The Relative Virtues of Bottom-up and Top-down Theories of Fair Use

The Relative Virtues of Bottom-Up and TopDown Theories of Fair Use

A Response to Abraham Bell and Gideon Parchomovsky, The Dual-Grant Theory of Fair Use, 83 U Chi L Rev 1051 (2016).

Pamela Samuelsont

made, the quantity and value of the materials used, and the degree in which the use may prejudice the sale, or diminish the profits, or supersede the objects, of the original work."5 For the next 135 years, courts weighed these factors in typical common-law fashion in a wide variety of cases.6 Not until 1976, however, did Congress choose limitation in US copyright law.7

the fair use

to codify

As codified, the fair use provision identifies six types of favored uses, namely, "criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research."8 Drawing on Folsom and the fair use case law, Congress directed courts to weigh four factors when determining whether a use was fair or infringing:

(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and (4) the effect of the use upon the potential market for or value of the copyrighted work.9

In the forty years since the fair use doctrine was codified, courts have grappled with fair use defenses in hundreds of cases. One of the great virtues of fair use is its flexibility, for it allows courts in a wide variety of contexts to balance the interests of copyright owners in having control over unfair exploitations of their works and the interests of follow-on creators and other users in having some freedom to make reasonable uses of copyrighted materials, some of which may have spillover benefits for the public at large. 10 Yet, because of the fact-intensive and case-by-case

nature of the common law of fair use, commentators often complain that it is unpredictable and incoherent."

as it has been interpreted by the US Supreme Court, among others, and with the constitutional policies that underlie US copyright law. This Essay explains why I think that Dual-Grant takes an unduly narrow view of the work that fair use does and should do in US copyright law, is blatantly inconsistent with existing case law in more ways than it acknowledges, fails to recognize important values found in many fair use cases, and would, if followed, have the unintended consequence of making fair use more unpredictable and incoherent than it is now.

I. DUAL-GRANT OFFERS A CRITIQUE OF THE MARKET FAILURE THEORY OF FAIR USE

In The Dual-Grant Theory of Fair Use, Professors Bell and Parchomovsky set forth a theory of fair use that, at least in places, resonates with the limited monopoly conception of fair use.17 The "dual grant" to which the authors refer is the idea that copyright law should recognize that users, as well as authors, have rights that the law should respect, and the related idea that fair use is a manifestation of this. 18This not-really-so-novel proposition will find a receptive audience among the generally progressive copyright professoriat.19

What is novel in Dual-Grant is the use it makes of the userrights concept as a basis for criticizing the "predominant" market failure theory of fair use.20 The market failure theory posits that uses should be deemed fair and noninfringing only if transaction costs or other impediments would make it infeasible to form a viable market to transact about the uses.21 The main defects of the market failure theory, in Bell and Parchomovsky's view, are that

it assumes that authors have the right to control all uses of their works in the absence of market failure and it ignores that courts in recent years have emphasized the importance of fair use as a means to enable socially beneficial uses, such as those that promote freedom of speech and truth seeking.22 The market failure theory also poses a risk of "eternal contraction" of fair use because technology can be expected to reduce transaction costs,23 Which would likely cause fair use to shrink accordingly. This would mean that copyright could no longer facilitate the social benefits that flow from public-regarding fair uses. This critique of the market failure theory of fair use is among the strongest and most persuasive parts of Dual-Grant, although it is worth noting that the market failure theory has had more salience in the law review literature than in the case law.24

To overcome the limitations of the market failure theory of fair use, Dual-Grant identifies three types of uses that should qualify for user-rights status: those that involve political speech, those that involve truth seeking, and those that involve criticism and commentary.25 The justification it offers for privileging these uses with the elevated status of user rights lies in the "widespread nonpecuniary follow-on benefits [the use would have] for nonusers."26 Under the dual-grant theory, these types of uses would be presumptively fair.27 Copyright plaintiffs could overcome this presumption only by presenting evidence of harm to the market "so extreme as to eliminate the incentives to create the work."28 The benefits that Dual-Grant anticipates would flow from adoption of

its two-step fair use analysis would be greater predictability, coherence, and ease of administration.29

