Ben Tauber *
Download a PDF version of this article here.
.Introduction
U.S. copyright law derives its authority from Article I, Section 8 of the Constitution (the “IP Clause”), which grants to Congress the power to assign copyrights in order to “promote the Progress of Science and useful Arts.”1 The primary goal of the Clause, and therefore of copyright law, is to safeguard the public benefit that results from the production of creative works.2 In doing so, the clause recognizes that artists must be incentivized to create through the ability to protect their works.3
Central to these dual principles is a market failure. Expressive works are costly to produce, involving significant expenditures of time, intellectual and creative effort, and, yes, money.4 However, they are incredibly cheap to reproduce. The digital age has made copying especially trivial; entire works can be copied in a couple of mouse clicks with little-to-no loss in quality.5 Therefore, without protections for expressive works, rampant reproduction would drive the costs of these works down to the marginal cost of copying (essentially zero).6 Without being able to recoup the costs of creating a work, most creatives would likely not bother to create in the first place.7 By providing artists, authors, and other creatives exclusive rights in their works, copyright law creates a market for the works that would otherwise not exist, thereby incentivizing creation.
However, copyright law requires a balance to be struck between too little and too much protection. Too little protection, and works would not get created because the prices of works would not justify the costs of creation. Too much protection, and works would not get created because the fear of getting sued over the smallest instances of accidental appropriation would make the costs of creation too high.8 For this reason, copyright law does not prevent every instance of copying. For example, the doctrine of scènes à faire excludes “incidents, characters or settings which are as a practical matter indispensable, or at least standard, in the treatment of a given topic” from copyright protection,9 and the doctrine of fair use allows authors to engage in limited copying of protected works for “criticism, comment, news reporting, teaching … scholarship, or research.”10 In this vein, courts have long recognized a doctrine known as de minimis—short for de minimis non curat lex, or “the law does not concern itself with trifles.”11 In the copyright context, the de minimis doctrine stands for the principle that there are instances of copying that are so minimal or inconsequential that they are not legally actionable.12
The current state of the de minimis doctrine, however, is messy, to say the least. Federal copyright law contains no mention of the doctrine, and the Supreme Court has not even addressed its existence, let alone its metes and bounds. Accordingly, the de minimis doctrine in its current form is a patchwork of judicial interpretation, fracturing along circuit borders.
The current legal patchwork contributes to widespread uncertainty regarding the doctrine, raising questions of exactly what it is and what it covers. While circuit splits necessarily lead to uncertainty, circuit splits in copyright law are especially concerning, as “creating inconsistent rules among the circuits would lead to different levels of protection in different areas of the country, even if the same alleged infringement is occurring nationwide.”13
The goal of this note is to provide a general survey of the current state of the doctrine while addressing what a unified de minimis framework could look like. First, this note outlines two conceptual grounds on which courts base the de minimis doctrine: the concept of “legal copying”—which implicates the doctrine of substantial similarity—as well as the concept of “trivial uses” of copyrighted works. This note argues that these concepts share a unifying “improper purpose” rationale: that a copier is liable only if the appropriation implicates the copyright holder’s legally protected copyright interests.
I then apply these concepts to a proposed two-part framework, utilizing it to address two current points of contention in copyright law. The first concerns the role of observability in substantial similarity analysis. Contrary to copyright law orthodoxy, the Second Circuit holds that copying a visual work in its entirety, but sufficiently obscuring it within a new work, constitutes non-actionable de minimis copying. As a component of substantial similarity, the observability doctrine rests on legally shaky grounds and may make it more difficult for plaintiffs to establish actionable infringement. The concept of “trivial use” may therefore be a more appropriate legal ground on which to base the observability doctrine. The second split concerns whether the de minimis doctrine may be applied to instances of sound recording infringement, often seen in cases of music sampling. The Sixth Circuit, well in the minority, holds that samples are per se infringing. This note argues, in accordance with the Ninth Circuit, that the Sixth Circuit approach is based on a poor understanding of the de minimis doctrine. Finally, this note concludes with a brief review of the general state of the de minimis doctrine among the several circuits as well as thoughts on what the future of the doctrine could look like.
I. Relation of the De Minimis Doctrine to Legal Copying
A plaintiff seeking to prove copyright infringement in the United States usually must establish two components. First, he must establish that copying actually occurred. In most circuits, however, this is not enough. The plaintiff must also establish that what was taken was substantial enough to be legally actionable.14 In other words, there must be “substantial similarity” between the two works.
As this note will discuss, the de minimis doctrine in copyright law has been poorly defined and inconsistently applied in the several circuits. One thing on which most courts agree, however, is that the doctrine is fundamentally linked to the concept of “legal” or “actionable” copying, and thereby to the concept of substantial similarity. At bottom, the term “de minimis” has been understood to mean copying “in fact” that nevertheless does not rise to the level of actionable—or “legal”—copying.
