with social or technological change. That political reality affects whether courts or legislatures (or other institutions, such as administrative agencies) are best situated to effect particular transitions in intellectual property law.15 Court-driven transitions are likely to be more incremental, as courts attend to their particular role in a democratic society.16 These transitions may not always be labelled as “change” by those who effect them; as much change is effected through judicial efforts to distinguish and reconcile past decisions as through their explicit repudiation. Legislators are more apt to stress what is new in any statutes they enact, although political exigencies sometimes compel even democratically elected legislatures to describe as a “clarification” that which we know is an instrument of change. A recent example of this can be found in the EU Digital Single Market Directive, where the EU legislature substantially expanded the scope of the communication to the public right as regards online content-sharing service providers, but noted in a Recital that it was “appropriate to clarify in this Directive that online content-sharing service providers perform an act of communication to the public or of making available to the public when they give the public access to copyright-protected works or other protected subject matter uploaded by

15

16

In some instances, a successful transition may require both institutions. Even when the legislature has sought to make what it sees as a clear transition, the success of that move (and the coherence of the result) depends upon proper judicial interpretation and application of the law. In turn, this is not simply a function of consistent readings by superior courts but also consistent application by lower courts. See supra Chapter 4 (Derclaye) (examining the success of national courts in interpreting EU design law). Of course, the undue intervention of appellate courts can also in certain circumstances retard the success of any substantive policy embodied in the law. See supra Chapter 3 (Austin). Moreover, the obligation to protect existing reliance interests may vary depending upon whether the change is the result of legislative amendment or judicial interpretation. Compare supra Chapter 1 (Arnold) with supra Chapter 2 (Bently). In the case of legislative change, the legislation may itself make provision for how to handle the treatment of rights that existed or events that occurred prior to the effective date of the new legislation. But as Richard Arnold described in his chapter, this is sadly not always done. See supra Chapter 1 (Arnold). The Court of Justice of the European Union (CJEU) has recognized that the failure may force courts to read the change in some circumstances as being prospective only. See Case C-21/18, Textilis Ltd v. Svenskt Tenn AB, EU:C:2019:199 (CJEU 2019). However, as Lionel Bently persuasively argues, the extent of real change often being effected by decisions of the Court of Justice counsels in favor of a less clear-cut line. He offers three suggestions: (1) treat such rulings as prospective only; (2) implement transitional measures as one would with legislative change; and (3) take account of reliance interests in formulating remedies. Which of these options to pursue may turn in part on whether the incorrect assumption about the interpretation of EU law was reasonable.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

494

Graeme B. Dinwoodie and Ansgar Ohly

their users.”17 Moreover, courts are constrained in their capacity to change intellectual property law by the cases or issues that come before them (although they may also be presented with issues, especially technological issues, in new or singular contexts that have not yet registered with the legislature). Legislatures paint on a blanker canvas, although (even setting aside political pressures) national parliaments operate subject to their own constraints.

42.2.3 International Commitments and Constraints Transitions are on occasion required because of international treaty provisions. Indeed, for several decades towards the end of the twentieth century the vast majority of copyright legislation in the United States (US) and elsewhere was motivated in part by international considerations.18 These considerations included implementing treaty commitments, but also concerns about maximizing protection for US works abroad by responding to reciprocity conditions in foreign legislation or consciously seeking to adopt a leadership role in the international community. The international intellectual property regime has over time come to intrude ever more intensely in national lawmaking, in part because that regime has itself undergone a transition. Once a commitment to national treatment backed up by a set of a few loosely enforced minimum standards, the international system has been made more effective by incorporation within the WTO dispute settlement system.19 And the range of obligations that international law imposes on countries has increased substantially. This has happened in part because global trade and digital communication technologies have made inter-dependence more significant and the need for international regulation more pressing. But it also reflects the range of international institutions that have turned their attention to intellectual property law, as IP reaches into so many areas of contemporary existence.20 These institutions may have regulation or protection of intellectual property as part of their core mission, but this is no longer necessarily so. Indeed, general international institutions and instruments – whether concerned with human rights or

17

18

19 20

See Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on Copyright and Related Rights in the Digital Single Market and Amending Directives 96/9/ EC and 2001/29/EC, OJ L 130, 17.5.2019, 92, Recital 64. These types of framing are not restricted to legislatures. As Barton Beebe recounts in his chapter, Frank Schechter sought to portray his proposal for dilution protection in a different light by selective quotation from the German Opinion from which he claimed to draw inspiration. Beebe corrects this deficiency by offering the first full-length translation of the Odol Opinion into English. See supra Chapter 39. See Graeme B. Dinwoodie, Incorporating International Norms in the Development of Contemporary Copyright Law, 62 Ohio State Law Journal 733 (2001). See supra Chapter 8 (Ricketson); supra Chapter 6 (Frankel). See supra Chapter 7 (Gervais).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

495

investment law – have become important features of the ever-complex international regime.21 While international law rarely dictates a particular policy choice, within the European Union (EU), Member States are substantially constrained by the dictates of EU law (both intellectual property instruments and general principles of EU law, including fundamental rights). Even when implementing Directives in national law rather than complying with directly effective Regulations, national lawmakers have scant room for maneuver (which some might say assists in ensuring coherence). The Court of Justice has adopted canons of interpretations, and indeed offered substantive interpretations themselves, that seem determined to further transitions commenced but not completed by the EU legislature.22 And the Court is apparently less constrained by political forces that have caused impasses in the legislative process.23 Thus, at the EU level too, the choice of making a transition through legislation or judicial activism is manifest. Although the EU legislature has created a sizeable body of intellectual property law (other than in patent law), much of the harmonization has been effected by the assertive nature of the Court of Justice case law. Indeed, the need to bring an issue within EU competence has clearly informed judicial interpretation of copyright Directives.24 21

22 23

24

Sometimes, intellectual property fits too uneasily into the general regime, prompting carve-outs of intellectual property. For example, the Hague Conference on Private International Law has struggled for over twenty years to try and incorporate intellectual property law within a system of jurisdiction and recognition of judgments. See supra Chapter 10 (Heinze). As Christian Heinze explains, these efforts have not borne fruit. Intellectual property is excluded from the scope of the recently concluded Hague Convention on the Recognition and Enforcement of Foreign Judgments in Civil or Commercial Matters. See Hague Convention on the Recognition and Enforcement of Foreign Judgments in Civil or Commercial Matters (concluded 2 July 2019), Art. 2(1)(m), at https://www.hcch.net/en/instruments/conventions/full-text/?cid=137. However, that failure did prompt two important “soft law” instruments (one emanating from the US and one from Europe). See American Law Institute, Intellectual Property: Principles Governing Jurisdiction, Choice of Law, and Judgments in Transnational Disputes (ALI, 2008); European Max-Planck-Group on Conflict of Laws In Intellectual Property (CLIP), Conflict of Laws in Intellectual Property: The CLIP Principles and Commentary (Oxford University Press, 2013). Annette Kur was an adviser on the ALI Project and a member of the CLIP working group. See supra Chapter 2 (Bently); supra Chapter 5 (Dreier). See Christophe Geiger, The Construction of Intellectual Property Law in the European Union: Searching for Coherence, in C. Geiger (ed.), Constructing European Intellectual Property Law (Edward Elgar, 2012), 3, 13 (noting the Court’s advantage in overcoming political impasse). See Case C-610/15, Stichting Brein v. Ziggo BV, EU:C:2017:99 (Advocate-General, CJEU 2017). The Advocate-General (Szpunar) noted in his Opinion in Ziggo that the Commission and several Member States had argued that liability for sites of the type at issue in that case should be resolved under the domestic legal systems of the Member States (using doctrines such as accessory liability). The Advocate-General rejected that argument for pragmatic reasons: “Such an approach would, however, mean that liability, and ultimately the scope of the copyright holders’ rights, would depend on the very divergent solutions adopted under the different national legal systems. That would undermine the objective of EU legislation in the

