“To expect a jury to understand legal rules in the same way and depth just because a judge patiently explains them borders on the fantastic.”

—Frank Upham[1]

I. Introduction

One of the high-profile copyright litigations of the 2010s—Oracle v. Google—has generated interest in a topic that previously received little attention from courts and commentators: the respective roles of judges and juries in fair use determinations. Finally reaching the Supreme Court in 2020, the Oracle v. Google litigation has centered on whether the application programming interfaces (APIs) of Oracle’s copyrighted Java programming language are protected by that copyright and, if they are protected by copyright, whether Google’s use of thirty-seven Java APIs was fair use.[2] An initial jury concluded that the APIs were protected by copyright and deadlocked on fair use; a trip to the Federal Circuit (because the case included patent claims) confirmed the jury’s copyrightability determination and remanded for the fair use question to be considered anew.[3] A second jury concluded that Google’s repurposing of the thirty-seven APIs was a fair use of Java, but the Federal Circuit disagreed, reversing the fair use determination and leaving Google liable for infringement.[4]

Because this is widely believed to be the first time that an appellate panel has reversed a jury’s finding of fair use,[5] and the court expressly did so on de novo review, the Federal Circuit’s ruling has triggered a lively discussion about the relative roles of judges and juries in fair use determinations—lively enough for the Supreme Court to order additional briefing on this issue.[6] Fair use is a mixed question of law and fact, so of course, there is the policy question of who resolves these mixed questions and how. But before the policy debate, there is the constitutional question: whether defendants have a Seventh Amendment right to jury determination of the overall fair use question. Presumably, if there is such a Seventh Amendment right, judges may only review such jury determinations on a “clearly erroneous” standard.

In any litigation, history can get simplified to the point of distortion, and people can stake out historical claims or arguments about institutional structure based on the substantive outcome they seek in a particular litigation. Both appear to have happened in Oracle v. Google on the question of the proper role of juries in fair use determinations. In its 2018 decision, the Federal Circuit had such an oversimplifying footnote that said, “Justice Joseph Story described fair use as a ‘question of fact to come to a jury’ in 1845.”[7]

In the service of arguing that Google’s use of the APIs was fair, the Supreme Court amicus brief of a respected trade association repeated this claim.[8] Both the Federal Circuit footnote and respected appellate litigators’ brief appear to have relied on an unusual reading of the case law from one commentator with all three—Federal Circuit footnote, amicus brief, and law review article—referring to Justice Story’s 1845 decision while riding circuit in Massachusetts, Emerson v. Davies.[9]

The problem is that Justice Story said no such thing in Emerson v. Davies. (We will get to what he did say below.)

What should be the proper standard for judicial review of jury fair use determinations, especially appellate review of jury determinations on fair use? This Article proposes that there is only a moderate case for a Seventh Amendment right to jury determination of § 107 fair use—and many junctures at which one could conclude that no such right exists. This Article then argues that, in the absence of a clear Seventh Amendment right, making fair use determinations “judge-centric” is both desirable to make fair use more predictable ex ante and consistent with the Supreme Court’s framework for mixed questions of law and fact in U.S. Bank National Ass’n v. Village at Lakeridge, LLC.[10]

II. Is There a Seventh Amendment Right to Jury Determination of Fair Use?

Do litigants have a right to jury determination of § 107 fair use under the Seventh Amendment? There is no question that an action for copyright infringement could be brought before a jury in eighteenth-century England, although there is also no question that most of the cases occurred in courts of equity.[11] And while Seventh Amendment analysis still turns on what was available to litigants in British and American courts before 1791, in modern Seventh Amendment jurisprudence, the question is not whether the cause of action could have been heard before a jury, but whether the “particular trial decision” or its “analogue” came before juries.[12]

Because fair use, as it is now codified in 17 U.S.C. § 107, did not exist before 1791, whether or not there is a Seventh Amendment right to have jury determination of fair use turns partly on what, if anything, is understood as the “analogue” of modern fair use in eighteenth-century English case law and partly on whether the Seventh Amendment threshold is met by that “analogue” sometimes coming before juries or regularly coming before juries. Like other commentators, I agree that the origin of fair use can be traced back to early English decisions under the Statute of Anne,[13] but I believe that there is no reliable basis to say that what we now understand to be questions of fair use went to juries at that time.

A. The Thin and Conflicting Historical Record

Professor Ned Snow is perhaps the most vocal advocate of the view that there is a Seventh Amendment right to jury determination of § 107 fair use. Snow claims that “[t]he early history of copyright law demonstrates that English common-law courts heard the issue of fair use, expressly reserving the issue for juries to determine”[14] and, more boldly, “[f]or over two centuries, courts have treated the issue of fair use in copyright law as a question for the jury.”[15] For anyone who learned that the fair use doctrine traces back to Justice Story’s 1841 decision in Folsom v. Marsh,[16] Snow’s historical claim is jarring. There is a case to be made that the Seventh Amendment guarantees jury determinations of fair use or elements thereof, but the case is much more equivocal than Professor Snow’s claim.

Early English cases—if filed in law, not equity—sent the general question of infringement to juries, sometimes framed in words and concepts that now sound like fair-use considerations. Professor Snow’s argument is that “fair use” was built into the primary finding of infringement, i.e. “[e]arly English copyright cases contemplated the sort of use that a defendant made of a work, deciding whether the use was permissible or not.”[17] In the Court’s Seventh Amendment jurisprudence, Snow would argue either that “fair use” was actually decided in these cases or that the eighteenth-century “analogue” to § 107 fair use was the infringement determination itself. To prove this, Professor Snow relies principally on language in two cases: Sayer v. Moore (1785)[18] and Cary v. Kearsley (1802), both from the King’s Bench.[19] (Of course, by the lights of Seventh Amendment jurisprudence, the 1802 decision should not count directly—it can only be evidence of what English courts might have been doing prior to 1791.)

Before we delve into what was actually said in these cases, it is helpful to remember that eighteenth-century English courts were divided into courts of law (the King’s Bench, Common Pleas, or Exchequer of Pleas) and courts of equity (Chancery).[20] A plaintiff could bring a copyright infringement action in a court of equity—which could award injunctive relief—and it appears most plaintiffs did just that.[21] Legal historians have yet to find any ruling or statement that a copyright infringement defendant in a court of equity had the right to remove the case to a court of law or to have any elements of the infringement action decided by a jury sitting in one of the courts of law.[22] Instead, as we will see below, it was just the opposite.

As to Sayer v. Moore and Cary v. Kearsley, both cases were before the King’s Bench (law), and both cases involved what we would now consider principally factual works—Sayer concerned sea charts and Cary concerned “road books [which] were an early form of road atlas.”[23] While there is much in these decisions that jives with our own thinking about copyright, there was also much going on that reflects both concerns incongruent with today’s range of copyrighted works[24] and ideas incompatible with our current doctrine.[25]

In Sayer, the plaintiffs had expended substantial monies in preparing their sea charts which used a “plain” projection, while defendant’s charts used the more modern Mercator projection.[26] From this difference alone, we know that the “expression” of the facts presented in the two sets of sea charts was quite different—indeed, an admiral of the Royal Navy testified to that effect.[27] Professor Snow quotes Lord Mansfield’s comments in Sayer as, “In all these [copyright cases,] the question of fact to come before a jury is, whether the alteration be colourable or not? . . . [T]he jury will decide whether it be a servile imitation or not.”[28] From those words, Snow concludes that Lord Mansfield thought “the issue of whether a defendant’s use was permissibly fair or impermissibly infringing to be one for the jury,” and that the case supports the argument that there is a Seventh Amendment right to overall determination of fair use.[29]

But one need not interpret Sayer v. Moore that way. Let us put to one side the complication that Lord Mansfield’s jury was a “special jury” comprised of “merchants who would be able to advise on the commercial aspects of the case.”[30] There is a complex history on how English courts used special juries and complex reasoning on how that can impact a Seventh Amendment analysis.[31] Separate from those issues, I believe that when a copyright scholar reads the entirety of what was testified and what Lord Mansfield said, it is evident that Mansfield’s prospective jury charge addressed what we now call the fact/expression dichotomy and the Arnstein-esque question of whether the defendant copied at all. Lord Mansfield says:

