Patrick Goold’s interesting new article, Unbundling the “Tort” of Copyright Infringement[1] (“Unbundling”) centers on a key lack of clarity that Professor Goold perceives in the cause of action for copyright infringement. The lack of clarity, he argues, afflicts threshold definitions of what constitutes actionable copying.
Under federal copyright law, to prove infringement the plaintiff copyright owner usually must first persuade the finder-of-fact that the plaintiff owns a valid copyright and that the defendant factually used the copyrighted work in one of the ways governed by statute.[2] Then the plaintiff must prove something more.[3] The copyright owner also bears what might be called a normative burden: The plaintiff must prove that the defendant has engaged in “improper appropriation”[4] by using the plaintiff’s copyrighted work to produce something “substantially similar”[5] to the plaintiff’s work of authorship.[6] Later, the copyright owner may also need to struggle with a normative claim by the defendant that her use of the plaintiff’s expression, even if “substantial,” should be permitted as “fair.”[7]
As Goold notes,[8] courts do not reliably define the normative part of the plaintiff’s cause of action in the same way. The terms “substantially similar” or “improper appropriation” receive somewhat varying interpretations. A related difficulty that Goold identifies typically arises on the defense side: He perceives inconsistency in how courts distinguish “fair” from “unfair” uses.[9] In Unbundling, Goold suggests that more clarity would result if the judiciary would follow his recommendations and identify within copyright infringement several individual, if currently inchoate, tort causes of action.[10]
Goold in previous work has made good use of tort doctrine to explore copyright law,[11] and his new project has intriguing possibilities. Today’s federal copyright statute runs for many pages that are often dense with complex language, yet its details nevertheless fail to resolve many cases. Sometimes fraying or uncertainty in core concepts is responsible for this failure. Similarly hoping to increase clarity, Pamela Samuelson is unpacking many of the ways that courts employ copyright’s “merger” doctrine.[12] She, along with other scholars, has begun unbundling and identifying several distinct defenses hiding behind the label “fair use.”[13] Given copyright’s many unsolved puzzles, determining whether the “substantial similarity” or “improper appropriation” aspect of infringement can be unbundled is a route that certainly has promise.
Unbundling argues that copyright infringement appears to be a unitary tort, but that it actually contains within itself five unarticulated subtorts. Goold suggests that if in fact we could unpack copyright into its component concerns, a “gallery of wrongs”[14] of the type other torts possess, there might well be a distinct-infringement test for each type of infringement.[15] Thus, he argues that discerning the different subtorts within the copyright-infringement bundle could help judges choose, and lawyers anticipate, the appropriate test for identifying infringing uses. This is an advance, so far as I know, over prior applications of the unbundling method in copyright scholarship.
He suggests that the standard copyright-owner interest in revenue be divided into two subtorts: protection from unauthorized consumer copying and protection from competitors diverting one’s potential customers. In addition, Goold suggests that within copyright lie three additional causes of action geared to protect copyright owners’ interests in privacy, in reputation, and in controlling “rivalrous”[16] uses.[17]
The five subtorts that Goold offers are: (1) consumer copying, which he identifies as the primary “wrong” with which copyright law is concerned,[18] followed by (2) diversion of customers by competitors,[19] (3) invasion of expressive privacy,[20] (4) injury to artistic reputation,[21] and (5) breach of creative control (by which he means interference with a rivalrous use).[22] For mainstream interpretations of copyright, Goold’s first two initial categories—customer copying and competitor/publisher diversion—are the standard concerns that copyright courts address. The other three are more controversial than the article indicates.
Of those three, a right to control rivalrous uses probably has the strongest claim to being based in the case law. Unfortunately, that right has probably inescapable definitional weaknesses that make it dangerously susceptible to expansion. The subtort to redress reputation injury flowing from misattribution of authorship has some intriguing possibilities, though its feasibility is also questionable (though for different reasons). As for the invasion of privacy subtort, it goes strongly against the grain of some recent copyright cases.
My own view is that Goold overstates the explanatory role of tort law.[23] But even were that not the case, the courts need to reach some kind of “settled” understanding on these various interests before a cause of action is created or definitively rejected, and that no such consensus on the three matters mentioned yet exists, whether they are viewed as forms of tort or otherwise. Goold’s work may nevertheless be an important step toward reaching closure on these and other open questions in copyright law.
Let us take the five categories of Goold’s subtorts in order.
I. Consumer Copying
Copyright’s familiar concerns lie with commercially significant copying by competitors and consumers. It is this commercial recompense that an author hopes for, and that serves as incentive for production. Goold takes an unconventional approach, though, dividing consumer from competitor copying.
Goold’s first major innovation is to put consumer copying as the copyright’s central tort within a tort, and to designate publisher diversion of consumers as a secondary concern of copyright law. This elevation of consumers as copiers is profoundly ahistorical. The United States borrowed its initial copyright scheme largely from the English Statute of Anne, which in turn evolved out of battles among publishers.[24] In the United States, the first federal Copyright Act prohibited only the unconsented “printing, reprinting, publishing and vending” of the copyrighted work.[25] As a physical matter in 1790, consumers could neither print nor reprint. Consumers might copy longhand, but that was unlikely to be commercially significant.
Copyright law’s focus on publisher behavior persisted well into the twentieth century, even after the start of modern home copying technology. For example, when in 1972 sound recordings were made federally copyrightable, Congress went out of its way to explain that home copyists could continue to safely use their tape recorders to make permanent copies of their favorite songs because (says the legislative history) copyright law only addresses commercial copying.[26] At least in 1972, Congress seemed to envisage no possible liability for private behavior, whether the private behavior was copying a work or performing or adapting it.[27] Only when the home product was physically replicated and sold commercially would copyright law take action.
