INTRODUCTION

Say we think a new book is going to libel us, and we ask a court for a preliminary injunction against the book's publication. We argue that we're likely to succeed on the merits of our libel claim, and that failure to enjoin the speech would cause us irreparable harm.

Too bad, the court will certainly say; a content-based preliminary injunction of speech would be a blatantly unconstitutional prior restraint. Maybe after a trial on the merits and a judicial finding that the speech is in fact constitutionally unprotected libel, we could get a permanent injunction, though even that's not clear.¹ But we definitely could not get a preliminary injunction, based on mere likelihood of success. Likewise for preliminary injunctions against obscenity and other kinds of speech, despite the fact that such speech, if ultimately found to be unprotected at trial, could be criminally or civilly punished.²

In copyright cases, though, preliminary injunctions are granted pretty much as a matter of course, even when the defendant has engaged in creative adaptation, not just literal copying. How can this be? True, the Supreme Court has held that copyright law is a constitutionally permissible speech restriction; though copyright law restricts what we can write or record or perform, the First Amendment doesn't protect copyright-infringing speech against such a restraint.³ But libel law and obscenity law are likewise constitutionally valid restrictions on speech, and yet courts refuse to allow preliminary injunctions there. The "First Amendment due process"⁴ rule against prior restraints applies even to speech that's alleged to be constitutionally unprotected. Why, then, not to allegedly infringing speech?⁵

We explore this question below. In Part I, we discuss the history of preliminary injunctions in copyright cases and the current law relating to such injunctions. In Part II, we develop our central thesis by explaining why copyright law is a speech restriction; why preliminary injunctions of speech are generally unconstitutional; and why, at least as a doctrinal and conceptual matter, it's hard to see how copyright law could be treated differently for First Amendment purposes. What's more, we argue, giving copyright law a free ride from the normal First Amendment due process rules risks discrediting those rules in other contexts.

In Part III, we step back and ask whether this inquiry has cast some doubt on the prior restraint doctrine itself--whether copyright law's tolerance of preliminary injunctions might be right, and the free speech doctrine's condemnation of such injunctions might be wrong. In Part IV, we discuss the implications of the collision between copyright law principles and free speech principles, and propose some changes that are needed to bring copyright law into line with constitutional commands. We conclude that permanent injunctions in copyright cases should generally be constitutional, and the same should go for preliminary injunctions in cases that clearly involve literal copying, with no plausible claim of fair use or of copying mere idea rather than expression. Other preliminary injunctions, though, should generally be unconstitutional.

In Part V, we briefly explore these questions with regard to other kinds of intellectual property--trademarks, rights of publicity, trade secrets, and patents. We conclude that the problem is not limited to copyright, and that at least in trademark and right of publicity cases, preliminary injunctions may sometimes run afoul of the First Amendment. Finally, in Part VI we say a bit about the practical prospects for revising the law along the lines we suggest.

I. STANDARDS GOVERNING PRELIMINARY INJUNCTIONS IN COPYRIGHT CASES

A. History

1. England. Preliminary injunctions have been a feature of English copyright law since its inception. The first modern English copyright act was the Statute of Anne in 1709.⁶ At that time, remedies were divided between courts of law and courts of equity. While the overlap between the two produced some disputes, it was well established that injunctive relief could be ordered by the Court of Chancery notwithstanding the existence of a legal remedy in damages.⁷ Lord De Grey's review of the early cases led him to conclude that injunctive relief could be granted in chancery "on a reasonable pretence, and a doubtful right, before the hearing of the cause."⁸ There seems little question that he was referring to what we would today call a preliminary injunction.

The issuance of such injunctions in England appears to have been rather common in the early days of copyright, and the standard for their issuance rather lenient. In Donaldson v. Beckett, Lord De Grey concluded that of the seventeen copyright decisions between 1710 and 1774, seven appeared to be "injunctions granted ex parte, upon filing the bill, with an affidavit annexed."⁹ Howard Abrams refers to several preliminary injunctions from the 1730s "issued on an ex parte basis and . . . not contested by the defendants."¹⁰ Why the defendants didn't contest these cases is not entirely clear; it may be that these were relatively obvious cases of counterfeiting. But it is notable that English courts regularly issued preliminary injunctions in copyright cases in which "the defendant [was] not so much as heard."¹¹

Furthermore, in cases where the defendant did receive notice of the proceeding in chancery and contested it, preliminary injunctions were still routinely issued. Indeed, in one case the Chancellor issued an injunction "till the hearing" at law on the basis of an expired copyright, on the grounds that the plaintiff might in addition be entitled to common law copyright protection.¹² English courts continued to apply the standard offered by Lord De Grey: where there was a doubtful claim, or a "fair question,"¹³ ex parte injunctions were available. Some courts characterized this as an effective presumption in favor of injunctive relief.¹⁴

On the other hand, some English courts proved reluctant to grant preliminary injunctions where novel or uncertain questions of copyright law were at issue. In Osborne v. Donaldson,¹⁵ for example, the court dissolved a preliminary injunction after determining that the plaintiff's statutory copyright had expired, holding over the then-unresolved issue of common law copyright for the law courts.¹⁶ However, it is noteworthy that the Chancellor in Osborne had apparently issued an ex parte preliminary injunction as a matter of course on the basis of an expired copyright, and only dissolved the injunction after receiving the defendants' answer.¹⁷ This suggests that a plaintiff was presumptively entitled to preliminary injunctive relief unless the defendant could show some reason why the court should refuse to grant the injunction.¹⁸

Further, English courts seemed inclined to presume the equivalent of "irreparable injury" to plaintiffs if copyright injunctions were not granted. In Wilkins v. Aikin,¹⁹ Lord Eldon wrote that equity must act in such cases because the legal right

cannot be made effectual by any action for damages; as, if the work is pirated, it is impossible to lay before a Jury the whole evidence as to all the publications, which go out to the world, to the Plaintiff's prejudice. A Court of Equity therefore acts, with a view to make the legal right effectual by preventing the publication altogether . . . .²⁰

One court went so far as to enjoin copyright infringement in a case where the only known pirated copy had already been sold and there was no evidence that any other copies existed.²¹ Similarly, in time-sensitive cases, the fact that the decision to grant or deny a preliminary injunction would effectively finally determine the rights of the parties was used as an argument in favor of granting the injunction rather than denying it.²² A defendant could defeat an application for an injunction by demonstrating that "irreparable injury would result from granting the injunction, unless the case is so clear that there can be no reasonable doubt with regard to the legal title . . . ."²³ It seems clear that in England before 1850, preliminary injunctions were the normal remedy in copyright cases.

2. United States. Despite the fact that United States copyright law was based largely on the English model,²⁴ early U.S. courts showed some reluctance to grant preliminary relief to copyright plaintiffs.²⁵ Of the reported copyright cases in the first fifty years of the republic, most refused to grant preliminary injunctive relief,²⁶ and on several occasions courts and commentators noted that the American rule seemed less favorable to plaintiffs than the English case law.²⁷ In particular, these early decisions inquired in detail into the merits of plaintiffs' copyright claims,²⁸ and seemed in some cases to resolve doubt on the merits against the plaintiff.²⁹

Early American copyright plaintiffs also faced substantial burdens in demonstrating that the balance of hardships favored an injunction, even when success on the merits was likely.³⁰ In Dun v. Lumberman's Credit Ass'n,³¹ the Supreme Court held that even though copying of protectable material was clearly established, a permanent injunction was not warranted because the infringer's work contained a large amount of material that was not copied, and enjoining publication of the infringer's book would impose too great a hardship.³² To similar effect is White v. Bender,³³ where the court refused to issue a preliminary injunction despite clear evidence of copying because the infringing work was not a "mere copy" and was unlikely to supersede the original.³⁴ In another case, a preliminary injunction was refused even though validity and infringement were clear, simply because the infringement was innocent and noncommercial, and the defendant was solvent.³⁵

In contrast to the early English practice, which presumed that damages would not adequately compensate injured plaintiffs,³⁶ early U.S. courts applied the "irreparable injury" rule³⁷ with some force to copyright plaintiffs. Plaintiffs who had made a clear showing of infringement were sometimes denied injunctive relief if the defendant presented evidence that it could pay adequate damages after the completion of the suit.³⁸ Courts also used the fact that granting or denying a preliminary injunction would effectively decide the case as a reason to deny the injunction.³⁹ This contrasts sharply with the English case of Prince Albert v. Strange, discussed above,⁴⁰ where the same factual circumstance was held to support rather than oppose the granting of a preliminary injunction.

At some point in the early part of this century, judicial hostility towards preliminary injunctions in copyright cases began to abate. In his 1917 treatise on copyright law, Arthur Weil noted that "the modern tendency is to issue temporary injunctions more freely than in earlier times."⁴¹ There is some evidence that this change corresponds roughly in time with a more general increase in the willingness of courts to grant injunctions, at least in labor cases.⁴² Whatever the cause, it seems clear that early U.S. courts were not nearly so willing to grant preliminary injunctions in copyright cases as either their English predecessors or their modern counterparts.

