Reading Together and Apart: Juries, Courts, and Substantial Similarity in Copyright Law

In 1994, the Honda Motor Company launched a new television commercial for its Honda del Sol convertible.1 Intended to highlight the car’s detachable rooftop, the commercial was developed over two years, at one point carrying the project name of “James Bob.”2 As described in a district court opinion, the commercial’s focus was “a fast-paced helicopter chase scene featuring a suave hero and an attractive heroine [in a Honda del Sol], as well as a menacing and grotesque villain”3 who jumped out of the helicopter onto the top of the car but was then sent flying when the hero detached the Honda’s roof.4 The hero wore a tuxedo, both characters spoke with a British accent, and the commercial featured bright, horn-driven music throughout.5

After Metro–Goldwyn–Mayer, Inc. (“MGM”), the owner of the rights in the James Bond film franchise, saw the commercial, it immediately demanded that Honda take the commercial off the air.6 When Honda refused, MGM filed suit, alleging, among other things, copyright infringement.7 In an attempt to resolve the dispute before the commercial’s Super Bowl airing, Honda changed the characters’ accents to American and altered the music.8 Still not appeased, MGM moved for a preliminary injunction.9

When I introduce the case in class, I typically begin by showing the commercial and asking students whether, if they had been in-house counsel for Honda, they would have approved going forward with the commercial. (The students have not, at that point, read the opinion in the case.) Inevitably, several students immediately voice a cautionary note, remarking that they would have expected a challenge from the owner of the copyright in the James Bond character10 because that character and the character in the commercial are, to use the language of copyright doctrine, “substantially similar.”

A momentary diversion for some background: In the United States, copyright law is said to provide an economic incentive to create expressive works. A creator of such a work, under this theory, will not have such an incentive if others are permitted to copy and distribute her work for sale.11 Because the copier need recoup only the costs of copying and distribution to profit, and not the costs of production, the copier can offer the work for sale at a lower price than that of the original author, which ostensibly dissuades the author from producing in the first place. Hence, by according the original author certain rights to control use of the work for a limited time, including the right to transfer that control to another, copyright law provides an incentive to create the work in the first place.12

Of course, the reality of creative production is sometimes very different. Some authors are motivated by attributional or reputational concerns and not by profit;13 they may not mind reproduction of their work so long as they are accurately credited or may want to limit copying for reasons other than purely economic ones.14 And regardless of the author’s motivation, U.S. copyright law does not allow unlimited control of a work by the copyright owner. Fair uses, de minimis uses, uses of the ideas a work embodies, certain kinds of sales and displays, and various other socially beneficial, lobbied for, or legally insignificant uses lie beyond the copyright owner’s reach. But if a use does not fall into one of these categories, copyright law is generally unconcerned with the copyright owner’s motivations in restricting the use.15 In every copyright infringement case, then, there are two fundamental questions to be answered: (1) Is the defendant even using the copyright owner’s work in the first place?16; and (2) if so, does that use fall into one of the categories that lies outside the scope of the owner’s legal right to control the work?17

It is the first question that gives rise to the doctrine of substantial similarity. In many cases, the answer to the question is simple: The defendant is, for example, unquestionably copying and distributing an entire musical recording or performing the plaintiff’s copyrighted stage play. But in other cases, the question becomes more complicated because the defendant has created a work that incorporates some, but not all, of the plaintiff’s work, has merely been inspired by the plaintiff’s work, or has simply created a work in the same genre as the plaintiff’s work. Thus, the two works may be similar on their face, but the question is whether they are similar enough to warrant a finding of infringement (that is, “substantially similar”).

So, to return to the Honda case, when students say, after viewing the commercial, that they think the commercial might infringe the copyright in the James Bond character, I then ask: “How do you know that the character in the Honda commercial is James Bond?” That question, in turn, leads us to compile a list of characteristics that signal James Bond, many of which seem entirely generic in isolation: wearing a tuxedo, speaking in a British accent, maintaining composure while engaging with villains, driving a fancy sports car, and being accompanied by an attractive woman.18 Yet despite the vagueness of the items on the list, most of the students “know” that the Honda protagonist is meant to be James Bond, and the only remaining question in their minds is whether Honda could mount a plausible fair use claim on the grounds that the commercial is engaging in parody.19

Rarely, however, do we explicitly discuss the process by which the students came to understand James Bond as a character or the process by which they compared the image of that character in their minds with the character they saw in the commercial.20 Nor do I spend much time querying the moral basis for the conclusion (by some students) that Honda’s failure to secure permission to include the character in the advertisement was wrongful, although this conclusion is much more contested among the group.21 In other words (and to my discredit), we do not attempt to rigorously unpack what they are doing as readers when they determine whether Bond and the Honda protagonist are “substantially similar” or when they inevitably go on to invoke the Austin Powers franchise, another “substantially similar” (but almost certainly fair) use of the James Bond character.