The Act that secures copy-right to authors guards against the piracy of the words and sentiments; but it does not prohibit writing on the same subject. As in the case of histories and dictionaries: in the first, a man may give a relation of the same facts, and in the same order of time; in the latter an interpretation is given of the identical same words. In all these cases the question of fact to come before a jury is, whether the alteration be colourable or not? [T]here must be such a similitude as to make it probable and reasonable to suppose that one is a transcript of the other, and nothing more than a transcript. So in the case of prints, no doubt different men may take engravings from the same picture. The same principle holds with regard to charts; whoever has it in his intention to publish a chart may take advantage of all prior publications. There is no monopoly of the subject here, any more than in the other instances; but upon any question of this nature the jury will decide whether it be a servile imitation or not.[32]

Mansfield says that when dealing with two factual works (“histories and dictionaries”), the jury must consider the possibility that independent creation produced similar works: “there must be such a similitude as to make it probable and reasonable to suppose that one is a transcript of the other, and nothing more than a transcript.”[33] In modern American copyright parlance, this is “probative similarity” as evidence of copying.[34] When Mansfield says “with regard to [sea] charts; whoever has it in his intention to publish a chart may take advantage of all prior publications,” he might have been addressing the copying of expression, but he definitely was addressing the copying of facts.[35]

And it is this language that Justice Story was quoting in Emerson v. Davies, i.e. the “question of fact to come to a jury” passage was just a very small portion of an approximately 130-word quotation of Lord Mansfield by Justice Story in a section of the Emerson opinion that was clearly about copying versus independent creation—exactly what Lord Mansfield was discussing in Sayer. To get a true sense of how much Justice Story was focused on the question of the defendant’s copying, not fair use, consider Story’s quotation of Lord Mansfield in context:

And the question then comes to this, whether [Davies] has, in substance, copied these pages, in plan, method, arrangement, illustrations and tables, from the plaintiff’s work, with merely colorable alterations and devices to disguise the copy, or whether the resemblances are merely accidental, and naturally or necessarily grew out of the objects and scheme of the defendant, Davies’s work, without any use of the plaintiff’s. If the defendant, Davies, had before him, at the time, the work of the plaintiff, and used it as a model for his own plan, arrangements, examples and tables, then I should say, following the doctrine of Lord Ellenborough, in Roworth v. Wilkes, that it was an infringement of the plaintiff’s copyright, notwithstanding the alterations and disguises in the forms of the examples and unit marks. Lord Mansfield, in Sayre v. Moore, cited 1 East, 361, 362, note, said: “In all these cases the question of fact to come to a jury, is, whether the alteration be colorable or not. There must be such a similitude as to make it probable and reasonable to suppose, that one is a transcript of the other, and nothing more than a transcript. So in the case of prints; no doubt different men may take engravings from the same picture. The same principle holds in regard to charts, that a man who has it in his intention to publish a chart, may take advantage of all prior publications. There is no monopoly in the subject here, any more than in the other instances. But upon a question of this nature the jury will decide, whether it be a servile imitation or not.”[36]

It is simply astonishing to say that Justice Story was talking about fair use in this passage, and those few words should not have been transmogrified—through a carefully worded law review footnote,[37] then a sloppy Federal Circuit footnote, then a distinguished appellate brief—into a proposition that was not Justice Story’s focus. I do not deny that some fair use considerations may have been in Mansfield’s thinking, but that should not overshadow how much the Sayer dispute was about the fact/expression dichotomy and the defendant having copied from the plaintiff little other than facts. It should not overshadow how much kindred issues (not fair use) were being discussed by Justice Story in the passage in question in Emerson v. Davies. Indeed, nowhere in Emerson v. Davies does Story say he himself believes that the test he set out in Folsom four years earlier was a “question of fact to come to a jury.”[38]

Similarly, the works at issue in the Cary v. Kearsley case were akin to road atlases (the plaintiff’s work was described as “an Itinerary, or Book of Roads”).[39] The defendant’s work had the same town names, same distances between towns, and even some errors that had appeared in the plaintiff’s work.[40] Lord Ellenborough noted that “in the defendant’s book there are additional observations, and in some part of the book I find corrections of misprinting,” and that “the defendant was authorised to use a work published as this of the plaintiff’s, to make extracts from it into any original work of his own”—a statement that makes a great deal of sense given the plaintiff’s work was largely a compilation of facts.[41] Indeed, that Lord Ellenborough was actually speaking to the extraction of facts from a fact-based work is supported by him comparing what the defendant did to “only using an erroneous dictionary.”[42] To the degree the plaintiff’s work had prose commentary, “it appeared that there was no entire particular paragraph transcribed.”[43]

Still, Lord Ellenborough and plaintiff’s counsel had a colloquy that definitely goes much further than someone extracting facts from a pre-existing work. When plaintiff’s counsel asks whether copying “a whole essay” would be a “piracy,” Lord Ellenborough directly says, “That would depend on the facts.”[44] But then he quickly deflects, switching away from whole copying to say “a man may fairly adopt part of the work of another.”[45] The comment does sound like a discussion of fair use considerations (“the question will be, Was the matter so taken used fairly”).[46]

Lord Ellenborough then made some comments about how two works compiling surveys and distances of various roads would be similar, saying he would convey all this thinking

to the jury, leaving them to say, whether what [was] taken or supposed to be transmitted from the plaintiff’s book, was fairly done with a view of compiling a useful book, for the benefit of the public, upon which there has been a totally new arrangement of such matter,—or taken colourable, merely with a view to steal the copy-right of the plaintiff[.][47]

Like Snow, Bill Patry believes that “[a]lthough Lord Ellenborough did not use the words ‘fair use,’ instead asking whether the matter was ‘used fairly,’ we find here the origins of fair use.”[48] But saying one detects the “origins” of fair use is not saying one detects “fair use,” just as saying amino acids are the origin of life is not saying a petri dish of amino acids is life.

As Patry correctly describes it, the interchange between counsel and bench in Cary is still “[o]n the basic question of infringement.”[49] Alexandra Sims goes further, finding “[i]t . . . curious that Cary v Kearsley is considered by some to represent the beginning of the fair dealing exception.”[50] In her careful “micro”-level “legal history” of the Cary case and its antecedents, copyright historian Isabella Alexander notes,

[I]t becomes clear that words like “originality” and “fair use” are simply not being used in the same way that we use them today. While this does not mean that the cases are of no value, it does mean that judicial uses of such words can be less easily deployed as precedent supporting contemporary argument.[51]

Indeed, modern courts still speak of prima facie infringement in terms that are also used in fair use analysis, i.e. “improper” appropriation and the metes and bounds of de minimis doctrine.[52] But that does not mean that a Ringgold inquiry into de minimis use[53] is a fair use analysis. Like me, Sims concludes that Cary “represents the beginning of a judicial recognition of fairness in relation to the use of factual materials in the creation of new works, but not fairness in the sense [of] using material for the purpose of review and criticism or even quotation.”[54]

Turning from these two cases, another way one could argue for a Seventh Amendment right to jury determination of fair use is through “fair abridgement” case law. As I said, everyone agrees that the question of copyright infringement could be brought before eighteenth-century courts of law and, therefore, juries; scholars also generally agree that the defense of “fair abridgement,” which sometimes arose in these cases, was the doctrinal ancestor to modern American fair use. Indeed, Folsom v. Marsh was a fair abridgement case.[55] Second, while fair abridgement was a judge-created doctrine (it does not appear in the Statute of Anne itself), it could well have been treated as an issue of fact that would have been decided by the jury when a copyright infringement claim was filed in the King’s Bench or Common Pleas during the eighteenth-century.[56] Third, one can believe that fair abridgement should be treated as an “analogue” to modern fair use for Seventh Amendment purposes.

Of course, one can question each step in this chain of reasoning. First, it is not clear that prior to 1790 there was any regular practice of juries deciding fair abridgement. The early case law is complicated by the fact that most actions were brought in equity (before the Chancery court), not in courts of law. As I said earlier, legal historians have not found any ruling or statement that an infringement defendant in a court of equity had the right to remove the case to a court of law or to have any elements of the infringement action decided by a jury.

As an equity court, Chancery could not empanel a jury, but a defendant sued in the Chancery court could demand that factual issues involving witnesses be sent to be tried before a jury in a court of law.[57] That sort of request is what happened in the 1740 Chancery case Gyles v. Wilcox, Barrow, and Nutt concerning a work that had, verbatim, copied a prior law book with “only some old statutes hav[ing] been left out” and Latin and French quotations being translated into English.[58] The Lord Chancellor noted that “[w]here books are colourably shortened only,” that would be an infringement of the copyright, whereas “a real and fair abridgement” that showed new “invention, learning, and judgment” would not be.[59] Confronted with a request to send the factual issues to a jury, the Lord Chancellor said:

Mr. Attorney General has said I may send it to law to be determined by a jury; but how can this possibly be done? [I]t would be absurd for the chief justice to sit and hear both books read over, which is absolutely necessary, to judge between them, whether the one is only a copy from the other.

