Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

The “Creating Around” Paradox

I. Introduction

In his article Creating Around Copyright, Joseph Fishman advances the provocative and significant thesis that the constraints imposed by copyright law exclusivity promote the creativity of subsequent follow-on authors.1×1. Joseph P. Fishman, Creating Around Copyright, 128 Harv. L. Rev. 1333 (2015). Copyright law is generally regarded as promoting creativity by enhancing the choices available to creators, as offering exclusive rights that make creative opportunities financially viable where they might otherwise seem economically infeasible or unattractive.2×2. See Dan L. Burk, Law and Economics of Intellectual Property: In Search of First Principles, 8 Ann. Rev. L. & Soc. Sci. 397, 399–401 (2012) (surveying arguments supporting exclusive rights in creative works). And much also has been said about the opportunities that are thus lost due to the restriction of copyright’s exclusivity, which shuts follow-on authors out of certain creative choices that are preempted by existing works.3×3. See, e.g., id. at 402–03 (surveying arguments critiquing exclusive rights in creative works). But Fishman argues, perhaps counterintuitively, that by limiting creative choices, copyright exclusivity actually enhances the output of follow-on authors by requiring them to “create around” existing works.

As evidence of this effect, Fishman proffers both anecdotal and experimental indications of creating around. Fishman notes, for example, that Star Wars creator George Lucas initially hoped to license movie rights to the existing Flash Gordon outer space character, but negotiations failed.4×4. See Fishman, supra note 1, at 1336. The inaccessibility of the Flash Gordon property forced Lucas to develop his own space opera, which ultimately proved to be at least as popular as, and by many measures far more successful than, the copyrighted work that inspired it. Fishman offers other examples of such failed licensing attempts that ultimately resulted in the creation of alternative works, such as the Nintendo video game character Mario, which was initially intended to be the cartoon character Popeye.5×5. See id. Fishman ties these examples to work in experimental psychology showing that constraint on creative choices may lead to greater creative output, and argues that the constraints on creation stemming from copyright exclusivity may function to further the production of new works.

Yet embedded in Fishman’s theory is a paradox that threatens to disable the putative benefits of creating around. Specifically, the conditions that are necessary for creating around are the same conditions that we would expect to lead to licensing of existing works, rather than the creation of new ones. In other words, it appears that creating around can only occur when we would expect it not to occur. This paradox draws out additional features of Fishman’s comparison to patent law. In this Response, I illuminate this problem, showing how the logic of Fishman’s argument leads inevitably to this paradox. This conundrum presents a problem for Fishman’s analysis, but I suggest it is not necessarily fatal to his proposition. I conclude with several suggestions as to how one might escape the creating around paradox, and argue that Fishman’s theory might yet inform our understanding of copyright law’s purposes.

II. A Constraint Conundrum

Fishman draws heavily on an analogy to patent law’s concept of “inventing around” in developing his theory of creating around in copyright.6×6. Id. at 1351–58. I have argued elsewhere that copyright law may foster technological inventing around in the same sense as that found in patent law, separate from the phenomenon that Fishman describes. See Dan L. Burk, Inventing Around Copyright, 109 Nw. U. L. Rev. Online 64 (2014). Patent law has long recognized and even encouraged the practice of inventing around, by which follow-on inventors skirt the perimeter of exclusive rights in a previous invention as defined in the written claims of a patent.7×7. See, e.g., F.M. Scherer & David Ross, Industrial Market Structure and Economic Performance 624–26 (3d ed. 1990) (discussing inventing around patent claims); Donald F. Turner, The Patent System and Competitive Policy, 44 N.Y.U. L. Rev. 450, 455 (1969) (same). Decisions of the United States Court of Appeals for the Federal Circuit have often touted inventing around as one of the beneficial features of patent law: patent exclusivity encourages follow-on inventors to develop innovative alternatives to existing inventions by requiring new inventors to avoid the zones of previously claimed rights.8×8. See, e.g., Yarway Corp. v. Eur-Control USA, Inc., 775 F.2d 268, 277 (Fed. Cir. 1985); State Indus., Inc. v. A.O. Smith Corp., 751 F.2d 1226, 1236 (Fed. Cir. 1985); Kimberly-Clark Corp. v. Johnson & Johnson, 745 F.2d 1437, 1457–58 (Fed. Cir. 1984).