Professor Said’s article, A Transactional Theory of the Reader in Copyright Law, highlights the importance of asking these questions, encouraging more deliberate attention—by scholars, courts, and juries—to the process of engaging with creative works as an active exercise between author and reader.22 Relying predominantly on the work of literary theorist Louise Rosenblatt, who offered a nuanced version of reader-response theory that limits the range of reasonable responses to those cued by the text,23 Professor Said joins those scholars who call for juries (and judges) to be provided with more guidance in copyright infringement cases, both to answer the question of what should be excluded in the comparison of two works and to shape the ultimate process of comparison.24 For example, as she persuasively suggests (using Rosenblatt’s terminology), juries should be allowed to rely on experts when they are engaged in the process of efferent reading (the process of extracting information from a text).25 Juries ostensibly engage in efferent reading when answering the second question (as I have framed it above) in a copyright infringement case—determining which aspects of the defendant’s use fall into categories that lie beyond the scope of the plaintiff’s rights. By contrast, Professor Said suggests, juries are in need of less assistance when they are engaged in the process of aesthetic reading (responding to the work as a whole).26 As applied to a copyright infringement case, this suggestion is designed to help answer the first question (again, as I have framed it)—whether the defendant is even making use of the plaintiff’s work. Ultimately, contends Professor Said, “[b]y ignoring the realities of how people actually engage with art, courts overestimate what jurors can do, and underestimate their need for guidance on both law and how to engage with art.”27

As my previous work suggests,28 I am sympathetic to Professor Said’s project writ large. Focusing our attention in copyright law on the audience aligns with the goal of making creative works available for others to read, interpret, and build upon. In this vein, a reader-response approach to copyright law can be particularly valuable when the meaning of a work is directly at issue, such as when the question is whether, in the fair use inquiry, the defendant has engaged in a transformative use of the plaintiff’s work. In such an inquiry, the decision-maker must determine whether transformation should be judged only with respect to the defendant’s intentions and efforts or also (or instead) by how the work is experienced by its various audiences.29 Reader-response theory thus allows the doctrine to expand the notion of how a work conveys meaning, making room for differences that might not exist with an author-centered approach to meaning.

But this then means that the question of whether two works are substantially similar must also, under reader-response theory, be similarly expansive, even though Professor Said’s approach would seem to favor a cabining of the range of interpretations. This is not, to be clear, a critique of Professor Said’s thoughtful work but rather an acknowledgment of the breadth of reasonable engagement with creative works.30 Courts can tell juries that in making a determination of substantial similarity they should or should not consider certain aspects of the two works, much as courts tell juries not to consider the prejudicial testimony just objected to in court. (In both cases, the admonition may represent the triumph of hope over experience.) But beyond that, it is unclear how we could ever truly say, except in the cases that are not likely to make it to a jury in the first place, that any jury was right or wrong in deeming two works to be similar or dissimilar—in part, because we are unlikely to agree on just what “substantially similar” means. Were the students who concluded that the Honda commercial “took too much” of the James Bond character right or wrong? What about the students for whom Jason Bourne or Ethan Hunt came more quickly to mind?31

The challenge is that copyright doctrine has never put much meat on the bones of “substantial similarity.” In many courts, substantial similarity is described only as “whether an average lay observer would recognize the alleged copy as having been appropriated from the copyrighted work.”32 In Arnstein v. Porter, a case involving a comparison between two musical compositions, the court framed the inquiry as “whether defendant took from plaintiff’s works so much of what is pleasing to the ears of lay listeners, who comprise the audience for whom such popular music is composed, that defendant wrongfully appropriated something which belongs to the plaintiff.”33 Scholars have critiqued the simplicity of these formulations,34 noting that juries are given little information on how to conduct this task beyond “general abstract statements” directing them to consider the significance of both qualitative and quantitative aspects of the works.35