The court is not under an indispensable obligation to send all facts to a jury, but may refer them to a master, to state them, where it is a question of nicety and difficulty, and more fit for men of learning to inquire into, than a common jury.[60]

The Lord Chancellor concluded that “rather than leave it to a jury” it was better to send the infringement versus fair abridgment problem to “two persons of learning and abilities in the profession of the law.”[61]

Again, this was a case of a court sitting in equity, and it does not directly show what a court of law would have done or felt compelled to do; it shows only one of England’s most learned jurists opining that with a copyright infringement case, even in the face of a fair abridgment defense and a request from defendant’s counsel, “The court [wa]s not under an indispensable obligation to send all facts to a jury.”[62] The Gyles v. Wilcox decision was taken up by Patrick Devlin in his widely cited analysis of how English courts allocated issues to juries in complex litigations before 1791.[63] Considered in this broader context, Professor Devlin concluded that the Lord Chancellor would have made the same decision “if there had been an action for damages instead of a bill for an injunction to stay the publication.”[64] It appears that copyright historians have not uncovered any record that would establish “when (if ever) the Chancery was obligated to send a case to law”[65]—and, therefore, a jury.

Second, even if the question of fair abridgment before a law court was typically decided by juries, one could reason that fair abridgment is an ancestor of fair use, without being what the Supreme Court calls a Seventh Amendment “analogue.”[66] Lastly, assuming fair abridgment is a Seventh Amendment “analogue” for fair use, one could argue that it is not enough that fair abridgment would have been decided by a jury when the action was filed in the King’s Bench or Common Pleas. Can we show that by 1791 there was any established practice of this happening, particularly when the bulk of eighteenth-century copyright claims in England were filed in courts of equity? All this makes for an equivocal case that there is a Seventh Amendment right to jury determination of the overall fair use inquiry.

B. Analysis Under the Court’s Seventh Amendment Jurisprudence

How the Seventh Amendment argument plays out also depends partly on the choice between a couple of prior occasions when the Seventh Amendment met intellectual property: the Court’s 1998 Feltner v. Columbia Pictures decision[67] and its 1996 Markman v. Westview Instruments decision.[68]

In Feltner, the Court considered whether there is a Seventh Amendment right to jury determination of statutory damages under 17 U.S.C. § 504(c).[69] The Court first noted that English “copyright suits for monetary damages were tried in courts of law, and thus before juries,”[70] both under the seventeenth-century common law and the 1710 Statute of Anne, which, itself, established statutory damages.[71] Justice Thomas then noted that some of the pre-1791 American copyright statutes specified that infringement actions would be tried at “law” or in an “action of debt,” both of which would go to a jury—and that three of those state statutes had statutory damage provisions.[72] This evidence, coupled with one 1789 Connecticut decision in which a jury awarded the copyright owner £100 (under a statute with statutory damages),[73] led the Court to conclude that “[t]he right to a jury trial includes the right to have a jury determine the amount of statutory damages, if any, awarded to the copyright owner.”[74] This contrasts with the Court’s determination a couple years later that appellate review of a jury’s determination of punitive damages “does not implicate . . . Seventh Amendment concerns,” because “the jury’s award of punitive damages does not constitute a finding of ‘fact.’”[75]

Two years before Feltner, the Court had engaged in a more elaborate Seventh Amendment analysis in Markman v. Westview Instruments, Inc.[76] The Markman Court concluded that the proper interpretation of a patent’s claims prior to the patent infringement trial “is exclusively within the province of the court.”[77] To do this, the Court had to conclude that patent claim construction was not a “guaranteed jury issue” under the Seventh Amendment.[78] The 1996 Markman decision highlights one critical piece of the Seventh Amendment machinery: while the Seventh Amendment guarantees “[t]he right of trial by jury thus preserved is the right which existed under the English common law when the Amendment was adopted,”[79] that does not mean every issue in such a trial must be decided by the jury. As the unanimous Markman Court noted:

[W]e ask, first, whether we are dealing with a cause of action that either was tried at law at the time of the founding or is at least analogous to one that was. If the action in question belongs in the law category, we then ask whether the particular trial decision must fall to the jury in order to preserve the substance of the common-law right as it existed in 1791.[80]

A few years later—and after Feltner—the Court restated the Markman Seventh Amendment inquiry in City of Monterey v. Del Monte Dunes: once the cause of action is one that would go to a jury, a court “must determine whether the particular issues of liability were proper for determination by the jury”[81] and such “issues that are proper for the jury must be submitted to it ‘to preserve the right to a jury’s resolution of the ultimate dispute.’”[82] This was in keeping with the Court’s reasoning in earlier cases that “the question whether a jury must determine [some issue] in a trial in which it must determine liability” turns on “whether the jury must shoulder th[at] responsibility as necessary to preserve the ‘substance of the common-law right of trial by jury’”[83] and that the “Amendment was designed to preserve the basic institution of jury trial in only its most fundamental elements.”[84] The City of Monterey v. Del Monte Dunes Court elaborated:

We look to history to determine whether the particular issues, or analogous ones, were decided by judge or by jury in suits at common law at the time the Seventh Amendment was adopted. Where history does not provide a clear answer, we look to precedent and functional considerations.[85]

In Del Monte Dunes, the Court could “find no precise analogue for the specific test of liability submitted to the jury in this case.”[86] It then turned to precedent and also came up empty-handed.[87]

Retreating to “functional” considerations, the Del Monte Dunes Court provided an analysis with intriguing parallels to the fair use question. The Court concluded that it had historically “described determinations of liability in regulatory takings cases as ‘essentially ad hoc, factual inquiries,’ requiring ‘complex factual assessments of the purposes and economic effects of government actions.’”[88] Given the “particular trial decision”[89] was an “essentially ad hoc, factual inquiry” the majority decided that it was a matter for the jury.[90] At the same time, the Court recognized that the overall question of a regulatory taking “is probably best understood as a mixed question of fact and law,” and that it was only deciding—with the particular trial history before it—“that it was proper to submit this narrow, fact-bound question to the jury.”[91] The Court made clear that it was not addressing Seventh Amendment rights in relation to the broader mixed question of fact and law involved in a regulatory taking.[92]

So, what do we make of these precedents in determining whether there is a Seventh Amendment guaranteed jury right for § 107 fair use determinations? In Feltner, the Court’s analysis was fairly straightforward: it concluded that pre-1790 juries were deciding the same “particular trial decision”—statutory damages claims in copyright disputes—based on the statutes.[93] There was no need to fall back on the other tests. So, if the Supreme Court could be convinced that “fair use” determinations were embedded in eighteenth-century jury infringement verdicts, the Court might decide things along a simple Feltner line of reasoning. That seems to be what Professor Snow is seeking.[94] But the eighteenth-century English and nineteenth-century American cases do not clearly establish this. Even if Cary v. Kearsley was a “fair use” case, the dispute was nonsuited and never went to a jury; the judges’ statements in Cary and Sayer were statements of individual judges, not something said or adopted by the King’s Bench court.[95] And in Gyles v. Wilcox, the Lord Chancellor was decidedly against having the jury decide issues that one might try to shoehorn into “fair use.”[96]

It is also worth noting that in Markman the Supreme Court showed its discomfort with a parallel “embeddedness” argument, e.g. that construal of a patent’s claims was implicit with a jury determination of patent infringement:

Markman seeks to supply what the early case reports lack in so many words by relying on decisions like Turner v. Winter, 1 T. R. 602, 99 Eng. Rep. 1274 (K. B.1787), and Arkwright v. Nightingale, Dav. Pat. Cas. 37 (C. P. 1785), to argue that the 18th-century juries must have acted as definers of patent terms just to reach the verdicts we know they rendered in patent cases . . . . But the conclusion simply does not follow.[97]

In other words, it may be an anachronism to think that a modern test or step—claim construction, fair use—was necessarily embedded in the old decisions.

If the Court is not convinced that § 107 fair use was necessarily embedded in eighteenth-century jury verdicts, then the parallels to Markman and Del Monte Dunes become interesting. By 1781, patent documents included detailed descriptions of the invention that we would now identify as “specifications”[98]—but without discrete, specific claims. So, turning to “analogue[s],” the Markman Court concluded:

The closest 18th-century analogue of modern claim construction seems, then, to have been the construction of specifications, and as to that function the mere smattering of patent cases that we have from this period shows no established jury practice sufficient to support an argument by analogy that today’s construction of a claim should be a guaranteed jury issue. Few of the case reports even touch upon the proper interpretation of disputed terms in the specifications at issue, and none demonstrates that the definition of such a term was determined by the jury.[99]

The Court concluded that judges were interpreting early patent documents, just as judges interpreted written documents generally, so any analogue to modern claim construction was handled by judge, not jury.[100] This might provide the basis to distinguish the fair use issue in Oracle v. Google from Markman, but reasonable minds can differ on where this leads. Unlike determining the metes and bounds of patent claims, fair use determinations are not principally about interpreting written documents; on the other hand, fair use determinations also rarely turn on the credibility of witnesses, where the jury’s role is preeminent.