Goold’s paper addresses a world greatly changed since 1972. The progress of home reprographic and distribution technology, in the form of computers, tape recorders, video recorders, internet linkages, and the rest, has been so rapid as to make a profound difference in consumer abilities to create, obtain, and transmit copyrighted material.
Goold implies the law has also greatly changed. Indeed, Congress has paid increasing attention to consumer copying and transmission. One example is copyright’s criminal provisions. In 1978, private copying for noncommercial purposes was essentially free from criminal sanction.[28] In recent years, however, amendments to the criminal law provisions have largely eliminated that safe haven.[29] Another change lies in Congress not only permitting copyright owners to physically encrypt their digital work, but also backing up the encryption with federal penalties for bypass.[30]
Yet Goold overstates by placing consumer copying at the center of 2017 copyright law. Congress made its last major overhaul of federal copyright law in 1976, and the vast majority of the 1976 provisions remain intact. Admittedly, there have been several high-profile suits concerning peer-to-peer networks, and the district courts have seen a flood of “idiosyncratic” suits[31] seeking to milk disproportionate statutory damages from downloaders.[32] But the latter suits, though numerous, are not a reliable focus for assessing copyright policy.[33] Putting them aside, the vast majority of copyright litigation addresses copying and adaptation by commercial companies and republishers, not copying by consumers. The shift of focus to consumer copying might be justified given the drastic increase in potential commercial significance that consumer behaviors now have, but Goold offers no sustained argument to that effect,[34] and no convincing evidence that consumer copying constitutes a more significant infringement than commercial copying does. If we are to ignore a doctrine’s historical roots, we expect some exploration of the costs and benefits of doing so.
It is hard to see what Goold gains by prioritizing consumer copying. He wants to solve problems in articulating the tests for infringement and fair use, but he exchanges those puzzles for one of the most debated questions in copyright today—whether and under what conditions a failure to pay a license fee should weigh against a private person’s claim to “fair use” treatment.[35] That hardly seems a profitable exchange. As set forth in Unbundling, this new subtort requires, among other things, “[a]pplying the basic incentive-access policy calculus,” under which a court decides whether “specific uses [are] of the type where wealth redistribution [is] necessary to ensure optimal incentives.”[36] And that is only one question in the infringement inquiry.
II. Competitor Diversion of Customers
For most of the copyright bar, Goold’s second proposed category—publisher diversion of customers—is the core of copyright law. So identifying this category is useful primarily in the context of a panoply of meaningful choices; its value depends on the value of how well its competing concepts are set out.
One correction needs to be made to Goold’s explanation of customer diversion by competitors. His article gives an example of this subtort using two lighthouses that steal each other’s customers.[37] The example might easily be misunderstood to suggest that copyright makes actionable any kind of commercial harm that involves competitors selling similar products or services, regardless of the presence or absence of borrowing. Goold should be wary of using an example that may lead readers to overlook the particular kind of causation that is essential to copyright.
A copyright plaintiff must prove that a defendant has somehow used the plaintiff’s work in a way that made a difference to the defendant[38]: Unless free riding on the plaintiff’s work was a “but-for cause” of what the defendant produced or did, no liability arises. There is a requirement that the defendant’s product borrow something from the plaintiff’s work.
This two-lighthouse example might be appropriate were Goold writing about patent law, which does empower suit against independent inventors who happen to provide a product identical to what is patented. By contrast, to make Goold’s example fit copyright law, one lighthouse would have to be taking advantage of the other’s efforts or resources in some way. But in Goold’s actual example, one lighthouse makes no use of anything owned by or produced by the other. In his example as constituted, the only causal link is one of harm. Harm is neither necessary nor sufficient to meet copyright’s causal link. What satisfies copyright’s “copying” element is not harm done to a plaintiff’s market, but rather some benefit the defendant has reaped that is causally due to his or her use of the plaintiff’s work. There is no free riding between the two lighthouses. And while proof of free riding is far from sufficient to prove copyright infringement, it remains an essential component of the prima-facie case.
III. Redress of Privacy Invasions
The third role Goold attributes to copyright is protection against privacy invasions.[39] This is particularly problematic. A myriad of cases reject privacy and other dignitary roles for federal copyright, as will be shown below. It is possible these cases are wrong, but Goold seeks to provide a descriptive account of copyright law.[40] Of particular relevance is the en banc decision of the Ninth Circuit that “the protection of privacy is not a function of the copyright law.”[41]
Until the 1976 Copyright Act became effective in 1978, private and otherwise unpublished manuscripts were largely handled by state rather than federal copyright. Only a few categories of unpublished material were even eligible for federal registration.
Certainly, in 1978, federal copyright law expanded to embrace all unpublished works, so long as they were written down, tape-recorded, or otherwise “fixed.”[42] But this expansion of federal reach did not change federal policy. The key document, the House Report for the 1976 Copyright Act, gives a number of reasons for bringing unpublished works into federal copyright, most having to do with simplicity, administrability, and uniformity.[43] Nowhere is there a hint that federal copyright was meant to adopt any privacy or dignity concerns that states may have injected into their common law or statutory protections for local authors.[44] Further, when post-1978 courts began giving too much deference to the desires of authors to control the first publication of their words, Congress responded by dialing down the deference owed.[45]
Conceivably Goold can make privacy hay out of our country’s accession to the Berne Convention, which caused changes in federal copyright law[46] that arguably introduced personal and emotional concerns into certain subparts of federal law.[47] However, the Berne changes are in tension with America’s iconoclastic free speech tradition, so trying to make them the foundation for a privacy cause of action within copyright will be (or at least should be) an uphill battle.