The Second Circuit, for instance, utilizes a two-part test for proving copyright infringement.15 The first part requires establishing that the work was actually copied, either by direct evidence of copying or indirectly through showing that the defendant had access to the plaintiff’s work as well as similarities between the two works that are probative of copying.16 Upon establishing actual copying, the plaintiff must still establish that the copying at issue arises to an actionable level by showing that the defendant’s work is substantially similar to the plaintiff’s work. Substantial similarity, in turn, “requires that the copying is quantitatively and qualitatively sufficient” to support a conclusion of actionable copying.17 Qualitative sufficiency turns on whether the defendant’s work copies the protected expression of the plaintiff’s work rather than unprotectable ideas. This can be done either by determining whether an ordinary observer would regard the works as essentially the same, or by more discerningly conceptualizing the works at different levels of abstraction and filtering out unprotectable components before comparing what is left.18 Quantitative sufficiency turns on how much of the plaintiff’s work was copied. The Second Circuit defines de minimis as copying that “has occurred to such a trivial extent as to fall below the quantitative threshold of substantial similarity.”19 An artist engaging in de minimis (i.e. non-actionable) copying therefore takes so little of a copyrighted work that the artist’s work is not substantially similar to the work from which he copied, making the work non-infringing.20
The Ninth Circuit’s copyright infringement framework differs slightly from the Second Circuit’s. Nevertheless, it also understands de minimis as copying that does not rise to the level of substantial similarity. Similar to the Second Circuit, a plaintiff proving copyright infringement in the Ninth Circuit must demonstrate actual copying, either through direct evidence or by showing access as well as “similarities probative of copying.”21 The plaintiff then must establish that the works are substantially similar, both “extrinsically” and “intrinsically.”22 The “extrinsic” test mirrors the Second Circuit’s “abstraction-filtration” test for qualitative similarity: it dissects the two works to their “constituent elements” and compares whether these elements share objective similarities.23 The “intrinsic” test mirrors the Second Circuit’s “ordinary observer” test: it asks whether the ordinary person would find the “total concept and feel” of the two works to be substantially similar.24 Although the Ninth Circuit test does not have an explicit quantitative component, the amount of copying still plays a role in determining substantial similarity through the “intrinsic” component. Accordingly, the Ninth Circuit defines de minimis copying as copying that is “so meager and fragmentary that the average audience member would not recognize the appropriation,” resulting in a lack of substantial similarity between the works at issue.25
Not all circuits recognize a separation between actual and legal copying, however. In the Eighth Circuit, for instance, substantial similarity need only be shown to establish actual copying in the absence of direct evidence. In other words, if a plaintiff can show (1) ownership of a valid copyright and (2) direct evidence of copying, then the defendant is liable for infringement, regardless of the degree of similarity between the works.26 The Seventh Circuit utilizes this framework as well.27 It is no surprise, therefore, that these circuits have not formally recognized a de minimis doctrine in copyright law.28
Alexander v. Take-Two Interactive Software, Inc. shows how these circuits refuse to grant judgments for defendants on grounds of de minimis copying. In this case, the U.S. District Court for the Southern District of Illinois addressed, inter alia, the question of whether a video game company’s copying of a tattoo artist’s copyrighted design on a digital rendering of wrestler Randy Orton constituted de minimis use.29 After acknowledging the absence of Seventh Circuit precedent on the existence of a de minimis defense, the court refused to recognize the defense, doubting its general viability “[g]iven the overlap between the defense and actionable copying, which Alexander is not required to prove to sustain her case in this circuit ….”30
There is an argument that the Seventh and Eighth Circuits get it right—that there is no separate “legal copying” requirement to establish infringement, and therefore no de minimis defense. The Copyright Act, after all, essentially provides that defendants are strictly liable for copyright infringement unless they can establish fair use.31 This argument is myopic, however. The rule that infringement occurs only upon a finding of legal copying is well established in copyright law dating to the mid-1800s.32 Because the “legally protected interest” of a copyright holder lies in his ability to profit from the original and creative elements of his copyrighted work, an appropriative work does not infringe if it does not impact the “lay public’s approbation” of the original work’s creative expression.33 Accordingly, a substantial similarity test is necessary to assess whether a new work is so different from an original work that the lay public cannot recognize the appropriation: “[i]f the public does not recognize the appropriation, then the copier has not benefitted from the original artist’s expressive content. Accordingly, there is no infringement.”34 Moreover, the presence of fair use in the Copyright Act evinces a Congressional intention to provide for liability only if a defendant’s copying crosses a threshold of impropriety. The concept of legal copying merely draws upon what fair use doctrine has established to be true: that copyright law is not immune from normative determinations of what makes copying legally actionable.35
Therefore, the distinction between factual and legal copying is grounded in a proper understanding of copyright law. And with recognition of legal copying necessarily comes a recognition of de minimis copying as the absence of such. Because courts have relied on the doctrine of substantial similarity to address the question of what and how much needs to be copied for an infringement to be actionable, substantial similarity is inextricably linked to the de minimis doctrine.36
II. Extending De Minimis to “Trivial Use”
A. Overview
Some circuits, including the Ninth, view the term “de minimis” as merely signifying a lack of substantial similarity.37 However, it is perhaps more accurate to say that substantial similarity is a manifestation of the general de minimis principle as applied to copyright law. Absent specific indication to the contrary, every law is undergirded by the principle that trifling violations are not actionable.38 The Copyright Act provides no such indication. Therefore, while substantial similarity analysis is one way of actualizing this principle, the de minimis maxim is not necessarily restricted to instances where the amount of copying was insubstantial. Indeed, the maxim may be applied as it is applied in other legal contexts: to dismiss cases involving injuries “not only small but also indefinite, so that substantial resources would have to be devoted to determining whether there was any loss at all.”39 While the main thrust of the de minimis doctrine concerns a lack of substantial similarity, the Court of Appeals for the Second Circuit has also argued that de minimis in copyright law “can mean what it means in most legal contexts: a technical violation of a right so trivial that the law will not impose legal consequences.”40 Whether de minimis applies in this context, however, is a subject of disagreement between the Second and Ninth Circuits.