B. The Modern Rule

The Copyright Act of 1976⁴³ authorizes courts to grant "temporary and final injunctions on such terms as [they] may deem reasonable to prevent or restrain infringement of a copyright."⁴⁴ Preliminary injunctive relief in non-copyright cases is nominally governed by the traditional four-factor preliminary injunction test, which asks: (1) whether the plaintiff is likely to succeed on the merits; (2) whether the plaintiff will suffer irreparable injury if the injunction is not granted; (3) whether the balance of hardships tips in the plaintiff's favor; and (4) whether granting the injunction would be in the public interest.⁴⁵

In actual application, though, courts have given some of these factors primacy over others, have permitted substitution among the factors (so that if the balance of hardships strongly favors the plaintiff, she need only show "substantial questions" going to the merits),⁴⁶ and have come close to eliminating the irreparable injury rule altogether.⁴⁷

Modern U.S. courts have treated plaintiffs seeking preliminary injunctions in copyright cases more favorably than plaintiffs in other areas of law, and certainly more favorably than they treated copyright plaintiffs a century ago.⁴⁸ The ostensibly four-factor test⁴⁹ collapses, in virtually all circuits, to a simple inquiry into likelihood of success on the merits. If that can be demonstrated, a preliminary injunction is the expected remedy.⁵⁰ Indeed, the Ninth Circuit has held that a preliminary injunction is the expected remedy in a copyright infringement case once a "reasonable likelihood of success" has been demonstrated.⁵¹

If the copyright owner demonstrates a likelihood of success, courts are virtually unanimous in dispensing with the need to show irreparable injury.⁵² While Goldstein argues that the presumption of irreparable harm is stronger for "substantially original" copyrighted works than for works receiving lesser protection,⁵³ in fact the presumption is so widespread that it has even been applied to telephone white pages⁵⁴ and the page numbers of West's court reporters,⁵⁵ neither of which presents a particularly strong case for copyright protection.⁵⁶ Interestingly, modern U.S. courts have returned to the old English rule that preliminary injunctions should be granted, not withheld, where the ruling on the preliminary injunction motion will effectively decide the case one way or the other.⁵⁷ And various courts have found both the defendant's solvency and the defendant's insolvency to favor the plaintiff.⁵⁸

The third factor--balance of the hardships--is similarly unimportant in copyright cases because courts virtually always weigh the balance of the hardships in favor of the copyright plaintiff if it has demonstrated a likelihood of success on the merits. The Second and Ninth Circuits, the most important in copyright cases because together they house the New York publishing industry and acting community, Hollywood, and Silicon Valley, do not even require hardship to the defendant to be balanced at all if the plaintiff can show that it is likely to succeed on the merits of its claim.⁵⁹ Even the remaining circuits, which nominally consider this factor in all cases, generally "tip the balance of hardships in the plaintiff's favor to the extent that it has shown a likelihood of success on the merits."⁶⁰ While some courts will give the benefit of the doubt to defendants if the defendant's product is mostly composed of noninfringing materials,⁶¹ even that rule is not universally accepted.⁶²

The final factor--the "public interest"--has also been effectively eliminated in copyright cases. Courts find it "virtually axiomatic that the public interest can only be served by upholding copyright protections,"⁶³ and accordingly grant preliminary injunctions if the plaintiff has shown a likelihood of success on the merits.⁶⁴ Efforts by defendants to assert a competing public interest in the wide dissemination of information generally fail.⁶⁵ As a result, courts which purport to apply a four-factor test in fact generally find a showing of probable success on the merits to be virtually conclusive in favor of granting an injunction.⁶⁶

The lenience of this standard is compounded by the definition of "success on the merits" in many copyright cases. A prima facie case of copyright infringement requires that the plaintiff prove that it owns a valid copyright and that the defendant has infringed the copyright by violating one of the exclusive rights enumerated in the Act.⁶⁷ But the first of these elements is trivial, at least in preliminary injunction cases: the plaintiff is entitled to rely on the presumption of validity that attaches to its copyright registration.⁶⁸ Even in the rare case in which the plaintiff has not registered a copyright, courts will sometimes overlook that fact and issue an injunction on an unregistered work.⁶⁹ Since registration of copyrights is essentially pro forma and does not entail any significant examination by the Copyright Office into originality or the propriety of ownership,⁷⁰ copyright plaintiffs have essentially cleared one of the two hurdles to injunctive relief simply by virtue of filing suit.⁷¹ The requirement that the plaintiff prove infringement, while more substantial, can be satisfied by proxy; plaintiffs need merely show access and sufficient similarities between the two works to raise an inference of copying (and therefore of infringement).⁷² Where plaintiffs have made a prima facie showing of likelihood of success, in fact, some courts hold that it is reversible error to deny a preliminary injunction.⁷³

To be sure, defendants have the opportunity to rebut the plaintiff's prima facie case at the preliminary injunction hearing. They can raise questions as to both the validity of the plaintiff's copyright and the scope of the protection it should receive; they can contest the inference of copying, or argue that the portions copied were not themselves protectable;⁷⁴ and they can (or at least they should be able to) raise statutory defenses such as fair use.⁷⁵ Still, it is striking that on most issues likely to be significant in determining the ultimate outcome of the copyright claim, it is the defendant and not the plaintiff who realistically bears the burden of proof in a preliminary injunction hearing.⁷⁶ Further, merely raising a plausible defense will not prevent a court from issuing a preliminary injunction; in a contested case, a substantial likelihood of success by the plaintiff may be enough for an injunction.⁷⁷

It is common practice in copyright cases to condition the grant of a preliminary injunction on the plaintiff's posting of a bond, which may be forfeited if the defendant ultimately prevails at trial.⁷⁸ These bonds can be substantial,⁷⁹ and arguably provide considerable solace to defendants faced with preliminary injunctions. Nonetheless, the practice in copyright cases seems to reverse the irreparable injury rule: where plaintiffs were traditionally presumed to be compensable by means of a damages award, preliminary injunctions are now routine and it is defendants who are entitled to "damages" if they are wrongfully enjoined.⁸⁰

Finally, there are other provisional remedies available to copyright plaintiffs besides preliminary injunctions. In appropriate cases, plaintiffs can obtain temporary restraining orders ex parte in accordance with Rule 65 of the Federal Rules of Civil Procedure.⁸¹ They may also be entitled to seize the allegedly infringing goods before trial,⁸² an act which in practice will have much the same effect as an injunction prohibiting distribution or sale. Such seizures are also often ex parte, a fact which has caused some writers--and the Federal Rules Advisory Committee--concern.⁸³

C. Summary

Several things are notable about the law of copyright preliminary injunctions. First, the modern U.S. practice is unquestionably more favorable to plaintiffs than to defendants, and it is more favorable in copyright cases than in most other contexts. Copyright plaintiffs can dispense with most of the normal requirements for obtaining a preliminary injunction, and disputed issues are regularly resolved in their favor.

Second, in many cases it is the defendant, not the plaintiff, who effectively bears the burden of proof with respect to such injunctions. Bond practice reinforces this shift--it is the defendant who is presumed to be compensable by money damages if an injunction wrongly issues, while plaintiffs are conclusively presumed to suffer irreparable harm if an injunction does not issue.

Finally, while the pro-plaintiff trend in copyright injunctions has historic antecedents in early English practice, they are not lineal antecedents. Early U.S. copyright practice differed markedly from both the traditional English rule and the modern U.S. rule. Only in the 1900s have preliminary injunctions come to be an expected part of U.S. copyright cases.

II.FREEDOM OF SPEECH, COPYRIGHT, AND INJUNCTIONS

A. Copyright Law as a Speech Restriction⁸⁴

Copyright law restricts speech: it restricts you from writing, painting, publicly performing, or otherwise communicating what you please.⁸⁵ If your speech copies ours, and if the copying uses our "expression," not merely our ideas or facts that we have uncovered, the speech can be enjoined and punished, civilly and sometimes criminally.⁸⁶

And copyright law applies to creative adaptation as well as to literal copying, so it's no answer to say that the copiers aren't themselves originating the speech. Writers are barred from creating works, even works based on real events, whose plots are too similar to what others have done.⁸⁷ Artists are forbidden from creating artworks that are too similar to others' art.⁸⁸ Musicians are restricted from using "vocal percussion effects" which are too similar to those of other artists.⁸⁹ Indeed, in some cases basing even a small bit of an otherwise original work on others' copyrighted expression can subject a defendant to liability, as when movie producers are held liable for including copyrighted posters, sculptures, or even furniture in brief scenes from the film.⁹⁰ Copyright law seriously restricts speakers' ability to express themselves the way they want.