Goold is right that some federal copyright cases hint that privacy is a legitimate copyright concern. But doctrinal development, demands of internal consistency, and copyright policy lean predominantly the other way. For example, consider the difficulty of inserting privacy concerns into copyright law in light of the historic distinction between ownable expression and unprotectable ideas and facts.[48] The Supreme Court has indicated that this freedom to copy facts may be essential to copyright’s constitutionality.[49] Therefore, no copyright cause of action would lie for uncomfortable facts gleaned from even the most private diary. In addition, a host of differences between copyright law and privacy rights would make it difficult to know how to shape a privacy claim in copyright law. Common law privacy rights are personal, so that, for example, they usually expire upon death,[50] whereas copyright not only survives an author’s death, but can be owned by someone other than the creator.
IV. Artistic Reputational Injury Occasioned by Misattribution of Altered Work
This may well be the most promising new sub-tort in Goold’s arsenal. Before discussing it, let us lay some groundwork.
Copyright is sometimes casually equated with plagiarism, but one of the distinctions between the two lies in the role that reputation plays. Plagiarism results from inaccurate attribution of authorship, usually by someone seeking an improper boost for his or her reputation (or grade) by making unacknowledged use of others’ language. Lack of attribution is the core of the wrong in plagiarism, and proper attribution is its cure.
In copyright, attribution has almost the opposite effect. Proper attribution tends to make copyright infringements more harmful rather than less, as identifying the true author of what an infringer is selling is likely simply to increase the infringer’s profit.
Say, for example, that someone without authorization mass-produces and sells a best-selling Stephen King novel. Sales of the infringing version without King’s name would be lower than the volume of sales that would result if the infringer accurately put King’s name on the cover. Similarly, if someone without authorization translates King’s novel into Spanish, this behavior too will infringe rights held by the copyright owner,[51] and again using King’s name would increase sales. Should the unauthorized translator name herself as sole author of the Spanish-language novel, sales are likely to be modest.
Goold does not fall into the trap of equating copyright with plagiarism. Goold’s article thus plays a valuable role by focusing on one narrow kind of reputational injury: that which flows from producing a degraded or distorted version of the plaintiff’s work and attributing it to plaintiff. An example might be the translator just mentioned: If she made drastic changes to Stephen King’s plot or characters, and nevertheless attributed the quite different work to King, his reputation as a skillful writer might suffer.
But even as to that fact pattern, the article leaves many issues unaddressed. Perhaps most obvious is the danger that the reputation inquiry will collapse into inquiries into whether the second work is truly of lesser quality than the copied work. There are many hazards, some of them sounding in the First Amendment, given that it may see judges involve themselves in deciding what counts as a ‘worthy’ or ‘unworthy’ adaptation of a work of art.[52]
Admittedly, joining the Berne Convention required the United States to adopt the so-called “moral rights” of attribution and integrity,[53] which involve reputation. Congress, however, has been leery of importing such inquiries into American law. The legislation implementing those rights, the Artists Visual Rights Act of 1990 (“VARA”),[54] is quite narrow. Among other limits, it does not give so-called rights of attribution or integrity to literary works or music, or to “any poster, map, globe, chart, technical drawing, diagram, model, applied art, motion picture or other audiovisual work, book, magazine, [or] newspaper.”[55] In this and other ways, Congress cabined the reputational and integrity rights within limits so strong as to make the rights almost useless.[56]
This narrowness should not surprise us. Fair use and the First Amendment are about searching for truth, not securing reputations against ridicule. In fact, in a major copyright opinion, the Supreme Court held that the very human desire to avoid being ridiculed is a reason that favors giving parodists a very free rein to quote and distort the copyrighted oeuvres they ridicule.[57]
VARA distinguishes ownership of integrity and moral rights from ownership of copyright.[58] Moreover, as Judge Frank Easterbrook has argued, the narrowness of VARA suggests Congress does not want broad moral rights inserted into copyright law generally.[59] Explicit limits should not be casually contravened by inserting a reputational right into the general infringement action “through the back door.”[60]
Goold needs to give greater attention to the practical consequences of embracing this sub-tort. The Supreme Court has refused to embrace attribution issues in federal trademark law[61] largely because of “serious practical problems.”[62] The Court rejected attribution inquiries despite plausible support for attribution questions in the statutory language of the trademark statute and despite the fact that giving consumers accurate information about source is a crucial concern of trademark law. One of the Court’s primary concerns was that a legal requirement of accurate attribution might require litigants and judges to engage in a fruitless “search for the source of the Nile and all its tributaries.”[63]
However suggestive it might be, the Supreme Court’s trademark opinion does not compel the exclusion of attribution issues from copyright law. In fact, the trademark dispute had focused on the labeling of a work in the public domain, and the Court opined that public-domain works were likely to present more difficult authorship issues than works whose copyrights are still valid.[64]
Nevertheless, given the accretive nature of culture, identifying the authors of works still in valid copyright can also be immensely difficult. Consider a child’s copyrighted stuffed toy, based on a copyrighted animated movie, which was based on a copyrighted story in English, which in turn was based on a copyrighted Russian-language version of a Ukrainian folk tale.
Conceivably there are ways to make explicit copyright recognition of the reputation/attribution issue somewhat feasible for literary works and other works not eligible for VARA (though VARA’s own limits raise doubts about doing so). For example, detailed rules might ameliorate some line-drawing problems, such as how to distinguish between authorial and non-authorial contributions to a work.[65] And Goold is correct that there have been hints in some copyright cases that courts will be particularly unsympathetic to defendants who not only make unauthorized changes to a work but also name the horrified copyright owner as its author. All told, of the three new sub-torts Goold advances (privacy, control, and reputation), this reputational sub-tort is the most normatively promising.