II. DUAL-GRANTRUNS AFOUL OF THREE MAJOR PRECEDENTS

By endorsing a presumption of fairness if the use qualifies for user-rights status, and by placing a heavy burden on the copyright owner to prove that its incentives to create would be destroyed to overcome that presumption, Dual-Grant takes issue, albeit sub silentio, in three respects with the Supreme Court's decision in Campbell v Acuff-Rose Music, Inc.30 For one thing, the Court in Campbell characterized fair use as an affirmative defense,31 which lower courts have interpreted to mean that defendants always bear the burden of proof in fair use cases.32 In recent scholarship, Professor Lydia Loren has shown that this interpretation of fair use is incorrect.33 Congress meant for fair use to be a defense to infringement in the sense that defendants must raise the issue when they believe their uses of someone else's work are fair, but defendants should not bear the burden of persuasion, as that would mean that if the evidence was in equipoise, the defendant would inevitably lose.34 At least one appellate court has recently agreed with Loren and characterized fair use as a defense, but not an affirmative defense.35 Dual-Grant is not consistent with what Campbell said about burden of proof issues or with how that decision has been understood over time.

A second and more significant respect in which Dual-Grant

deviates from Campbell arises from the Supreme Court's repudiation of presumptions in fair use cases in favor of a "sensitive balancing of interests."36 Third, the Court in Campbell recognized that courts in fair use cases should pay attention to evidence of substantial adverse impacts on the market,37 but did not go so far

29 See id at 1085-87. 30 510 US 569 (1994). 31 32 See, for example, Cambridge University Press v Patton, 769 F3d 1232, 1259, 1280 (11th Cir 2014) (finding that Georgia state officials had the burden of proving a fair use defense).

Id at 590.

33 See Lydia Pallas Loren, Fair Use: An Affirmative Defense?, 90 Wash L Rev 685,

696-704 (2015).

34 See id at 699-704. 35 See Lenz v Universal Music Corp, 801 F3d 1126, 1133 (9th Cir 2015). The Supreme Court asked the solicitor general to offer his views on whether the Court should hear the appeal, but ultimately denied certiorari. Universal Music Corp v Lenz, 137 S Ct 416 (2016).

36 See Campbell, 510 US at 584-85. 37 See id at 590. The Court was also much more favorably disposed toward parodies as fair uses, see id at 579-85, than Dual-Grant is, but at least the article acknowledges

as to require evidence of the total destruction of authorial incentives, as the Dual-Grant theory does.38 Given these deviations from Campbell, it will be surprising if courts adopt the dual-grant theory unless and until the Supreme Court itself does so.

Another Supreme Court decision with which Dual-Grant is inconsistent is Harper & Row, Publishers, Inc v Nation Enterprises,39 but at least the article says so directly.40 Dual-Grant proffers this decision as an exemplar of uses of copyrighted materials for "political speech" purposes that should be presumptively fair4l and asserts that President Gerald Ford would have written his memoirs regardless of the risk that The Nation would scoop Time and publish excerpts from Ford's discussion of the Nixon pardon, so that the dual-grant presumption of fair use ought to have prevailed.42 Bell and Parchomovsky are not, of course, the first copyright scholars to have criticized the outcome or reasoning of the H&R decision.43 Their dual-grant theory of fair use may spark new interest in questioning that decision, although courts will continue to treat H&R as good law until the Court reverses itself.

Dual-Grant also takes aim at American Geophysical Union v Texaco Inc,44 which held that photocopying of scientific and technical articles for research purposes was an infringement, not a

this deviation from Campbell. Bell and Parchomovsky, 83 U Chi L Rev at 1101-04 (cited in note 3).

38 See Bell and Parchomovsky, 83 U Chi L Rev at 1086 (cited in note 3). 39 471 US 539 (1985) ("H&R"). 40 See Bell and Parchomovsky, 83 U Chi L Rev at 1090-91 (cited in note 3). 41 See id. I regard H&R as a news reporting case, not as a political speech case. It was only incidental to the outcome and reasoning of the H&R decision that the author of the work in question was a politician. See H&R, 471 US at 557 & n 6. Had The Nation instead excerpted parts of a forthcoming Harry Potter novel or of a biography of Bruce Springsteen, the outcome and the analysis of the Supreme Court's decision would have been the same; yet these would not be "political speech" cases. Dual-Grant does not explain why its characterization of H&R is more appropriate than the conventional understanding of it as a news reporting case. See, for example, Michael J. Madison, Fair Use and Social Practices, in Peter K. Yu, ed, 1 Intellectual Property and Information Wealth: Issues and Practices in the Digital Age 177, 189-91 (Praeger 2007) (describing the result of the case as a disagreement about journalistic practices). Some news reporting is presumably presumptive fair use under Bell and Parchomovsky's "truth seeking" category of presumptively fair uses. See Bell and Parchomovsky, 83 U Chi L Rev at 1091-96 (cited in note 3). 42 See Bell and Parchomovsky, 83 U Chi L Rev at 1091 (cited in note 3). The Second Circuit's decision made clear that Time was willing to publish the excerpts for which it had contracted, The Nation's scoop notwithstanding, but Harper & Row refused to go ahead with this, perhaps to make its "harm" claim stronger. See Harper & Row, Publishers, Inc u Nation Enterprises, 723 F2d 195, 199, 208 (2d Cir 1983).