To be sure, copyright doctrine requires the concept of substantial similarity—otherwise, as the Second Circuit noted in Nichols v. Universal Picture Corp., “a plagiarist would escape by immaterial variations.”36 At the other extreme, however, it simply cannot constitute copyright infringement to be inspired by another’s work, or to use its themes, even if those who originated the themes would experience some level of pique in not being credited for that use. Two sports writers can write a column putting forward the same theory of why the team named the favorite to win the Super Bowl failed to achieve its goal. The second writer may be excoriated for being unoriginal, but unless she uses similar enough phrasing, structure, or arrangement, her faults are journalistic, not legal, much as the student who copies wholesale a theory for a term paper without crediting her source has committed academic misconduct but not necessarily copyright infringement.

It is the area between these two poles, however, that remains undefined, and perhaps inevitably so. Without a substantial overhaul of copyright doctrine to cabin the ways in which two works can be deemed legally similar, how can we say that a jury has answered the question incorrectly?37 We can use Professor Said’s inventive Appendix to her article, in which she demonstrates the difference between efferent reading and aesthetic reading by having the reader engage with a set of song lyrics, a poem, and a painting, as a jumping-off point. Consider, for example, the lyrics to the Carla Bruni song, originally written in French but translated into English by Professor Said for the benefit of readers not familiar enough with the language of the original.38 Once Professor Said alerts the reader to the availability of the translation, the translation then becomes for some the work under discussion, which she uses to introduce the concepts of efferent and aesthetic reading to her audience.39 These readers will see a translation and its original as substantially similar only if they believe (or are told) that the translation is an accurate one—meaning that they will experience the work in the same way as they would if they were able to read the poem in its original language.40 Without that knowledge, those readers would visually compare the text of the original to the text of the translation and conclude that they are largely dissimilar.

But for readers who understand the language of the original, the similarity question will no doubt be influenced by their assessment of the choices made by the translator. Should the sound of a clock in the last line of the Bruni lyrics, for example, be rendered in the translation as written for a French-speaking audience (“tic tac tic tac”) or rendered to be more familiar to an English-speaking audience (“tick tock tick tock”)? Should “ranger les souvenirs” be translated as “tidy up my memories” or “store my memories”? Are these two renderings substantially similar because they are “close enough”? Should this difference have any legal relevance?

These questions also arise in the second example in Professor Said’s Appendix. Here, Professor Said provides context for the reader to better appreciate the poem by Adrienne Rich—namely, that it invokes the themes of an earlier poem by Gerard Manley Hopkins.41 Professor Said’s layered description of the themes and metaphors of Rich’s poem provides a starting point for the reader to engage with the work,42 but is a reader who does not see these metaphors or who sees more dissimilarity than similarity necessarily wrong, even if we acknowledge that some interpretations are too unmoored from the text to be reasonable? What about a viewer who cannot appreciate the visual perspective in the Lawrence painting that Professor Said describes, to take the Appendix’s third example,43 or who does not perceive color in the same way as other viewers?

Ultimately, we cannot escape interpretation in the law, and most interpretations are only as correct as the last body to pronounce them so.44 The jury functions as readers not only when it compares the two works at issue but also when it hears opening and closing statements, attempts to understand jury instructions, or listens to the statements made by fellow jurors during deliberations. Engaging with the works at issue is reading, to be sure, but in any litigated case, it’s reading all the way down. Professor Said is particularly persuasive, then, when she contends that the first step should be to make the jury responsible only for comparing the two works, leaving the judge to follow up by deciding whether that similarity is legally cognizable.45 Where I am, perhaps, more skeptical than Professor Said is in my lack of confidence that further guidance will result in any real difference in case outcomes. Any case that makes it to a jury will be one about which reasonable readers can disagree.46