V. The Right of Creative Control
This last sub-tort on Goold’s list, involving creative control, is particularly controversial when it is asserted in a context where the alleged lack of control interferes with no identifiable commercial plan or activity of the copyright owner.[66] Yet the notion may have staying power. Some kinds of control—such as the right to be the sole maker of a movie or other derivative work, or the right of first publication—do have the commercial resonance that can matter to incentives.
Still, a ‘right of control’ per se would be intolerably expansive. To proffer it requires detailed specification of limits. For example, on a normative level, one can doubt whether copyright should curtail harmless and speech-related behavior simply because the sequence of words, sounds, images or symbols are being deployed in a way their author doesn’t like. On a descriptive level, I do not see strong support for a right to control harmless behavior in the cases or statute. Much of the evidence is to the contrary. For example, Goold draws support from an early piece by Register David Ladd,[67] but Ladd’s views were largely opposite to those the contemporaneous Supreme Court adopted.[68]
The Constitutional purpose of copyright is to provide economic incentives.[69] A ‘right of control’ sub-tort severed from economic harm has a weak connection to incentives.
Conclusion
Goold’s article clears some necessary territory. However, too much underbrush remains to justify the article making recommendations as to how courts should actually proceed.
One can understand scholars feeling pressure to find and announce solutions, particularly while “improper appropriation” in copyright law remains highly problematic. Over seventy years have passed since negligence law received from Judge Learned Hand a clear structure for its somewhat parallel category, “unreasonable behavior.”[70] (As if to rub it in, Judge Hand opined on a crucial distinction in copyright that “[n]obody has ever been able to fix that boundary, and nobody ever can.”[71]) And lack of conceptual clarity leads to some jarring results.[72] Yet most infringement litigation proceeds along unsurprising paths.
As for “fair use,” no emergency threatens there either. Scholarship has refined the doctrine’s contours, and recent caselaw has made fair use liberties available to a wide range of publicly valuable but unauthorized applications of copyrighted work. Thus, for example, Goold quotes Larry Lessig’s famous quip from 2004 that fair use amounts to a mere “right to hire a lawyer,”[73] though Goold concedes in the footnotes that not all share this view. Fair use has changed greatly since—so greatly that some commentators argue that the doctrine transitioned “from weak reed to powerful shield in a decade’s time[.]”[74]
Unbundling is a notable venture for a new scholar. It shows a range of knowledge and an independent cast of mind. Admittedly, I remain unpersuaded by its particular effort to make visible the hidden causes of action that lie within federal copyright, particularly when Goold argues for protecting reputation, privacy, and a broad right to control—three issues not directly concerned with commercial incentives. These three interests have some basis in the statute and caselaw, but their substantive support is sufficiently weak that I am puzzled by Goold’s choice to focus on them. Nevertheless, his basic method—investigating whether a variety of distinct interests are being protected by copyright law, and assessing whether their several natures might mandate specialized tests for infringement or assessing fair use—is one that many in the field could profitably emulate.
By disentangling various distinct interests that are often bundled together in a blurry and confusing way in copyright cases, Goold places an important set of questions on the agenda of copyright judges, legislators, and commentators: Are these interests that copyright law does or should serve? If not, how would unambiguously excluding them change current copyright practice? If, by contrast, copyright does or should explicitly embrace them, what changes could be triggered by unambiguously recognizing these interests, particularly regarding the infringement inquiry and weighing the fair use factors? Goold’s article gives urgency and clarity to this important set of inquiries.
*Copyright © by Wendy J. Gordon 2017. Wendy J. Gordon is William Fairfield Warren Distinguished Professor at Boston University and Professor of Law at the BU School of Law.
I thank my husband, Michael Zimmer, for his superb editorial eye; Oren Bracha for his perceptive comments; the University of Texas School of Law at Austin for its generous hospitality while I worked on this piece; and Alexander Piala (UT class of 2018) for his excellent research assistance. In addition, I am grateful to the wondrous resources of BU’s Pappas Law Library, particularly the good offices of Stefanie Weigmann and Shira Megerman.
[1]Patrick R. Goold, Unbundling the “Tort” of Copyright Infringement, 102 Va. L. Rev. 1833, 1833–34 (2016) [hereinafter Goold, Unbundling].
[2]Making use of the copyrighted work is usually known as “actual copying” or “copying in fact.” “Copying in fact” or “actual copying” are terms of art. They address whether someone has factually borrowed or used the copyrighted work in question. In this context, the opposite of “copying” is relying solely on other sources or on independent creation. (In this broad sense of “copying,” to read a copyrighted book chapter over the radio is to “copy” it. In some other contexts, however, to “copy” may literally mean “to reproduce.”) Although to “copy” (in the context of the plaintiff’s case in chief) means essentially “to use,” not all unauthorized uses of copyrighted work are unlawful. For example, the Copyright Act gives everyone the liberty to copy ideas. 17 U.S.C. § 102(b) (2012). For another example, while the Act gives copyright owners some rights over public performance, the Act does not reach private performance. Id. § 106(4). Thus reading a copyrighted book aloud in one’s living room cannot violate the performance right. Id. See also id. § 101 (definition of “publicly”).
[3]VMG Salsoul, LLC v. Ciccone, 824 F.3d 871, 877 (9th Cir. 2016) (noting that “proof of actual copying is insufficient to establish copyright infringement . . .”).