43 See Bell and Parchomovsky, 83 U Chi L Rev at 1090 n 157 (cited in note 3) (citing

five articles critical of H&R).

44 60 F3d 913 (2d Cir 1994).

fair use.45 Under the dual-grant theory, the scientists' research (or, as Dual-Grant conceives it, truth-seeking) purpose should have made the use presumptively fair. Unless the challenged photocopying would destroy incentives to produce the articleswhich the authors guess would be unlikely-the use should have been found to be fair.46 The willingness and ability of the Copyright Clearance Center (CCC) to offer photocopy licenses was, according to Dual-Grant, "largely beside the point,"47 even though this was the key issue for the Second Circuit majority.48

Professors Bell and Parchomovsky are less willing to affirm the public-regarding nature of fair use when cases involve nonprofit educational contexts,49 even though three of the six statutorily favored uses-"teaching ... fall within that general rubric, and the fair use provision goes on to favor "nonprofit educational purposes" as a consideration in its articulation of the purpose factor.so Although the authors approve of the Second Circuit's digital books rulings because of the public benefits flowing from these digital libraries,51 they express skepticism about the electronic course reserves fair use defense in

Cambridge University Press v Becker.52 The authors recognize that education is a public good and that enabling students to have electronic access to copies of in-copyright materials may serve research purposes.53 Yet the authors balk at supporting fair use in the Becker case, saying that "if universities and other educational institutions were permitted under the scope of fair use to include complete copies of mandatory, copyrighted course materials in course portals, it is likely that authors' incentives to create at least some of these works would be in jeopardy."54

With all due respect, I beg to differ. The Dual-Grant theory should pay more attention to the findings of fact in the Becker case and think harder about the electronic course reserves issue. The overwhelming majority of the claimed infringements in Becker were book chapters, written by academic authors, often from decades-old edited collections.55As someone who has written numerous chapters for edited books, I know what motivates scholars like me to create those book chapters: we write these chapters mostly as favors to colleagues who got a book contract, and we hope that the book chapter will enhance our reputation as good scholars. Our motivation to write these chapters would not be diminished, and might well be enhanced, by their appearance on electronic course websites. When instructors at universities such as Georgia State assign academic-authored book chapters as readings for courses, we are overjoyed, as it means someone knows our work exists and finds it valuable.

When Cambridge University Press (CUP) decided to publish books in which the relevant book chapters appeared, it typically did not pay the authors for individual book chapters, and it was not expecting to make money from sales or licensing of those chapters, but from the sale of books. Evidence in the Becker case has shown that revenues from digital licensing of book chapters has been a trivial source of income to publishers.56

It is important to realize that the Becker lawsuit is a test case, funded by the Association of American Publishers and the Copyright Clearance Center (CCC), whose goal has been to set a precedent against Georgia State University that would give the publishers considerable leverage to insist that every college and university take a license from CCC to use in-copyright materials for electronic course websites.57 If CUP succeeds in the lawsuit, none of these revenues are likely to flow to the authors of the book chapters.58 The Eleventh Circuit agreed with the district court in Cambridge University Press v Patton59 that the educational and noncommercial nature of the uses weighed in favor of fair use,60 and as long as the market harm was either nonexistent or relatively small, the challenged uses should be fair.61

IV. SOME PUBLIC-REGARDING USES Go UNRECOGNIZED IN DUALGRANT

Professors Bell and Parchomovsky elevate the political speech, truth telling, and critical comment types of fair uses to "user rights" status, but they do not address some other situations that seem plausibly eligible for similarly privileged status. DualGrant seems quite indifferent, for example, to freedom of expression interests of authors who draw on others' works in expressing themselves in the creation of new works, even though this too seems consistent with the constitutional purpose of copyright. They mention, although they do not endorse, a fair use case involving visual artist Jeff Koons, who made copies of a photograph taken by photographer Andrea Blanch as part of a large-scale

293-93, CUP has filed a new appeal of the fair use ruling to the Eleventh Circuit. See generally Brief of Appellants Cambridge University Press, Oxford University Press, Inc., and Sage Publications, Inc., Cambridge University Press v Becker, No 16-15726 (11th Cir filed Nov 18, 2016).