Patent inventing around is expected to occur as the product of bargaining breakdown.9×9. Martin J. Adelman, The Supreme Court, Market Structure, and Innovation: Chakrabarty, Rohm and Haas, 27 Antitrust Bull. 457, 464 (1982). Typically it would be in the interest of the patent owner to license her existing property so as to forestall the development of a competing substitute. At the same time, it would likely be in the interest of a competitor to license the existing patented invention rather than invest in developing a new substitute, so long as the license comes at a cost lower than that required to develop that substitute. The fact that inventing around occurs indicates that the parties’ valuations of the patent were sufficiently divergent that licensing failed, and so developing a substitute became the preferable alternative for the competitor.

Fishman’s scenario of creating around presumably relies on a similar bargaining breakdown. To take one of his specific examples, he points out that George Lucas’ Star Wars was developed because licensing negotiations for the existing Flash Gordon property failed; if Flash Gordon had been available, Lucas would presumably have used it instead.10×10. See Fishman, supra note 1, at 1336. In the language of transaction cost economics, the creator must decide whether to “make or buy” the creative work — whether to invest in production capacity to generate the needed material or to license it from elsewhere.11×11. I have discussed this literature in some depth previously; see Dan L. Burk & Brett H. McDonnell, The Goldilocks Hypothesis: Balancing Intellectual Property Rights at the Boundary of the Firm, 2007 U. Ill. L. Rev. 575, 583–90. I am grateful to Professor Rebecca Tushnet for pointing out the “make or buy” parallel. See Rebecca Tushnet, IPSC Part 7: Copyright Limitations, Rebecca Tushnet’s 43(B)log (Aug. 8, 2014 3:10 PM), http://tushnet.blogspot.com/2014/08/ipsc-part-7-copyright-limitations.html [http://perma.cc/XDC8-ECTH]. The economic literature tells us that the choice will rest upon the relative costs of developing the material versus licensing it, including the likely transaction costs of each alternative.12×12. See Burk & McDonnell, supra note 11, at 578–80 (summarizing the literature on transaction cost analysis). In Fishman’s terms, if the transaction cost of licensing is prohibitive, a creator like Lucas will turn to creating around.

Indeed, the choice whether to make a new work rather than licensing the old seems to depend on the somewhat counterintuitive requirement that Fishman’s proposition would only be expected to function in the presence of pervasive and ongoing bargaining breakdown. If Flash Gordon is unavailable at a competitive licensing price, the follow-on creator who is determined to stick with a space opera project may develop Star Wars, because that is the next best alternative. But consider the situation once Star Wars comes into existence: both Flash Gordon and Star Wars must be unavailable at a competitive licensing price before yet another follow-on developer will create around the two existing properties to develop, say, Battlestar Galactica.13×13. Although not without being accused of infringement by the owners of Star Wars. See Twentieth Century-Fox Film Corp. v. MCA, Inc., 715 F.2d 1327 (9th Cir. 1983) (copyright infringement suit against the creators of Battlestar Galactica). And thereafter, all three properties will have to be unavailable for licensing before additional space opera projects are independently developed via Fishman’s expectation of creating around.

Speaking more generally, then: If alternative works are available for license at a reasonable cost, creators will be inclined to use an existing work rather than create their own. But of course if the work is unavailable, the resultant creating around necessarily means that there will then be a new alternative to the work that has been created around, and that the alternative might subsequently be available for licensing. If it is not available either, successive alternate works may be developed via creating around, but as the constellation of similar works grows larger, the likelihood increases that one or more will be made available for licensing. Thus future creators will become less and less likely to create around and more and more likely to license one of the existing substitute works, unless for some reason all of them are unavailable at a licensing cost less than the cost of new development.