Because the Ninth Circuit views de minimis as merely a lack of substantial similarity, it does not consider the de minimis doctrine to properly cover uses of copyrighted works that are technically infringing, but overall trivial or inconsequential in effect.41 For example, in Design Data Corp. v. Unigate Enterprise, Inc., the Court of Appeals for the Ninth Circuit overturned the district court’s finding that the defendant’s intentional download of the plaintiff’s computer program was de minimis because there was no evidence that the defendant installed or used the program.42 The circuit court stated: “In light of the overwhelming thrust of authority, which upholds liability even under circumstances in which the use of the copyrighted work is of minimal consequence, it was error to grant summary judgment on the basis that UE’s download of SDS/2 constituted a de minimis infringement.”43
The Court of Appeals for the Ninth Circuit derives this interpretation from its reading of federal copyright law: “[c]opyright is a creature of statute, and the only rights that exist under copyright law are those granted by statute …. [N]owhere in the [Copyright] Act [of 1976]’s numerous and detailed provisions is there any exception for the de minimis use of a concededly infringing work, i.e., for a ‘technical violation.’ The Act defines a copyright infringer as anyone who violates any of the exclusive rights of the copyright owner.”44
This understanding of copyright law is overly formalistic and fails to account for de minimis as a judicial filter for frivolous copyright claims. It is true that the 1976 Act largely wrote the common law of copyright out of existence.45 Nevertheless, it is not true that the Act contains the beginning, middle, and end of all American copyright law. For example, the Act does not dictate how to assess whether a use has infringed on an author’s copyright; consequently, several circuit courts have devised their own judicial tests for determining infringement.46 Courts may properly step in and supplement their own reasoning when a statute is ambiguous or fails to explain an aspect of the law. De minimis comfortably falls within this prerogative. Contrary to the Ninth Circuit’s assertion, a broader understanding of de minimis does not create a new right, nor does it supersede what the Act defines as a copyright infringer. De minimis simply stands for the principle that there are some instances of technical copying that do not rise to the level of infringement. While much of this principle is exercised through substantial similarity analysis, the principle naturally extends to trivial use as well. As Judge Pierre Leval of the Second Circuit elegantly puts it:
Trivial copying is a significant part of modern life. Most honest citizens in the modern world frequently engage, without hesitation, in trivial copying that, but for the de minimis doctrine, would technically constitute a violation of law. We do not hesitate to make a photocopy of a letter from a friend to show to another friend, or of a favorite cartoon to post on the refrigerator. Parents in Central Park photograph their children perched on José de Creeft’s Alice in Wonderland sculpture. We record television programs aired while we are out, so as to watch them at a more convenient hour. Waiters at a restaurant sing “Happy Birthday” at a patron’s table. When we do such things, it is not that we are breaking the law but unlikely to be sued given the high cost of litigation. Because of the de minimis doctrine, in trivial instances of copying, we are in fact not breaking the law. If a copyright owner were to sue the makers of trivial copies, judgment would be for the defendants. The case would be dismissed because trivial copying is not an infringement.47
The Ninth Circuit’s rigid approach makes infringers out of average people engaging in everyday activities and empowers copyright trolls. According to the Ninth Circuit, someone who prints out a copyrighted picture on the internet to put in a personal scrapbook may but for the grace of the copyright owner be sued for infringement. If that person’s fair use argument fails, she may be held liable and forced to pay statutory damages. Indeed, in Bell v. Wilmott Storage Services, LLC, something very similar happened. The Court of Appeals for the Ninth Circuit found that the hosting of a copyrighted image on a server violated the copyright holder’s public display right even though “the image was only accessible to those users who conducted a reverse image search48 … or those who knew the precise address of the image database archiving the photograph.”49 In doing so, it overturned a district court’s finding that the defendant’s actions constituted de minimis infringement on grounds of triviality. In other words, the defendant in this case was liable for infringement for “publicly displaying” an image inaccessible to almost every person on Earth. In fact, the defendant was unaware that the image was on its server at all.50
To explain the absurdity of this outcome by non-internet analogy, imagine a copyrighted painting hangs on a wall behind a locked door in a restaurant’s dining room.51 No one can see the painting, and the restaurant owner does not even know the painting is there. The painting can be viewed only if someone (a) somehow finds the key that unlocks the door, or (b) uses a device that tells the user where every copy of an image is in the world upon the user telling the device the exact image to look for, then finds the painting amid the multitude of copies. According to the Ninth Circuit’s reasoning, this should result in liability for the restaurant owner. This patently ridiculous outcome cannot be the regime that Congress intended to create in passing the Copyright Act: one in which every person must get a license for even the most trivial uses of copyrighted material or risk being sued. And a thorough application of the de minimis doctrine empowers courts to do more than shrug their shoulders and rigidly apply the text of the Act.