57 See, for example, Meredith Schwartz, Georgia State Copyright Case, What You Need to Know-and What It Means for E-reserves (Library Journal, May 17, 2012), archived at http://perma.cc/6FQC-PJ78.

It has been a long-standing and very common practice among academic publishers to insist that authors of books and book chapters assign copyrights to the publishers, which is why publishers, not authors, brought this lawsuit. See Authors, Keep Your Copyrights. You Earned Them. (Authors Guild, Aug 13, 2015), archived at http://perma.cc/B2FK9PNZ.

59 769 F3d 1232 (11th Cir 2014). 60 61

Id at 1263-68. Id at 1275-81.

work that featured some portions of the photograph in a new context.62

Bell and Parchomovsky are right that the fair use ruling in Blanch v Koons63 did not depend on the existence (or not) of market failure.64 The Second Circuit was deferential not only to the artist's felt need to use parts of the Blanch image, but also to the public interest in having access to this stimulating artwork.65 This latter interest would seem to reflect "nonpecuniary [ ] benefits" that affected "nonusers" of the work, on which Bell and Parchomovsky rest the justification for their dual-grant theory.66 Yet, Dual-Grant would not confer user rights on creative reusers,67 even though this is at odds with the post-Campbell transformative use cases.68

Dual-Grant also takes no stance on the fair use status of copying computer program object code during a reverse engineering process for purposes such as extracting information necessary to create an interoperable program. Appellate courts have deemed such uses to be fair,69 even though Congress in 1976 could not have foreseen that these types of uses might be fair. 70 Members of the general public enjoy significant benefits when this kind of reverse engineering is lawful because it makes development of competing and complementary products more available in the marketplace.71 One might have hoped that Dual-Grant would move beyond the First Amendment-related uses of copyrighted materials to recognize social benefits of some technological fair uses that affect the availability of new noninfringing works to be available to the public.

Dual-Grant provides little or no guidance about how the theory would have courts deal with run-of-the-mill fair use cases that involve more ordinary uses, such as those aimed at setting historical context,72 making incidental uses of copyrighted materials in later works,73 showing images in comparative advertising,74 and making copies of an artist's work for her portfolio.75 As a bottomup scholar of copyright law, I know that these kinds of putative fair uses are very common. The dual-grant theory does not directly say so, but it would seemingly consign them to a market failure test,76 although to the extent that fair use cases involve commercial uses, Dual-Grant would presume those uses to be unfair.77

Under the dual-grant theory, personal uses would come under case-by-case, and market-bytechnology-by-technology, market scrutiny,78 rather than enjoying, as in Sony,79 a presumption that such uses are fair unless copyright owners prove demonstrable harm to the market. This consignment of personal uses to the vagaries of the common law would achieve less predictability for fair use than the authors of Dual-Grant seemed to promise.

incredibly important whether putative fair uses are "transformative" or "non-tranformative."80 Litigants in many cases struggle mightily to have the challenged uses designated as "transformative" because Campbell directs much more nuanced and defensefriendly analyses of fair use factors in transformative-use cases.81 For instance, publishers characterized as nontransformative the practice of certain patent lawyers who made exact copies of scientific and technical articles for submission to the US Patent & Trademark Office (PTO) in connection with their clients' applications for patents on technologies for which the articles were disclosed as prior art.82 Because these submissions were for a different purpose than the works' original purpose, the patent lawyer defendants were able to persuade the court to deem the use as transformative.83 The characterization of the use as transformative was a key element in the lawyers' successful fair use defense.84

Under the dual-grant theory, lawyer-submissions-of-priorart copying might be valorized as truth-seeking uses that would qualify for Dual-Grant's strong presumption of fairness.85 Because scientists will continue to write papers about their experiments and theoretical findings even if patent lawyers are allowed to make copies of those articles to send to the PTO, the lawyers would likely prevail in a similar case under the dual-grant theory.86

But what about Dorling Kindersley (DK)'s use of Grateful Dead posters in its book featuring a cultural history of the band?87 If DK could persuade the court to consider this use a truthseeking one-a bit of a stretch-then the strong presumption of fairness would kick in, and the publisher would win.88 But if Bill Graham Archives could persuade the court that the use was

commercial and not public-regarding, then the Dual-Grant's presumption against fair use would seemingly kick in.89 Bill Graham Archives, which owns copyrights in the posters, would then probably win, especially given its willingness to license such uses.90 The outcome of the case would depend on which conception of the use the courts found persuasive.