[4]For courts using this terminology, see, for example, Muller v. Anderson, 501 F. App’x 81, 83 (2d Cir. 2012) (“[T]he plaintiff must prove . . . improper appropriation.”); Walker v. Time Life Films, 784 F.2d 44, 48 (2d Cir. 1986) (“Walker must show . . . that his expression was ‘improperly appropriated’ . . . .” (quoting Hoehling v. Universal City Studios, 618 F.2d 972, 977 (2d. Cir. 1980)); Arnstein v. Porter, 154 F.2d 464, 468 (2d Cir. 1946) (noting the plaintiff must show that “the copying . . . went so far as to constitute improper appropriation”).
[5]For courts employing this terminology, see, for example, Almeda Mall, L.P. v. Shoe Show, 649 F.3d 389, 391 (5th Cir. 2011) (concluding that “the trade name SHOE SHOW is not substantially similar to THE SHOE DEPT. . . .”); Cavalier v. Random House, 297 F.3d 815, 822 (9th Cir. 2002) (“Copying may be established by showing that the infringer had access to plaintiff’s copyrighted work and that the works at issue are substantially similar in their protected elements.”); Warner Bros. v. Am. Broad. Cos., 720 F.2d 231, 239 (2d Cir. 1983) (“The basic issues concerning the copyright infringement claim are whether the Hero and Superman works are substantially similar so as to support an inference of copying . . . .”).
[6]Infringement can occur through producing an unauthorized physical reproduction, an unauthorized derivative work, or an unauthorized public performance. 17 U.S.C. § 106(1), (2), (4), (6). In addition, a defendant can infringe by distributing or publicly displaying an unlawfully made copy or even, under some circumstances, by distributing or publicly displaying a lawfully made copy the defendant does not own. See id. § 106(3) (distribution right) as modified by id. § 109(a) (first sale doctrine). See also id. § 106(5) (public display right) as modified by id. § 109(c). Additional causes of action exist under the statute, but they are best seen as paracopyright and distinct from copyright infringement per se. See, e.g., id. §§ 1201–04 (civil and criminal penalties applicable to, inter alia, unauthorized circumvention of physical copy-restraints such as encryption).
[7]Id. § 107 (“[T]he fair use of a copyrighted work . . . is not an infringement of copyright.”). See generally Wendy J. Gordon, Excuse and Justification in the Law of Fair Use: Transaction Costs Have Always Been Only Part of the Story, 50 J. Copyright Soc’y U.S.A. 149 (2003) (common law flavored introduction to the complex “fair use” doctrine).
[8]Goold, Unbundling, supra note 1, at 1836–37.
[9]Id. Fair use doctrine is a vehicle for evaluating whether a substantial borrowing might, in the context of a particular fact pattern, nevertheless be normatively entitled to go forward without permission and payment. The fair use doctrine can render even exact copies noninfringing. 17 U.S.C. § 107 (listing “multiple copies for classroom use” as a possible focus for fair use); Sony Corp. of Am. v. Universal City Studios, 464 U.S. 417, 449–51 (1984) (home videotaping of entire TV programs can be fair use when done for purposes of time-shifting).
Goold is quite right that ordinarily “fair use” arises in a defendant’s case, and that the defendant will usually bear some or all of the burden of persuasion. Goold, Unbundling, supra note 1, at 1836. However, the statute itself does not identify “fair use” as an affirmative defense. Rather, the Copyright Act’s key provision on fair use, 17 U.S.C. § 107 (2012), says only that fair uses are “not an infringement of copyright,” a phrase that could be legitimately interpreted to place a burden of proving unfair use on plaintiffs as part of their case in chief. In some cases, some portion of a burden to prove unfair use has indeed been placed on plaintiffs, whether implicitly or explicitly.
The right to make fair uses of others’ work is an important part of the public’s liberties to use others’ copyrighted expression. This set of liberties is sometimes identified with “users’ rights.” See CCH Canadian Ltd. v. Law Soc’y of Upper Can., [2004] 1 S.C.R. 339, 364 (Can.) (“User rights are not just loopholes.” (quoting David Vaver, Copyright Law 171 (2000)).
[10]Goold, Unbundling, supra note 1, at 1838, 1898.
[11]See Oren Bracha & Patrick R. Goold, Copyright Accidents, 96 B.U. L. Rev. 1025, 1027–1029 (2016). This article explores how copyright law should treat defendants who, after a good faith but fruitless attempt to locate and pay any copyright holders, take the risk of publishing, and then learn they have copied a substantial amount of copyrighted material. One might say such defendants have “accidentally” copied a copyrighted work, a metaphor that Bracha and Goold take seriously and deploy to good effect.
[12]Pamela Samuelson, Reconceptualizing Copyright’s Merger Doctrine, 63 J. Copyright Soc’y U.S.A. 417, 417–19 (2016).
[13]See, e.g., Pamela Samuelson, Unbundling Fair Uses, 77 Fordham L. Rev. 2537, 2539–41 (2009) (arguing that the success of fair use defense falls into common patterns); Lydia Pallas Loren, Redefining the Market Failure Approach to Fair Use in an Era of Copyright Permission Systems, 5 J. Intell. Prop. L. 1, 7–8, 16–18 (1997) (recognizing fair use rights as expansive and exploring how fair use doctrine impacted rights given to the copyright owner and user over time); 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, 1627–35 (1982) (exploring fair use as a response to, inter alia, high transaction costs, external benefits, noncommercial uses, nonmonetizable values, and antidissemination motives).
[14]Goold argues that copyright infringement is oddly “singular,” as opposed to the multiple causes of action that characterize other areas of tort. Goold, Unbundling, supra note 1, at 1838–39. He also contends that copyright lacks the explicit “gallery of wrongs” structure found in many other areas of tort. Id. at 1855–56. These claims are puzzling.