In Harper & Row, Publishers, Inc. v. Nation Enterprises,⁹¹ the Supreme Court made clear that copyright law is substantively constitutional: the First Amendment does not shield speech that infringes another's copyright.⁹² Copyright, the Court said, is itself an "engine of free expression" because it "supplies the economic incentive to create and disseminate ideas."⁹³

But the Court strongly implied that this rationale would not justify restrictions on copying facts or ideas, or on certain kinds of commentary historically protected by the fair use doctrine. The Court characterized "the Copyright Act's distinction between copyrightable expression and uncopyrightable facts and ideas" and "the latitude for scholarship and comment traditionally afforded by fair use" as "First Amendment protections."⁹⁴ It pointed out that "[n]o author may copyright his ideas or the facts he narrates,"⁹⁵ and cited Justice Brennan's concurrence in New York Times Co. v. United States,⁹⁶ which said that "[c]opyright laws are not restrictions on freedom of speech as copyright protects only form of expression and not the ideas expressed."⁹⁷ And the Court stressed that it would be an "abuse of the copyright owner's monopoly" for copyright law to become "an instrument to suppress facts."⁹⁸

Harper & Row thus suggests that the lines between using others' expression and using their ideas and between some fair uses and other uses are of First Amendment significance. Speech communicating facts and ideas in a way that is "substantially similar" to someone else's expression, and that is not a fair use, is constitutionally unprotected.⁹⁹ Speech that communicates the same facts and ideas in other ways, or speech that borrows expression but falls within the fair use exception, is constitutionally protected.¹⁰⁰ Enjoining or punishing noninfringing speech would thus be an unconstitutional restraint of First Amendment protected speech. So would reducing or eliminating the protections afforded users by the idea-expression dichotomy and the fair use rules.¹⁰¹

Such a dividing line makes constitutional sense. Free speech law may tolerate certain limits on how someone expresses an idea or a fact, but it ought not tolerate restrictions on communicating ideas and facts as such, or on criticism or parody that must in some measure borrow another's words.¹⁰² When you express an idea that someone else pioneered or discuss facts that others uncovered or use their expression while commenting on it, you might be free riding on their hard work, but it is a free riding we must allow.¹⁰³

Thus, we will assume what Harper & Row strongly suggested, and what lower courts have reaffirmed:¹⁰⁴ speech that copies another's expression (setting aside fair use) can be restricted, but speech that borrows only facts or ideas, or that fairly uses expression, is constitutionally protected.

B. Prior Restraint Doctrine

1. Speech May Not Be Restrained Until a Final Judicial Determination That It's Unprotected by the First Amendment. As we've argued above, copyright law is a speech restriction. Accordingly, injunctions against distributing a supposedly infringing work are injunctions restraining speech; and preliminary injunctions restraining speech are generally considered unconstitutional "prior restraints."¹⁰⁵

We stress here that we're referring to preliminary injunctions, not permanent injunctions.¹⁰⁶ A permanent injunction, entered following a final determination that the speech is unprotected, is generally seen as constitutional.¹⁰⁷ "The special vice of a prior restraint," the Court has held, "is that communication will be suppressed . . . before an adequate determination that it is unprotected by the First Amendment."¹⁰⁸ After speech is conclusively judicially determined to be unprotected--because it is obscene, for example--a permanent injunction (such as the ones at issue in Kingsley Books, Inc. v. Brown¹⁰⁹ or Pittsburgh Press Co. v. Pittsburgh Commission on Human Relations¹¹⁰) would be no more troubling on constitutional grounds than a civil or criminal penalty.¹¹¹ "[T]he order will not have gone into effect before [the court's] final determination that the [speech was] unprotected."¹¹² The Court's occasional dicta suggesting that all injunctions are prior restraints are therefore somewhat erroneous overgeneralizations.¹¹³

But when the injunction is entered prior to this final judicial determination, it is almost always treated as an unconstitutional prior restraint.¹¹⁴ This is so even when the injunction is entered for the seemingly laudable purpose of preserving the status quo pending the final determination of whether the speech is protected.¹¹⁵

Many of the quintessential unconstitutional prior restraint schemes were in fact prophylactic mechanisms that were aimed at preventing unprotected speech, but that ended up at least temporarily restricting protected speech as well. Thus, traditional licensing systems made it illegal for people to print anything, protected or not, unless they had licenses.¹¹⁶ Traditional prescreening censorship systems made it illegal for people to distribute anything, protected or not, until the censors gave it their approval.¹¹⁷ The injunction in Near v. Minnesota¹¹⁸ banned a newspaper that had printed libelous material in the past from operating, even to publish material that was not libelous.¹¹⁹ Such restraints are presumptively unconstitutional, because they may restrict protected speech as well as unprotected speech.¹²⁰

The same goes for preliminary injunctions that restrict speech, as the Court held in Vance v. Universal Amusement Co.¹²¹ Vance struck down a scheme which had let state courts preliminarily enjoin alleged obscenity "based on a showing of probable success on the merits and without a final determination of obscenity."¹²² Such injunctions, the Court held, were unconstitutional because they allowed "prior restraints of indefinite duration on the exhibition of motion pictures that [had] not been finally adjudicated to be obscene."¹²³ The Court continued: "That a state trial judge might be thought more likely than an administrative censor to determine accurately that a work is obscene does not change the unconstitutional character of the restraint if erroneously entered."¹²⁴

Vance's wording and logic are relevant to all speech restrictions, not just obscenity law, and lower courts have in fact struck down preliminary injunctions in other contexts on this very ground.¹²⁵ Preliminary injunctions, no matter what sort of speech is being enjoined, rest not on "[a court's] final determination that the [speech is] unprotected,"¹²⁶ but on a mere finding of a likelihood that the speech is unprotected.¹²⁷

And this likelihood that the court must find need not be near certainty: a preliminary injunction can generally issue in a copyright case on a finding of "sufficiently serious questions going to the merits to make them fair grounds for litigation and a balance of hardships tipping decidedly in its favor."¹²⁸ In the Seventh Circuit, when the balance of hardships tips towards the plaintiff, even a "better than negligible chance that [the plaintiff] will succeed on [the merits]"¹²⁹ may be enough.

Under the copyright preliminary injunction cases, then, a court may enjoin the speech even if it recognizes that there's a serious possibility that the speech will ultimately be proven to be constitutionally protected--for instance, even if it concludes that there are "sufficiently serious questions going to the merits" of a fair use defense.¹³⁰ While the injunction is in force, the enjoined person may not say what he wants to say, even if the speech would ultimately prove to be constitutionally protected--under the collateral bar rule, the enjoined person may be held in contempt for violating the injunction even if he can prove at the contempt hearing that the speech is in fact protected by the First Amendment.¹³¹

What's more, as we discussed in detail in Part I.B, the preliminary injunction standard actually applied in copyright cases is even more lenient to the plaintiff than the traditional equity rules for preliminary injunctions--courts presume that the copyright plaintiff can show irreparable harm and that the public interest and the balance of hardships tip in its favor.¹³² Courts also suggest that "[t]he overarching consideration in granting preliminary relief is, of course, to preserve the status quo,"¹³³ which generally suggests a preference in close cases for restricting the publication of the defendant's speech. Such restraints should be unconstitutional, at least unless (as we'll discuss below) there's a framework for getting a truly prompt final determination of whether the speech is protected--something current preliminary injunction practice does not provide.

2. Why This Rule Makes Sense. The above summary of the prior restraint doctrine is not entirely free from controversy--many still claim that the doctrine bans all injunctions of future speech, whether or not the injunction is preceded by a holding that the speech is constitutionally unprotected.¹³⁴ Nonetheless, we believe our summary best fits the precedents. Subject to two caveats we'll mention below,¹³⁵ the case law does indeed allow permanent injunctions of unprotected speech, entered after a final judicial finding that the speech is unprotected, but doesn't allow restraints entered before such a finding.

And this rule makes sense out of what is otherwise a puzzling distinction. Most prior restraints--except the rare ones that actually involve seizing speech products¹³⁶--are only enforced through subsequent punishment. An injunction doesn't physically keep you from speaking, but only makes you subsequently liable for violating the injunction. Conversely, the deterrent force of traditional subsequent punishments, such as criminal penalties or civil liability, restrains speakers from speaking in the first place (as it's intended to do).¹³⁷

The focus on final judicial determination that the speech is unprotected solves this puzzle. Traditional subsequent punishments, such as civil or criminal liability, punish speech only when, at the end of the trial, the speech is found to be constitutionally unprotected. But judicial or administrative preliminary restraints are actually more restrictive because while they are in force they restrain even speech that may ultimately prove to be protected. In addition, since preliminary injunctions are often easier to get than final determinations on the merits and are adopted with less evidence and less deliberation, the danger that the court will get it wrong and mistakenly restrict protected speech is even greater.¹³⁸

Of course, this requirement of a final judicial determination makes it harder to prevent constitutionally unprotected, and quite possibly harmful, speech. For instance, instead of preliminary enjoining something that looks like it's probably a constitutionally unprotected libel, and thus averting the harm that the potential libel will do, the court must wait until trial and a final determination that the speech is indeed libelous. By then, the speech will likely have been published and widely disseminated, and damages will be an uncertain and incomplete remedy. Nonetheless, prior restraint law embodies a judgment that such prophylactic injunctions are generally too grave a burden on free speech rights.¹³⁹

Some have argued that the prior restraint doctrine should turn on other distinctions. Vincent Blasi argues that all injunctions, including permanent ones, should be presumptively barred because they "rely upon adjudication in the abstract, . . . encourage regulatory agents to overuse the power to regulate, and . . . adversely affect audience reception of controversial messages."¹⁴⁰ Justice Brennan likewise criticized even permanent injunctions, entered after a final adjudication that the speech is unprotected, on the grounds that the ultimate decision as to whether the speech is protected should be made by juries, the traditional administrators of civil and criminal liability, and not by judges, the traditional issuers of injunctions.¹⁴¹

We ultimately agree with Martin Redish's responses to these arguments,¹⁴² both as a theoretical matter and because the Redish responses better reflect the prior restraint case law. In any event, though, the Brennan and Blasi theories would support narrower, not broader, judicial power to issue preliminary injunctions. Our criticisms of preliminary injunctions in copyright cases should thus be shared even by those who adhere to the Brennan or Blasi schools.