This example demonstrates the risk that the dual-grant theory's competing dual presumptions, some for and some against fair use, would make fair use law less predictable than it is now. This would undermine at least one of the goals that Bell and Parchomovsky aim to achieve through recommending judicial adoption of their dual-grant theory.91

Among the possible implications of a "user rights" conception of fair use that Dual-Grant does not consider is whether userrights fair uses should be protected against contractual overrides.92 Under the dual-grant theory, perhaps contracts could not override user-rights fair uses, at least as to mass-market end user license terms.93 Perhaps contractual overrides would, under their theory, be more enforceable as to uses that would have been fair or otherwise privileged but for the contractual restriction.94 Or perhaps those who seek to make user-rights fair uses should be entitled to bypass technical protection measures used by copyright owners or bring lawsuits to insist on their right to exercise political speech or other user-rights fair uses.95 Perhaps the authors can address these questions in subsequent work.

The theoretical contribution that Dual-Grant makes to the law review literature will further burnish, deservedly so, the reputations of its coauthors as ambitious scholars who articulate novel theories aimed at bringing coherence to the law of copyright. The article and its theory are stimulating and worthy of serious engagement. In the end, however, neither is persuasive for reasons set forth in this Essay. And I predict courts will ignore this theory, in no small part because it repudiates important rulings by the Supreme Court, including Campbell's rejection of the utility of presumptions in fair use cases.

Having studied the fair use case law from the bottom up, I can say with confidence that the fair use case law has been moving in a generally very positive direction, in keeping with the constitutional purpose of copyright and the Sony dictum that would deem as fair any use that does not cause demonstrable harm to the market for copyrighted works or undermine incentives to create works of authorship in the first place. Sony has set forth a simple rule that has general application in fair use cases and not just to the subset of user-rights fair use cases to which the dualgrant theory would apply.

--- FOOTNOTES ---

INTRODUCTION

Some conceptions of the fair use limitation of US copyright law have their groundings in the case law out of which the doctrine emerged., (I call these the "bottom-up" approaches.) Other theories of fair use have sprung from the very bright minds of copyright scholars whose collective goal has generally been to bring some needed coherence to the common law of fair use.2 (I call these "top-down" approaches.) The Dual-Grant Theory of Fair Use is the latest of the top-down theories to have appeared in the literature.3

The concept of fair use was first articulated in an 1841 copyright decision by Justice Joseph Story in Folsom v Marsh.4 In deciding whether a use was fair or infringing, Story thought that courts should consider "the nature and objects of the selections

t Richard M. Sherman Distinguished Professor of Law, University of California,

Berkeley School of Law.

1 See, for example, Harper & Row, Publishers, Inc u Nation Enterprises, 471 US 539, 550-51 (1985) (endorsing the "implied consent" theory of fair use, under which a use will be deemed fair if a reasonable author would have agreed to it), citing Folsom V Marsh, 9 F Cases 342, 344-349 (CCD Mass 1841). See also Alan Latman, Fair Use of Copyrighted Works, Study No 14, Subcommittee on Patents, Trademarks, and Copyrights of the Senate Committee on the Judiciary, 86th Cong, 2d Sess 6-18 (1960) (reporting on the author's empirical study of the fair use case law and recommending codification of the fair use limitation). 2

See, for example, William W. Fisher III, Reconstructing the Fair Use Doctrine, 101

Harv L Rev 1659, 1766-83 (1988) (articulating a semiotic democracy theory of fair use).

See Abraham Bell and Gideon Parchomovsky, The Dual-Grant Theory of Fair Use,

83 U Chi L Rev 1051, 1059 (2016).

9 F Cases 342 (CCD Mass 1841). For an excellent discussion of this case, see generally R. Anthony Reese, The Story of Folsom v. Marsh: Distinguishing between Infringing and Legitimate Uses, in Rochelle Cooper Dreyfuss & Jane C. Ginsburg, eds, Intellectual Property Stories 259 (Foundation 2006).

5 Folsom, 9 F Cases at 348. 6 7 See Copyright Act of 1976 § 107, Pub L No 94-553, 90 Stat 2541, 2546, codified as

See Latman, Fair Use of Copyrighted Works at 8-14 (cited in note 1).

amended at 17 USC § 107.