Just as the notion of “claims” in a patent did not exist in 1791, § 107 fair use doctrine had not crystallized as of 1791. If the Seventh Amendment question is whether the evidence shows an “established jury practice sufficient to support an argument by analogy,” we are stuck with conflicting evidence on whether the pre-1791 copyright issues arguably analogous to fair use were thought of as best handled by jury (Cary v. Kearsley) or judge (Gyles v. Wilcox). Stronger evidence of judges handling an analogue to § 107 fair use determinations arguably starts in the early, middle nineteenth-century, when Justice Story presides over Folsom.[101]

Alternatively, if we do not have enough pre-1791 evidence that juries were deciding anything analogous to fair use, perhaps we are thrown to a situation like Del Monte Dunes where we must “look to precedent and functional considerations.”[102] This is where Del Monte’s observations about mixed questions of law and fact—and preserving jury determinations of “narrow, fact-bound question[s]”—has promising implications for the mixed question of fair use.[103] Del Monte suggests endorsement of the modern § 107 fair use framework guaranteeing a right to jury determination of all the facts going into fair use analysis while preserving de novo appellate review of the final determination of law.

Intuitions will differ as to the right answer to these questions. For myself, I agree with Alexander that with close attention to legal history “[w]e ought therefore to be wary of statements which purport to have traced a constant line of authority across 200 years”[104] and that whatever the legal history answer is, it will not be as convincing as the other policy arguments for de novo appellate review.

III. A Mixed Question of Fact and Law—and (Almost) Everyone Agrees We Would Benefit from Greater Clarity

The question of appellate review of jury determinations of fair use should also be considered within the general framework for appellate review of mixed questions of fact and law that the Court provided in its 2018 U.S. Bank National Ass’n v. Village at Lakeridge, LLC decision.[105] The U.S. Bank decision gave us an analysis that is likely to be grist for years of commentary and discussion:

Mixed questions are not all alike. . . . [S]ome require courts to expound on the law, particularly by amplifying or elaborating on a broad legal standard. When that is so—when applying the law involves developing auxiliary legal principles of use in other cases—appellate courts should typically review a decision de novo. But . . . other mixed questions immerse courts in case-specific factual issues—compelling them to marshal and weigh evidence, make credibility judgments, and otherwise address what we have (emphatically if a tad redundantly) called ‘multifarious, fleeting, special, narrow facts that utterly resist generalization.’ And when that is so, appellate courts should usually review a decision with deference.[106]

There is no slight in calling this an “it all depends” approach: in summarizing its analysis, the Court said, “In short, the standard of review for a mixed question all depends—on whether answering it entails primarily legal or factual work.”[107]

Not surprisingly, this framework can point in opposite directions when it comes to fair use. On the one hand, courts repeatedly tell us that fair use determinations are “case-by-case.”[108] Focusing on that, one could conclude that the U.S. Bank test “suggests a deferential review” of jury, fair-use determinations.[109] But there is also no question that fair-use jurisprudence is a body of “auxiliary legal principles” developed through precedent, suggesting de novo review of the overall fair-use conclusion.

Do fair-use determinations “entail[] primarily legal or factual work”?[110] Overall, fair-use determinations are not as “factual sounding” as the question at issue in U.S. Bank (whether a transaction was an arm’s length transaction between strangers);[111] a fair-use determination is a legal conclusion more than a “factual inference[] from undisputed basic facts.”[112]

Perhaps a better approach is to ask how much we want or need guiding principles going forward to answer the sort of question the court is confronting. On this count, the mixed question in U.S. Bank had not been the subject of much juridical analysis: the Court noted that “[o]ur own decisions, arising in a range of contexts, have never tried to elaborate on the established idea of a transaction conducted as between strangers; nor, to our knowledge, have lower courts.”[113] That description is the opposite of fair-use jurisprudence, where the Court’s decisions have elaborated extensively on the statutory framework of fair use—as have the appellate courts over and over and over again. In other words, in U.S. Bank, a deferential standard was appropriate because “there [wa]s no apparent need to further develop ‘norms and criteria,’ or to devise a supplemental multi-part test, in order to apply the familiar term,”[114] but with fair use cases—from the Court’s 1984 Sony v. Universal City Studios decision[115] through the Second Circuit’s recent Fox News v. TVEyes case[116]—there has very much been the need to further develop norms, criteria, and considerations.[117]

In Acuff-Rose, the very Supreme Court opinion in which the majority announced that fair use determinations are “not to be simplified with bright-line rules,”[118] Justice Kennedy’s concurrence candidly recognized that “[t]he common-law method instated by the fair use provision of the copyright statute presumes that rules will emerge from the course of decisions.”[119] From this perspective, de novo review of overall fair use determinations is proper because of the systemic need to “clarify legal principles” and “provide guidance to other courts resolving other disputes.”[120]

Developing and clarifying legal doctrine has provided a justification for de novo review in other areas of the law. In its 1991 Salve Regina College v. Russell decision, the Court cited situations in which “probing appellate scrutiny will . . . contribute to the clarity of legal doctrine” as situations that could warrant de novo review.[121] In its 2018 Straits Financial LLC v. Ten Sleep Cattle Co., the Seventh Circuit gave another U.S. Bank-inspired formulation of the problem: “deference ends when the district court ventures beyond ‘case-specific factual issues’ and reaches broader legal conclusions applicable to a whole class of future cases.”[122] In another post-U.S. Bank case, the Ninth Circuit was confronted with the question “whether any exception exists to the general rule that a corporate records custodian may not assert a Fifth Amendment privilege to refuse production of corporate documents.”[123] The appellate panel concluded that this was a legal question, but noted that even if it were a mixed question of fact and law, the court “would nonetheless review the matter de novo because ‘applying the law [would] involve[] developing auxiliary legal principles of use in other cases.’”[124]

It seems virtually impossible to deny that the fair use case law consciously develops “auxiliary legal principles” that contribute to ex ante clarity on copyright liability and “provide guidance to other courts resolving other disputes.”[125] If it was not widely believed that the fair-use determination in Oracle v. Google involves “broader legal conclusions applicable to a whole class of future cases” and “develop[s] ‘auxiliary legal principles of use in other cases,’”[126] the Supreme Court would not have eleven amicus briefs filed on behalf of at least forty individuals, companies, and organizations addressing the fair-use question substantively (separate from all the other amicus briefs in the case).[127]

The U.S. Bank Court also noted that deference is appropriate when the jury or trial judge’s determination involves facts that “utterly resist generalization.”[128] But that is not what happens in a fair-use analysis. Fair-use analysis relies on placing facts into categories that are generalizations: “criticism,” “comment,” “educational use,” “parody,” “unpublished,” “factual works,” and “commercial use” are just some of the generalizations we use in a fair-use inquiry in relation to the basic facts.[129]

Maintaining and developing those generalizations (and what facts fall into each) helps make fair use reliable to both users and creators ex ante; those generalizations keep fair use from descending completely into an “ad hoc” approach[130] so “vague”[131] and “unpredictable”[132] as to threaten the status of copyright as a form of property.[133] And, as Morris Arnold has noted, “legal systems which delegate law-finding functions to laymen cannot be anxious about producing a systematic substantive law, since they give professionals so few occasions to review the legal significance of facts.”[134]

How corrosive to systematic, predictable fair use would jury determinations be if such determinations were scrutinized only for clear error? Obviously, we do not know. David Nimmer has commented that jury determinations of copyright statutory damages post-Feltner have “debunked [his] fears” of jury inconsistency and that “when juries have been called upon to award statutory damages outside of the music performance context, they have proven able to produce figures that appellate courts ultimately approve.”[135] But the fair use doctrine serves a very different purpose than damage calculations: few people shape their behavior ex ante based on the amount of statutory damages likely to be awarded, but we hope and believe that people do shape their behavior ex ante based on whether or not they will be legally liable for copyright infringement—and knowing that often depends on some confidence about the contours of fair use.