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

496

Graeme B. Dinwoodie and Ansgar Ohly

42.2.4 New Agents of Transition The choice presented between judicial and legislative change is not new. But it may no longer fully capture the agents of change in intellectual property law. The nature, scope and enforcement of intellectual property rights are increasingly determined by commercial actors, particularly by online intermediaries. The prediction that the internet would result in wholesale disintermediation has proven more complex than anticipated.25 The choices made by such actors, often on a transnational basis that reflects the nature of their business, are sometimes the reason for changes in intellectual property law on the ground. Yet, those choices are – unlike legislative or judicial choices – not readily subject to public scrutiny, and made in a crucible of self-interest. It is an ongoing challenge to ensure scrutiny of such ordering, and provisions designed to ensure the public structuring of such behavior are often more aspirational than effective.26 As intellectual property claims proliferate, those transnational private actors increasingly rely on technological mechanisms to make their policy choices scaleable.27 At their peak, Google received around a billion takedown requests a year, roughly 20 million requests per week. Responding to such a large number of requests requires automation, potentially further obscuring public scrutiny.28 And, of course, those requests are not all individual, manual requests; automation can generate the requests just as much as it can process them. Artificial intelligence may over time further improve the consistency of response, but that too is dependent upon scrutiny. Indeed, artificial intelligence systems are being deployed throughout

25 26

27

28

relatively abundant field of copyright, which is precisely to harmonise the scope of the rights enjoyed by authors and other rightholders within the single market. That is why the answer to the problems raised in the present case must, in my view, be sought rather in EU law.” See id., para. AG3. See supra Chapter 9 (Ricolfi). See supra Chapter 12 (Khan); see generally Graeme B. Dinwoodie, Private Ordering and the Creation of International Copyright Norms: The Role of Public Structuring, 160 Journal of Institutional and Theoretical Economics 161, 189 (2004) (“If open government is a mantra of liberal democratic polities, then transparent private ordering must equally be the obligations of those who purport (beneficently) to arrive at the same results through private governance.”). It will as a result be interesting to see how EU Member States implement and enforce the provision of the recent Single Market Directive which regulates the obligation that “Member States shall provide that online content-sharing service providers put in place an effective and expeditious complaint and redress mechanism that is available to users of their services in the event of disputes over the disabling of access to, or the removal of, works or other subject matter uploaded by them.” In particular, Article 17(9) requires that “Complaints submitted under the mechanism . . . shall be processed without undue delay, and decisions to disable access to or remove uploaded content shall be subject to human review.” See Digital Single Market Directive, Art. 17(9). See supra Chapter 12 (Khan). Indeed, such transparency in the use of automation might be important to ensure protection of important social values at the edge of intellectual property. See supra Chapter 15 (Richardson).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

497

the intellectual property eco-system, including in efforts to make IP offices more efficient and predictable in processing applications.29

42.2.5 External Forces Norms external to intellectual property law might drive a transition. As noted above, intellectual property law is increasingly shaped by international obligations, and these might not necessarily be provisions of international intellectual property law narrowly understood. The recent decision of an investment arbitral panel assessing the legality of Uruguay’s prohibition on the use of tobacco trademarks relied in part on the content of draft regulations prepared under the auspices of the Framework Convention on Tobacco Control (FCTC), which was adopted by the World Health Organization in 2003.30 The dispute-settlement panel of the World Trade Organization tackling the TRIPS compatibility of Australia’s similar plain-packaging legislation likewise clearly was influenced by these instruments.31 So too fundamental human rights norms, both at the national and international level, might cause or dictate the nature of a transition. And intellectual property law remains situated against a backdrop of general principles of private law which can occasionally be relied upon to effect change in intellectual property rules.32 The nature of transitions that emanate from these different external sources might appropriately vary. International law tends to confine the type of change that must occur rather than dictate a particular transition, and it might also limit the extent to which transitions can occur without regard to the protection of vested reliance interests. But international law, even after TRIPS, has more limited enforcement potential; the US still has not changed either its copyright or trademark law to comply with adverse decisions handed down by the WTO dispute settlement body. In contrast, national fundamental rights commitments often have a constitutional character that makes them impervious to legislative over-ride, but as a result national courts tend to avoid relying on constitutional provisions if possible. (In contrast, the 29 30 31

32

See supra Chapter 13 (Gangjee). See Philip Morris Brands v. Uruguay, ICSID Case No. ARB/10/7 (2016). See Australia – Certain Measures Concerning Trademarks, Geographical Indications and Other Plain Packaging Requirements Applicable to Tobacco Products and Packaging, Reports of the Panels, WT/DS435/R, WT/DS441/R, WT/DS458/R, WT/DS467/R (June 28, 2018); see also Australia – Certain Measures Concerning Trademarks, Geographical Indications and Other Plain Packaging Requirements Applicable To Tobacco Products And Packaging, WT/DS435/ AB/R, WT/DS441/AB/R (Appellate Body 9 June 2020). Many such rules are then later accommodated within the body of intellectual property law. See supra Chapter 22 (Hilty) (discussing the source of the exhaustion principle in German copyright law); see also British Leyland Motor Corp. v. Armstrong Patents Co [1986] UKHL 7 (relying on UK land law principle of non-derogation from grant to create an exception from copyright for car spare parts). Indeed, some scholars have argued that reference to other areas of private law is important to prevent the ever-increasing fragmentation of private law. See supra Chapter 36 (Norgaard).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

498

Graeme B. Dinwoodie and Ansgar Ohly

Court of Justice of the European Union (CJEU) draws on fundamental rights a lot, perhaps to allow the Court to create a uniform EU law solution.) General principles of national private law both have direct enforcement capacity in litigation and can be legislatively overriden.

42.2.6 Theoretical Metrics Finally, scholars and policy makers have over recent years developed new theoretical justifications for intellectual property law, such as the growing importance of trade theories of comparative advantage at the international level and law and economics at the national level. These changes have altered what we must find to justify changes and the metrics we apply to assess the wisdom or success of that change. As to the latter, as the intellectual property community has grown transnationally allowing an ever more vibrant comparative debate, we have encountered new alternatives from which to source possible reforms and against which to test our efforts at transition.33

42.2.7 Transitions in Transition Each of these developments is itself a transition. And each represents another variable to consider in determining how best to make changes in intellectual property law. This dynamic environment – transitions in the means of transition – is seen not only where technological innovation occurs. It is a phenomenon found elsewhere. Consider globalization. Globalization may prompt the need for a change in the law by rendering territorial assumptions unstable.34 But it also offers a new means of effecting change as it makes global institutions, multinational companies, and territorially unconfined technologies options through which to implement change. And, even at the subglobal level, transborder activity has generated innovations in the creation and enforcement of regional unitary rights, a process that brings its own sets of issues.35

42.3 coherence 42.3.1 Why Coherence? While there can be no doubt that intellectual property law is subject to constant transition, it is less evident that it should be coherent. To be sure, construing the law 33

34 35

See supra Dogan, “Annette Kur: Toward Understanding” (discussing role of Annette Kur). This is not to say that all countries, even with a common background, must inevitably converge around a single norm. As David Llewelyn notes, when legislation and legal decisions from a multitude of jurisdictions are available almost instantaneously, it is important to guard against the over-ready transfer of concepts from one legal system to another. See supra Chapter 41 (Llewelyn). See supra Chapter 10 (Heinze) (discussing the shift from territoriality to universality). See supra Chapter 11 (Torremans).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