Professor Redish does suggest that preliminary injunctions might be permissible if "a strong likelihood exists that the government will be able to establish that the challenged expression is [substantively unprotected]," though he notes the "problems plaguing the use of preliminary injunctive relief against expressive activity."¹⁴³ One could support this position on the grounds that due process, including First Amendment due process, is flexible enough to tolerate a certain risk of error when the restraint is only temporary and when it is entered by a judge rather than an administrative body. We think, though, that this is inconsistent with the case law--especially with Vance--and with the core principle that emerges from Professor Redish's own work and that the Court has validated: that it is improper for speech to be restricted by a "preliminary restraint imposed prior to a decision on the merits of a final restraint."¹⁴⁴

3. Two Possible Exceptions. There are two possible exceptions to the basic rule that speech should not be restrained until a final judicial determination that it's unprotected. On the side of greater protection of speech, there are a few Supreme Court cases that suggest that speech may not be enjoined even after a finding that it is indeed unprotected. Thus, for instance, in Curtis Publishing Co. v. Butts,¹⁴⁵ a libel case, the Court suggested in passing (and in dictum) that it had "rejected all manner of prior restraint on publication, despite strong arguments that if the material was unprotected[,] the time of suppression was immaterial."¹⁴⁶ There certainly has been a long tradition in American law--beginning with the 1839 New York case of Brandreth v. Lance¹⁴⁷--that libels ought never be enjoined, and some state constitutions have been read as barring such injunctions.¹⁴⁸ Nonetheless, neither the Court nor lower courts have gone far with the Curtis Publishing dictum, and some lower courts have in fact authorized permanent injunctions against libels.¹⁴⁹

On the side of greater tolerance for injunctions, some cases have allowed prior restraint schemes that do temporarily restrict the distribution of speech, so long as the government promptly gets a final judicial determination of whether the speech is indeed unprotected.¹⁵⁰ As the Court held in Freedman v. Maryland,

Any restraint imposed in advance of a final judicial determination on the merits must . . . be limited to preservation of the status quo for the shortest fixed period compatible with sound judicial resolution . . . . [T]he procedure must . . . assure a prompt final judicial decision, to minimize the deterrent effect of an interim and possibly erroneous [restraint].¹⁵¹

These cases, which establish what we'll call "the Freedman exception," have been quite controversial,¹⁵² and though they're firmly established in the obscenity context, it's not clear that they would be extended elsewhere. For instance, we strongly doubt that the Court would (or should) uphold a requirement that all books--or even all non-time-sensitive books--be presented to a Defamation Prevention Board for screening for libel,¹⁵³ even if the scheme guaranteed a prompt judicial determination of whether the speech was in fact libelous. Likewise, we doubt that courts would uphold a system under which the Postmaster General could even temporarily deny use of the mails to publications he thought were libelous.¹⁵⁴ On the other hand, the Court has suggested that the Freedman standards may apply in at least some non-obscenity contexts, such as injunctions against offensive demonstrations.¹⁵⁵

Thus, preliminary injunctions must at the very least be subject to the Freedman v. Maryland principles: the preliminary injunction must be immediately appealable, with the assurance of a prompt decision (not just a prompt hearing) about whether the speech is actually unprotected--actually obscene, libelous, or infringing--rather than merely likely to be unprotected.¹⁵⁶ And if the Freedman exception is limited to the obscenity context, then all other preliminary injunctions of speech would be unconstitutional unless the court were to simultaneously determine on the merits that the speech was indeed unprotected.

To summarize, even if it's permissible to restrict speech that falls within certain unprotected categories, it shouldn't be permissible to restrict speech, for several months or perhaps even just for several days, simply on the grounds that it might fall within an unprotected category. Or at least so says Vance and the remainder of the Court's prior restraint jurisprudence.

C. Ex Parte Temporary Restraining Orders

The Court's skepticism towards content-based preliminary injunctions extends in spades to ex parte temporary restraining orders (TROs). Ex parte restraints on speech are generally forbidden, at least unless one can show that it's impossible to give the target notice and an opportunity to participate. The Court held this in Carroll v. President and Commissioners of Princess Anne,¹⁵⁷ where it struck down a ten-day ex parte TRO that banned--in order to prevent violent clashes--a march by the racist National States Rights Party.¹⁵⁸ Even when speech may be enjoined, the Court said, both sides must be present in court to ensure that an injunction is issued only when genuinely necessary and is drafted as narrowly as possible.¹⁵⁹ When only one side is heard, the Court concluded, "there is insufficient assurance of the balanced analysis and careful conclusions which are essential in the area of First Amendment adjudication."¹⁶⁰

It might conceivably be desirable to permit a narrow exception to the general ban on TROs if giving notice would pose unusual dangers--for instance if it would create a risk that an infringer would escape and take the infringing goods with him to be distributed elsewhere.¹⁶¹ Nonetheless, even if such an exception were to be recognized, it could only be applied in extreme cases. It can't be enough to show that a speaker will, before the injunction is entered, continue engaging in his allegedly harmful speech: this argument could be made for most attempts to enjoin supposedly unprotected speech, and would justify ex parte orders delaying the publication of supposedly libelous books or newspaper articles, ex parte prohibitions of news coverage that supposedly interferes with a fair trial, and so on. Certainly lower courts have interpreted Carroll as prohibiting such ex parte speech restrictions.¹⁶²

D. Can Preliminary Injunctions of Copyright Infringement Be Distinguished from Preliminary Injunctions of Other Speech?

So far, we've applied standard First Amendment doctrine to copyright law and concluded that preliminary injunctions in copyright cases pose serious free speech problems. But could it be that copyright law is somehow specially immune from the normal concerns surrounding other speech restrictions? We've often heard this view among copyright lawyers. While copyright law is clearly a speech restriction, to many it lacks that speech restriction flavor. It doesn't sound like censorship, just people enforcing their lawful property rights. Still, while many have this intuition, is there some specific reason underlying it, some reason that can justify setting aside the normal First Amendment procedural guarantees?

1. Property Rights. The argument that copyright law should be exempted from standard First Amendment procedural rules because it protects property rights strikes us as a non sequitur.¹⁶³ Free speech guarantees can't be avoided simply by characterizing a speech restriction as an "intellectual property law." After all, one could plausibly view libel law as protecting a person's property interest in his reputation, or a company's property interest in its product's reputation--some courts have indeed done so.¹⁶⁴ Likewise, a bill introduced in the 104th Congress would have declared the United States flag copyrighted, and would have imposed "criminal penalties for the destruction of a copyrighted flag."¹⁶⁵ Congress can't get around the First Amendment merely by characterizing otherwise protected speech¹⁶⁶ as treading on a property interest, as the flag copyright bill sought to do.¹⁶⁷ One still has to ask whether these intellectual property laws are unconstitutional speech restrictions (something the Supreme Court has in fact asked in its free speech/intellectual property cases),¹⁶⁸ and, if they are substantively constitutional, what procedural protections the First Amendment nonetheless requires.

The First Amendment does not, of course, license people to trespass on private real estate in order to speak.¹⁶⁹ Trespass laws are generally applicable to all conduct (speech or otherwise) and operate without regard to the communicative impact of the speech, so it makes sense not to view them as speech restrictions for purposes of the prior restraint doctrine.¹⁷⁰ But content-based laws, specifically targeted at speech, must be seen as speech restrictions regardless of whether one frames them as "property" rules. They may be substantively valid speech restrictions, but calling them property rules doesn't justify exempting them from the normal First Amendment procedural principles, especially when they ban people from saying a particular thing anywhere, at any time, and not just on others' land.¹⁷¹ Indeed, if the First Amendment rules didn't apply to intellectual property laws, the Court's entire discussion of the First Amendment issue in Harper & Row¹⁷² would be not only unnecessary but affirmatively misleading.

Furthermore, consumption of intellectual property, unlike consumption of tangible property, is "nonrivalrous"--one person's use of a work does not prevent others from using it as well.¹⁷³ This makes intellectual property sufficiently unlike tangible property that some courts that have faced the issue directly have even concluded that copyrights and trademarks are not "property" within the meaning of the Takings Clause.¹⁷⁴ Whether or not that's correct, the nonrivalrous aspect of intellectual property infringement weakens the property rights argument. Generally speaking, writing graffiti on someone else's building damages the building owner in a different way than making a copy of a book injures the author. One might perhaps think of both as using another's private property to facilitate one's speech,¹⁷⁵ but graffiti, unlike unauthorized copying, interferes with the owner's use of his own property. This isn't to say that copyright law is unjustifiable or an unconstitutional speech restriction, only that the constitutional questions surrounding it can't be avoided merely by invoking the tangible property cases.

2. Private Enforcement. Because copyright law is largely enforced by private litigation rather than government prosecution, some argue that it's much less likely to turn into an engine of government censorship.¹⁷⁶ But of course libel law is also enforced almost exclusively by private litigation.¹⁷⁷ Despite this, courts correctly ban preliminary injunctions in libel cases, even those brought by people who aren't public officials.¹⁷⁸

Libel law is a government-imposed restriction that uses government power to restrict speech that people want to see restricted;¹⁷⁹ this is troublesome enough that the Court has insisted on strict procedural safeguards to make sure such a restriction is applied correctly. The same should be true for copyright cases.¹⁸⁰

3. Content-Neutrality. It's also incorrect to argue that intellectual property law is content-neutral and should therefore be subject to laxer rules.¹⁸¹ Copyright liability turns on the content of what is published.