B. Justifications
In allowing for trivial uses of copyrighted material, the Second Circuit delivers much more natural results. For instance, in Knickerbocker Toy Co. v. Azrak–Hamway International, Inc., the Court of Appeals for the Second Circuit relied on the trivial use understanding of de minimis to reject a toy manufacturer’s infringement claim against a competitor who printed a copyrighted picture of the manufacturer’s product for use as an in-office sample.52 In American Geophysical Union v. Texaco, Inc., the Court of Appeals for the Second Circuit suggested that photocopying for individual use may not constitute infringement under the de minimis doctrine.53 Judge Leval has argued that a private person using a video tape recorder to time-shift a TV program “for a one-time noncommercial viewing” engages in mere de minimis copying.54 And in a one-off case out of the Fifth Circuit, the District Court for the Southern District of Texas ruled that the unauthorized display of a single advertisement that accidentally remained in one store (out of 23,000) for less than one month after the expiration of the parties’ licensing agreement was “so trivial as to fall below the threshold required for actionable copying.”55
Congress itself has recognized that some trivial forms of copying are not actionable. Stubbornly, however, it has tried to fit this conclusion within the confines of fair use. But doing so is like trying to fit a square peg in a round hole. In the Senate and House Reports accompanying the Copyright Act of 1976, for instance, Congress explicitly acknowledged “the making of a single copy [in braille] or phonorecord by an individual as a free service for a blind person” as an example of fair use.56 Why exactly this constitutes fair use is a mystery, however. For one, mere translation into braille, without more, is not transformative for purposes of fair use; “[s]ince the [Copyright] Act of 1870, it had been clear that translation was among the rights (now called ‘derivative’) that belong exclusively to the author.”57 As for the second fair use factor, the Congressional Reports do not make any distinction between braille translations of unpublished and published works, or of “creative” and “factual” works.58 The third fair use factor is of even less help, as the Congressional Reports contemplate the translation of an entire work.59 Presumably, Congress’ determination hinges on the fourth fair use factor, evidenced by its emphasis on the braille translation as a “free service.”60 But this does not paint a complete picture, as the Senate Report also states that the making of multiple free copies for the same purpose would not constitute fair use.61 While at least one later court has more generously applied the fair use factors to works made accessible to the print-disabled,62 the determining factor in the Congressional Reports seems to be the number of copies made. This is supported by the Senate’s additional determination that “a single copy reproduction of an excerpt from a copyrighted work by a calligrapher for a single client” would not constitute infringement.63 But the Copyright Act does not carve out an exception for single copies, in the fair use provision nor anywhere else.64 Therefore, if making a single braille translation for a blind person is not an infringement, the trivial use application of the de minimis doctrine much better explains why that is the case.65
Implicit in the trivial use application of the de minimis doctrine is the same understanding underlying the substantial similarity application: that “the manner of copying might impinge so little on the copyright owner’s legitimate interests” as to be nonactionable.66 It is crucial to remember that copyright law is designed to protect an author’s interest in personal gain—both monetarily and reputationally—from her creative work. Copyright law generally recognizes that some forms of infringement are not actionable because they do not implicate these legally protected interests. While this is most commonly seen through the doctrines of fair use and substantial similarity,67 “trivial” de minimis use is but another extension of the same principle.68 The making of a singular free copy of a copyrighted work, while not necessarily fair use, may nonetheless be so trivial so as to not implicate the original author’s protected copyright interests.
Accordingly, the Ninth Circuit approach is out of step with both Congressional understanding and a common-sense approach to copyright law. A proper application of the de minimis doctrine requires acknowledging that it covers trivial uses of whole copyrighted material.
C. A Note on Accounting for Triviality Through Damages
It may be argued that triviality is already properly accounted for by the damages component of copyright law. 17 U.S.C. § 504 lays out two potential remedies for copyright infringement. Primarily, a copyright owner may recover the “actual damages suffered by him or her as a result of the infringement.”69 However, if the copyright owner cannot establish a “causal connection between the alleged infringement and some loss of anticipated revenue,”70 then he may only recover statutory damages “in a sum of not less than $750 or more than $30,000 as the court considers just.”71 Accordingly, the Copyright Act contemplates a scenario where infringement has been established but the harm to the copyright owner cannot be calculated.
However, there are several reasons why section 504 cannot adequately account for trivial uses of copyrighted works. First, the justification for awarding statutory damages is not the same as that for finding de minimis infringement. Damages are levied on those guilty of infringement, while someone who engages in de minimis copying is by definition not an infringer. Statutory damages accounts for actionable injuries that are difficult to quantify, while the de minimis doctrine accounts for injuries that are so trivial to not be legally actionable.72 Second, because plaintiffs have a right to a jury trial on the amount of statutory damages awarded under section 504(c),73 resolving all cases of trivial use at the damages stage risks wasting significant judicial resources on essentially insignificant matters.74 The de minimis doctrine exists broadly to prevent the devotion of substantial resources to litigation whose costs outweigh the benefits.75 Third, the amount of money rewarded by the statutory damages provision is not nominal. Even the statutory minimum of $750 can be a significant sum to an individual defendant. Someone who engages in de minimis copying should not have to pay damages to remedy a wholly insignificant injury; statutory damages are not a consolation prize for copyright owners who experienced trivial harm.