499

as a coherent system is one of the primary goals of the Continental European doctrinal methodology, particularly of German “Rechtsdogmatik”. But this methodology is a child of the age of civil law codifications. Outside the application of civil law codes and in an age of swift technological and economic change it is less clear that coherence can or even should be achieved. When the German Civil Code entered into force in 1900, intellectual property law was not included,36 and there were no plans to create an “Intellectual Property Code”. Patents, trade marks and copyright were seen as specific areas of law that catered to rather different needs, and the great international conventions (the Paris and Berne Conventions) distinguished between copyright and industrial property.37 While today it has become common to regard intellectual property as a distinct area of economic law, few practitioners or academics can claim to be experts on all subareas of intellectual property law. Patent law and copyright law in particular have become extremely diverse. Indeed, one may even doubt the coherence of the respective treatment of ICT patents and biotechnology patents, or copyright in traditional categories of work and copyright in software.38 Moreover, modern intellectual property policy has partly lost its codification impetus. It reacts to problems created by the swift technological and economic change and increasingly produces sui generis and ad hoc solutions. The thicket of EU Regulations and Directives in intellectual property law is strikingly different from the well-ordered garden which the drafters of the great civil law codifications

36

37

38

There are probably two reasons why intellectual property was not included. First, the drafters drew on the Roman law notion of property and were sceptical about “novel” rights which did not have any classical roots. Second, the German patent, trade mark and copyright Acts had just entered into force and were not in need of any reform. See Ansgar Ohly and Andreas Sattler, Bürgerliches Recht und geistiges Eigentum, in Diethelm Klippel, Martin Löhnig and Ute Walter (eds.), Grundlagen und Grundfragen des Bu¨rgerlichen Rechts (Gieseking, 2016), 165, 168–69. See Paris Convention for the Protection of Industrial Property, March 20, 1883, as last revised at Stockholm, July 14, 1967, 21 U.S.T. 1583, 828 U.N.T.S. 305; Berne Convention for the Protection of Literary and Artistic Works, September 9, 1886, as last revised at Paris, July 24, 1971, 1161 U.N.T.S. 3. See Dan L. Burk and Mark A. Lemley, The Patent Crisis and How the Courts Can Solve It (University of Chicago Press, 2009). Of course, courts in software copyright cases frequently protest that they are simply applying traditional principles. See, e.g., Computer Assocs. Int'l, Inc. v. Altai, Inc., 982 F.2d 693 (2d Cir. 1992). But the application clearly intensifies particular issues and makes the claim of equivalence hard to sustain. See Lotus Dev. Corp. v. Borland Int’l, Inc., 49 F.3d 807, 820 (1st Cir. 1995) (“Of course, one could still read the statute as a congressional command that the familiar doctrines of copyright law be taken and applied to computer programs, in cookie cutter fashion, as if the programs were novels or play scripts. Some of the cases involving computer programs embody this approach. It seems to be mistaken on two different grounds: the tradition of copyright law, and the likely intent of Congress.”). Indeed, insofar as courts have strived to approach copyright law horizontally, one effect of including software within copyright might have been to alter the way we think about traditional subject matter, applying lessons developed specially for software more generally.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

500

Graeme B. Dinwoodie and Ansgar Ohly

hoped to create. The new EU Directive on Copyright in the Digital Single Market is an example in point:39 rather than consolidating the copyright acquis, it reacts to topical issues such as platform liability or text and data mining in a pointillist way. Nevertheless, coherence has been one of Annette Kur’s great research themes, and rightly so. Intellectual property law aims at the right balance of protection and freedom. Creating incentives for inventions and innovations and for creations, protecting authors’ interests and safeguarding undistorted communication in markets is just as important as making sure that the public domain is not unduly restricted. It is one of the foremost tasks of research in intellectual property law to search for an equilibrium,40 to detect areas where this balance is threatened and where developments in one field of intellectual property law interfere with other parts of the intellectual property ecosystem. The more intellectual property rights expand, the more they tend to overlap and the more there is need for coherence.

42.3.2 The Normative Framework: IP in a Multi-Level Legal System The normative framework of intellectual property and unfair competition law is a multi-level system,41 consisting of international treaties, of Union law in the EU or federal law in the US, and of national law in Europe or (some minimal) state law in the US.42 Intellectual property law has been one of the pioneers of international harmonization since the late nineteenth century. The territorial limitation of IP rights may result in obstacles to international trade and can in particular restrict fundamental freedoms in the internal market created by EU law. From the perspective of coherence, internationalization and Europeanization create both opportunities and threats. On the one hand, treaties and EU law set international standards and create international coherence.43 At the international level, intellectual property law is more coherent than many other areas of the law, which is also reflected in the academic discourse, where European and international discussion have transcended the confines of traditional comparative law methodology. But international harmonization also creates an incentive for the competition of systems and creates challenges for traditional national doctrines which have lost their justification. 39

40 41

42

43

See Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on Copyright and Related Rights in the Digital Single Market and Amending Directives 96/9/ EC and 2001/29/EC, OJ L 130, 17.5.2019, 92. See supra Chapter 18 (Yu). On the genesis and meaning of this concept, see Axel Metzger, Extra Legem intra ius: Allgemeine Rechtsgrundsa¨tze im Europa¨ischen Privatrecht (Mohr Siebeck, 2009), 115–119. Most US intellectual property law is now a federal matter. With the enactment of federal trade secret protection, publicity rights remain the principal right protected only at state level. Not only in a substantive, but also in a methodological sense. See supra Chapter 6 (Frankel), discussing the role of the Vienna Convention rules of treaty interpretation for the coherence of intellectual property law.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

501

Indeed, the slow-moving and hard-to-change international system may embed particular approaches in national law long beyond their policy utility. Thus, although several national policymakers would like to take advantage of the informational opportunities presented by mass digitization of copyrighted works, some of the desired mechanisms run headlong into the prohibition on formalities introduced by the Berne Convention in 1908.44 Of course, the coherence created by international treaties must not be overestimated. Treaties are the results of political compromise, and most of their provisions, apart from clear rules such as the term of copyright law, are necessarily vague. The contracting states enjoy considerable freedom, not only of legislating within the broad frame set by the treaties, but also “freedom of relative legal characterisation”, namely the freedom to reach results prescribed by the treaties using different legal concepts.45 At the EU level, the Member States enjoy a similar, if narrower, margin for legislative creativity, as EU Directives are only binding with respect to the result without dictating the means to reach that result.46 But in the EU these margins of appreciation are constantly narrowing, given the legislative tendency to move from minimum standards to full harmonization and the preference of the CJEU for the autonomous interpretation of legal concepts. This begs a fundamental question of what room the Member States still have to pass legislation which concretises the broad framework of EU law. In copyright law, in particular, the CJEU has interpreted concepts such as “communication to the public” or “fair compensation” in such great detail that domestic statutory provisions which had initially been within the boundaries of EU law now conflict with EU law as interpreted by the CJEU. The Court has arguably repeatedly overstepped the line between interpretation and lawmaking by harmonizing the law in areas in which the Member States had been unable to reach a political compromise.47 This causes frictions, as the CJEU can only answer the questions referred to it by national courts, which are embedded in the context of that particular legal system, and as the CJEU only decides a very limited number of cases, which it cannot pick and choose. Hence, more than once, bad references have made bad EU law. Arguably the preliminary reference system is in need of a reform which takes better account of the proactive nature of the CJEU jurisprudence. One option is the creation of a specialist IP court within the CJEU system. To some extent this has just been realised at the level of the General Court, as only six of its ten chambers will 44 45

46

47

See Berne Convention, Art. 5. See supra Chapter 8 (Ricketson) (citing Miha´ly Ficsor, The Law of Copyright and the Internet: The 1996 WIPO Treaties, their Interpretation and Implementation (Oxford University Press, 2002), 499). Consolidated Version of the Treaty on the Functioning of the European Union, OJ C 326, 26.10.2012 (TFEU), 47, Art. 288. See supra Chapter 5 (Dreier).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

502

Graeme B. Dinwoodie and Ansgar Ohly

deal with trade mark, design, and other intellectual property matters in the future.48 More radical, a transition from a court which only gives preliminary rulings to a full appellate court might be considered. If the attempts at creating a Unified Patent Court are successful despite all obstacles, and if it is a success, it may serve as an example for future development in other areas of IP. More cautious changes could concern the procedure of the preliminary reference system. For example, the CJEU, like the US Supreme Court,49 could be allowed to reject references50 and thereby to exercise self-restraint by limiting its purview to general questions such as the protection of fundamental rights and freedoms.