Perhaps this is because, to use Rosenblatt’s terminology, most of the tasks readers are asked to accomplish in a copyright infringement case involve aesthetic reading rather than efferent reading. Determining whether two works are similar because they both invoke the same genre or tropes is not a task that can be accomplished in the same way that one can count the number of times the word “green” appears in a poem or decide whether a film falls into the broad category of “romantic comedy” or “violent psychological thriller.”47 Assessing whether two works are similar enough to warrant an inference that the defendant copied from the plaintiff is a slightly more efferent task insofar as the goal is clearly identified (did the defendant copy or not?), but accomplishing that task still requires an assessment of how similar is similar enough. Identifying the ways in which two works are similar does not mean that those are the only ways in which the works could be similar, or that those similarities are in any way significant. Guidance on how to engage these questions can be helpful, but it can ultimately be no more determinative than a cultural critic who hopes to persuade her readers of the value of her expertise.48

The problem, then, with the substantial similarity inquiry is not that juries are given insufficient information about the process of deciding— it’s that the doctrine has never made the endgame clear. How, ultimately, do we know when a defendant’s work has taken too much? When Walter Benjamin wrote of a work’s “aura,” criticizing reproductions that shared many aesthetic or material qualities with the original as nevertheless deficient, perhaps he was capturing something of the conclusion for at least some readers.49 We can issue calls for more in the way of juror education: how to understand the nuances of a particular genre, whether to read a work quickly or with care and time, or whether to bring any prior familiarity with either work into the analysis. We can attempt to focus the analysis by asking jurors to decide whether the works are economic or aesthetic substitutes for one another, although that is not likely to be particularly useful.50 But in the end, isn’t a juror likely to be motivated most by the fact that she “knows” that the man in the commercial is supposed to be James Bond and that such a use is somehow “improper,” regardless of whatever guidance she has been given? What kind of guidance as to similarity, beyond encouragement to engage in the task more ardently, could yield a different result?

Perhaps, then, the question we should be focusing our attention on is not simply how jurors and judges read but also how jurors and judges as readers become scholars and critics, attempting to persuade others that their reading is correct. If the question of copyright infringement is decided by a jury, our reader cannot be a collection of individuals, each coming to his or her own conclusion. The group must cohere around an agreed-upon result, if not an agreed-upon rationale, subject to invalidation only if some other reader (the trial judge or an appellate panel) deems the result to be so unreasonable as to warrant reconsideration.51 Some jurors may have deep familiarity with the works or genres at issue in a case; others may be experiencing the works for the first time. Reader-response theory has helped copyright law to understand the validity of individual experience for individual reception of a work, but copyright doctrine has not yet sufficiently explored what the theory can tell us for how groups engage in the process of reading and interpretation as a collective body and the process by which an interpretation gains traction in a community to become canon or cultural knowledge.52

We are, as Professor Said thoughtfully reminds us, engaged in constant acts of reading.53 Her calls for more open conversations, both inside and outside the courtroom, about how we read are well worth heeding. The process by which we read, engage, and interpret, both together and apart, should be the next set of experiences that copyright law should take into account.

  1. [1].     Metro–Goldwyn–Mayer, Inc. v. Am. Honda Motor Co., 900 F. Supp. 1287, 1291–92 (C.D. Cal. 1995).

  2. [2]. Id. at 1291.

  3. [3]. Id. at 1292.

  4. [4]. Ibpimin, Honda del Sol Commercial, YouTube, (last visited Mar. 23, 2017).

  5. [5]. Id.

  6. [6]. Metro–Goldwyn–Mayer, Inc., 900 F. Supp. at 1292.

  7. [7]. Id.

  8. [8]. Id.

  9. [9]. Id.

  10. [10]. Scholars disagree on whether one can have a copyright in a character qua character, or whether copyright in a character derives from the copyright in the work of which the character is a part. See generally Zahr K. Said, Fixing Copyright in Characters: Literary Perspectives on a Legal Problem, 35 Cardozo L. Rev. 769 (2013).

  11. [11]. Copyright law speaks in terms of authorship, which refers to the owner of the rights in the work and not necessarily the individual who created the work. See 17 U.S.C. § 201(b) (2012) (“In the case of a work made for hire, the employer or other person for whom the work was prepared is considered the author for purposes of this title, and, unless the parties have expressly agreed otherwise in a written instrument signed by them, owns all of the rights comprised in the copyright.”).

  12. [12]. Shyamkrishna Balganesh, Foreseeability and Copyright Incentives, 122 Harv. L. Rev. 1569, 157781 (2009).