It's true that copyright law draws no ideological distinctions: just like libel law, obscenity law, and fighting words law, it applies equally to speech advocating democracy, speech advocating communism, and speech with no ideological message at all. But while this might make the law viewpoint-neutral, it doesn't make it content-neutral,¹⁸² and it doesn't avoid the prohibition on preliminary injunctions, which applies even to viewpoint-neutral restrictions such as libel law.

4. Subject Matter of the Jeopardized Speech. Some suggest that injunctions in copyright cases pose less of a threat to free speech because they typically involve nonpolitical matters; after all, would it be such a big deal if a court erroneously concluded that Battlestar Galactica infringes the plot of Star Wars?¹⁸³ When the risk of error or chill falls only on such nonpolitical material, the argument goes, there's no need for special procedural protections.

This argument, though, runs counter to fifty years of First Amendment precedents that give art, entertainment, comedy, and the like the same level of protection that is given to political speech.¹⁸⁴ The Court has never suggested that the risk of erroneous restriction is less important as to these materials than as to political advocacy. Vance v. Universal Amusement Co.¹⁸⁵ itself was an obscenity case, involving movies that were probably no more socially valuable than the subject of the typical copyright lawsuit; nonetheless, the Court concluded that their display could not be preliminarily enjoined so long as they had not been definitively found to be constitutionally unprotected.

And of course quite a few copyright cases do involve political or other socially significant speech--as did Harper & Row¹⁸⁶ and the cases that involved the videotapes of the Kennedy assassination¹⁸⁷ and the Los Angeles riot¹⁸⁸--and quite a few libel cases involve material that seems to be as much "entertainment" as a typical movie or novel.¹⁸⁹ A law that precludes news organizations from publishing the film of an assassination--or the audiotape of a White House intern¹⁹⁰--certainly affects "core" protected speech.

5. Copyright Law Furthers Free Speech Values. Nor can copyright law be exempted from the general prior restraint rules on the grounds that "copyright itself [can] be the engine of free expression."¹⁹¹ Copyright law's speech-enhancing effect, coupled with the specific constitutional authorization for copyright law,¹⁹² may justify holding copyright law to be a substantively valid speech restriction, though even this is in large part true because copyright law is limited in the powers it gives copyright owners.¹⁹³ But the point of procedural rules, such as the prior restraint doctrine, is precisely to make sure that even substantively valid speech restrictions don't end up restricting speech that should remain protected.

Preliminary injunctions may be entered without a final adjudication that the speech is unprotected, and thus may restrict speech that is not in fact infringing, frustrating free expression rather than furthering it. And there's no reason to think that applying the prior restraint doctrine would substantially diminish copyright law's benefits to the marketplace of ideas: damages remedies and permanent injunctions entered after a final adjudication would still be available, and would give creators plenty of incentive.

We're also generally skeptical of distinguishing supposedly speech-furthering restrictions from other restrictions. Many kinds of speech restrictions may be seen as furthering free speech values in some way. Justice White argued that libel law furthers free speech: libel law, he claimed, was needed in part because "virtually unrestrained defamatory remarks about private citizens [may] discourage them from speaking out and concerning themselves with social problems."¹⁹⁴ Some have argued that pornography tends to "silence" women, which might suggest that obscenity law may serve First Amendment values.¹⁹⁵ Similarly, some have argued in favor of banning racist speech on the grounds that it silences minorities.¹⁹⁶ Others have claimed that restrictions on the speech of the wealthy further free speech values by preventing well-funded speech from "drowning out" other speech.¹⁹⁷ Less controversially, licensing requirements for use of a public forum serve free speech by preventing concurrent parades or demonstrations that would interfere with each other.

Some of these arguments (for instance, the one about licensing requirements) may justify substantive speech restrictions, and some might not.¹⁹⁸ But even when the arguments support substantive restrictions, they don't justify exemption from the normal procedural rules that ensure that the substantive rules are accurately applied; and they certainly don't justify any special suspension of the procedural rules for copyright law and not for other supposedly speech-enhancing restrictions.

6. The Copyright and Patent Clause. Unlike libel and obscenity law, copyright laws have a specific textual hook in the Constitution: Article I, Section 8 of the Constitution authorizes Congress "[t]o promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries."¹⁹⁹ In Harper & Row, the Court mentioned this as one reason why copyright law is a constitutionally permissible speech restriction.²⁰⁰

But the existence of the congressional power can't exempt copyright law from all First Amendment scrutiny. The Copyright and Patent Clause grants power to Congress, but the point of the Bill of Rights is to restrain the federal government in the exercise of its enumerated powers. In exercising its other powers, Congress is subject to First Amendment constraints: For instance, the government has the enumerated power to run the post office, but this doesn't mean it can refuse to carry communist propaganda;²⁰¹ the government has power to regulate interstate commerce, but this doesn't mean it can impose content-based restrictions on the interstate distribution of newspapers. Likewise, in exercising its copyright power, Congress is bound by the Fourth, Fifth, and Sixth Amendments.²⁰² Copyright law must be bound by the First Amendment too.

That is not to say that the presence of the Copyright and Patent Clause is irrelevant to the First Amendment inquiry. We agree that it would be unsound to read the First Amendment as entirely eliminating the copyright power created by the Framers only two years earlier.²⁰³ The Copyright and Patent Clause does represent the Framers' judgment that "copyright itself [can] be the engine of free expression,"²⁰⁴ so courts ought not, in their zeal to protect speech, eviscerate the incentive that copyright law provides. This was a good reason for the Harper & Row Court to conclude that copyright law is substantively constitutional.

But this argument hardly shows that copyright law ought to be free of the traditional procedural protections available in all other First Amendment cases.²⁰⁵ A general ban on preliminary injunctions wouldn't eviscerate or even greatly diminish the incentive provided by copyright law, just as such a ban doesn't eviscerate libel law or obscenity law; it would merely require the law to be enforced in a slightly different way.

7. Importance of the Government Interest. The interest promoted by copyright law--the interest in providing an incentive for the dissemination of ideas--is, even without regard to its constitutional status, quite important. One might even go so far as to call it "compelling," a term of constitutional significance.

But many speech restrictions--for instance, those aimed at protecting individual reputation, preventing harms caused by the distribution of child pornography, protecting national security, and preventing violent acts incited by violent advocacy--are justified by important (perhaps even compelling) interests, too. The requirement that speech be finally adjudicated to be unprotected before it can be enjoined doesn't prevent these important interests from being served; it just requires that they be served through a certain set of procedures. True, the procedures might in some cases make the process of vindicating these interests a bit less effective. But the general First Amendment judgment underlying the prior restraint doctrine is that this extra burden is justified by the interest in preventing the punishment of constitutionally protected speech.

8. Irreparable Harm. Courts often stress that copyright infringements presumptively cause "irreparable injury," harm that can't be remedied by an eventual damages award.²⁰⁶ And to a certain extent that's true: preventing illegal behavior is generally more effective than is trying to compensate for the harms after they have occurred.²⁰⁷

But this is true for virtually all kinds of harmful speech; for some kinds, such as libel, it's even more true than for copyright infringement. Copyright law is aimed primarily at ensuring that authors are economically rewarded so that they and others will continue to create new works of authorship²⁰⁸--damages can generally reward authors relatively adequately and are often not terribly hard to estimate. Reputation, on the other hand, once sullied can never be perfectly repaired. Damages are only a highly imperfect palliative, may be impossible to get from an indigent defendant,²⁰⁹ and in any event are very hard to estimate, even roughly. Further, one might reasonably contend that enriching plaintiffs is not really the point of libel law, as it is with copyright: certainly one sees many cases of copyright owners licensing their rights in exchange for money, while it is harder to imagine people regularly selling others the right to defame them. Yet despite all this, libels may not be preliminarily enjoined.

And even if the harm caused by copyright infringement really were "irreparable," this shouldn't matter. The point of the prior restraint doctrine is that prevention of potentially harmful speech isn't allowed without a final judicial finding that the speech is unprotected.²¹⁰ This necessarily causes some harm to the government interest (and to the private plaintiff's interest), whether it's the interest in protecting people's reputations, in preventing the moral harms caused by obscenity, or in preventing competitive injury to copyright owners. But under the prior restraint doctrine, it's a harm that must be borne. It certainly is borne by libel victims; why not copyright owners?