42.3.3 Towards a Doctrine of EU Pre-emption? In US law, the doctrine of federal pre-emption plays an important role regulating the relationship between intellectual property law and state unfair competition law,51 a theme which is also addressed repeatedly in this book. In the EU, there is as yet no elaborated doctrine of pre-emption. In areas that have been fully harmonized, such as the law of registered trade marks, design law and ever-increasing areas of copyright law, the national courts of the Member States must interpret their law in accordance with the EU Directives and with the case law of the CJEU. But outside the area of full harmonization, for instance as regards protection against unfair product imitation by unfair competition law,52 the EU legislature has dodged the issue of overlaps.53 Most Directives explicitly state that other regimes of protection such as 48

49

50 51

52

53

See Court of Justice of the European Union, Press Release No. 111/19 (September 19, 2019), at https://curia.europa.eu/jcms/upload/docs/application/pdf/2019-09/cp190111en.pdf. On the relationship between the US Supreme Court and the (specialized) Court of Appeals for the Federal Circuit, see Rochelle C. Dreyfuss, Percolation, Uniformity and Coherent Adjudication: The Federal Circuit Experience, 66 SMU L. Rev. 505, 526 (2013). See supra Chapter 5 (Dreier). See Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (1989); Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225 (1964); Compco Corp. v. Day-Brite Lighting, Inc., 376 U.S. 234 (1964); see also 17 U.S.C. § 301 (statutory pre-emption). For a comparative overview, see Ansgar Ohly, The Freedom of Imitation and Its Limits – A European Perspective, 41 International Review of Intellectual Property and Competition Law 506–524 (2010). Indeed, the trend appears to be toward full cumulation of rights – see Case C-168/09, Flos SpA v. Semeraro Casa e Famiglia SpA, ECLI:EU:C:201:29 (CJEU 2011) – and an openness to national unfair competition protection even in the face of principled denial of EU intellectual property protection. See Lego (C-48/09) [2010] ETMR 63, para. 61 (“the position of an undertaking which has developed a technical solution cannot be protected – with regard to competitors placing on the market slavish copies of the product shape incorporating exactly the same solution – by conferring a monopoly on that undertaking through registering as a trade mark the three-dimensional sign consisting of that shape, but can, where appropriate, be examined in the light of rules on unfair competition”) and the Opinion of Advocate-General Sánchez-Bordona in Case C-833/18, Brompton Bicycle Ltd. v. Chedech, ECLI:EU:C:2020:79, para. 100: (“I should add that the possible refusal to grant copyright protection would not preclude reliance on other provisions laid down to combat slavish or parasitic imitations.”) But

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

503

unfair competition law remain applicable.54 In its early Beele judgment,55 the CJEU held that the principle of free movement of goods did not prevent Dutch courts from prohibiting slavish imitation on the basis of unfair competition law, at least when it caused a likelihood of confusion. This approach may have been legitimate as long as harmonization remained partial, but under modern EU law a more robust doctrine of pre-emption may be required.56 A possible normative anchor is provided by the fundamental freedoms.57 The free movement of goods, for example, can be restricted if necessary for the protection of intellectual property law, but restrictions must be proportionate. Where EU law already provides a balanced system of protection through secondary law, additional national protection, in particular by means of unfair competition law, can become disproportionate. The closer the net of EU intellectual property law becomes, the more EU law “tightens the noose on national regulatory deviations”.58 But the exact scope of EU pre-emption, even if it were accepted, may be difficult to determine. Examples in point are the tort of passing off or its equivalents in Continental European systems. The Unfair Commercial Practices Directive59 (UCPD) protects consumers against misleading practices, but it does not directly protect the interests of traders. Does it pre-empt national causes of action which are provided to original product manufacturers against competitors who cause confusion by selling imitated products?60 The UCPD protects consumers, whereas Art 10bis (3)(i) of the Paris Convention and many national doctrines of unfair competition mainly have the

54

55 56 57

58 59

60

see also the Cofemel case, discussed infra at 42.3.4, where the CJEU held that an overlap of copyright and design protection could “be envisaged only in certain situations”. See, e.g., Recital 40 of Directive (EU) 2015/2436 of the European Parliament and of the Council of 16 December 2015 to Approximate the Laws of the Member States Relating to Trade Marks (Trade Marks Directive), OJ L 336, 23.12.2015, 1. Case C-6/81, Industrie Diensten Groep v. J.A. Beele Handelmaatschappij, [1982] ECR 707. For an outline of a theory of EU pre-emption, see supra Chapter 37 (Husovec). See Annette Kur, Nachahmungsschutz und Freiheit des Warenverkehrs – der wettbewerbsrechtliche Leistungsschutz aus der Perspektive des Gemeinschaftsrechts, in Hans-Jürgen Ahrens et al. (eds.), Festschrift fu¨r Eike Ullmann (Beck, 2006), 717–735. As Martin Husovec points out – see supra Chapter 37. Directive 2005/29/EC of the European Parliament and of the Council of 11 May 2005 Concerning Unfair Business-to-Consumer Commercial Practices in the Internal Market, OJ L 149, 11.6.2005, 22. As suggested by Matthias Leistner – see supra Chapter 38. So far, the UCPD has not been applied to passing off cases in England, although this has been suggested in the legal literature – see Phillip Johnson and Johanna Gibson, The “New” Tort of Passing Off, 131 Law Quarterly Review 474–494 (2015). In Germany, the Federal Supreme Court has held that § 4(3)(a) of the Unfair Competition Act, which prohibits the sale of imitated products that cause confusion, is outside the scope of the UCPD – see BGH, Case I ZR 124/06, GRUR 2010, 80 para. 17 – Exzenterzähne [Eccenter Teeth]. This view is shared by the prevailing opinion in literature – see Helmut Ko¨hler in Helmut Ko¨hler, Joachim Bornkamm and Jo¨rn Feddersen, Gesetz gegen den unlauteren Wettbewerb, 37th ed. (Beck, 2019), § 4 para. 3.16a; Ansgar Ohly in Ansgar Ohly and Olaf Sosnitza, UWG – Kommentar, 7th ed. (Beck, 2016), § 4 para. 3/10.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

504

Graeme B. Dinwoodie and Ansgar Ohly

original manufacturer’s interest in mind. This militates against the pre-emptive effect of the UCPD. On the other hand, it is true that national judges sometimes tend to find confusion when their hunch tells them that the defendant should not have imitated. The normative concept of an average, reasonably circumspect and well-informed consumer, which is adopted by the UCPD, arguably provides the superior analytical framework. Ultimately, the EU probably needs harmonization of unfair competition rules which protect competitors. Ironically, Brexit may make it more likely that the remaining EU Member States reach a consensus in this respect.