First, much of the activity denominated “copyright infringement” is socially useful and not morally wrongful in itself; the “wrong” lies in disobeying the law which requires the purchase of a license. To label every breach of law a “wrong” is technically a correct usage, but the usage nevertheless dilutes the force of the word.
Second, copyright has a full (one might say over-full) gallery of distinct breaches, full of detailed differences. The Copyright Act sets out many subcategories of copyright infringement that it explicitly distinguishes, both in terms of types of works protected, 17 U.S.C. § 102(a), and in terms of the typology of rights that attach to owning a particular kind of work, id. §§ 106, 106A. The Act then makes each type of work and type of right subject to particularized exemptions and defenses. Id. §§ 108–122. A good illustration of the lines the statute draws can be seen by comparing the acts that can infringe copyright in a “sound recording” with the acts that can infringe copyright in a “musical work.” Id. §§ 102(a), 106, 114. The copyright statute thus certainly seems to display a “gallery” of causes of action.
Nevertheless, the importance of Goold’s point depends neither on whether copyright infringement is a ‘wrong’ in a meaningful sense, nor on the question of whether copyright infringement is unitary or internally diverse. All Goold needs to show is that making some additional distinctions among types of breach would be useful. That question his article skillfully raises.
[15]Goold, Unbundling, supra note 1, at 1838–39.
[16]Id. at 1870–71. Intangible patterns like works of authorship are usually considered nonrivalrous because, as Goold explains, “one person’s use does not affect the use of another.” Id. at 1870. Despite the possibility of infinitely replicating a book or song, physical inexhaustibility does not guarantee that one person’s use will not affect another person’s profit. Goold essentially tries to identify occasions on which the rivalrous aspect predominates. For further discussion of what might constitute a “rivalrous use,” and of the category’s ambiguities, see Part IV.
[17]I am glad Goold did not propose a subtort to vindicate publisher, as distinct from authorship, interests. In my view, publishers’ claims under copyright should be related to the publishers’ role in incentivizing creative expression, and no deference should be paid in copyright cases to supporting publishers’ noncreative activities. See Wendy J. Gordon, Authors, Publishers and Public Goods: Trading Gold for Dross, 36 Loy. L.A. L. Rev. 159, 197–198 (2002).
[18]Goold, Unbundling, supra note 1, at 1857.
[19]See id. at 1860.
[20]See id. at 1865.
[21]See id. at 1867.
[22]See id. at 1869.
[23]Copyright reflects the influence of a number of common law doctrines to which Goold gives little serious attention—for example, unjust enrichment law (also known as restitution) provides for recovery to volunteers who confer benefits without contract. Such cases provide significant insight into when and why the common law might be unwilling to require beneficiaries to pay for benefits others create. Yet Goold dismisses the relevance of restitution law, largely on the ground that giving it analogic significance “would potentially justify [copyright] owners claiming reward every time the work is enjoyed.” Id. at 1854. Goold’s position is deeply puzzling, given that unjust enrichment law gives recovery far more sparingly than does tort law. See, e.g., Wendy J. Gordon, Of Harms and Benefits: Torts, Restitution, and Intellectual Property, 21 J. Legal Stud. 449, 450 (1992) (discussing how duties “to guard against harm are far more common than duties to provide or pay for benefits”). Goold cites some of my work on this topic, but fails to explain adequately why he disagrees about the influence restitution law would have. Goold, Unbundling, supra note 1, at 1854 n.150.
[24]See Oren Bracha, Owning Ideas: The Intellectual Origins of American Intellectual Property, 1790–1909, 36–42, 58 (2016).
[25]Copyright Act of 1790, 1 Stat. 124 (1790) (repealed 1802).
[26]117 Cong. Rec. 34,748–49 (1971) (“Mr. KAZEN: Am I correct in assuming that the bill protects copyrighted material that is duplicated for commercial purposes only? Mr. KASTENMEIER: Yes.”).
[27]Jessica Litman, Campbell at 21/Sony at 31, 90 Wash. L. Rev. 651, 662 (2015).
[28]As initially enacted in 1976, Section 506 provided, in pertinent part, that criminal penalties applied to someone who “infringes a copyright willfully and for purposes of commercial advantage or private financial gain.” An Act for the General Revision of the Copyright Law, Pub. L. No. 94-553, § 506, 90 Stat. 2586 (1976).
[29]First, the original term, “private financial gain,” has recently been redefined to expand the reach of the criminal provisions. Today, 17 U.S.C. § 101 (2012) states, “The term ‘financial gain’ includes receipt, or expectation of receipt, of anything of value, including the receipt of other copyrighted works.”
Second, criminal liability under the copyright statute has been expanded to include, inter alia, behavior such as bypassing encryption. See id. §§ 1201, 1204. Third, Section 506 itself covers additional activity that consumers might well engage in, including “the reproduction or distribution, including by electronic means, during any 180–day period, of 1 or more copies or phonorecords of 1 or more copyrighted works, which have a total retail value of more than $1,000.” Id. § 506(a)(1)(B).
[30]See, e.g., id. § 1204 (creating criminal penalties for violation).
[31]Matthew Sag, IP Litigation in U.S. District Courts: 1994–2014, 101 Iowa L. Rev. 1065, 1077 (2016).
[32]Id. at 1075–80. The primary goal of these suits is “creating an independent litigation revenue stream that is unrelated to compensation for the harms of infringement and that is unconcerned with deterrence.” Id. at 1076. There also had been a wave of suits against end-users “to ‘educate’ the public about filesharing and to reinforce that education with deterrence.” Id. Those suits essentially ended in 2008. Id. at 1075[MD to edit this part]ucat”nce”,5–76olation)idn’is view.case)hat not available in the westlaw/lexis version].o analyzing the cla.