9. Ideas Can Be Expressed in Another Way. Copyright law, properly applied, generally restricts expression and not ideas or facts. Even a preliminarily enjoined speaker may be able to communicate the gist of his message, so long as he expresses it differently enough. Thus, some have suggested that copyright law is not a speech restriction because it leaves open adequate alternate channels of communication by allowing defendants to choose different words to express their ideas.²¹¹

But obscenity law likewise restricts expression more than idea. If a city holds up the display of a sexually suggestive movie, without providing the Freedman v. Maryland protections,²¹² the filmmaker might still be able to communicate his ideas by just cutting or softening some of the more explicit scenes. Nonetheless, the prior restraint doctrine applies to obscenity law; and this makes sense because, as the Court pointed out in Cohen v. California,²¹³ "one can[not] forbid particular [expression] without also running a substantial risk of suppressing ideas in the process."²¹⁴

Furthermore, while a damages award or a permanent injunction will by definition punish or restrict only copying of expression, preliminary injunctions may well enjoin even speech that will ultimately turn out to copy only facts or ideas. This is the nature of preliminary relief: because the test for entering a preliminary injunction turns on a reasonable likelihood of success on the merits, some fraction of preliminary injunctions will eventually prove to have enjoined speech that's not infringing--speech that copies only facts or ideas, or speech that's a fair use.²¹⁵ So preliminary injunctions in copyright cases, like preliminary injunctions in libel cases, may often suppress facts and ideas, and not just particular modes of expression.

10. "The Only Question Is Who Gets to Do the Publishing." Some might argue that copyright law is unlike other restrictions because it doesn't actually suppress speech: the expression involved will in due course be published, but by the copyright owner and to his financial benefit. An injunction won't deprive the public of the speech; it will just cause the speech to come from a different person. As with the right of publicity, the argument would go, "the only question is who gets to do the publishing."²¹⁶

In cases of nonliteral copying, though, this factual claim will often be wrong. Nonliteral copiers, even when they use another's expression, are by definition creating something different from (albeit in some ways similar to) the copyright owner's work.²¹⁷ Sometimes the differences are quite dramatic; copying can be infringement even when the copier adds a great deal, and thus only part of the copy is substantially similar to the original.²¹⁸ A satire of the O.J. Simpson trial based on Dr. Seuss's The Cat in the Hat is clearly quite different from the original Seuss book, and yet it was preliminarily enjoined as a likely infringement.²¹⁹ The movie 12 Monkeys was vastly different from a drawing on which the visual design of one of its scenes was based, and yet the movie's distribution was preliminarily enjoined as a likely infringement.²²⁰ And when the expression is enjoined, the public will only see what the copyright owner is saying, not what the would-be copier wanted to say.

What's more, even if the copying is literal or close to it, a copyright injunction may well prevent the expression from reaching the public altogether, at least for the many decades that the copyright lasts; some copyright owners use copyright law precisely for this purpose.²²¹ The whole point of a copyright injunction, as opposed to a suit for damages, is not merely to redistribute the profits, but to prevent dissemination of defendant's work.²²² There might be good reasons, in the copyright owner's mind or even the public's, to restrict distribution of the new work, but it seems hard to argue that such an injunction isn't really a speech restriction.

11. No Need for Timeliness. Some might argue that brief, temporary injunctions of movies or books pose little danger to free speech because such media, unlike newspapers or demonstrations, aren't particularly time-sensitive: it doesn't much matter whether you see a movie or read a book today or a week or two later.

This is a plausible argument, and is perhaps the best defense (though maybe still not an adequate defense) of the Court's toleration of prescreening systems for books and movies, so long as such systems really do create only a brief delay.²²³ It might justify some preliminary injunctions in copyright cases, and perhaps even libel cases, involving non-time-sensitive media.

But even for these injunctions, Freedman v. Maryland imposes certain requirements which copyright injunctions don't now satisfy--preliminary injunctions in copyright cases can last for many months, hardly the brief restraint that Freedman contemplates.²²⁴ And the timeliness argument can't justify any preliminary injunctions of more time-sensitive publications, such as newspaper articles or television programs or even articles in weekly magazines.

12. Tradition: "A Copyright Question, Not a Free Speech Question." Finally, we've heard some people argue that it's a mistake--a form of "constitutional law imperialism"--to disrupt, on recently discovered First Amendment grounds, a balance carefully established over 200 years. Whether preliminary injunctions should be available in copyright cases should be seen, the argument goes, as a copyright law question, not as a free speech question.

Constitutional law imperialism, however, isn't our invention. The theory of Marbury v. Madison is that all U.S. laws are subject to the U.S. Constitution;²²⁵ and as the Court has clearly held, no "formulae for the repression of expression" can claim "talismanic immunity from constitutional limitations."²²⁶ Tradition alone cannot prevent constitutional scrutiny, as the Court made clear with respect to obscenity law and libel law, which had long been seen as not "rais[ing] any Constitutional problem"²²⁷ but which were ultimately held to be restrained by free speech principles.²²⁸ Sometimes, as in Harper & Row, the Court will conclude that this or that speech restriction passes constitutional muster; but all speech restrictions, regardless of their historical provenance, must be tested by First Amendment standards.

Moreover, as we discussed in Part I.B, the tradition of routinely issuing preliminary injunctions in U.S. copyright cases is a relatively recent one, dating back largely to the early 1900s. Even the tradition of applying copyright law to adaptations and abridgments, as opposed to literal or very nearly literal copying, is mostly a creature of the 1900s.²²⁹ This broader procedural and substantive copyright protection may in fact be justifiable, but not on the grounds that "we've always done it this way, so it must be constitutional." If copyright law is to be exempted from the normal free speech rules, it should be because of some substantive difference between copyright and other speech restrictions, not simply because it has not historically received much constitutional scrutiny.

E. Against Special Pleading for Copyright

We thus see no compelling normative reason to treat copyright differently from other speech restrictions, restrictions that are likewise substantively valid but that nonetheless require certain procedural safeguards. And we see a good reason not to treat copyright more favorably than other speech restraints.

The First Amendment demands sacrifices from many who earnestly believe in the legitimacy of their favorite speech restrictions. Exempting restrictions such as copyright law--which is largely identified with rich and powerful interests--from the "normal" rules of the First Amendment throws the legitimacy of free speech protection into question. Partisans of restrictions on bigoted speech, for instance, routinely defend the restrictions they propose by pointing out that not all speech restrictions are looked upon with disfavor;²³⁰ while we disagree with their substantive proposals, they are right to demand that existing exceptions to the First Amendment get the same scrutiny as the proposed ones. Likewise, libel victims who are denied preliminary injunctions in their lawsuits against publishers deserve an explanation for why the same publishers routinely get preliminary injunctions in their lawsuits against supposed infringers.

Any special preference for copyright law must thus be justified by some substantial difference between copyright and other speech restrictions. Where, as here, no such difference exists, favoritism for a particular kind of speech restriction risks corroding public respect for First Amendment law more generally. And, ironically, publishers and producers--the very people who often benefit from the way copyright law now ignores First Amendment protections--have the most to lose from any corrosion of First Amendment protection outside copyright.²³¹

III.DOES PRIOR RESTRAINT DOCTRINE MAKE SENSE?

A. Questioning Prior Restraint Doctrine

Thus, under orthodox prior restraint doctrine, many (though, as Part IV describes, not all) preliminary injunctions in copyright cases are unconstitutional. But could it be that the law of copyright preliminary injunctions is right, and prior restraint doctrine is wrong?

Why not tolerate some preliminary restraints on speech, even given that such restraints will sometimes restrict speech that will ultimately be found constitutionally protected? The restriction on protected speech will only be temporary, the argument goes, and will be for a good cause: without the restriction, the underlying government interest, be it defending national security, protecting reputation, or stimulating creative expression, wouldn't be furthered as effectively. Sure, under a prior restraint regime, some books or movies or newspaper stories will have to be delayed until their truthfulness or absence of infringement is ultimately adjudicated; but as a result, libel law and copyright law, both worthy rules, will be better vindicated.

We think this is a credible argument. Prior restraint doctrine does cause real harms to government interests, and needs a more thorough defense than the Court has given for it in the past.²³² And certainly the copyright experience shows that in at least one area of the law prior restraints are commonplace, even uncontroversial, and yet the sky hasn't fallen.

Nonetheless, we think the ban on prior restraints, including preliminary injunctions, makes practical as well as doctrinal sense. The free speech costs of preliminary injunctions generally do exceed the benefits--especially in copyright cases.

B. Free Speech Cost of Delay

To begin with, the temporary delay created by preliminary injunctions is still a significant speech restriction: the time between the issuance of a preliminary injunction and a decision on the merits is rarely less than a year, and often more. When the work is tied to a particular current event, such as an election campaign, trial, scandal, or policy debate, this cost is dramatic--by the time the injunction is lifted, the work might have become largely pointless. Other works, too, may reflect quickly-changing trends and public interests: delaying the release of a movie or a musical work by a year, for example, may deprive it of a large part of its productive lifespan, and may cost millions of dollars.

Furthermore, the collateral bar rule precludes the defendant from disobeying the injunction and then attacking its constitutionality in court later.²³³ If a court enjoins you, you can be jailed or fined for violating the injunction, even if a court later finds that the injunction was improperly issued. So the injunction is in fact likely to delay the speech: defendants can't be expected to make a judgment that this injunction really is worth disobeying in order to get the information out.²³⁴ And the delay may well become permanent; many cases settle on terms favorable to the plaintiff once a preliminary injunction has been issued.²³⁵

C. The Risk that Preliminary Injunctions Will Restrict Speech That Ultimately Proves Protected

The risk of delay is, of course, relevant primarily for those works that are ultimately found to be protected by the First Amendment. If preliminary review almost always correctly predicted the result on the merits, then relatively little protected speech would be lost.