42.3.4 Common Features of Intellectual Property Law It is next to impossible to construe intellectual property law as a coherent whole without first identifying the common features of all intellectual property rights.61 The possibly most fundamental question is whether intellectual property is “property” at all.62 Article 17(2) of the EU Charter of Fundamental Rights may be too simplistic in stating “Intellectual property shall be protected” and in thereby aligning intellectual property with the other forms of property protected by virtue of Article 17 (1) of the Charter.63 One can look at this issue from the perspective of legal theory and social sciences. The Blackstonian notion of property as the “sole and despotic dominion which one man claims and exercises over the external things of the world”64 may seem attractive to proponents of strong protection, but applied to intellectual property it may result in a misguided ontological view of a world where intangible, pre-existing things are allocated to an owner.65 The structure of intellectual property differs from property in land or in movables, and at least on the Eastern side of the Atlantic66 more research will have to be conducted into its foundations. While the label “property” is often abused to extend the scope of protection, modern property theories have demystified the concept. If used analytically rather than rhetorically, the notion of property may well have its doctrinal virtues.67 But does the classification of a right as a “property right” also have practical consequences? An interesting test case discussed in this volume is the law of trade 61

62

63

64 65 66

67

For an attempt, see the contributions in Ansgar Ohly (ed.), Common Principles of European Intellectual Property Law (Mohr Siebeck, 2012). See Ole-Andreas Rognstad, Property Aspects of Intellectual Property (Cambridge University Press, 2018). See Christophe Geiger, Intellectual Property Shall be Protected!? Article 17 (2) of the Charter of Fundamental Rights of the European Union: A Mysterious Provision with an Unclear Scope, 31 European Intellectual Property Rev. 113 (2009). William Blackstone, Commentaries on the Law of England, Book II, Ch. 1. See supra Chapter 17 (Peukert). For an overview of US property theories and their application to intellectual property, see Michael Goldhammer, Geistiges Eigentum und Eigentumstheorie (Mohr Siebeck, 2012). See Michael A. Carrier, Cabining Intellectual Property through a Property Paradigm, 54 Duke L.J. 1 (2004).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

505

secrets, which has recently undergone reform both in the EU and in the US. While some authors think that classifying trade secrets as “property” leads to overprotection,68 others find that it has “surprising virtues”.69 The EU Directive70 states that it “should not create any exclusive right to know-how or information protected as trade secrets”.71 But the protection afforded by the Directive has some of the hallmarks of property. In particular, the definition of “infringing goods” in Article 2(4) is dangerously broad.72 The holder of the trade secret can stop the marketing and manufacturing of these products even if not their construction, but only their marketing benefits from the trade secret. In this respect, trade secret protection is a typical hybrid between intellectual property and unfair competition law: it has some of the elements of property while lacking others.73 The notion of “property” may help to explain this, but, as an agnostic realist might observe, the classification is of little practical value.74 The EU Directive will result in greater coherence among EU jurisdictions, but a look at the national implementations shows that these differ and that fundamental questions such as, again, the relationship to unfair competition law have not been finally resolved.75 68

69

70

71 72 73

74

75

See Gustavo Ghidini and Valeria Falce, Trade Secrets as Intellectual Property Rights: A Disgraceful Upgrading – Notes on an Italian “Reform”, in Rochelle C. Dreyfuss and Katherine J. Strandburg (eds.), The Law and Theory of Trade Secrecy (Edward Elgar, 2011), 141–151; Tanya F. Aplin, Confidential Information as Property? , 24 King’s Law Journal 172–201 (2013). See Mark A. Lemley, The Surprising Virtues of Treating Trade Secrets as IP Rights, 61 Stanford L. Rev. 311 (2008). See Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the Protection of Undisclosed Know-How and Business Information (Trade Secrets) Against their Unlawful Acquisition, Use and Disclosure, OJ L 157, 15.6.2016, 1. See Trade Secrets Directive, ibid., Recital 16. See supra Chapter 19 (Lee). For an account of trade secrets as hybrids, see Jerome H. Reichman, Legal Hybrids between the Patent and Copyright Paradigms, 94 Columbia L. Rev. 2432, 2519–2557 (1994). One area where the classification might matter after all is private international law. The application of the lex loci protectionis principle, which governs the choice of law in intellectual property disputes (for the EU, see Article 8(1) of the Rome II Regulation), favors the interests of persons receiving the information, because they can rely on the limitations and exceptions of trade secret law which apply in the country in which they act. The interests of the trade secret holder are better served by the application of the law at his or her seat, which, in the EU, may well be the result if Article 6(2) of the Rome II Regulation with its reference to Article 4 of the Rome II Regulation is applied. It is not the nature of trade secret law, however, that should determine the outcome, but the policies of trade secret law and an analysis of the interests involved. See Rochelle C. Dreyfuss and Linda Silberman, Misappropriation on a Global Scale: Extraterritoriality and Applicable Law in Transborder Trade Secrecy Cases, available at https:// papers.ssrn.com/sol3/papers.cfm?abstract_id=3086819, and Josef Drexl, in Jan von Hein (ed.), Mu¨nchener Kommentar zum BGB, vol. xii, Internationales Privatrecht II, 7th ed. (Beck, 2018), Internationales Lauterkeitsrecht, paras. 183–186. Jens Schovsbo, supra Chapter 20, compares the former diversity of legal approaches in the EU to a smörgåsbord, which allows eaters to pick and choose according to one’s tastes, and argues that it has given way to New Nordic Cuisine, where everyone gets the same dish, although the dishes are indebted to local ingredients.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

506

Graeme B. Dinwoodie and Ansgar Ohly

One common feature of intellectual property rights is that they are not ends in themselves; rather, they serve a larger purpose. Thus, the introduction of new rights requires justification. The mere fact that personal and industrial data may have significant economic value, for example, is not a sufficient reason for introducing exclusive rights in data.76 The link between right and function may be more evident in the US, where Congress has a purpose-bound power to legislate in the area of patent and copyright law,77 than in Continental Europe, where at least copyright law was predominantly justified by deontological rather than utilitarian considerations.78 Meanwhile, however, the CJEU stresses that only the “specific subject matter” of intangible goods merits protection79 and uses the functions of intellectual property rights as topoi when balancing protection with freedom. This is particularly evident in trade mark law.80 EU copyright law in contrast is, according to the InfoSoc Directive,81 committed to a high level of protection.82 Some economic rights, particularly the reproduction right, may be overreaching. In an analogue world, reproductions were made by publishers, who produced books as marketable products. But in the digital world it is the users who reproduce. Many copyright scholars are concerned about this, look for ways of limiting protection in a way proportionate to the function of copyright law, and find a model in trade mark law.83 They agree that delimitation by purpose should be regarded as a fundamental principle applying to all intellectual property rights.84 One source of inspiration to copyright law scholars is the doctrine of “trade mark use”, which has served as a filter in several jurisdictions to limit trade mark infringements to cases where the trade mark is “used as a mark” rather than as an ornament or a description. In the same vein, it could be argued that the copyright

76 77 78

79

80

81

82 83

84

See supra Chapter 16 (Drexl). See US Constitution, Art. I, § 8, Clause 8. See Jane C. Ginsburg, A Tale of Two Copyrights: Literary Property in Revolutionary France and America, 64 Tulane L. Rev. 991–1031 (1990); Alain Strowel, Droit d’auteur et copyright: divergences et convergences: e´tude de droit compare´ (E. Bruylant, 1993); Christian Gero Stallberg, Urheberrecht und moralische Rechtfertigung (Duncker & Humblot, 2006). See Joined Cases C-403/08 and C-419/08, Football Association Premier League Ltd et al. v. QC Leisure et al. and Karen Murphy v. Media Protection Services Ltd., ECLI:EU:C:2011:631, paras. 106–108; Case C-128/11, UsedSoft GmbH v. Oracle International Corp., ECLI:EU:C:2012:407, paras. 62, 63. See, e.g., Case C-323/09, Interflora Inc. et al. v. Marks & Spencer plc et al., ECLI:EU: C:2011:604, paras. 54–65. Directive 2001/29/EC of the European Parliament and of the Council of 22 May 2001 on the Harmonisation of Certain Aspects of Copyright and Related Rights in the Information Society, OJ L 167, 22.6.2001. Recitals 9 and 11 of the InfoSoc Directive, ibid. See the chapters in P. Bernt Hugenholtz, Copyright Reconstructed: Rethinking Copyright’s Economic Rights in a Time of Highly Dynamic Technological and Economic Change (Wolters Kluwer, 2018). See supra Chapter 23 (Dusollier).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