III. Proposed Framework
As a whole, de minimis is a doctrine designed to identify cases of non-actionable copying before, and differently from, a fair use inquiry. Because de minimis analysis “must balance the interests protected by the copyright laws against the stifling effect that overly rigid enforcement of these laws may have on the artistic development of new works,”76 it should be fact-sensitive, with an eye towards two principal considerations. First, courts should look at the degree to which the plaintiff’s work was copied by the defendant. This consideration overlaps completely with substantial similarity analysis, and is merely a way of verbalizing the principle that an appropriative work that is not substantially similar to the work it appropriates is non-infringing. Second, courts should determine whether the defendant’s use of the plaintiff’s work implicates the plaintiff’s legally protected copyright interests. As part of this determination, courts may examine how the plaintiff’s work manifests in the defendant’s work. This prong is aimed at fully actualizing the inherent principle of de minimis as a legal concept: that the law will not address trivial injuries.
This framework is not meant to represent anything more than a very low bar to establishing infringement. Substantial considerations of whether certain technically infringing uses are infringing as a matter of law are properly addressed through fair use analysis. However, as previously established, there are instances where the fair use factors do not result in a correct outcome, or where the alleged infringement is so trivial to not merit such an inquiry. It is in this zone that the de minimis doctrine has a proper role.
IV. On “Observability” in Substantial Similarity Analysis
The use of quantitative sufficiency in de minimis analysis raises a question of how exactly quantitative sufficiency is determined. Are courts to only look at how much of the plaintiff’s work was taken, or may they also examine the quantity of copied material in the defendant’s work? Most circuit courts that have addressed the question, including the Court of Appeals for the Second Circuit, have definitively stated that what matters for substantial similarity is the significance of the copying as to the plaintiff’s work.77 Indeed, common sense dictates that “no plagiarist” should be able to “excuse the wrong by showing how much of [the plaintiff’s] work he did not pirate.”78 A rule to that effect would allow a plagiarist to escape copyright liability by burying significant copied material within his own, lengthy work. However, the de minimis doctrine as practiced in the Second Circuit contains an element that seemingly goes against this rule.
A. The Second Circuit Test
The defining de minimis case in the Second Circuit is Ringgold v. Black Entertainment Television, Inc., in which the Court of Appeals for the Second Circuit considered whether a work of art appearing in the background of a sitcom episode for a total of 26.75 seconds constituted de minimis use.79 Although it agreed that courts are to refer to the quantity of the plaintiff’s work taken in determining substantial similarity, the Ringgold court nevertheless incorporated into its de minimis analysis an “observability” component for infringement of visual works, which looks at “the length of time the copied work is observable in the allegedly infringing work and such factors as focus, lighting, camera angles, and prominence.”80 Since Ringgold, the observability doctrine in the Second Circuit has expanded. Courts in the Second Circuit now recognize several additional observability factors, including recognizability by a lay observer of the plaintiff’s work,81 distance at which the plaintiff’s work is perceived,82 whether the work is in the foreground or background,83 and whether and to what extent the “dialogue, action, and/or camera work in the secondary work calls the viewer’s attention to the copyrighted work.”84 Courts in the Second Circuit have also applied observability factors to cases of sound recording and musical composition infringement.85
B. The Ninth Circuit Approach
The Court of Appeals for the Ninth Circuit has never directly considered whether observability is a proper component of analyzing the substantial similarity of visual works. Arguably, the court has implicitly rejected the observability doctrine, holding that de minimis only applies “to the amount or substantiality of the copying—and not the extent of the defendant’s use of the infringing work.”86 Where the Ringgold court assessed whether the poster at issue was sufficiently observable to cross the threshold of substantial similarity, the Ninth Circuit instead sees “no place” for a de minimis inquiry where the “degree of copying [is] total.”87
However, it may be that the Ninth and Second Circuits use different language to ultimately get to the same place.88 Like the Ninth Circuit’s intrinsic test for substantial similarity, the Second Circuit’s observability doctrine also asks whether the “average lay observer” would recognize the appropriation.89 And the Ninth Circuit’s “total concept and feel” test can conceivably incorporate observability principles. Indeed, in VMG Salsoul v. Ciccone, the Court of Appeals for the Ninth Circuit concluded that the defendant’s sampling of a copyrighted sound recording constituted de minimis infringement in part because the sample was “transposed” and “overlaid … with sounds from many other instruments,” and was therefore “easy to miss.”90 It is true that the court also accounted for the fact that the sample was only 0.23 seconds long,91 and it is therefore unclear whether the court would have held differently had a longer portion of the plaintiff’s song been sampled. Nevertheless, the court undoubtedly incorporated factors relating to “aural observability” in its analysis, suggesting that its approach is not entirely unfriendly to the observability doctrine.