  13. [13]. See generally Rebecca Tushnet, Economies of Desire: Fair Use and Marketplace Assumptions, 51 Wm. & Mary L. Rev. 513 (2009); Rebecca Tushnet, Payment in Credit: Copyright Law and Subcultural Creativity, 70 Law & Contemp. Probs. 135 (2007).

  14. [14]. For example, an author may refuse to authorize creation of a sequel to her work (e.g., in the form of fan fiction) not because she is worried about future income streams but because she wants to control the context in which her work appears. See generally Darren Waters, Rowling Backs Potter Fan Fiction, BBC News, (last updated May 27, 2004, 12:11 PM) (noting that J.K. Rowling generally approves of fan fiction so long as it is noncommercial and not X-rated). Creative Commons licenses allow authors who prefer to announce preemptive waivers of certain rights to do so while preserving their ability to impose limitations on uses of their work by others. See Licensing Types, Creative Commons, (last visited Mar. 24, 2017). For example, a Creative Commons license might allow reproduction and distribution of a work so long as the work is attributed to the original author. Id. U.S. copyright law does not, in most instances, explicitly include attribution as a right of a copyright owner, but copyright owners can require attribution in connection with their right to control other uses of their work.

  15. [15]. One exception is the doctrine of copyright misuse, which has not gained as much traction in copyright law as it has in patent law. When judges have invoked the doctrine, it has been to call out copyright claims seen as overly strategic. See, e.g., Omega S.A. v. Costco Wholesale Corp., 776 F.3d 692, 696–706 (9th Cir. 2015) (Wardlaw, J., concurring in judgment); Assessment Techs. of WI, LLC v. WIREdata, Inc., 350 F.3d 640, 647 (7th Cir. 2003).

  16. [16]. I recognize the implications of using the words “use” and “user” to refer to the activity of engaging with creative works but adopt those words here as shorthand for a range of activities. See Jessica Litman, Readers’ Copyright, 58 J. Copyright Soc’y U.S.A. 325, 331 n.22 (2011) (discussing implications of “users” versus “readers”); cf. Zahr K. Said, A Transactional Theory of the Reader in Copyright Law, 102 Iowa L. Rev. 605, 609 n.14 (2017) (describing decision to use “reading” as the term for engagement with a work); Rebecca Tushnet, Worth a Thousand Words: The Images of Copyright, 125 Harv. L. Rev. 683, 751–52 (2012) (discussing implications of “users” versus “readers”).

  17. [17]. Cf. Shyamkrishna Balganesh, The Normativity of Copying in Copyright Law, 62 Duke L.J. 203, 208 (2012) (“In the second step, referred to as the substantial-similarity analysis, the court applies the level of scrutiny, i.e., thickness, so chosen to its analysis of the two works—the original and the defendant’s copy—to determine whether the copying does indeed rise to the level of an infringement. Although no doubt a factual determination, this step operates as a subjective evaluation of the different parts of the two works and of their relative contributions to the overall significance of the work, both as a quantitative and qualitative matter, in order to assess whether the copying amounted to a ‘wrong.’” (footnote omitted)).

  18. [18]. The protagonist of the commercial requests neither a shaken nor a stirred martini.

  19. [19]. In this regard, the students correctly predict the course of the court’s analysis. The court determined that Honda was not likely to succeed on the merits of a fair use defense. Metro–Goldwyn–Mayer, Inc. v. Am. Honda Motor Co., 900 F. Supp. 1287, 1300–01 (C.D. Cal. 1995).

  20. [20]. Some of this might be affected by the fact that both processes engage with visual works rather than texts, assuming that the students’ knowledge came from the films rather than from the Ian Fleming novels. Tushnet, supra note 16, at 746 (“In MGM, words are abstract, whereas visuals are concrete, and thus the scope of an audiovisual work’s copyright is broader than the scope of a written work’s copyright.”); see Gaiman v. McFarlane, 360 F.3d 644, 660–61 (7th Cir. 2004) (“The description of a character in prose leaves much to the imagination, even when the description is detailed.... A reader of unillustrated fiction completes the work in his mind; the reader of a comic book or the viewer of a movie is passive.”).

  21. [21]. Cf. Balganesh, supra note 17, at 251 (“The similarity analysis thus operates as copyright’s correlative framework, in which the defendant’s actions are examined through the lens of a right-duty relationship between the plaintiff and the defendant.”).

  22. [22]. See generally Said, supra note 16.