507

owner should only be protected against the “use of the work as a work”.85 Interestingly, trade mark law may be moving in a different direction, as the CJEU seems to be discarding the notion of “trade mark use” and appears to conduct the necessary balancing exercises partly in assessing infringement, and partly in applying exceptions. But this systematic issue aside, delimitation by function may be one of the fundamental principles that could contribute to the necessary equilibrium between protection and freedom while at the same time creating coherence. Among the principles which undoubtedly apply to all intellectual property rights is the doctrine of exhaustion. In the analogue world, it achieves a fair trade-off between the necessary protection of the right owner and the freedom of trade.86 But exhaustion only works well where tangible goods are traded. In the digital world, exhaustion soon reaches its limits. Unlike the US courts,87 the CJEU, driven by the aim to secure free trade, tried to transpose exhaustion into the online world by allowing exhaustion when “used” software was sold online.88 While the purpose is understandable, the reasoning of the Court is doctrinally problematic. Recently, the CJEU held that the UsedSoft rule cannot be extended to other kinds of digital works.89 Although this creates new incoherences, the judgment makes perfect sense. Indeed, “outside the world of physical goods, the principle of exhaustion’s scope of application may be exhausted”,90 both for doctrinal reasons and because of the practical consideration that right owners can prevent online exhaustion by technical means. A crucial question for the future will be whether the balance of interest underlying analogue exhaustion will still have anything to contribute, or whether untrammelled freedom of contract should prevail. While this is mainly a copyright issue, similar problems might arise in patent law, for example when data, as the German Federal Supreme Court decided, can in some instances be direct results of patented processes and, hence, subject to exhaustion.91 Another horizontal issue in intellectual property law is the regime of jurisdiction and applicable law. The traditional choice of law rule in intellectual property law is the lex loci protectionis rule, which is closely connected with the principle of territoriality. Annette Kur, along with the other members of the CLIP project, has argued in favour of relaxing the strict lex loci protectionis rule in online cases and of 85 86

87

88 89

90 91

See supra Chapter 24 (Rognstad). Although this trade-off may be thwarted by regulatory law, as Vincenzo di Cataldo points out. See supra Chapter 21. See Vernor v. Autodesk, 621 F.3d 1102 (9th Cir. 2010); Capitol Records, LLC v. ReDigi Inc., 910 F.3d 649 (2nd Cir. 2018). See Case C-128/11, UsedSoft GmbH v. Oracle International Corp., ECLI:EU:C:2012:407. See Case C‑263/18, Nederlands Uitgeversverbond et al. v. Tom Kabinet Internet BV et al., ECLI: EU:C:2019:1111. As Reto Hilty puts it – see supra Chapter 22. BGH, Case X ZR 33/10, GRUR 2012, 1230 – MPEG-2-Videosignalcodierung [MPEG-2 Video Signal Encoding] (headnote translated in 44 International Review of Intellectual Property and Competition Law 602 (2013)).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

508

Graeme B. Dinwoodie and Ansgar Ohly

allowing one central legal order with the closest connection to the dispute to be identified. While the principles of territoriality and lex loci protectionis remain justified, they should not be seen as absolute.92 Instead, their effects should be limited in an appropriate way, either at the level of jurisdiction or at the level of choice of law. One question which has arisen in the recent case law of the CJEU is whether there can and should be coherence between both regimes.93 In Nintendo v. BigBen, the CJEU was prepared to relax strict territoriality considerations and to determine the applicable law not on a country-by-country basis but by an overall assessment of the defendant’s conduct.94 The question remains whether similar considerations should apply to jurisdiction in cases of infringement of EU trade marks, as the German Federal Supreme Court recently held,95 or whether the claimant should retain the benefit of suing in every country in which the jurisdiction has effects.96

42.3.4 Overlaps A test case for coherency is the suite of intellectual property overlaps. When two rights protect the same interest under different conditions, with a different scope or with different exceptions, there is a risk of incoherence. But not every overlap between intellectual property rights necessarily has this consequence. The rights may have different functions, and if the scope of the right and the exceptions are determined in accordance with these functions, an overlap may not be objectionable. According to the principles formulated by Annette Kur, overlaps are acceptable if, first, the conditions and the scope of protection of every right reflect their justification, and, second, the rights are adequately balanced in themselves.97 Unsurprisingly, several contributors to this volume discuss issues of IP overlaps. Negative effects caused by overlaps can be prevented either by exclusionary rules or 92 93 94

95 96

97

See supra Chapter 10 (Heinze). See supra Chapter 11 (Torremans). See Joined Cases C-24/16 and C-25/16, Nintendo Co. Ltd. v. BigBen Interactive GmbH et al., ECLI:EU:C:2017:724, paras. 87–111. BGH, Case I ZR 164/16, GRUR 2018, 84 paras. 15–48 – Parfummarken [Perfume marks]. As Annette Kur suggested – see Annette Kur, Die “Parfummarken”-Entscheidung des BGH – De-facto Abschaffung der deliktischen Zuständigkeit im Unionsmarkenrecht?, GRUR 2018, 358–364 – and as the CJEU recently confirmed in Case C-172/18, AMS Neve Ltd. v. Heritage Audio SA et al., ECLI:EU:C:2019:674. See Annette Kur, Cumulation of IP Rights Pertaining to Product Shapes: An “Illegitimate Offspring” of IP Law?, in Gustavo Ghidini and Luis Mariano Genovesi (eds.), Intellectual Property and Market Power: ATRIP Papers 2006–2007 (Eudeba, 2008), at 613, 616, and Probleme im Überschneidungsbereich von Marken und Designs, MarkenR 2017, 185. See also Estelle Derclaye and Matthias Leistner, Intellectual Property Overlaps: A European Perspective (Hart Publishing, 2011), 314–333; Artur Geier, Schutzkumulationen: Angriff auf die Gemeinfreiheit oder legitimer Schutz Geistigen Eigentums? (Mohr Siebeck, 2015), 135–218.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

509

by a “multi-lever approach” which allows intangible subject matter to be protected by more than one right, but which at the same time restricts the scope of the right to its proper function and interprets exceptions to prevent what Annette Kur calls “asymmetric convergence”. The overlap between copyright and trade marks, such as seen in the Norwegian Vigeland case, is a good vehicle through which to explore this concept.98 When the sculptor’s works entered the public domain, the city of Oslo tried to have pictures of some of the most famous sculptures (such as the Angry Boy) registered as trade marks. A robust doctrine of public domain might favour an exclusionary rule under which the trade mark registration of works which have fallen out of copyright is prevented. Potential users do not have to worry about infringement, and an evergreening of copyright is effectively prevented.99 On the other hand, trade mark law provides a toolbox which may allow more nuanced solutions. A flexible concept of infringement and exceptions, such as the new EU law provision which allows the use of non-distinctive features, may also prevent trade marks from being turned into eternal copyright while at the same time allowing the proper use of works of art as signs. There is even a surprising parallel in patent law, where the modern tendency of excluding categories of gene patents from protection is criticised by those who advocate a reasonable restriction of scope as the more nuanced solution.100 Essentially the same question arises in many situations of overlap. A classical case, discussed by several authors,101 is the protection of product shapes by design rights, copyright, trade marks and unfair competition law. The range of national approaches to the copyright/design overlap reaches from full cumulation in France, over partial cumulation in legal systems which only allow copyright protection for highly individual or aesthetic works, to those which apply a severability test102 or restrict the term of copyright protection in industrial designs, as the Berne Convention allows.103 In the EU, this has caused a conflict between the goal of harmonization and what may be seen as a reasonable competition of systems. While initially Article 17 of the Designs Directive permitted the EU Member States to set the level of originality for works of applied arts individually, the CJEU, in its recent Cofemel judgment,104 held that this provision has been derogated by the judicial harmonization of the concept of work in EU law and that there is only one EU law threshold of originality, which is independent of aesthetic merit. The CJEU conceded that there must not be a full overlap between copyright and design protection, 98 99 100 101 102 103