At the end of the day, this may be the crux of the deference-or-de novo debate in relation to fair use. If one views fair use as a “highly subjective inquiry,”[136] then deference to jury and trial court decisions may best serve one’s vision of fair use. If one expects fair use to be “exceedingly difficult to predict”[137] and “so case-specific that it offers precious little guidance about its scope to artists, educators, journalists, Internet users, and others,”[138] then one might lessen appellate oversight of jury and trial judge conclusions on fair use. To paraphrase the language of a 1988 Supreme Court decision upholding an abuse-of-discretion standard of review, if one believes that fair use “questions that arise in litigation are not amenable to regulation by rule because they involve multifarious, fleeting, special, narrow facts that utterly resist generalization,”[139] then deference to jury determinations makes more sense.

In contrast, if one wants a fair use ecosystem in which there are some areas of more objective ex ante rule-like norms exempting certain reasonably defined categories of behavior from copyright liability—if one believes (as Justice Kennedy clearly did) that much of the conduct exempted by § 107 fair use is “amenable to regulation by rule”—one is likely to prefer de novo review as a means to clarify what is and is not fair use.[140]

IV. Conclusion

Lawyers often make procedural or systemic arguments when they believe particular procedures or organizational principles will produce the substantive outcome they want at that moment.[141] One gets the sense that this is what has happened in Oracle v. Google: the appellate court’s reversal of a jury finding of fair use has triggered a cadre of people insisting that juries should decide fair use—when one suspects that the same folks would be preaching a different gospel, or be noticeably silent, had circumstances been reversed (the jury found against fair use and the judges overrode that conclusion in favor of fair use). Indeed, that is what happened when a judge reversed a jury conclusion against fair use in a 2017 district court case[142]—hardly anyone noticed.

In 2010, copyright scholar Ronan Deazley wrote “definitive readings as to the meaning, impact, and legacy of the Statute of Anne, and the case law that it prompted, remain as elusive as ever.”[143] Perhaps that is an overstatement—there are things we definitely know about eighteenth-century copyright—but it does speak to the modesty with which one should approach the historical record. Considering what English jurists said and did in 1740, 1774, 1785, and 1802, there is probably no clear answer on whether there should be Seventh Amendment right to jury determination of modern § 107 fair use.

Meanwhile, a strong case can be made that within the framework for mixed question of fact and law that the Court gave us in U.S. Bank National Ass’n v. Village at Lakeridge, LLC, fair use is a mixed question where “probing appellate scrutiny will . . . contribute to the clarity of legal doctrine”[144] whereas deference to jury determinations would likely exacerbate the “significant ex ante uncertainty” of fair use.[145]

  1. Frank Upham, Mythmaking in the Rule-of-Law Orthodoxy, in Promoting the Rule of Law Abroad 75, 87 (Thomas Carothers ed., 2006).

  2. Oracle Am., Inc. v. Google LLC, 886 F.3d 1179, 1185–86 (Fed. Cir. 2018), cert. granted, 140 S. Ct. 520 (mem.) (2019).

  3. Id. at 1185.

  4. Id.

  5. Reply Brief for the Petitioner at 10, Google LLC v. Oracle Am., LLC, 140 S. Ct. 520 (mem.) (2019) (No. 18-956), 2019 WL 2024847, at *10 (“It is essentially unheard of for a court of appeals to reverse a jury’s finding of fair use under the Copyright Act.”).

  6. Google LLC v. Oracle Am., Inc., 140 S. Ct. 2737 (mem.) (May 4, 2020) (“The parties [are] directed to file supplemental letter briefs addressing the appropriate standard of review for the second question presented, including but not limited to the implications of the Seventh Amendment, if any, on that standard.”); see also Jonathan Band, Supreme Court Orders Additional Briefing on Standard of Review in Google v. Oracle, Disruptive Competition Project (May 4, 2020), http://www.project-disco.org/intellectual-property/050420-supreme-court-orders-additional-briefing-on-standard-of-review-in-google-v-oracle/ [https://perma.cc/AV2Q-858K].

  7. Oracle, 886 F.3d at 1194 n.3.

  8. Brief of Amicus Curiae American Intellectual Property Law Ass’n in Support of Neither Party at 26, Google LLC v. Oracle Am., Inc., 140 S. Ct. 2737 (mem.) (May 4, 2020) (No. 18-956), 2020 WL 224321, at *26.

  9. Emerson v. Davies, 8 F. Cas. 615, 623–24 (C.C.D. Mass. 1845) (No. 4,436).

  10. U.S. Bank Nat’l Ass’n v. Vill. at Lakeridge, LLC, 138 S. Ct. 960, 965, 967 (2018).

  11. H. Tomás Gómez-Arostegui, What History Teaches Us About Copyright Injunctions and the Inadequate-Remedy-at-Law Requirement, 81 S. Cal. L. Rev. 1197, 1222–23, 1273–74 (2008) (discussing how copyright suits arose in courts of chancery after 1660 and, later, in courts of law in the mid-1700s—after the passage of the Statute of Queen Anne); Isabella Alexander, Sayer v Moore (1785), in Landmark Cases in Intellectual Property Law 59, 66 (Jose Bellido ed., 2017) [hereinafter Alexander, Sayer] (in the 1780s, “the Court of Chancery tended to be the forum of choice for copyright plaintiffs of the time”); H. Tomás Gómez-Arostegui, Equitable Infringement Remedies Before 1800, in 81 Research Handbook on the History of Copyright Law 195, 196 (Isabella Alexander & H. Tomás Gómez-Arostegui eds., 2016) [hereinafter Gómez-Arostegui, Equitable Infringement] (“For most litigants before the year 1800, the sine qua non of a copyright or printing-patent suit was the injunction.”).

  12. See infra Sections II.A, II.B.

  13. The Statute of Anne; April 10, 1710: 8 Anne, c. 19 (1710), Avalon Project, https://avalon.law.yale.edu/18th_century/anne_1710.asp [https://perma.cc/E3JG-5QD5] (last visited Nov. 2, 2020).

  14. Ned Snow, Who Decides Fair Use—Judge or Jury?, 94 Wash. L. Rev. 275, 279 (2019).

  15. Id. at 276.

  16. Folsom v. Marsh, 9 F. Cas. 342, 349 (C.C.D. Mass. 1841) (No. 4,901).

  17. Snow, supra note 14, at 292.

  18. See id.; Sayre v. Moore (1785) 102 Eng. Rep. 138; 1 East 358 (K.B.). The Sayre case is reported in a footnote to Cary v. Longman (1801) 102 Eng. Rep. 138; 1 East 358 (K.B.). The spelling of Sayre/Sayer is contested; many think “Sayer” is the (more) correct form.

  19. Cary v. Kearsley (1803) 170 Eng. Rep. 679; 4 Esp. 168 (K.B.).

  20. The Chancery court also had some jurisdiction as a court of law called “petty bag,” but (a) the plaintiff or defendant would need to be an officer of the Chancery (a kind of jurisdictional privilege held by officers of Chancery court) and (b) the Chancery court had “no power to summon a jury.” 3 J.H. Thomas, A Systematic Arrangement of Lord Coke’s First Institute of the Laws of England, at xiv–xv, 328 n.109 (Hargrave et al. eds., 1818).

  21. Alexander, Sayer, supra note 11, at 66.

  22. E-mail from H. Tomás Gómez-Arostegui, IP Fac. Scholar & Professor of L., Lewis & Clark L. Sch., to Justin Hughes, Professor of L., Loy. L. Sch. (Jan. 10, 2020, 01:46:05 EST) (on file with author).

  23. Isabella Alexander, 'Manacles upon Science’: Re-Evaluating Copyright in Informational Works in Light of 18th Century Case Law, 38 Melbourne U. L. Rev. 317, 326 (2014) (exploring a line of cases ending in Cary v. Kearsley and noting that the litigations concerned “[r]oad books [which] were an early form of road atlas”); Alexander, Sayer, supra note 11, at 66.

  24. For example, Isabella Alexander points out that the copyright disputes about maps and sea charts often seemed to turn on “who had produced the ‘better’ map—and better meant more accurate, more scientific in appearance, more in tune with Enlightenment ideology and more useful to the commercial and military needs of the expanding empire.” Alexander, Sayer, supra note 11, at 80–81. That dynamic does not “map” easily onto our current copyright disputes about music, film, fashion, television shows, etc.

  25. One of those doctrines being, for example, the sweat of the brow justifications for copyright. See id. at 79 (explaining that, in the 1793 Heather v. Moore case, counsel told the jury that copyright law was intended to “secure to ‘persons who had been at great labour and expence, the fruits of their labour’”).