C. Challenges and Outcomes of the Observability Doctrine
Under the observability doctrine, a defendant’s work that furtively incorporates a plaintiff’s entire copyrighted work may still be held non-infringing.92 Although the observability doctrine is now established Second Circuit precedent,93 the Ringgold court seems to have invented it out of whole cloth, citing no legal basis for its existence. Indeed, where legal precedent has established that courts may only look at the significance of copying in relation to the plaintiff’s work to determine substantial similarity, a doctrine that contemplates how visually important the plaintiff’s work is within the defendant’s work explicitly goes the other way.94
One challenge of the observability doctrine is its unique applicability to visual (and perhaps aural) works. The Ringgold court provides no justification for why it singles out visual works as subject to this different standard. It is true that such factors simply do not translate to instances of entire literary works copied into other literary works, or appropriation of computer code. But difficulty of application is not a good reason for applying different tests to different kinds of works; in fact, it is a good reason for doing the opposite.
It is also true, however, that visual works have unique properties with unique considerations. It is much easier for a copyrighted painting to be significantly obscured within a movie than it is for a copyrighted essay to be significantly obscured within a book. Instances of visual work infringement as seen in Ringgold and Sandoval are also much more likely to be incidental (and accidental) than a comparative instance of literary work infringement. This provides some justification for having a test that is more forgiving to the defendant in these cases.95
In any event, the observability doctrine adds uncertainty and burden brought on by producing an even more fact-sensitive, case-by-case determination of substantial similarity. While plaintiffs generally must establish that a potentially infringing work is substantially similar to theirs, a plaintiff in the Second Circuit must additionally establish that the plaintiff’s work was sufficiently observable to a lay audience viewing the defendant’s work—even if the defendant undisputedly appropriated the plaintiff’s entire work. In doing so, the plaintiff must not only show that the two works are substantially similar, but that the defendant’s use of the plaintiff’s copyrighted material is substantially similar.
For instance, in Solid Oak Sketches, LLC, v. 2K Games, Inc., the District Court for the Southern District of New York held that a video game’s display of copyrighted tattoos on in-game renderings of professional basketball players constituted de minimis infringement.96 Although the tattoos were rendered in their entirety, their display was “small and indistinct,” “out of focus,” and indiscernible due to the “quick and erratic movements” of the player characters.97 The case therefore turned not on the qualitative or quantitative copying of the plaintiff’s work, but on the defendants’ use of the copyrighted tattoos.98
How exactly a plaintiff goes about establishing substantially similar use is unclear. Courts addressing this component as a matter of law at the motion-to-dismiss or summary judgment phases have no clear guidelines for how to determine observability—which factors to consider or prioritize. Indeed, as observability is a fact-sensitive inquiry, no one factor is dispositive in every case. For instance, “the length of time the copied work is observable” seems to play a primary role in observability analysis,99 but courts in the Second Circuit have determined that even very brief displays of plaintiffs’ works may be actionable when the works are “conspicuously displayed.”100 What is the outcome, then, when a copyrighted work appears in a film for no more than three seconds, in the background, but brightly lit and in focus? Two different judges may reasonably reach two different conclusions. Moreover, because “a higher quantity of copying is required to support a finding of substantial similarity” when the plaintiff’s work is not “wholly original,”101 two copyrighted works displayed at the exact same level of observability in a defendant’s work may nevertheless differ as to whether their respective uses are de minimis.
Substantial similarity analysis has long been used to determine how much of a plaintiff’s copyrighted work appears in a defendant’s work, and whether what was taken constitutes a significant part of the original work. Simply put, it determines whether a plaintiff’s work was actionably copied by the defendant. If a potentially infringing work copies an artist’s entire work, substantial similarity necessarily exists between them. While de minimis can be understood partly as a lack of substantial similarity, that understanding cannot support an inquiry that analyzes how a plaintiff’s work appears in an infringing work. If the Second Circuit’s observability doctrine is to rest on solid legal ground, it must be grounded in something other than substantial similarity.
D. Resolution Through the Proposed Framework
The Second Circuit’s observability doctrine—while poorly grounded in principles of substantial similarity—finds a natural home within the above-proposed framework, which incorporates a broader understanding of the de minimis doctrine to encompass trivial copying that does not implicate the copyright holder’s legally protected interests. In other words, two identical “uses” of a copyrighted work (in terms of quantitative and qualitative taking) may vary as to whether they are de minimis depending not on the degree of substantial similarity (which in this case would be equal) but rather on how they implicate the original author’s protected interests through the work’s observability in the appropriating works. While it may be legally unorthodox to look at the extent of use in a defendant’s work to determine substantial similarity, it makes perfect sense to do so if one is trying to discern whether a certain use of copyrighted material only trivially implicates a copyright owner’s legally protected interests. A pinball machine in the background of a scene may not be fair use, but it is certainly trivial enough to not implicate the copyright owner’s interest in benefitting from his unique design. A picture accidentally hosted on a nigh-inaccessible web address may not be fair use, but it is certainly trivial enough to not implicate the photographer’s interest in ensuring people associate the picture with him.
Utilizing this framework also ameliorates the problem related to uncertainty. Of course, the case-by-case nature of all substantial similarity determinations makes it impossible to draw a clear line between what copying is substantially similar and what is not. However, removing the additional dimension of observability from the purview of substantial similarity adds more consistency to that doctrine. Moreover, rather than getting bogged down in legally unstable notions of “substantially similar use,” the proposed framework gives plaintiffs a more coherent idea of where their case stands and what they must establish to clear the de minimis threshold.