104

See supra Chapters 26 (Dreyfuss), 25 (Calboli and Ginsburg), and 27 (Senftleben). See supra Chapter 27 (Senftleben). See supra Chapter 26 (Dreyfuss). See supra Chapters 38 (Leistner), 31 (Quaedvlieg), 29 (Rosén) and 33 (Tischner). As US law does – see supra Chapter 28 (McKenna). See the country reports in Estelle Derclaye (ed.), The Copyright/Design Interface – Past, Present and Future (Cambridge University Press, 2018). Case C-683/17, Cofemel – Sociedade de Vestuário SA v. G-Star Raw CV, ECLI:EU:C:2019:721, paras. 47–48.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

510

Graeme B. Dinwoodie and Ansgar Ohly

but has failed to provide guidelines for the demarcation. As an aside, Cofemel is an interesting study object for the institutional aspects of transition and coherence, as the CJEU took the liberty of overruling a statutory provision, based on the consideration that it was derogated by the judicial harmonisation of the concept of work. One approach to preventing overlaps is the functionality exclusion. It was first developed in the US, but it has also made its way into EU trade mark and design law. In the US in particular, it has been justified in part by the consideration that trade mark rights should not give a competitive advantage beyond the opportunity of establishing a reputation under the mark.105 Functionality can lead to satisfactory results when its application is based on a fact-sensitive and market-oriented analysis, but there is also the risk of formalism.106 On this point, giving weight to the availability of design alternatives – which may be less determinative after TrafFix – leaves room for judicial idiosyncrasies and lax reasoning. One of the risks of overlaps, as mentioned above, is “asymmetric convergence”. When one and the same subject matter is protected by several rights, the owner will tend to cherry-pick and rely on the right whose exceptions do not apply in the case at hand. The cases of perfume flacons which attract copyright protection are examples in point.107 Such cases may be symptoms of an imbalance which occurs when the scope of a right is constantly extended whereas the narrow catalogue of exceptions is not.108 In the relationship between copyright and patent law, the EU legislature has tried to create convergence by incorporating copyright exceptions into patent law in the Agreement on the Unified Patent Court.109 Where such a horizontal coordination between the areas of intellectual property law is missing, it is up to the courts to decide whether there can be a spillover of exceptions110 or whether there are overarching principles such as, in the EU context, the free movement of goods, on the basis of which the courts can give precedence to the regime which has a closer connection to the particular case.111 The CJEU has sometimes interpreted

105

106 107

108 109 110

111

See Qualitex Co. v. Jacobson Products Co., Inc., 514 U.S. 159, 165 (1995); Graeme B. Dinwoodie, The Death of Ontology: A Teleological Approach to Trademark Law, 84 Iowa L. Rev. 611, 690 (1999). The decision of the US Supreme Court in TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23, 33 (2001) has been read as furthering a “channeling rule” even where competitive concerns are not implicated. See Mark P. McKenna, (Dys) functionality, 48 Hous. L. Rev. 823, 827–28 (2011) (“Functionality, according to the Supreme Court, is not simply about competitive need for product features; it is also, even primarily, intended to police the boundary between trademark and patent law by channeling protection of useful product features exclusively to the patent system.”). See supra Chapters 30 (Janis) and 3 (Austin). See Case C-337/95, Parfums Christian Dior SA v. Evora BV, [1997] ECR I-6013; BGH, Case I ZR 256/97, GRUR 2001, 51 – Parfumflakon. See Annette Kur, Ersatzteilfreiheit zwischen Marken- und Designrecht, GRUR 2016, 20, 25. See supra Chapter 32 (Metzger). As Irene Calboli and Jane Ginsburg discuss for the relationship between copyright and trade mark law – see supra Chapter 25. Case C-337/95, Parfums Christian Dior SA v. Evora BV, [1997] ECR I-6013.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

Conclusion – Transition and Coherence in IP Law

511

exceptions in a narrow and textual way,112 whereas it has adopted a broader, marketoriented perspective113 in others.114 Its apodictic reasoning in such cases has been criticized as unsatisfactory. Again, some cases demonstrate the limitations of a judicial system based on preliminary rulings, as questions which only relate to one intellectual property right may not allow the CJEU to see the whole picture.115

42.3.5 Intellectual Property and (Un-)fairness “Fairness” is the core concept of unfair competition law, but it arguably also pervades intellectual property law. This is particularly true for trade mark law or the broader “law of distinctive signs”, which also encompasses unregistered signs and geographical indications116 and which in itself is an area striving for coherence.117 The limitations of trade mark rights explicitly refer to “fairness” or “honest practices”, but considerations of fairness also inform the CJEU’s case law on trade mark functions.118 The extended protection of well-known marks against dilution and, in the EU, against misappropriation, is predominantly justified by considerations of fairness. Indeed, in German law, which in turn influenced Schechter’s plea for the introduction of anti-dilution protection in the US,119 extended protection of well-known marks was granted on the basis of unfair competition law and general tort law before the EU Trade Mark Directive was implemented into German law.120 The theoretical underpinnings of anti-dilution and anti-misappropriation protection, however, are far from stable. In US law, the actionability of “reaping without sowing”121

112 113

114 115

116 117 118

119 120 121

Case C-500/14, Ford Motor Company v. Wheeltrims srl, ECLI:EU:C:2015:680. See Joined Cases C-397/16 and C-435/16, Acacia Srl v. Pneusgarda Srl and Audi AG et al. v. Porsche AG, ECLI:EU:C:2017:992. See supra Chapter 33 (Tischner). Ford v. Wheeltrims (supra note 112) is an example in point. The decision of the CJEU not to extend the “repair clause” from design law to trade mark law may well be justified, as it may be preferable to allow the distribution of spare parts showing the OEM’s trade mark on the basis of a broad interpretation of the exceptions provided by trade mark law – see Kur, note 108, at 25, 28–29. However, the questions referred to the CJEU did not prompt the Court to interpret the trade mark law exceptions. But had the CJEU attached greater significance to coherence, it would have been entitled to clarify the relation to trade mark law exceptions by obiter dicta. See supra Chapter 34 (Knaak). See supra Chapter 35 (von Mühlendahl). See Cases 236/08 to 238/08, Google France SARL et al. v. Louis Vuitton Malletier SA et al., ECLI:EU:C:2010:159, paras. 75–98; Case C-323/09, Interflora Inc. et al. v. Marks & Spencer plc et al., ECLI:EU:C:2011:604, paras. 54–64. See supra Chapter 39 (Beebe). LG Elberfeld, Case 13 O 89/24, GRUR 1924, 204 – Odol; RG, RGZ 115, 401, 410 – Salamander II. See International News Service v. Associated Press, 248 U.S. 215, 239–240 (1918); see also Rudolf Callmann, He Who Reaps Where He Has Not Sown: Unjust Enrichment in the Law of Unfair Competition, 55 Harv. L. Rev. 595 (1942).