  26. Sayre v. Moore (1785) 102 Eng. Rep. 138, 140; 1 East 358, 362 (K.B.).

  27. On behalf of the defendant, Admiral Campbell testified

    that there were only two kinds of charts, one called a plain chart, which was now very little used; the other, which is the best, called the Mercator, and which is very accurate in the degrees of latitude and longitude. That this distinction was very necessary in the higher latitudes, but in places near the Equator it made little or no difference. That the plaintiffs’ maps were upon no principle recognized among seamen, and no rules of navigation could be applied by them; and they were therefore entirely useless.

    Id.; see also Alexander, Sayer, supra note 11, at 67 (describing testimony of witnesses as to differences).

  28. Snow, supra note 14, at 276 n.1 (quoting Sayre, 102 Eng. Rep. at 140).

  29. Id. at 294–95.

  30. Alexander, Sayer, supra note 11, at 66. Lord Mansfield was especially keen on the use of such special juries composed of experts from the relevant commercial field. See James Oldham, English Common Law in the Age of Mansfield 20–22 (2004).

  31. See, e.g., James C. Oldham, The Origins of the Special Jury, 50 U. Chi. L. Rev. 137 (1983); Patrick Devlin, Jury Trial of Complex Cases: English Practice at the Time of the Seventh Amendment, 80 Colum. L. Rev. 43 (1980).

  32. Sayre, 102 Eng. Rep. at 140 (emphasis added).

  33. Id.

  34. See, e.g., Johnson v. Gordon, 409 F.3d 12, 17–18 (1st Cir. 2005) (“[A] sufficient degree of similarity exists between the copyrighted work and the allegedly infringing work to give rise to an inference of actual copying. We have referred to that degree of similarity as ‘probative similarity.’” (quoting Lotus Dev. Corp. v. Borland Int’l, Inc., 49 F.3d 807, 813 (1st Cir. 1995), aff’d, 516 U.S. 233 (1996)); Peel & Co. v. Rug Mkt., 238 F.3d 391, 394 (5th Cir. 2001) (“As direct evidence of copying is rarely available, factual copying may be inferred from (1) proof that the defendant had access to the copyrighted work prior to creation of the infringing work and (2) probative similarity.”); Alan Latman, “Probative Similarity” as Proof of Copying: Toward Dispelling Some Myths in Copyright Infringement, 90 Colum. L. Rev. 1187 (1990).

  35. Sayre, 102 Eng. Rep. at 140.

  36. Emerson v. Davies, 8 F. Cas. 615, 618, 623–24 (C.C.D. Mass. 1845) (No. 4,436).

  37. Snow, supra note 14, at 276 n.1 (“Emerson v. Davies, 8 F. Cas. 615, 623–24 (C.C.D. Mass. 1845) (No. 4,436) (Story, J.) (describing fair use as a ‘question of fact to come to a jury’ (quoting Sayre, 102 Eng. Rep. at 139 n.(b))) . . . .”).

  38. Emerson, 8 F. Cas. 615.

  39. Cary v. Kearsley (1802) 170 Eng. Rep. 679, 679; 4 Esp. 168, 168 (K.B.).

  40. Id. (explaining that the plaintiff “proved that several mistakes and errors, which had crept into the plaintiff’s book in the printing, were copied verbatim into the defendant’s”).

  41. Id.

  42. Id. at 679, 4 Esp. at 169.

  43. Id. at 680, 4 Esp. at 169.

  44. Id. at 680, 4 Esp. at 170 (emphasis added).

  45. Id. (emphasis added).

  46. Id.

  47. Id. at 680, 4 Esp. at 170–71.

  48. William F. Patry, Patry on Fair Use § 1:11, at 41 (2019 ed. 2019).

  49. Id. And Patry does not use Cary v. Kearsley to prove that the issue of “fair use” went to juries. See id. at 40–42. While there was a jury in the Cary case, it never actually reached a verdict, because “[t]he counsel for the Plaintiff consented to be nonsuited.” Cary, 170 Eng. Rep. at 680; 4 Esp. at 171.

  50. Alexandra Sims, Appellations of Piracy: Fair Dealing’s Prehistory, 2011 Intell. Prop. Q. 3, 21.

  51. Alexander, supra note 23, at 324–25, 332, 352.

  52. See Ringgold v. Black Ent. Television, Inc., 126 F.3d 70, 74–75 (2d Cir. 1997); Castle Rock Ent., Inc. v. Carol Publ’g Grp., Inc., 150 F.3d 132, 137 (2d Cir. 1998) (quoting Laureyssens v. Idea Grp., Inc., 964 F.2d 131, 139–40 (2d Cir. 1992)).

  53. See Ringgold, 126 F.3d at 76–77 (holding that defendant’s use of a poster of a quilt as the background in a television show was not de minimis).

  54. Sims, supra note 50, at 21.

  55. Matthew Sag, The Prehistory of Fair Use, 76 Brook. L. Rev. 1371, 1375 (2011) (citing Folsom v. Marsh, 9 F. Cas. 342 (C.C.D. Mass. 1841) (No. 4,901)). Matthew Sag argues that the fair use doctrine should be “understood as the continuation of a long line of English fair abridgment cases, dating back to the beginning of statutory copyright law in 1710.” Id. at 1373. Sag highlights the “substantial continuity between fair abridgment in the premodern era and fair use in the United States today.” Id. However, Sag himself notes that the transition from fair abridgment in mid-1700s to the modern fair use doctrine was both “gradual” and “haphazard.” Id. at 1411. He also warns that “[d]espite the continuity and congruity between the premodern and contemporary understanding[s,] . . . the parallels should not be overstated.” Id. at 1404.

  56. Id. at 1380–81. On this step, neither Sayer nor Cary were actually fair abridgement cases, so they are no help. See Sayre v. Moore (1785) 102 Eng. Rep. 138, 139–40 n.(b); 1 East 358, 361–62 (K.B.); Cary v. Kearsley (1802) 170 Eng. Rep. 679, 679–80; 4 Esp. 168, 168–71 (K.B.).

  57. 1 W.S. Holdsworth, A History of English Law 452 (3d ed. 1922) (“If the parties descend to issue, this court (the Chancery) cannot try it by jury, but the Lord Chancellor . . . delivereth the record by his proper hands into the King’s Bench to be tried there . . . and after trial had, to be remanded into the Chancery, and there judgment to be given.” (alteration in original) (quoting Edwardo Coke, The Fourth Part of the Institutes of the Laws of England; Concerning the Jurisdiction of Courts. 80 (The Lawbook Exch., Ltd. 2002) (1817))).

  58. Glyes v. Wilcox (1740) 26 Eng. Rep. 489, 489, 491; 2 Atk. 141, 144 (Ch.).

  59. Id. at 490, 2 Atk. at 143.

  60. Id. at 490–91, 2 Atk. at 143–44.

  61. Id.

  62. Id. at 489–90, 2 Atk. at 143–44.

  63. Devlin, supra note 31, at 72–73. Court citations to the Devlin article include the U.S. Supreme Court in its Seventh Amendment decision in Markman discussed infra Section II.B; see also Markman v. Westview Instruments, Inc., 517 U.S. 370, 382 n.7 (1996).

  64. Devlin, supra note 31, at 73 (“[The Lord Chancellor] refused a jury. It is difficult to believe that if there had been an action for damages instead of a bill for an injunction to stay the publication, the Chancellor would not have reached the same conclusion. To send to a jury for comparison two learned and legal books would be farcical.”).

  65. Gómez-Arostegui, Equitable Infringement, supra note 11, at 211–12. Based on his extensive review of pre-1800 English cases, Gómez-Arostegui concludes that where Chancery sent a factual dispute to a court of law, “[i]f an issue of fact alone was involved, then the case would usually return to the Chancery soon after a verdict,” . but if there was an unresolved question of the law, “the [jury] verdict would then be followed by an argument decided at the ‘Bar’ by all the judges of the particular court.” Id. at 212. As one 1799 lawyer’s guide to the King’s Bench [a law court] put it, “it has become the practice for the jury, when they have any doubt as to the matter of law, to find a special verdict, stating the facts, and referring the law arising thereon to the decision of the court.” 2 William Tidd, The Practice of the Court of King’s Bench in Personal Actions 799 (London, 2d ed. 1799). This suggests that Google would have had a Seventh Amendment right to specific jury findings on fair use factors, although if the Oracle v. Google jury had made specific findings on the first and fourth § 107 factors that were favorable to Google, it seems likely that the Federal Circuit panel would have reversed such findings on a clearly erroneous standard. See Oracle Am., Inc. v. Google LLC, 886 F.3d 1179, 1204, 1210 (Fed. Cir. 2018).