  23. [23]. Id. at 618. See generally Louise M. Rosenblatt, The Reader, The Text, The Poem: The Transactional Theory of the Literary Work (1978).

  24. [24]. See, e.g., Shyamkrishna Balganesh et al., Judging Similarity, 100 Iowa L. Rev. 267, 289 (2014) (contending that juries are influenced in their substantial similarity analysis by factors that “ought to be kept out of the copyright system altogether”); Mark A. Lemley, Our Bizarre System for Proving Copyright Infringement, 57 J. Copyright Soc’y U.S.A. 719 (2010) (arguing in favor of enhanced expert guidance at both stages of the infringement analysis and noting that jurors are not likely to be fully equipped to filter out nonprotectable elements of the work at the second stage).

  25. [25]. Said, supra note 16, at 622, 636–40.

  26. [26]. Id. at 638–43.

  27. [27]. Id. at 608.

  28. [28]. See generally Laura A. Heymann, Everything Is Transformative: Fair Use and Reader Response, 31 Colum. J.L. & Arts 445 (2008).

  29. [29]. Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 582 (1994) (noting that the threshold fair use question in that case was “whether a parodic character may reasonably be perceived”); see also Cariou v. Prince, 714 F.3d 694, 707 (2d Cir. 2013) (“What is critical is how the work in question appears to the reasonable observer, not simply what an artist might say about a particular piece or body of work.”). See also generally Andrew Gilden & Timothy Greene, Fair Use for the Rich and Fabulous?, 80 U. Chi. L. Rev. Dialogue 88 (2013) (cautioning that focus on community response should not be limited to privileged communities).

  30. [30]. Like Professor Said, I focus here on copyrighted works that would typically be described as having aesthetic appeal, even though copyright also extends to works, such as software, that are not engaged with by audiences in the same way as other works.

  31. [31]. See Said, supra note 16, at 612 (“No scholarship has attempted to articulate a broad theory of what copyright case law’s judge and jury do as readers when they address the question of infringement.”). For the uninitiated, see, e.g., The Bourne Identity (Universal Studios 2002); The Bourne Supremacy (Universal Studios 2004); The Bourne Ultimatum (Universal Studios 2007); see also Mission: Impossible (Paramount Pictures 1996); Mission: Impossible 2 (Paramount Pictures 2000); and Mission: Impossible 3 (Paramount Pictures 2006).

  32. [32]. Ideal Toy Corp. v. Fab-Lu Ltd., 360 F.2d 1021, 1022 (2d Cir. 1966); see also Peter Pan Fabrics, Inc. v. Martin Weiner Corp., 274 F.2d 487, 489 (2d Cir. 1960) (noting that although the two designs at issue in the case were not identical, “the ordinary observer, unless he set out to detect the disparities, would be disposed to overlook them, and regard their aesthetic appeal as the same. That is enough; and indeed, it is all that can be said...”).

  33. [33]. Arnstein v. Porter, 154 F.2d 464, 473 (2d Cir. 1946).

  34. [34]. See, e.g., Balganesh et al., supra note 24, at 269 (“Copyright law thus seems to assume that the question of substantial similarity can continue to remain a simple comparison of the two works, even in the face of extensive factual evidence that bears directly on the dispute in question.”); Tushnet, supra note 16, at 717, 719 (suggesting that doctrine of substantial similarity for visual works “just is; there is no way to break it down or describe it” and just because “two works produce the same emotional state in a viewer does not mean that they are the same”); id. at 738–39 (proposing that substantial similarity should be abandoned entirely and that “[a] reproduction right that is truly a reproduction right would cover only pure copying and copying so nearly exact that observers would be inclined to see two works as the same”).

  35. [35]. Balganesh et al., supra note 24, at 274 (citing Final Charge to Jury and Special Verdict Form at 14, Oracle Am., Inc. v. Google, Inc., 750 F.3d 1339 (2014) (No. C 10-03561 WHA)).

  36. [36]. Nicholas v. Universal Pictures Corp., 45 F.2d 119, 121 (2d Cir. 1930).

  37. [37]. Cf. Zahr K. Said, Reforming Copyright Interpretation, 28 Harv. J.L. & Tech. 469, 474 (2015) (“[I]n copyright litigation, the works at issue must be interpreted for legal purposes, and here it is indisputable not only that a fixed meaning may attach but that frequently for an outcome to be reached, it must attach so that copyright doctrines can be applied.”).