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

512

Graeme B. Dinwoodie and Ansgar Ohly

has been restricted to specific cases,122 and the history of EU trade mark law shows that the broad protection afforded to trade marks with a reputation was adopted following the Benelux model without much discussion or analysis.123 The EU Directive on comparative advertising has its merits, but in some respects it also results in overprotection and an undue restriction of free commercial speech.124 Despite its shaky foundations, one virtue of an anti-misappropriation provision in trade mark law is that it allows courts to openly and honestly balance the arguments for and against protection without cloaking it by theories of pre- or post-sale confusion.125 The evident weakness of the concept of fairness lies in its inherent vagueness. “We are all against unfairness, just as we are all in favour of mother and apple pie”,126 as Lord Justice Jacob wryly observed, but it is far from clear whether “fairness” can be operationalised as a legal concept. When unfair competition law emerged as a distinct field of law in Continental Europe in the late nineteenth century, there may have been a widespread intuition of what “honest practices in trade and commerce” are,127 but in a modern diverse and liberal society and in times of swift technological and economic changes, it becomes much more difficult to agree on standards of honesty. How standards of fairness can be identified, how they can be made operational, how they affect the distribution of power between legislation and judiciary, how their application may be governed by statutory multi-factor tests and, finally, whether heuristics of fairness can also be embedded into artificial intelligence – these are questions for further research. To those who know Annette Kur, it will come as no surprise that she has started a research project on fairness in intellectual property law, from which we can expect many instructive thoughts on the questions asked above. Most of the contributions to this volume were inspired by her work on transition and coherence. They demonstrate impressively her impact on European and international intellectual property research.128 Along with the other contributors, we dedicate this book to her: Ad multos annos, Annette!

122

123

124

125 126

127 128

For the “hot news doctrine” in a modern context, see National Basketball Association v. Motorola, Inc., 105 F.3d 841 (2d Cir. 1997). See Andreas Sattler, Dilution of Well-known Trademarks – An Analysis of its Foundations in Germany and the European Union, 3 Zeitschrift fu¨r Geistiges Eigentum/Intellectual Property Journal 304–328 (2011). See supra Chapter 41 (Llewelyn); see also Annette Kur, Lionel Bently and Ansgar Ohly, Sweet Smells and a Sour Taste – The ECJ’s L’Oréal decision, available at https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=1492032. See supra Dogan, “Annette Kur: Toward Understanding”. See L’Oréal SA v. Bellure NV, [2007] EWCA Civ 968, para. 160 per Jacob, LJ (to be fair, he used the term “misappropriation” rather than “fairness”). See supra Chapter 36 (Norrgard). As is stressed by Stacey Dogan, “Annette Kur: Toward Understanding”, and Peter Yu, Chapter 18. See also the homage by Dieter Stauder, “Greetings to Annette Kur from the Second Floor”.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529.053

CAMBRIDGE INTELLECTUAL PROPERTY AND INFORMATION LAW

Titles in the Series (formerly known as Cambridge Studies in Intellectual Property Rights) Graeme W Austin, Andrew F Christie, Andrew T Kenyon and Megan Richardson Across Intellectual Property: Essays in Honour of Sam Ricketson Brad Sherman and Lionel Bently The Making of Modern Intellectual Property Law Irini A. Stamatoudi Copyright and Multimedia Products: A Comparative Analysis Pascal Kamina Film Copyright in the European Union Huw Beverly-Smith The Commercial Appropriation of Personality Mark J. Davison The Legal Protection of Databases Robert Burrell and Allison Coleman Copyright Exceptions: The Digital Impact Huw Beverly-Smith, Ansgar Ohly and Agnès Lucas-Schloetter Privacy, Property and Personality: Civil Law Perspectives on Commercial Appropriation Catherine Seville The Internationalisation of Copyright Law: Books, Buccaneers and the Black Flag in the Nineteenth Century Philip Leith Software and Patents in Europe Geertrui van Overwalle Gene Patents and Clearing Models Lionel Bently, Jennifer Davis and Jane C. Ginsburg Trade Marks and Brands: An Interdisciplinary Critique Jonathan Curci The Protection of Biodiversity and Traditional Knowledge in International Law of Intellectual Property Lionel Bently, Jennifer Davis and Jane C. Ginsburg Copyright and Piracy: An Interdisciplinary Critique Megan Richardson and Julian Thomas Framing Intellectual Property: Legal Constructions of Creativity and Appropriation 1840–1940 Dev Gangjee Relocating the Law of Geographical Indications Andrew Kenyon, Megan Richardson and Ng-Loy Wee-Loon The Law of Reputation and Brands in the Asia Pacific Region Edson Beas Rodrigues, Jr The General Exception Clauses of the TRIPS Agreement: Promoting Sustainable Development Annabelle Lever New Frontiers in the Philosophy of Intellectual Property Sigrid Sterckx and Julian Cockbain Exclusions from Patentability: How the European Patent Office is Eroding Boundaries Sebastian Haunss Conflicts in the Knowledge Society: The Contentious Politics of Intellectual Property Helena R. Howe and Jonathan Griffiths Concepts of Property in Intellectual Property Law Rochelle Cooper Dreyfuss and Jane C. Ginsburg Intellectual Property at the Edge: The Contested Contours of IP Normann Witzleb, David Lindsay, Moira Paterson and Sharon Rodrick Emerging Challenges in Privacy Law: Comparative Perspectives Paul Bernal Internet Privacy Rights: Rights to Protect Autonomy Peter Drahos Intellectual Property, Indigenous People and Their Knowledge Susy Frankel and Daniel Gervais The Evolution and Equilibrium of Copyright in the Digital Age Edited by Kathy Bowrey and Michael Handler Law and Creativity in the Age of the Entertainment Franchise Sean Bottomley The British Patent System and the Industrial Revolution 1700–1852: From Privileges to Property Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529

Susy Frankel Test Tubes for Global Intellectual Property Issues: Small Market Economies Jan Oster Media Freedom As a Fundamental Right Sara Bannerman International Copyright and Access to Knowledge Andrew T. Kenyon Comparative Defamation and Privacy Law Pascal Kamina Film Copyright in the European Union (second edition) Tim W. Dornis Trademark and Unfair Competition Conflicts Ge Chen Copyright and International Negotiations: An Engine of Free Expression in China? David Tan The Commercial Appropriation of Fame: A Cultural Critique of the Right of Publicity and Passing Off Jay Sanderson Plants, People and Practices: The Nature and History of the UPOV Convention Daniel Benoliel Patent Intensity and Economic Growth Jeffrey A. Maine and Xuan-Thao Nguyen The Intellectual Property Holding Company: Tax Use and Abuse from Victoria's Secret to Apple Megan Richardson The Right to Privacy: Origins and Influence of a Nineteenth-Century Idea Martin Husovec Injunctions against Intermediaries in the European Union: Accountable but Not Liable? Estelle Derclaye The Copyright/Design Interface: Past, Present and Future Magdalena Kolasa Trade Secrets and Employee Mobility: In Search of an Equilibrium Péter Mezei Copyright Exhaustion: Law and Policy in the United States and the European Union Graham Greenleaf and David Lindsay Public Rights: Copyright’s Public Domains Ole-Andreas Rognstad Property Aspects of Intellectual Property Elena Cooper Art and Modern Copyright: The Contested Image Paul Bernal The Internet, Warts and All: Free Speech, Privacy and Truth Sebastian Felix Schwemer Licensing and Access to Content in the European Union: Regulation between Copyright and Competition Law Daniela Simone Copyright and Collective Authorship: Locating the Authors of Collaborative Work Emily Hudson Drafting Copyright Exceptions: From the Law in Books to the Law in Action Graeme Gooday and Steven Wilf Patent Cultures: Diversity and Harmonization in Historical Perspective Graeme W. Austin, Andrew F. Christie, Andrew T. Kenyon and Megan Richardson Across Intellectual Property: Essays in Honour of Sam Ricketson Mark Burdon Digital Data Collection and Information Privacy Law Edited by Niklas Brunn, Graeme B. Dinwoodie, Marianne Levin and Ansgar Ohly Transition and Coherence in Intellectual Property Law: Essays in Honour of Annette Kur

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108688529