V. The “Sampling” Circuit Split
A. The Sixth Circuit Stance
Another question raised by the de minimis doctrine is whether it can be applied to cases of infringement involving music sampling. In Bridgeport Music, Inc. v. Dimension Films, the Court of Appeals for the Sixth Circuit evaluated whether the use of a sample from a sound recording in a rap song constituted de minimis infringement where “a two-second sample from the guitar solo was copied, the pitch was lowered, and the copied piece was ‘looped’ and extended to 16 beats.”102 Despite agreeing in principle with the district court’s finding of de minimis infringement, the Court of Appeals for the Sixth Circuit nevertheless found for the plaintiff, holding that the de minimis doctrine (and substantial similarity analysis as a whole) does not apply to infringements of copyrighted sound recordings.103
The Court of Appeals for the Sixth Circuit primarily based its rationale on its reading of section 114 of the Copyright Act of 1976, which defines the scope of exclusive rights in sound recordings. Section 114(b) states in part:
The exclusive right of the owner of copyright in a sound recording [to prepare derivative works based upon the copyrighted work] is limited to the right to prepare a derivative work in which the actual sounds fixed in the sound recording are rearranged, remixed, or otherwise altered in sequence or quality. The exclusive rights of the owner of copyright in a sound recording [to reproduce the copyrighted work in copies or phonorecords and prepare derivative works based upon the copyrighted work] do not extend to the making or duplication of another sound recording that consists entirely of an independent fixation of other sounds, even though such sounds imitate or simulate those in the copyrighted sound recording.104
As the court reasoned, because section 114(b) gives a sound recording owner the exclusive right to rearrange, remix, or otherwise alter the sounds fixed in the recording—and because the text only draws a line at an entirely independently created soundalike—the sound recording owner therefore “has the exclusive right to sample his own recording.”105 Moreover, being that “the only way to infringe on a sound recording is to re-record sounds from the original work,” any re-recording is guaranteed to be substantially similar because it is an exact copy of the original.106
The court provided other, more values-based reasons for its conclusion. For one, sampling necessarily appropriates something of value, no matter how minuscule the sample or how buried it is in the new song.107 Sampling allows producers to avoid hiring studio musicians, thereby threatening the ability of musicians to make a living.108 Secondly, even if cost is not a factor, a producer may still believe that the inclusion of a certain sample would add to a record some material creative element that would not exist in the sample’s absence.109 Moreover, because sampling lifts sounds directly from the copyrighted medium, sampling is more of a “physical taking rather than an intellectual one.”110 Finally, the fact that sampling requires conscious intent to appropriate a copyrighted recording suggests that the de minimis doctrine, which is largely meant to weed out cases of incidental copying, is less appropriate in cases of sound recording infringement.111
The Court of Appeals for the Sixth Circuit’s holding that plaintiffs alleging sound recording infringement need only establish actual copying to succeed on their claims represents a clear departure from how courts have approached every other category of infringement, standing in stark contrast to the established understanding that actionable infringement requires the allegedly infringing work to be substantially similar to the work from which it appropriates.112 It is no surprise, therefore, that courts outside the Sixth Circuit have largely declined to follow Bridgeport’s rule.113
B. The Ninth Circuit Stance
In VMG Salsoul, the Court of Appeals for the Ninth Circuit explicitly declined to follow Bridgeport, creating a circuit split. The court based its argument on four grounds. As an initial matter, the rule that infringement occurs only upon a finding of substantial similarity is well established in copyright law.114
Second, nothing in section 114 of the Copyright Act suggests an intention to expand the scope of exclusive rights granted to sound recording copyright owners beyond those granted to owners of other copyrights. If anything, section 114(b) is intended to limit the rights of a sound recording copyright owner.115 Indeed, the portion of section 114(b) interpreted by the Bridgeport court begins by stating what a sound recording copyright owner’s derivative works right is limited to.116 The text continues by outlining where a sound recording copyright owner’s exclusive rights “do not extend.”117
That section 114 limits, rather than expands, the scope of sound recording copyrights is confirmed by the legislative history of the Copyright Act. The House Report clarifies the approach of the Act as “set[ting] forth the copyright owner’s exclusive rights in broad terms in section 106, and then to provide various limitations, qualifications, or exemptions in the 12 sections that follow. Thus, everything in section 106 is made subject to sections 107 through 118, and must be read in conjunction with those provisions.”118 It makes little sense to read an expansion of rights into a provision that clearly limits them. Taken as a whole, section 114(b) merely stands for the principle that a sound recording copyright owner’s rights do not extend beyond the actual sounds fixed in the recording to the underlying composition.