One might therefore expect the development of copyright “thickets” in which successively denser entitlements are clustered around an existing property, crowding out new follow-on creations resembling works already created. Such thickets have been a concern in the patent literature, where dense clusters of exclusivity are believed to hamper follow-on innovation.14×14. See, e.g., Carl Shapiro, Navigating the Patent Thicket: Cross Licenses, Patent Pools, and Standard Setting, in 1 Innovation Policy and the Economy 119, 121 (Adam B. Jaffe et al. eds., 2001). But it is difficult to detect or even imagine a similar concern in copyright. What I have described above with regard to alternative space operas seems not so much the development of tangled and impenetrable copyright thickets as it does the development of families of creative works that we term genre.15×15. See John M. Swales, Genre Analysis 33 (Michael H. Long & Jack C. Richards, eds.) (1990) (discussing various definitions of genre). Indeed, copyright law goes out of its way to foster genre, for example via the scènes à faire doctrine that ensures stock characters and situations remain readily available in the public domain,16×16. See Jessica Litman, The Public Domain, 39 Emory L.J. 965, 1021–22 (1990) (explaining the role of the scènes à faire doctrine in ameliorating the problem of overlapping copyright claims). or via the merger doctrine that provides that if there are only a limited number of ways to express an idea, the vehicles of expression are not subject to copyright.17×17. See Dan L. Burk, Owning E-Sports: Proprietary Rights in Professional Computer Gaming, 161 U. Pa. L. Rev. 1535, 1565–67 (2013) (describing how genre fits within copyright doctrine).

Indeed, Fishman recognizes the need for copyright law to foster alternatives, arguing in favor of a thoughtfully designed fair use defense18×18. See Fishman, supra note 1, at 1381. and objecting to the rigid approach some courts have taken to music sampling.19×19. See id. at 1384–85. But such alternatives must be unavailable for creating around to occur. This suggests that Fishman’s expectation of creating around would most likely be realized in a counterfactual regime where a single copyright owner always held broad, concentrated, and monopolistic rights in a particular work, so as to preclude other alternatives that might be readily licensed instead. In the patent context, this type of broad monopoly is typically thought of as the “prospect” entitlement structure associated with the theories of Professor Edmund Kitch, who famously articulated an explanation as to why patent rights are broadly exclusive.20×20. Edmund W. Kitch, The Nature and Function of the Patent System, 20 J.L. & Econ. 265 (1977). But patent law expects inventing around to be the exception — even the marginal exception — rather than the rule.21×21. See Adelman, supra note 9, at 464. Indeed, patent law’s doctrine of equivalents ensures that any inventing around must skirt the claims of existing inventions by a wide margin. See Paul N. Katz, The Doctrine of Equivalents and Its Impact on “Designing Around”, 4 Fed. Cir. B.J. 315, 320–23 (1994). A similar rule of broad copyright entitlements would gut Fishman’s argument, ensuring relatively few licensing substitutes by simultaneously ensuring that creating around seldom occurs. Hence the paradox that the conditions necessary for creating around copyright will also tend to prevent it.

III. Maintaining Creative Constraint

If the entitlement structure of copyright does not deter alternatives, then we would expect multiple licensing alternatives to be available at a cost that makes further creating around unattractive, hamstringing the creating around hypothesis. And yet Fishman offers some specific examples of creating around, indicating that it occurs at least some of the time.22×22. See Fishman, supra note 1, at 1336. Are these marginal or outlying cases, or might they be representative of a more common effect that he postulates? It seems to me that at least three plausible hypotheses might be advanced as to how creating around might still occur in the face of the paradox I have just described.