But why should we expect this to happen? For a preliminary injunction, a court need not conclude that the plaintiff is 95% likely to succeed on the merits, or even 51% likely. The threshold likelihood of success varies depending on the other equitable factors, such as the balance of hardships; if a court concludes that the balance of hardships tips far enough in the plaintiff's favor, it need only find "sufficiently serious questions going to the merits to make them fair grounds for litigation,"²³⁶ or, in the Seventh Circuit, a "better than negligible chance that [the plaintiff] will succeed on [the merits]."²³⁷ And recall that in copyright cases, courts generally presume that the balance of hardships and the public interest tip in the plaintiff's favor.²³⁸

When even a low probability of success on the merits can yield a preliminary injunction, we should expect that in many cases even speech that would ultimately be found protected will be preliminarily enjoined. If a court concludes, for instance, that a plaintiff should get a preliminary injunction even though he has only a 33% chance of success on the merits (perhaps because the balance of hardships favors him), this in effect represents a judgment that it is better that two constitutionally protected works be temporarily enjoined than one copyright infringement remain unenjoined. Perhaps this is a sound judgment as a matter of pure copyright law (though even there one might question the wisdom of the tradeoff²³⁹), but it is particularly troubling from a First Amendment perspective.

Moreover, temporary restraining orders may be issued on only a few hours of reflection, and preliminary injunctions must generally be issued within ten days of the TRO.²⁴⁰ The judge has limited time for contemplation. The parties have limited time for briefing. Preparation for a typical copyright trial, even a bench trial, generally takes many months; the arguments about why one work isn't substantially similar in its expression to another, or about why it's a fair use of another, are often sophisticated and fact-intensive, and must be crafted with a good deal of thought and effort. Unless we believe that the lion's share of that time is wasted--in which case we should just make preliminary relief automatically final, and save everyone a lot of time and money--it seems likely that omitting this effort will mean that preliminary review will often reach different results than would be available on final judgment.²⁴¹

The decision whether to grant a preliminary injunction is also often explicitly driven by the court's goal of "preservation of the status quo,"²⁴² which in copyright cases tends to refer to the status quo existing before the defendant's publication. Thus, where the ultimate result is in doubt, many judges may often lean in favor of an injunction; even if they don't explicitly cite this factor, the status-quo-preserving culture of preliminary injunction law may push them in this direction. This in effect further lowers the threshold likelihood of success needed for a preliminary injunction, thus increasing the number of cases in which a preliminary injunction will restrict speech that ultimately proves to be protected.

We do not claim that courts will always be ready to grant preliminary injunctions in close cases. Some judges may be slow to issue such injunctions: courts sometimes suggest that the "preliminary injunction is an extraordinary remedy . . . which is at best used sparingly, if at all."²⁴³ But others may take seriously the notion that even a "better than negligible" chance of success on the merits is enough for an injunction if the balance of hardships weighs sufficiently in the plaintiff's favor. This is the nature of discretionary remedies--some courts will exercise their discretion sparingly and others generously. We claim only that in some considerable number of cases, though of course not in all cases or in all courtrooms, preliminary injunctions will lead to the preliminary restraint of speech that is quite likely to ultimately prove constitutionally protected.

D. The Dangers of Judicial Discretion

The First Amendment difficulty is further exacerbated by the deeply discretionary nature of the preliminary injunction balance. One central tenet of First Amendment law is that speech restrictions should rest on standards that are as definite and nondiscretionary as possible.²⁴⁴ And while some uncertainties and judgment calls are of course inevitable, the Court has recognized that greater discretion brings a greater chance that even seemingly viewpoint-neutral speech restrictions will be enforced in viewpoint-discriminatory ways.²⁴⁵

Copyright's fair use test and substantial similarity of expression test are already rather vague,²⁴⁶ but adding the preliminary injunction balance exacerbates the problem. The preliminary injunction framework requires a judge to guess the likelihood of success on the vague merits tests, and then also to "weigh" this likelihood against the balance of hardships (itself a rather vague factor). Even conscientious judges may find it hard to make this judgment uninfluenced by their sense of the value or fairness or correctness of the defendant's speech and of the plaintiff's speech.

First Amendment law has almost entirely avoided such explicit balancing, "weigh all the factors" tests. Though tests such as the ones for obscenity and libel are hardly the brightest of lines, they are at least attempts to create more precise rules. The balancing, as Melville Nimmer pointed out, is generally done at the time the categories are defined, rather than on a case-by-case basis,²⁴⁷ so judgment calls are normally made more or less openly by the Supreme Court, rather than implicitly by many lower court judges exercising their largely unexplained discretion.

The main First Amendment test that involves explicit, unconstrained balancing is the Pickering standard, applicable where the government, acting as employer, imposes restrictions on what its own employees say on matters of public concern.²⁴⁸ The doctrine there specifically calls for "balanc[ing] . . . the interests of the [employee], as a citizen, in commenting upon matters of public concern" against "the interest of the State, as an employer, in promoting the efficiency of the public services it performs through its employees."²⁴⁹ The Court has acknowledged that "such particularized balancing is difficult,"²⁵⁰ and this seems to be an understatement. From all we've seen of the lower court decisions, the test is essentially indeterminate in all but the easiest cases. This indeterminacy is troubling even when the government is acting as employer, a context in which the Court properly accepts greater speech restrictions; it is especially improper for rules imposed by the government acting as sovereign.²⁵¹

E. The Comparatively Small Marginal Benefit of Preliminary Injunctions Relative to Damages

The argument against the prior restraint doctrine is that preliminary injunctions are necessary to adequately vindicate the government interest. We might really need preliminary injunctions in (for instance) libel cases, because people's reputations would otherwise be dramatically underprotected. After all, once a libel is published, there's no way to restore a reputation; damages are an extremely poor substitute. Likewise for preliminary injunctions in national security cases.

The Court has disagreed with this argument even in the libel and national security contexts,²⁵² but we need not go into this broader question here, because the pro-injunction argument has less force in copyright cases than it does in the other cases where preliminary injunctions are generally forbidden. Copyright law is primarily intended to provide a financial incentive that encourages people to contribute to the marketplace of ideas.²⁵³ Allowing damages awards and permanent injunctions but disallowing preliminary injunctions would not dramatically undercut this goal.

Copyright owners would still be rewarded fairly well by damages awards. Perhaps they'd be somewhat less rewarded than they would be if they could get preliminary injunctions, but there's no reason to think that the difference would be especially great. Of course damages, including consequential damages such as the loss of indirect business opportunities, would often not be easy to determine, but no more so in copyright than in any other field. Damages awards are thus a less restrictive--because focused on actual infringers, rather than merely possible infringers--but pretty much equally effective alternative to preliminary injunctions.²⁵⁴

In copyright preliminary injunction cases, courts often say that denying the injunction would cause "irreparable harm," but this is a legal fiction, not a judgment about economic reality.²⁵⁵ The theory is that copyright is property, property rules are supposed to give one an entitlement to exclusive use, not just compensation for use, and therefore an injunction is the proper way to vindicate the interest. The irreparable harm judgment doesn't rest on any particularized finding that awarding damages would inadequately compensate the copyright owner, or would substantially interfere with the government interest in providing people an incentive to create.

It may be, of course, that damages will be hard to calculate because the goods in question are unique. Arguments of this sort are often used to bolster the property rule nature of intellectual property rights.²⁵⁶ But while this may justify permanent injunctions, it doesn't necessarily justify preliminary injunctions, especially where speech is at stake.²⁵⁷ Courts make difficult valuation decisions in all sorts of cases. They must make them even when granting preliminary injunctions in copyright suits, because such injunctions must be accompanied by a bond to compensate the defendant in case the injunction is found to have been erroneously entered.²⁵⁸ This calculation is as important and as difficult as the calculation of the ultimate damages; the bond is itself a damages measure, albeit one aimed at compensating the defendant rather than the plaintiff, and the amount of the bond also affects the course of litigation by influencing the settlement discussions. If courts are going to engage in the difficult process of valuation anyway, it's hard to see why they can do it for purposes of calculating bonds but not damages.

We respect the law's judgment that copyright is a form of property, and we sympathize with the outrage that people might feel when something that they've been told is theirs is nonetheless used by others without their consent, even if they are eventually compensated in some measure. But it's important to recognize that intellectual property rights, unlike other property rights, are a form of content-based, government-imposed speech restriction. The mere fact that the restriction is denominated a "property right" should not exempt it from conventional First Amendment scrutiny,²⁵⁹ or justify government action, such as a preliminary injunction, that may end up restricting speech which ultimately proves to be constitutionally protected.²⁶⁰ This is especially so where, as a practical matter, damages would be a fairly effective remedy.

The primary instance where damages are not a relatively effective remedy is where copyright law is used not to assure economic reward but to protect the copyright owner's "moral right" to control how his work is used. A copyright owner might be outraged by having his work adapted in a certain way, or by having it linked with some other work, cause, or person, or (in case of unpublished works) by having it published at all.²⁶¹ Damages here would probably do little to palliate this outrage caused by loss of artistic control.