Third, the Bridgeport court’s interpretation of section 114(b) rests on a logical fallacy. The court relied on the provision that a sound recording copyright owner’s rights do not extend to “another sound recording that consists entirely of an independent fixation of other sounds” to conclude that a sound recording copyright owner’s rights must extend to all sound recordings that do not consist entirely of an independent fixation of other sounds.119 It is logically fallacious to infer the inverse of a conditional from the conditional. The correct inference to make from the provision is that if a sound recording copyright owner’s rights extend to a potentially infringing work, then that work does not entirely consist of an independent fixation of other sounds.120 This inference is not the same as the court’s inference. The correct inference merely restates section 114(b)’s instruction: sound recording copyright holders cannot claim that independently created “soundalikes” infringe on their copyright. It does not logically follow that all sound recordings that are not completely independently created are therefore per se infringing. Indeed, some of them may be protected by the same principles of substantial similarity delimiting every other exclusive right granted by the Copyright Act.121
Lastly, the Bridgeport court’s value propositions rest on a flawed understanding of copyright law. By arguing that sampling appropriates something of value by “free riding” on the work of musicians, the court extended the protective reach of copyright to the effort an artist expends on her art. But copyright law does no such thing. Copyright law protects only the expressive aspects of a copyrighted work, not the “fruit of the [artist]’s labor.”122 If what is taken from a copyrighted work is so insubstantial as to not appropriate the work’s expressive aspects, the taking is not actionable.
C. Relation to the Proposed Framework
The Court of Appeals for the Ninth Circuit’s argument is more persuasive because it rests on more principled understandings of statutory interpretation, logical reasoning, and the goals of copyright law. Section 114 merely defines the scope of a sound recording copyright, much like section 113 defines the scope of a copyright in pictorial, graphic, and sculptural works.123 Neither provision eases the burden of a copyright owner to prove substantial similarity. The fact that section 114 does not create a carve-out for substantial similarity is of no consequence. Substantial similarity is nowhere to be found in the Copyright Act; it is a judicial test established to examine whether an instance of alleged copying falls within copyright law’s proscriptive bounds. That does not make the doctrine any less legitimate.
The same principles underlying the proposed framework that apply to all other instances of infringement apply to infringements of sound recordings; sound recording copyright owners have the same legally protected interests as any other copyright owner.124 At bottom, courts must determine whether an instance of sound recording infringement implicates these interests. They should do so both by conducting a substantial similarity analysis as well as by determining whether the infringed recording appears in the new work in a more-than-trivial way.
Conclusion
The following hypothetical illustrates the current state of the de minimis doctrine. Imagine that you are a photographer taking a picture of a city park. In the background of the picture, partially obscured by a crowd of people, is a sculptor’s copyrighted sculpture. You did not intend to capture the sculpture—it just happened to be in the frame. You publish and sell your picture. Upon seeing your picture, the sculptor sues you for copyright infringement and requests a bench trial. If you are being sued in Illinois or Missouri, the district court will likely find that infringement has occurred, as direct evidence of copying can be seen by the clear presence of the sculpture in your picture. Your only recourse, then, is fair use, which would likely turn on whether your picture “transforms” the sculpture as well as the effect on the market for the sculpture (or licensed photographs of the sculpture). If you are being sued in California, the district court may find that the two works have a different “total concept and feel,” and are therefore not substantially similar. If you are being sued in New York, the district court may first determine whether the presence of the sculpture meets the quantitative threshold for actionable copying by referring to observability factors. The partial obscurity as well as the lack of focus on the sculpture may lead to a finding of de minimis copying. If the sculpture is sufficiently observable, the court may nevertheless determine that the unintentional inclusion of the sculpture is de minimis on trivial use grounds, as it does not implicate the protected interests of the sculptor. Of course, the court may also find non-infringement on fair use grounds. In this hypothetical, Second Circuit jurisprudence provides you with three potential off-ramps to liability. The Ninth Circuit (at least definitively) provides two. The Seventh and Eighth Circuits provide only one.
While some apparent circuit splits—like the difference in approach regarding observability—are perhaps more illusory than concrete, it is nevertheless the case that the in-flux status of the de minimis doctrine is resulting in an inconsistent application of copyright law across the circuits. The framework proposed by this note attempts to alleviate the confusion regarding the proper scope of the de minimis doctrine, while providing concrete prongs for its application.
Overall, a broad understanding of the de minimis doctrine has a proper role in copyright law as a defense against infringement, in accordance with copyright law’s central balancing act of encouraging creation without crossing a line into draconian over-restriction. A proper de minimis framework is not a mere technicality. It should affirmatively guard the gates of copyright law, ensuring that lawsuits go forward only if the copying at issue has improperly harmed what copyright law is meant to protect: the incentive structures that foster artistic creation. Principles of observability and trivial use are welcome judicial tools to filter out cases that do not surmount this bar.
While the de minimis doctrine should be applied broadly, it should also be applied rarely. De minimis should not be an easy workaround to a comparatively more difficult fair use inquiry. Both doctrines have unique roles within copyright law, serving different functions at different stages of analysis. Moreover, the de minimis doctrine should not turn establishing actionable copying into an onerous affair. Instead, de minimis should be used to dispose of the uncommonly raised “[q]uestions that never need to be answered,”125 which nevertheless sometimes find their way into courtroom doors.
Unfortunately, the only ways to definitively harmonize the interrelated concepts of substantial similarity, legal copying, and de minimis use across the circuits are (a) for Congress to amend the Copyright Act, or (b) for the Supreme Court to rule on the correct application of copyright law. Neither possibility seems especially likely. Courts should therefore take it upon themselves to expand their understanding of the de minimis doctrine in copyright law to properly account for the its full scale and scope. In the meantime, it is useful to clarify the doctrine by understanding exactly where and how its fault lines manifest, so that plaintiffs in different circuits are aware of their obligations and defendants are aware of the protections afforded to them.