First, it may be that classic economic assumptions about rational bargaining simply do not operate in this space. For example, if creators are overly attached to their creations, they may not be willing to license them at any reasonable price, prompting creating around in situations where we would rationally expect licensing to occur. Much of Fishman’s thesis relies upon psychological experiments regarding creativity in the face of constraint;23×23. See id. at 1362–69. a nascent literature on bargaining and creative works suggests that copyrightable works could be subject to an “endowment effect” that causes creators to overvalue them and to license less often than we might otherwise anticipate.24×24. Christopher Buccafusco & Christopher Jon Sprigman, The Creativity Effect, 78 U. Chi. L. Rev. 31 (2011); Christopher Buccafusco & Christopher Sprigman, Valuing Intellectual Property: An Experiment, 96 Cornell L. Rev. 1 (2010). Such overvaluing might generate the bargaining breakdown necessary to creating around.

A second possibility is that potential licensors and licensees might systematically fail to reach agreements due to the “muddy” entitlement structure of copyright.25×25. See Dan L. Burk, Muddy Rules for Cyberspace, 21 Cardozo L. Rev. 121 (1999) (describing the characteristics of “muddy” or indeterminate rules in copyright). Fishman complains that copyright’s substantial similarity doctrine is overly vague, creating fuzzy boundaries around copyrighted properties, and so putatively making it difficult for follow-on creators to design around the uncertain borders of existing works.26×26. See Fishman, supra note 1, at 1386. But fuzzy boundaries can be a feature just as well as they can be a bug, deterring overenforcement of copyright against follow-on creators.27×27. See Burk, supra note 25, at 144 (discussing the benefit to follow-on creators of uncertainty in the boundaries of substantial similarity); see also Jason Scott Johnston, Bargaining Under Rules Versus Standards, 11 J.L. Econ. & Org. 256 (1995) (showing that the uncertainty of incomplete entitlements can lead to efficient bargaining outcomes). Significantly for Fishman’s analysis, Professor Jason Johnston’s discussion of bargaining under uncertainty relies on the threat of an unauthorized but probabilistically permissible taking by one of the parties asserting an interest in a property right. See id. at 258. Rather than creating an impediment to the development of alternatives, such a boundary may be impeding licensing that would otherwise replace the investment in creation of new works, thus facilitating creating around. This possibility suggests that Fishman might rescue his larger creating around hypothesis by reconsidering the subsidiary need for copyright clarity.

Third, copyright law does not operate in a vacuum. Most of the examples that Fishman offers, such as those I have considered above, are commercial properties that also function as trademarks. Nintendo’s Mario and the various characters from Star Wars embody valuable reputational capital quite apart from their value as creative copyrighted works, and that additional value may radically change the mechanics of licensing. Deciding whether to “make or buy” a trademark is a complicated calculus to add to copyright negotiations,28×28. See Dan L. Burk & Brett H. McDonnell, Trademarks and the Boundaries of the Firm, 51 Wm. & Mary L. Rev. 345 (2009) (discussing transaction cost analysis of trademarks). and may deter licenses that we might expect to be negotiated were copyright alone at issue. The added trademark value of Flash Gordon or Popeye may be more than follow-on creators seeking a copyright license want to pay. Thus, creating around may be as much a product of trademark or related law as it as a creature of copyright.

IV. Conclusion

Copyright law assumes that a large number of creative alternatives are available to express any given idea. Patent law, on the other hand, does not depend on multiple alternatives, but rather tends toward singular exclusivity. By drawing on patent doctrine to formulate a theory of creative constraint in copyright, Fishman advances a theory in which alternatives must be both available and unavailable: as more alternatives arise, licensing becomes more likely than creating around. I have suggested some ways out of this paradox, and expect that a functional theory of creating around would adopt some explanation along these or similar lines in order to offer a viable account as to how copyright fosters creativity under constraint.


* Chancellor’s Professor of Law, University of California, Irvine.