  66. Intellectual property historians have noted the discontinuity between early and modern concepts. For example, Sag notes,

    [Brad] Sherman and [Lionel] Bently draw the distinction between modern and premodern copyright, using 1850 as a rough dividing line. . . . The premodern-modern distinction is useful in this context, as the categories of intellectual-property law were quite fluid up until the mid-nineteenth century. And only after the mid-1800s did the drive to internationalize copyright (a movement that culminated in the Berne Convention) significantly influence copyright law.

    Sag, supra note 55, at 1373 n.13. Sherman and Bently suggest that most of our modern concepts of copyright law can be seen evolving in the premodern era, but that evolution is messy and the caselaw from that era is, at best, a rough prototype of modern copyright law. Brad Sherman & Lionel Bently, The Making of Modern Intellectual Property Law: The British Experience, 1760–1911, at 192–93 (1999). Sag also acknowledges that “the general questions that courts ask in fair use cases have remained largely constant,” even so, “the answers have changed markedly.” Sag, supra note 55, at 1411. He gives a perfect example of what would have been fair abridgment in 1800, but not today: “No court today would entertain the notion that an abridgment that includes no critical commentary and merely retells the original work’s story is anything other than an infringement.” Id.

  67. Feltner v. Columbia Pictures Television, Inc., 523 U.S. 340 (1998).

  68. Markman v. Westview Instruments, Inc., 517 U.S. 370 (1996).

  69. Feltner, 523 U.S. at 342.

  70. Id. at 348–49.

  71. Id. at 349.

  72. Id. at 350.

  73. Id. at 350–51 (citing Hudson & Goodwin v. Patten, 1 Root 133, 134 (Conn. Super. Ct. 1789)).

  74. Id. at 353.

  75. Cooper Indus., Inc. v. Leatherman Tool Grp., Inc., 532 U.S. 424, 437 (2001).

  76. Markman v. Westview Instruments, Inc., 517 U.S. 370, 376–79 (1996). The influence of Markman in Seventh Amendment analysis has been considerable. As Paul Kirgis noted, Markman “has become the leading opinion governing the roles of judge and jury under the Seventh Amendment.” Paul F. Kirgis, The Right to a Jury Decision on Questions of Fact Under the Seventh Amendment, 64 Ohio St. L.J. 1125, 1136 (2003).

  77. Markman, 517 U.S. at 372.

  78. Id. at 380.

  79. Id. at 376 (quoting Balt. & Carolina Line, Inc. v. Redman, 295 U.S. 654, 657 (1935)).

  80. Id. (citations omitted).

  81. City of Monterey v. Del Monte Dunes at Monterey, Ltd., 526 U.S. 687, 718 (1999).

  82. Id. (quoting Markman, 517 U.S. at 377).

  83. Tull v. United States, 481 U.S. 412, 425–26 (1987) (quoting Colgrove v. Battin, 413 U.S. 149, 157 (1973)). In contrast to Feltner, in Tull the Court concluded “the trial court and not the jury should determine the amount of penalty” under penalty provisions of the Clean Water Act, even assuming that there was a right to trial by jury on the basic question of liability. Id. at 427.

  84. Galloway v. United States, 319 U.S. 372, 392 (1943).

  85. Del Monte Dunes, 526 U.S. at 718.

  86. Id.

  87. Id. at 719.

  88. Id. at 720 (citations omitted) (first quoting Lucas v. S.C. Coastal Council, 505 U.S. 1003, 1015 (1992); and then quoting Yee v. City of Escondido, 503 U.S. 519, 523 (1992)).

  89. Markman v. Westview Instruments, Inc., 517 U.S. 370, 376 (1996).

  90. Del Monte Dunes, 526 U.S. at 720–21 (quoting Lucas, 505 U.S. at 1015); Markman, 517 U.S. at 376.

  91. Del Monte Dunes, 526 U.S. at 721.

  92. Id. at 722 (“For these reasons, we do not attempt a precise demarcation of the respective provinces of judge and jury in determining whether a zoning decision substantially advances legitimate governmental interests. The city and its amici suggest that sustaining the judgment here will undermine the uniformity of the law . . . by subjecting all land-use decisions to plenary, and potentially inconsistent, jury review. Our decision raises no such specter.”).

  93. Feltner v. Columbia Pictures Television, Inc., 523 U.S. 340, 348–50 (1998).

  94. See supra Section II.A.

  95. My thanks to H. Tomás Gómez-Arostegui for this point. E-mail from H. Tomás Gómez-Arostegui, Professor of L., Lewis & Clark L. Sch., to Justin Hughes, Professor of L., Loy. Marymount Univ. (Jan. 9, 2020, 15:16 EST) (on file with author); Cary v. Kearsley (1803) 170 Eng. Rep. 679, 680; 4 Esp. 168, 171 (K.B.).

  96. Gyles v. Wilcox (1740), 26 Eng. Rep. 489, 491; 2 Atk. 141, 144 (Ch.).

  97. Markman v. Westview Instruments, Inc., 517 U.S. 370, 381 (1996).

  98. Id. at 379 (“At the time relevant for Seventh Amendment analogies, in contrast, it was the specification, itself a relatively new development, . . . that represented the key to the patent.”).

  99. Id. at 379–80 (footnote omitted) (citations omitted).

  100. Id. at 381–83. The Court also noted “[t]he construction of written instruments is one of those things that judges often do and are likely to do better than jurors unburdened by training in exegesis.” Id. at 388.

  101. Folsom v. Marsh, 9 F. Cas. 342, 349 (C.C.D. Mass. 1841) (No. 4,901).

  102. City of Monterey v. Del Monte Dunes at Monterey, Ltd., 526 U.S. 687, 718 (1999).

  103. Id. at 721.

  104. Alexander, supra note 23, at 361.

  105. U.S. Bank Nat’l Ass’n v. Vill. at Lakeridge, LLC, 138 S. Ct. 960, 967 (2018).

  106. Id. (citations omitted).

  107. Id.

  108. Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 577 (1994) (a § 107, fair-use determination “calls for case-by-case analysis”). Courts have further cited Acuff-Rose’s language in their decisions. See, e.g., Am. Soc’y for Testing & Materials v. Public.Resource.Org, Inc., 896 F.3d 437, 451 (D.C. Cir. 2018); Authors Guild v. Google, Inc., 804 F.3d 202, 213 (2d Cir. 2015); Cambridge Univ. Press v. Patton, 769 F.3d 1232, 1271–72 (11th Cir. 2014); Bouchat v. Baltimore Ravens Ltd. P’ship, 619 F.3d 301, 308 (4th Cir. 2010); Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d 1146, 1163 (9th Cir. 2007); Sarl Louis Feraud Int’l v. Viewfinder, Inc., 489 F.3d 474, 482 (2d Cir. 2007); Mattel, Inc. v. Walking Mountain Prods., 353 F.3d 792, 800 (9th Cir. 2003) (courts “engage in a case-by-case analysis and a flexible balancing of relevant factors”); Calkins v. Playboy Enters. Int’l, Inc., 561 F. Supp. 2d 1136, 1140 (E.D. Cal. 2008) (“In determining whether a use is fair, courts engage in a case-by-case analysis and a flexible balancing . . . .”).

  109. Snow, supra note 14, at 305–06. But Snow also recognizes the line of argument from U.S. Bank supporting de novo review on the grounds: “To be sure, appellate courts often develop legal principles that guide the fair-use analysis in future cases.” Id. at 304.

  110. U.S. Bank Nat’l Ass’n v. Vill. at Lakeridge, LLC, 138 S. Ct. 960, 967 (2018).

  111. Id. at 967–68 (“Thus the mixed question [in this case] becomes: Given all the basic facts found, was Rabkin’s purchase of MBP’s claim conducted as if the two were strangers to each other? That is about as factual sounding as any mixed question gets.”).

  112. Id. at 968 (quoting Comm’r v. Duberstein, 363 U.S. 278, 291(1960)). In contrast, any of the factor or sub-factor determinations within the § 107 framework will often be “factual inference[s] from undisputed basic facts,” i.e. commercial or noncommercial? How much of the plaintiff’s protected expression was appropriated? Id.

  113. Id.

  114. Id. (quoting Brief for Petitioner at 53, U.S. Bank Nat’l Ass’n v. Vill. at Lakeridge, LLC, 138 S. Ct. 960 (2018) (No. 15-1509)).

  115. Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 425 (1984).