17 USC § 107. 17 USC § 107.

9 10 See Glynn S. Lunney Jr, Fair Use and Market Failure: Sony Revisited, 82 BU L Rev 975, 977 (2002) (viewing fair use as an "arbiter between two competing public interests").

There are two different approaches one can take when developing an overarching theory of fair use. One is a top-down theoretical approach, such as the one that Professors Abraham Bell and Gideon Parchomovsky have taken in Dual-Grant. The other is a bottom-up approach, which involves reading the fair use cases closely and attempting to discern whether there is, in fact, any unifying principle that underlies the common law of fair use.

My article Unbundling Fair Uses was the product of this kind of bottom-up approach, based on my reading of more than three hundred fair use cases in the order in which they were decided.12 One conclusion of that paper was that fair use law is much more coherent and predictable than its critics often say. 13 After reflecting on the fair use cases and the policy-relevant clusters in which they typically fall, I offered a bottom-up theory of fair use as a manifestation of the limited monopoly conception of copyrightl4 that is reflected in this statement from the Supreme Court's Sony Corp of America v Universal City Studios, Inc15 decision: "[A] use that has no demonstrable effect upon the potential market for, or the value of, the copyrighted work need not be prohibited in order to protect the author's incentive to create. The prohibition of such []uses would merely inhibit access to ideas without any countervailing benefit." 16 Although Justice John Paul Stevens made this statement in the context of considering the fairness (or lack of it) of private, noncommercial copying, the observation strikes me as a generally sound principle of fair use law. in limited monopoly Unbundling Fair Uses may not be as elegant as the overarching top-down theory endorsed in Dual-Grant, but it does have the virtue of accounting for a much wider range of contexts in which fair use has been raised. My theory is also consistent with fair use law

theory endorsed

The bottom-up

11 See, for example, Bell and Parchomovsky, 83 U Chi L Rev at 1053 (cited in note 3) (describing fair use as "amorphous and vague, and ... notoriously difficult to apply"); Michael W. Carroll, Fixing Fair Use, 85 NC L Rev 1087, 1092-1121 (2007) (criticizing the application of the fair use doctrine as providing insufficient ex ante clarity).

12 Pamela Samuelson, Unbundling Fair Uses, 77 Fordham L Rev 2537, 2542 n 29

(2009). 13 14

Id at 2541. Id at 2617 (endorsing the limited monopoly theory of copyright as a grounding for

the fair use doctrine).

15 464 US 417 (1984). Id at 450-51. 16

Bottom-Up and Top-Down Theories ofFair Use

17 See Bell and Parchomovsky, 83 U Chi L Rev at 1117 (cited in note 3). 18

Id at 1055-56.

19 For an early conception of copyright as conferring rights on users, see L. Ray Patterson and Stanley W. Lindberg, The Nature of Copyright: A Law of Users'Rights 191-222 (Georgia 1991). See also generally Abraham Drassinower, Taking User Rights Seriously, in Michael Geist, ed, In the Public Interest: The Future of Canadian Copyright Law 462 (Irwin Law 2005); David Vaver, Copyright Defenses as User Rights, 60 J Copyright Society USA 661 (2013). Other copyright professionals and copyright industry representatives are less receptive to the user-rights conception of fair use. See, for example, David R. Johnstone, Debunking Fair Use Rights and Copyduty under U.S. Copyright Law, 52 J Copyright Society USA 345 (2005).

20 See Bell and Parchomovsky, 83 U Chi L Rev at 1060-66 (cited in note 3). Part IV of Dual-Grant discusses several other theories of fair use and compares them with the dual-grant theory. Id at 1107-17. Given the space limits of this Essay, I will not address how the authors deal with the other theories.

21 This theory is widely attributed to Wendy J. Gordon, Fair Use as Market Failure: A Structural and Economic Analysis of the Betamax Case and Its Predecessors, 82 Colum L Rev 1600, 1614-22 (1982).

22 See Bell and Parchomovsky, 83 U Chi L Rev at 1054-55 (cited in note 3). 23

Id.

24 The Supreme Court in Sony, for instance, made no reference to market failure, and its analysis depended instead on a conception of copyright as a limited monopoly grant. See Sony, 464 US at 450-51. Having read more than three hundred fair use cases, I can attest that market failure is almost never cited as a factor in fair use cases. See Samuelson, 77 Fordham L Rev at 2579 (cited in note 12) (noting that courts generally do not analyze these cases in terms of market failure). Because of this, I disagree with Bell and Parchomovsky's characterization of the market failure theory as the "predominant" justification. Bell and Parchomovsky, 83 U Chi L Rev at 1060 (cited in note 3).