[33]Id. at 1077 (“[P]olicymakers should be cautious about extrapolating from current trends in this context. . . .”). In my view, Congress and the courts are more likely to pull back these suits, which essentially abuse the system, than to treat them as a model.
[34]Goold, supra note 1, at 1857–59.
[35]See, e.g., Wendy J. Gordon, The Concept of “Harm” in Copyright, in Intellectual Property and the Common Law 452, 477–80 (Shyamkrishna Balganesh ed., 2013) (discussing whether failure to pay license fees constitutes a harm under various theories); Loren, supra note 13, at 6 (discussing rejection of fair use claims where the copyright owners established licensing systems).
[36]Goold, supra note 1, at 1858. See also id. passim.
[37]Goold, supra note 1, at 1860. Aside from the problem of identifying “customers” who would pay for a non-excludable good such as a shining light, the hypothetical does not involve any copying, borrowing, free-riding, or other use of the first lighthouse’s resources by the newcomer lighthouse. This cannot be a copyright example, then, for copyright suits cannot proceed without proof that the defendant has gained something from the plaintiff’s work. See, e.g., Wendy J. Gordon, Copyright and Tort as Mirror Models: On Not Mistaking for the Right Hand What the Left Hand Is Doing, in Comparative Law and Economics 311, 323–25 (Theodore Eisenberg & Giovanni B. Ramello eds., 2016) (discussing “copying” element in cause of action as requiring evidence that defendant gained output, defined broadly as appearance, quantity and cost of production, after coming into contact with plaintiffs work). Goold places little emphasis on the example; he is simply doing a rhetorical turn on a classic public-goods illustration. But just a sentence or two before the example, he commits real error, in stating that the reason why copyright law allows suit against publishers has nothing to do with a desire to allow authors to “internalize” some of the benefits their efforts give others. Goold reserves that concern with internalizing only for consumer copying. Goold, supra note 1, at 1859–60.
[38]See Sheldon v. Metro-Goldwyn Pictures Corp., 81 F.2d 49, 53 (2d Cir. 1936) (“[I]t makes no difference how far the play was anticipated by works in the public demesne [public domain] which the plaintiffs did not use.”). In a now-classic formulation, the Second Circuit emphasized that the key factual question in copyright is “whether the defendants actually used” the copyrighted work. Id. at 53. The court notes that:
. . . [I]f by some magic a man who had never known it were to compose anew Keats’s Ode on a Grecian Urn, he would be an “author,” and, if he copyrighted it, others might not copy that poem, though they might of course copy Keats’s. . . . [J]ust as he is no less an ‘author’ because others have preceded him, so another who follows him, is not a tort-feasor unless he pirates his work . . . .
Id. at 54 (emphasis added) (citations omitted). A plaintiff cannot complain of copying by the defendant, so long as the copying is not from the plaintiff’s work. Id. at 54.
[39]Goold, Unbundling, supra note 1 at 1865.
[40]Id. at 1898–99.
[41]Garcia v. Google, Inc., 786 F.3d 733, 745 (9th Cir. 2015) (en banc) (quoting Bond v. Blum, 317 F.3d 385, 395 (4th Cir. 2003)). This statement about privacy is arguably dicta, but was nevertheless a matter to which the opinion gave serious consideration. The en banc court continues:
“To the contrary, the copyright law offers a limited monopoly to encourage ultimate public access to the creative work of the author.” Bond v. Blum, 317 F.3d 385, 395 (4th Cir. 2003); see also Monge v. Maya Magazines, Inc., 688 F.3d 1164, 1177 (9th Cir. 2012) (quoting Bond and “pointedly” noting copyright cases are analyzed “only under copyright principles, not privacy law”)
Likewise, authors cannot seek emotional distress damages under the Copyright Act, because such damages are unrelated to the value and marketability of their works.
Id. at 745.
[42]See 17 U.S.C § 301 (2012) (pre-empting state copyright law); id. § 102.
[43]See H.R. Rep. No. 94-1476, at 129–31 (1976).
[44]Id.
[45]In 1992, the fair use provision was amended to decrease the negative impact of a work’s being unpublished; the statute now states that “[t]he fact that a work is unpublished shall not itself bar a finding of fair use if such finding is made upon consideration of all the above factors.” 17 U.S.C § 107.
[46]See, e.g., Visual Artists Rights Act of 1990 (“VARA”), title VI of the Judicial Improvements Act of 1990, Pub. L. No. 101–650, § 603 (a), Dec. 1, 1990, 104 Stat. 5128 (1990) (primarily codified at 17 U.S.C. §§ 101 & 106A). These sections comprise the key portions of VARA, which was part of the US effort to implement its earlier accession to the Berne Convention.
[47]But see 17 U.S.C § 104(c) (discussing effect of accession to Berne Convention and noting that rights “shall not be expanded or reduced by virtue of, or in reliance upon, the provisions of the Berne Convention, or the adherence of the United States thereto.”).
[48]See id. § 102(b); Feist Publications.v. Rural Tel. Serv. Co., 499 U.S. 340, 356 (1991) (“Section 102(b) is universally understood to prohibit any copyright in facts.”).
[49]Eldred v. Ashcroft, 537 U.S. 186, 219–21 (2003).
[50]See, e.g., Hendrickson v. Cal. Newspapers, 48 Cal.App.3d 59, 62 (Cal. Ct. App. 1975) (“It is well settled that the right of privacy is purely a personal one . . . . the right does not survive but dies with the person”).