  38. [38]. Said, supra note 16, at 615, 648–49.

  39. [39]. Id. at 629–30.

  40. [40]. See Sonia Colina, Evaluation/Assessment, in 2 Handbook of Translation Studies 43, 45 (Yves Gambier & Luc van Doorslaer eds., 2011) (“Reader-response approaches evaluate the quality of a translation by determining whether readers respond to it in the same way readers respond to the source.... It is not difficult to see the problems involved in trying to measure reader-response; one in fact wonders whether it is actually possible to determine whether two responses are equivalent, given that even monolingual texts can trigger non-equivalent reactions from slightly different groups of readers.” (citation omitted)); Robert Kirk Walker & Ben Depoorter, Unavoidable Aesthetic Judgments in Copyright Law: A Community of Practice Standard, 109 Nw. U.L. Rev. 343, 375 (2015) (“A work’s ‘total concept and feel’ is inherently subject to the quirks of individual taste and sophistication, and so decisions of law based on it are no more likely to be consistent than decisions subscribing to the doctrine of avoidance.”). U.S. copyright law today considers a translation to be a derivative work. See 17 U.S.C. § 101 (2012) (defining “derivative work” to include translations). But see Stowe v. Thomas, 23 F. Cas. 201, 208 (C.C.E.D. Pa. 1853) (“A translation may, in loose phraseology, be called a transcript or copy of her thoughts or conceptions, but in no correct sense can it be called a copy of her book.”).

  41. [41]. Said, supra note 16, at 615–16.

  42. [42]. Id.

  43. [43]. See id. at 616, 650.

  44. [44]. Cf. Brown v. Allen, 344 U.S. 443, 540 (1953) (“We are not final because we are infallible, but we are infallible only because we are final.”) (Jackson, J., concurring in the result).

  45. [45]. Said, supra note 16, at 641, 644 (“By asking juries to answer questions about their experiences of a work of art, the court could determine whether what juries found similar about two works was in fact protectable or not.”).

  46. [46]. Alfred C. Yen, Copyright Opinions and Aesthetic Theory, 71 S. Cal. L. Rev. 247, 293 (1998) (“Explicit references to the attitudes of actual people [in the substantial similarity analysis] does direct a judge’s attention away from her own subjective appraisal of works, but answers will rarely be found because actual people will likely disagree over the interpretation of works.”).

  47. [47]. See Said, supra note 16, at 629.

  48. [48]. Id. (noting that an efferent reading, as an objective exercise, “can be wrong,” but an aesthetic reading, as a subjective process, is not likely to yield “a set of correct answers”; indeed, an aesthetic reading “need merely to chart an interpretive course that responds to cues in the text, rather than cues imagined by the reader”).

  49. [49]. Walter Benjamin, The Work of Art in the Age of Its Technological Reproducibility: Second Version, in The Work of Art in the Age of Its Technological Reproducibility, and Other Writings on Media 19, 22–31 (Michael W. Jennings et al. eds., 2008).

  50. [50]. The strong similarity of the melodic lines of “He’s So Fine” and “My Sweet Lord” does not mean that a performer’s set list would include either one interchangeably. See generally Bright Tunes Music Corp. v. Harrisongs Music, Ltd., 420 F. Supp. 177 (S.D.N.Y. 1976), aff’d sub nom. ABKCO Music, Inc. v. Harrisongs Music, Ltd., 722 F.2d 988 (2d Cir. 1983).

  51. [51]. See generally Michael S. Pardo, Group Agency and Legal Proof; Or, Why the Jury Is an “It,” 56 Wm. & Mary L. Rev. 1793 (2015) (discussing concept of jury as single entity with agency separate from that of its individual members).

  52. [52]. See Janice A. Radway, Reading the Romance: Women, Patriarchy, and Popular Literature 8 (1984) (“The analytic focus [of a cultural analysis of romance novels] must shift from the text itself, taken in isolation, to the complex social event of reading....”).

  53. [53]. See Said, supra note 16, at 607–10, 618 n.52.


Vice Dean and Professor of Law, William & Mary Law School.

Many thanks to Mark Badger and Andrew Gilden for their helpful comments.