But we don't believe that the interest in protecting owners' supposed moral rights to control their works is weighty enough to justify preliminary injunctions. We're not even sure that, standing alone, it would be weighty enough to justify subsequent punishment. Recall that Harper & Row accepted copyright law largely because of the economic incentive that it provides for creative expression, and thus because of its contribution to the marketplace of ideas;²⁶² the moral rights rationale has not been similarly validated by the Court as a justification for restricting speech, and we doubt that it should be.²⁶³ Free speech requires us to suffer considerable indignities, for instance from statements of opinion that are injurious to our reputations, or from accurate statements of fact that are nonetheless unfair or emotionally distressing.²⁶⁴

Indeed, in many areas copyright law itself refuses to give owners categorical control over their works. Congress has chosen to enact only a limited and highly qualified moral rights law.²⁶⁵ The fair use doctrine has been interpreted to protect unauthorized parodies of a copyright owner's works, despite the "moral" objection the owner likely has to the publication of such a parody.²⁶⁶ Similarly, the copyright owner has no right to prevent a newspaper from publishing a devastating review of her new book, or even from using snippets of the book in the review to bolster the critique.²⁶⁷

In any event, even where certain speech restrictions, such as restrictions on false and defamatory statements of fact, are tolerated, procedural protections like the ban on preliminary injunctions are required. We don't see why authorial dignity is so important that it's entitled to special protection where other dignity interests are not.

F. Free Speech Benefit Incidental to Allowing Temporary Publication Even of Infringing Material

What if a court refuses to enter a preliminary injunction on prior restraint grounds, but the speech is ultimately found to be unprotected? In libel and obscenity cases, we might in retrospect wish that the speech had been preliminarily enjoined. False statements of fact and obscenity are theoretically of no constitutional value,²⁶⁸ so the fact that they could be distributed before the final adjudication on the merits is all minus and no plus.

But works that build on another's expression, such as The Nation's excerpt of the Ford memoirs,²⁶⁹ actually do have constitutional value even if they infringe a copyright. They are new works that may enlighten or entertain in ways the originals didn't; their original, noninfringing components often dwarf the infringing snippet, such as in the movie cases, where full-length, original feature films have been preliminarily enjoined because they included a shot of a copyrighted chair, poster, or sculpture in one scene.²⁷⁰ Tolerating dissemination of these potentially infringing works before the final adjudication on the merits increases the range of material available to the public. While withholding a preliminary injunction may impose some cost on the government interest underlying copyright law (albeit a cost substantially mitigated by the availability of damages), there is at least some countervailing benefit to the marketplace of ideas.

This is not to say that copyright law is substantively unsound. The social benefit of allowing damages awards and even permanent injunctions in copyright cases may justify the free speech cost of restricting creative adaptations. But the fact that the creative adaptations, even if ultimately found to be infringing, have constitutional value diminishes the aggregate social damage done by the continuing infringement, and therefore weakens the argument that this speech must be suppressed before a final determination that it really is infringing.

G. Conclusion

It thus seems to us that, at least with respect to preliminary injunctions in copyright cases, the traditional ban on prior restraints is sound. Allowing preliminary injunctions imposes more free speech cost, and creates fewer copyright benefits, than might at first appear.

Other situations might present closer calls. For instance, in libel cases, where damages are indeed a grossly inadequate substitute for prevention of publication, the prior restraint doctrine does dramatically burden the government interest. We think that on the whole the law is nonetheless right not to allow preliminary injunctions even in libel cases, but this is at least a tougher question. Likewise, in at least some narrow category of national security contexts, some sort of preliminary injunctions are inevitable; the court in United States v. Progressive, Inc.²⁷¹ may have been right to enjoin publication of H-bomb secrets,²⁷² though we don't wish to debate that issue here.

Copyright law, though, presents a comparatively weak case for preliminary injunctions and the concomitant suppression of speech that ultimately proves to be protected. It is thus all the more remarkable that courts have applied prior restraint doctrine in the libel, obscenity, and even national security contexts, but not to copyright law.

IV.MODIFYING INTELLECTUAL PROPERTY REMEDIES IN LIGHT OF FREE SPEECH DOCTRINE

We thus believe that the First Amendment requires some significant modifications to the law of preliminary injunctions in intellectual property cases. Ex parte TROs should almost always be forbidden;²⁷³ preliminary injunctions should likewise generally be prohibited in cases involving nonliteral copying.²⁷⁴ Still, a considerable range of injunctive relief may remain permissible.

A. Injunctions in Non-Speech Cases

Most obviously, our argument applies only when the work falls within the constitutional definition of "speech." Certain types of functional expressive works may not qualify as speech at all. Reproduction and distribution of computer object code, for instance, could be preliminarily enjoined with no First Amendment difficulty.²⁷⁵

B. Permanent Injunctions

As we discuss above, we don't believe that prior restraint doctrine prohibits (or should prohibit) permanent injunctions of unprotected speech, entered after a full consideration of the merits, whether at trial or on summary judgment.²⁷⁶ There's some controversy about this, especially in libel cases, where some courts refuse to enter even permanent injunctions.²⁷⁷ Still, we think that permanent injunctions are not a materially greater burden on speech than is the risk of financial or even criminal liability; and if one opposes both permanent injunctions and damages liability as undue restraints on free speech, then one's quarrel is with the Supreme Court's recognition of a copyright exception in Harper & Row,²⁷⁸ not with our particular proposal. Permanent injunctions against copyright infringement should thus be constitutionally permissible.²⁷⁹

Our approach is therefore consistent with the traditional view that intellectual property should be protected by means of a property rule (a presumptive entitlement to injunctive relief) rather than a liability rule.²⁸⁰ We would simply require a final showing of entitlement before the property rule is invoked.

C. Preliminary Injunctions in Piracy and Other "Easy" Cases

The danger of preliminary injunctions is that they may temporarily suppress speech that ultimately proves to be protected. If a judge can, at the time of the preliminary injunction hearing, make a reliable finding that the speech is constitutionally unprotected, rather than just that it might be unprotected, then there should be no problem with issuing the preliminary injunction.

This could happen whenever the defendant has made identical or nearly identical copies of the plaintiff's works, and there is no claim of fair use, but only some other copyright defense (such as that defendant has a supposedly valid license, or that plaintiff's copyright wasn't properly renewed, or something along those lines). Most cases alleging outright piracy, as well as most claims that the defendant has exceeded the boundaries of its license, would fall within this category.

In this situation, the defendant's conduct is not constitutionally protected speech, even if the defendant ultimately proves that it's not a copyright infringement. Harper & Row, as we read it, divides uses of another's work into two areas: on the one hand, use of another's idea or facts, or fair use of another's expression, which is constitutionally protected; on the other hand, use of another's expression that's not a fair use, which falls within the "copyright exception" to the First Amendment.²⁸¹ So long as Congress and the courts restrict speech that's within this copyright exception, there's no First Amendment problem.²⁸² And while the First Amendment imposes certain limits on copyright law--the requirement of copying, the idea-expression dichotomy, and some aspects of the fair use doctrine--it does not constitutionalize every nuance of the Copyright Act.

Thus, Congress could, for instance, abolish the registration requirement for a lawsuit, or bring some lapsed copyrights back into private hands,²⁸³ without violating the First Amendment.²⁸⁴ Likewise, courts can, pursuant to the Copyright Act's authorization,²⁸⁵ preliminarily enjoin speech that clearly falls within the "copyright exception" without violating the First Amendment. So long as there's a reliable determination on the merits that the speech does fall within the exception--rather than just a finding that the speech might fall within the exception--courts may act even based on only a tentative determination of other, constitutionally irrelevant, factors related to the copyright claim. This may inadvertently lead to temporary suppression of speech that ultimately proves to be noninfringing, but it won't lead to temporary suppression of speech that is constitutionally protected.

Under this approach, the initial determination that the defendant's speech is constitutionally unprotected will take place without a jury trial, but First Amendment law generally doesn't require that speech restrictions be ratified by a jury--just that they be made after a full consideration on the merits. For instance, petty criminal speech restrictions, those that result in a maximum sentence of no more than six months in jail, may be imposed without a jury;²⁸⁶ in civil free speech cases where there's no disputed issue of material fact, damages can be awarded without a jury; and given that the Seventh Amendment isn't incorporated against the states,²⁸⁷ a state could conceivably dispense with juries altogether in civil cases, even ones involving free speech, without violating the First Amendment.

This category of "easy cases" isn't necessarily limited to ones in which no speech-related argument is raised by the defendant. If a counterfeiter claims fair use in a situation where it's clearly not applicable, courts may still be able to dispose of the fair use claim without a full trial using the normal standards for pretrial decisionmaking: judgment on the pleadings, partial summary judgment on the issue of liability, and the like. The First Amendment doesn't preclude summary judgment in libel cases,²⁸⁸ and it shouldn't in copyright cases either.

Of course, not all cases will fit into this category. Where the defendant raises a credible argument worthy of trial that he didn't copy the work, that he copied only the unprotectable elements, or that his copying was protected by fair use, he is entitled to a full hearing on the merits, and courts should not impose a preliminary injunction before that hearing. But the fact that many of the most obvious cases of copyright infringement won't implicate the prior restraint rule at all should make our proposal less worrisome to copyright owners.

D. Applying the Freedman Exception to Copyright Law?

Freedman v. Maryland²⁸⁹ suggests that there's some room for preliminary injunctions against possibly protected speech, so long as the injunctions create only relatively slight delay, something on the order of