25 See Bell and Parchomovsky, 83 U Chi L Rev at 1087-99 (cited in note 3). These categories overlap to some degree with three of the six favored statutory uses in 17 USC § 107, namely, "criticism, comment, [and] news reporting." 17 USC § 107. Political speech is, of course, not the same as news reporting. Bell and Parchomovsky believe this type of use should have special status as well because of the importance of political speech as a First Amendment interest. Bell and Parchomovsky, 83 U Chi L Rev at 1088-89 (cited in note 3). They do not ground the truth seeking purpose in the text of § 107. See id at 1091-96.

26 Bell and Parchomovsky, 83 U Chi L Rev at 1085 (cited in note 3). 27 See id. 28

Id at 1086.

III. DUAL-GRANT OVERLOOKS IMPORTANT CONSIDERATIONS AFFECTING EDUCATIONAL USES

, scholarship, [and] research"-

Id at 931-32.

45 46 See Bell and Parchomovsky, 83 U Chi L Rev at 1093 (cited in note 3). 47

Id. These authors are not the first copyright scholars to have criticized the Texaco decision. See, for example, Ann Bartow, Educational Fair Use in Copyright: Reclaiming the Right to Photocopy Freely, 60 U Pitt L Rev 149, 197-202 (1998).

48 See Texaco, 60 F3d at 929-32. Texaco maybe the clearest example of the difference in outcome of a pure market failure approach to fair use as compared with the dual-grant theory.

49 See Bell and Parchomovsky, 83 U Chi L Rev at 1098 (cited in note 3) (suggesting that educational uses should not be prima facie fair uses unless they involved political speech or other public-regarding uses or "when the use was small enough that it could be justified by a transaction cost argument").

17 USC § 107. The statute also indicates that "multiple copies for classroom use"

may be fair. 17 USC § 107.

51 See Authors Guild, Inc u Google Inc, 954 F Supp 2d 282, 293-94 (SDNY 2013), affd, 804 F3d 202 (2d Cir 2015) (finding that it was fair use to make copies of research library books for purposes of indexing their contents and serving up snippets to users); Authors Guild, Inc u HathiTrust, 755 F3d 87, 102-03 (2d Cir 2014) (finding that it was fair use for libraries to have digital copies of research library books for text mining purposes). These cases are discussed in Bell and Parchomovsky, 83 U Chi L Rev at 1094-96 (cited in note 3), although the authors focused on the district court, not the Second Circuit, decisions.

52 863 F Supp 2d 1190 (ND Ga 2012), revd, Cambridge University Press v Patton, 769 F3d 1232 (11th Cir 2014). See also Bell and Parchomovsky, 83 U Chi L Rev at 1094-96 (cited in note 3).

53 See Bell and Parchomovsky, 83 U Chi L Rev at 1094-96 (cited in note 3). 54

Id at 1096. They say the Becker case "calls for a more cautious approach" and requires a "more contextual approach," but don't come out directly to say that Cambridge University Press should win. Id at 1095-96. Notice that Bell and Parchomovsky focus here on authorial incentives, not on the incentives of publishers to commercialize authorial works.

55 See Cambridge University Press v Becker, 2016 WL 3098397, *86-87 & n 52. 56 See Patton, 769 F3d at 1275-81. See also Becker, 2016 WL 3098397 at *12, 14-15 (describing the findings of fact in the district court decision on remand). After once again losing its copyright claims as to all but a few of the challenged course reserve uses, id at

62 See Bell and Parchomovsky, 83 U Chi L Rev at 1067-68 (cited in note 3), citing

Blanch u Koons, 467 F3d 244 (2d Cir 2006).

63 467 F3d 244 (2d Cir 2006). 64 See Bell and Parchomovsky, 83 U Chi L Rev at 1067-68 (cited in note 3). 65 See Blanch, 467 F3d at 253-55. 66 Bell and Parchomovsky, 83 U Chi L Rev at 1085 (cited in note 3). 67 Dual-Grant does address fan fiction, which it would allow as long as the uses were

noncommercial and the fan fiction work was made public. See id at 1106-07.

68 See Pamela Samuelson, Possible Futures of Fair Use, 90 Wash L Rev 815, 817-39

(2015) (discussing Campbell's impact on fair use cases).