  116. Fox News Network, LLC v. TVEyes, Inc., 883 F.3d 169, 174 (2d Cir. 2018).

  117. For a tip of the vast scholarly iceberg trying to figure out principles governing fair use, see Jiarui Liu, An Empirical Study of Transformative Use in Copyright Law, 22 Stan. Tech. L. Rev. 163 (2019); Rebecca Tushnet, Content, Purpose, or Both?, 90 Wash. L. Rev. 869 (2015); Pamela Samuelson, Unbundling Fair Uses, 77 Fordham L. Rev. 2537 (2009); Barton Beebe, An Empirical Study of U.S. Copyright Fair Use Opinions, 1978–2005, 156 U. Pa. L. Rev. 549 (2008); Matthew Sag, God in the Machine: A New Structural Analysis of Copyright’s Fair Use Doctrine, 11 Mich. Telecomms. & Tech. L. Rev. 381 (2005); Michael J. Madison, A Pattern-Oriented Approach to Fair Use, 45 Wm. & Mary L. Rev. 1525 (2004); Justin Hughes, Fair Use Across Time, 50 UCLA L. Rev. 775 (2003); William F. Patry, The Fair Use Privilege in Copyright Law (2d ed. 1995); William W. Fisher III, Reconstructing the Fair Use Doctrine, 101 Harv. L. Rev. 1659 (1988).

  118. Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 577 (1994) (holding that a § 107 fair use determination “is not to be simplified with bright-line rules, for the statute, like the doctrine it recognizes, calls for case-by-case analysis”).

  119. Id. at 596 (Kennedy, J., concurring) (emphasis added) (citation omitted).

  120. U.S. Bank Nat’l Ass’n v. Vill. at Lakeridge, LLC, 138 S. Ct. 960, 968 (2018). There would be less need for judges to guide the development of law if one believed that the law did not evolve. Morris Arnold proposed that eighteenth-century British and American juries “had much more power to decide questions both of law and fact than do modern ones.” Morris S. Arnold, A Historical Inquiry into the Right to Trial by Jury in Complex Civil Litigation, 128 U. Pa. L. Rev. 829, 833 (1980). Arnold connected this to a natural law vision of the legal system: “[S]uch large latitude to laymen is jurisprudentially significant because it seems to reveal a society at home with the notion that law and right are changeless truths discoverable by lawyers and laymen alike. In such a world law is a fact like any other.” Id. at 835. That, of course, is very different from our positivist mindset and our recognition that copyright law is very much changing and evolving.

  121. Salve Regina Coll. v. Russell, 499 U.S. 225, 233 (1991). Admittedly, the Court expressed this as a negative. Id. (“[D]eferential review of mixed questions of law and fact is warranted when it appears that the district court is ‘better positioned’ than the appellate court to decide the issue in question or that probing appellate scrutiny will not contribute to the clarity of legal doctrine.”).

  122. Straits Fin. LLC v. Ten Sleep Cattle Co., 900 F.3d 359, 368 (7th Cir. 2018) (quoting U.S. Bank, 138 S. Ct. at 967). In ABS Global, Inc. v. Inguran, LLC, 914 F.3d 1054, 1065 (7th Cir. 2019), the Seventh Circuit noted that de novo review is appropriate when the determination being reviewed “requires the application of one of those broad legal standards,” a description that applies to fair use.

  123. In re Twelve Grand Jury Subpoenas, 908 F.3d 525, 527 & n.1 (9th Cir. 2018), cert. denied, 140 S. Ct. 289 (2019).

  124. Id. at 527 n.1 (quoting U.S. Bank, 138 S. Ct. at 967).

  125. Cf. U.S. Bank, 138 S. Ct. at 967–68.

  126. Straits Fin., 900 F.3d at 368 (quoting U.S. Bank, 138 S. Ct. at 967).

  127. This is just a rough count of briefs addressing the substantive fair use question, not the copyrightability and § 102(b) issues in the litigation. These numbers are drawn from the amicus briefs lists compiled by Jonathan Band. Jonathan Band, Broad Support for Google in the First Round of Supreme Court Briefing, Disruptive Competition Project (Jan. 14, 2020), https://www.project-disco.org/intellectual-property/011420-broad-support-for-google-in-the-first-round-of-supreme-court-briefing/ [https://perma.cc/GZ3Q-ZNS2]. The individuals, companies, and organizations filing these briefs demonstrate a widespread belief that the fair-use determination will entail “broader legal conclusions applicable to a whole class of future cases” (American Intellectual Property Law Association, American Library Association, Association of College and Research Libraries, Association of Research Libraries, the Andy Warhol Foundation, the Computer & Communications Industry Association, Creative Commons, the Electronic Frontier Foundation, Etsy, Internet Association, Microsoft, Mozilla, Medium, Open UK, Protocol Labs, Python Software Foundation, Reddit, Rimini, the Robert Rauschenberg Foundation, the Software Preservation Network, the Software Freedom Law Center, Tidelift, Wikimedia, and seventeen law professors). Id.

  128. U.S. Bank, 138 S. Ct. at 967 (quoting Pierce v. Underwood, 487 U.S. 552, 562 (1988)).

  129. Some of these categories are mentioned in 17 U.S.C. § 107 (i.e. “criticism,” “comment”) and others have been developed through precedent (“parody,” “unpublished work”). For scholarly explanations of these different categories and generalizations, see sources cited supra note 117. See also Justin Hughes, Fair Use and Its Politics – at Home and Abroad, in Copyright Law in an Age of Limitations and Exceptions (Ruth Okediji ed., Cambridge University Press, 2016) (describing how these generalizations become de facto rules); Neil Weinstock Netanel, Making Sense of Fair Use, 15 Lewis & Clark L. Rev. 715, 727, 762, 769 (2010).

  130. William Patry, Barton Beebe’s Fair Use Study, Patry Copyright Blog (May 11, 2007, 7:41 AM), http://williampatry.blogspot.com/2007/05/barton-beebes-fair-use-study.html [https://perma.cc/A2E9-W8CV].

  131. Mark A. Lemley, Should a Licensing Market Require Licensing?, 70 Law & Contemp. Probs. 185, 185–86 (2007).

  132. Tushnet, supra note 117, at 871.

  133. Joseph P. Liu, Fair Use, Notice Failure, and the Limits of Copyright as Property, 96 B.U. L. Rev. 833 (2016) (reasoning that the failure of fair use to provide predictability to less sophisticated parties is grounds to stop treating copyright as a property right in relation to certain classes of users).

  134. Arnold, supra note 120, at 835; see also S.F.C. Milsom, Historical Foundations of the Common Law 30, 32, 36–37 (1969).

  135. David Nimmer, Juries and the Development of Fair Use Standards, 31 Harv. J.L. & Tech. 563, 589–90 (2018).

  136. Snow, supra note 14, at 306, 330–31.

  137. Neil Weinstock Netanel, Copyright’s Paradox 66 (2008) (“Given the doctrine’s open-ended, case-specific cast and inconsistent application, it is exceedingly difficult to predict whether a given use in a given case will qualify . . . .”).

  138. Michael W. Carroll, Fixing Fair Use, 85 N.C. L. Rev. 1087, 1090 (2007).

  139. Pierce v. Underwood, 487 U.S. 552, 560–62 (1988). The issue before the Court was the proper standard of review for determination that the Government’s position (in a litigation) was not substantially justified. Id. (quoting Maurice Rosenberg, Judicial Discretion of the Trial Court, Viewed from Above, 22 Syracuse L. Rev. 635, 662 (1971)).

  140. Rosenberg, supra note 139.

  141. Making institutional or constitutional arguments to support preferred policy outcomes is hardly unique to copyright. See, e.g., Timothy Meyer & Ganesh Sitaraman, Trade and the Separation of Powers, 107 Cal. L. Rev. 583, 643–44 (2019) (recognizing that arguments to limit Congress’ institutional role in trade agreements are intended to “circumscribe[] the number of access points that local economic interests have to the policy-making process,” and “that invoking institutional arguments is indeed about substantive policy preferences”); Neal Devins, The Judicial Safeguards of Federalism, 99 Nw. U. L. Rev. 131, 137 (2004) (“[E]lected officials invoke federalism when it comports with their substantive policy preferences, but they otherwise do not care about the federal-state balance.”).

  142. Corbello v. DeVito, 262 F. Supp. 3d 1056, 1068, 1076–77, 1079 (D. Nev. 2017).

  143. Ronan Deazley, The Statute of Anne and the Great Abridgement Swindle, 47 Hous. L. Rev. 793, 816 (2010).

  144. Salve Regina Coll. v. Russell, 499 U.S. 225, 233 (1991).

  145. Carroll, supra note 138, at 1095.