[51]The affected rights appear in: 17 U.S.C. § 106(1) (reproduction), § 106(2) (preparing derivative works), and § 106(3) (distribution to the public). See 17 U.S.C. § 106.
[52]The canonical caution to judges on this topic belongs to Justice Holmes: “It would be a dangerous undertaking for persons trained only to the law to constitute themselves final judges of the worth of pictorial illustrations, outside of the narrowest and most obvious limits.” Bleistein v. Donaldson Lithographing Co., 188 U.S. 239, 251 (1903).
[53]Berne Convention for the Protection of Literary and Artistic Works, art. 6bis, ¶(1), Sept. 9, 1886.
[54]The Visual Artists Rights Act of 1990, title VI of the Judicial Improvements Act of 1990, Pub. L. No. 101–650, 104 Stat. 5089, 5123–33 (1990) (primarily codified at 17 U.S.C. §§101 & 106A).
[55]17 U.S.C. § 101 (definition of “work of visual art,” with an additional provision defining what “[a] work of visual art does not include”).
[56]See id. § 106A. See also the remarkably narrow definition of the type of work covered by the attribution and integrity right. Id. at § 101 (definition of “work of visual art”).
[57]Campbell v. Acuff-Rose Music, 510 U.S. 569, 592 (1994).
[58]17 U.S.C. §106A(b) & (e)(2).
[59]Lee v. A.R.T. Co., 125 F.3d 580, 582–83 (7th Cir. 1997) (“It would not be sound to use § 106(2) to provide artists with exclusive rights deliberately omitted from the Visual Artists Rights Act.”).
[60]Id. at 582 (criticizing a theory “about what counts as a derivative work,” because adopting the theory would imply that “the United States has established through the back door an extraordinarily broad version of authors’ moral rights, under which artists may block any modification of their works of which they disapprove”).
[61]Dastar Corp. v. Twentieth Century Fox Film Corp, 539 U.S. 23, 36–37 (2003) (refusing to allow plaintiffs to use the Lanham Act—the federal trademark statute—to try questions regarding proper attribution of authorship).
[62]Id at 35.
[63]Id. at 36.
[64]Id. at 35. (“Without a copyrighted work as the basepoint, the word ‘origin’ has no discernable limits.”) Yet difficulties in determining authorial origin can arise regardless of a work’s legal status; even copyrighted works often contain multiple sources. Also, Dastar has been generally applied to bar trademark claims of false authorial attribution in cases involving copyrighted works as well as those involving public-domain works. See, e.g., Antidote Int’l Films v. Bloomsbury Publ’g, PLC, 467 F. Supp. 2d 394, 398 (S.D.N.Y. 2006) (holding that Dastar bars false advertising claims concerning authorship).
[65]For example, Congress or the courts might spell out how much ‘new matter’ must be added before a claim of attribution is warranted, for example, or how to treat a copyright owner’s claim that involves misattribution when the actual creator is not the person who owns the copyright. However, Goold’s advice against formal recognition of the sub-torts could make it difficult to specify such rules. Goold, Unbundling, supra note 1, at 1895–98 (recommending against statutory implementation).
[66]Sony Corp. of Am. v. Universal City Studios, 464 U.S. 417, 451 (1984) (discussing how lack of commercial harm weighted heavily in giving fair use treatment to home copyists of television programs).
[67]David Ladd, The Harm of the Concept of Harm in Copyright: The Thirteenth Donald C. Brace Memorial Lecture (Apr. 13, 1983), in 30 J. Copyright Soc’y U.S.A. 421, 422 (1983).
[68]Sony, 464 U.S. at 451.
[69]The Constitution gives Congress power to grant to authors and inventors exclusive rights over their works in order to “promote the Progress of Science and useful Arts.” U.S. Const. art. I, § 8, cl. 8.
[70]United States v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947) (“[T]he owner’s duty, as in other similar situations, to provide against resulting injuries is a function of three variables: (1) The probability that [a boat tethered at a mooring] will break away; (2) the gravity of the resulting injury, if she does; (3) the burden of adequate precautions.”). Judge Hand’s BPL test (so named for his variables: burden, probability, and loss) may not answer all questions, but it has stayed at the center of negligence discussions for over a half-century.
[71]Nichols v. Universal Pictures Corp., 45 F.2d 119, 121 (2d Cir. 1930). The boundary in question divides ideas (which the law makes freely copyable) from expression (over which copyright owners have some legal rights of control). See 17 U.S.C. § 102(b) (2012). Part of the purpose of determining “substantial similarity” or “improper appropriation” is to decide if the defendant has copied only ideas (which is permitted and indeed encouraged) or if the defendant has also copied expression.
[72]See, e.g., Wendy J. Gordon, How Jury in the ‘Blurred Lines’ Case was Misled (March 17, 2015, 2:30 PM), http://www.newsweek.com/jury-blurred-lines-case-was-misled-314856 (exploring jury finding of infringement in the Blurred Lines case). Goold cites to rulings in the Blurred Lines case. Goold, Unbundling, supra note 1, at 1835 n.4 (citing Williams v. Bridgeport Music, No. LA CV13-06004 JAK, 2015 WL 4479500, at *1–2 (CD. Cal. July 14, 2015)) (determining, inter alia, that the defendants’ motion for a new trial should be denied)).
[73]Goold, Unbundling, supra note 1, at 1837 (quoting Lawrence Lessig, Free Culture: How Big Media Uses Technology and the Law to Lock Down Culture and Control Creativity 187 (2004)).
[74]Rebecca Tushnet, Content, Purpose, or Both?, 90 Wash. L. Rev. 869, 872–73 (2015) (discussing doctrinal improvements in the strength of fair use since Lessig wrote).