69 See, for example, Sega Enterprises Ltd u Accolade, Inc, 977 F2d 1510, 1520-28 (9th

Cir 1993).

70 The closest category is "research," but Congress did not contemplate that this would include reverse engineering by commercial firms in the process of developing software. The kind of research that was in contemplation at the time was of the sort at issue in Williams & Wilkins Co u United States, 487 F2d 1345 (Ct Cl 1973), in which the court found it to be fair use for National Institutes for Health librarians to make photocopies of scientific articles for nonprofit research purposes. Id at 1362.

71 See Sega, 977 F2d at 1526-27.

V. DUAL-GRANT IGNORES MANY FAIR USE CONTEXTS AND MAY MAKE FAIR USE MORE UNPREDICTABLE

Dual-Grant also ignores the risk that the authors' theory, if taken seriously in litigation, may make fair use disputes more unwieldy. When one elevates certain kinds of fair uses to a special status, as Dual-Grant would do, there will inevitably be fights about whether any particular case being litigated qualifies for that special status. Since Campbell, for instance, it has become

72 See, for example, Bill Graham Archives v Dorling Kindersley Ltd, 448 F3d 605, 615 (2d Cir 2006) (finding it to be fair use to reproduce Grateful Dead concert posters in a cultural history book).

73 See Samuelson, 77 Fordham L Rev at 2575-76 (cited in note 12). 74 See generally, for example, Sony Computer Entertainment America, Inc v Bleem, LLC, 214 F3d 1022 (9th Cir 2000) (holding that an advertisement using comparative screenshots from console-run and computer-run video games likely constituted fair use).

75 See, for example, Neri v Monroe, 2014 WL 793336, *9 (WD Wis) (granting summary judgment to the defendants based on their fair use defense as to photographs that they used in highlighting their own contributions to the interior design).

76 See Bell and Parchomovsky, 83 U Chi L Rev at 1071 (cited in note 3). 77 See id at 1058. 78 See id at 1103-06 (noting that the strength of justification for fair use will vary

with the particular technology and social mores).

79 See Sony, 464 US at 449-51.

80 See generally Matthew D. Bunker and Clay Calvert, The Jurisprudence of Transformation: Intellectual Incoherence and Doctrinal Murkiness Twenty Years after Campbell v. Acuff-Rose Music, 12 Duke L & Tech Rev 92 (2013-2014) (discussing the rise of transformative use after Campbell and its creation of difficulties for the courts).

I discuss the different purpose fair use cases in Samuelson, 90 Wash L Rev at 845-

50 (cited in note 68).

82 See, for example, American Institute of Physics v Winstead PC, 2013 WL 6242843,

*6 (ND Tex).

Id. I have suggested that the different purpose fair use cases be recognized as a category separate from the more conventional types of transformative uses. Samuelson, 90 Wash L Rev at 854-55 (cited in note 68).

84 See, for example, American Institute of Physics, 2013 WL 6242843 at *6. 85 See Bell and Parchomovsky, 83 U Chi L Rev at 1093 (cited in note 3). 86 See id. 87 See Bill Graham Archives, 448 F3d at 607. 88 See Bell and Parchomovsky, 83 U Chi L Rev at 1091-96 (cited in note 3).

89 See id at 1058. 90 See Bill Graham Archives, 448 F3d at 607. 91 See Bell and Parchomovsky, 83 U Chi L Rev at 1057 (cited in note 3) ("The key to creating a viable dual-grant strategy ...is producing a workable dividing line between privileged uses and protected rights.").

92 Courts might, for instance, decide not to enforce contract provisions that purported to bind users not to criticize the copyright owner or the work at issue. For commentary on the enforceability of IP policy-related license restrictions, see Mark A. Lemley, Beyond Preemption: The Law and Policy of Intellectual Property Licensing, 87 Cal L Rev 111, 14344 (1999).

93 For a discussion of fair use restrictions, see Samuelson, 90 Wash L Rev at 859-62

(cited in note 68).

94 See, for example, Vernor v Autodesk, Inc, 621 F3d 1102, 1111-16 (9th Cir 2010) (enforcing a license restriction on transfers of purchasers' copies of computer software, although but for the license restriction, such transfers would have been lawful under 17 USC § 109(a .

95 See Samuelson, 90 Wash L Rev at 859-62 (cited in note 68) (discussing fair use as

a justification for bypassing technical protection measures).

CONCLUSION