Duke Law Journal
VOLUME 48 NOVEMBER 1998 NUMBER 2

FREEDOM OF SPEECH AND INJUNCTIONS IN INTELLECTUAL PROPERTY CASES

MARK A. LEMLEY
EUGENE VOLOKH††
 

ABSTRACT

Preliminary injunctions against libel, obscenity, and other kinds of speech are generally considered unconstitutional prior restraints. Even though libel may inflict truly irreparable harm on its victim, the most a libel plaintiff can hope for is damages, or perhaps a permanent injunction after final adjudication, not preliminary relief. Professors Lemley and Volokh argue the same rule should apply to preliminary injunctions in many copyright, trademark, right of publicity and trade secret cases.
They note 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, or justify government action that restricts speech which ultimately proves to be constitutionally protected. This is especially so because in most cases, damages would be a relatively effective remedy. The Court's prior restraint doctrine and sound First Amendment policy suggest that preliminary injunctions in intellectual property cases are often (though not always) unconstitutional.

TABLE OF CONTENTS

Introduction
I. Standards Governing Preliminary Injunctions in Copyright Cases
          A. History
                    1. England
                    2. United States
          B. The Modern Rule
          C. Summary
II. Freedom of Speech, Copyright, and Injunctions
          A. Copyright Law as a Speech Restriction
          B. Prior Restraint Doctrine
                    1. Speech May Not Be Restrained Until a Final Judicial Determination That It's Unprotected by the First Amendment
                    2. Why This Rule Makes Sense
                    3. Two Possible Exceptions
          C. Ex Parte Temporary Restraining Orders
          D. Can Preliminary Injunctions of Copyright Infringement Be Distinguished from Preliminary Injunctions of Other Speech?
                    1. Property Rights
                    2. Private Enforcement
                    3. Content-Neutrality
                    4. Subject Matter of the Jeopardized Speech
                    5. Copyright Law Furthers Free Speech Values
                    6. The Copyright and Patent Clause
                    7. Importance of the Government Interest
                    8. Irreparable Harm
                    9. Ideas Can Be Expressed in Another Way
                    10. "The Only Question Is Who Gets to Do the Publishing"
                    11. No Need for Timeliness
                    12. Tradition--"A Copyright Question, Not a Free Speech Question"
          E. Against Special Pleading for Copyright
III. Does Prior Restraint Doctrine Make Sense?
          A. Questioning Prior Restraint Doctrine
          B. Free Speech Cost of Delay
          C. The Risk that Preliminary Injunctions Will Restrict Speech That Ultimately Proves Protected
          D. The Dangers of Judicial Discretion
          E. The Comparatively Small Marginal Benefit of Preliminary Injunctions Relative to Damages
          F. Free Speech Benefit Incidental to Allowing Temporary Publication Even of Infringing Material
          G. Conclusion
IV. Modifying Intellectual Property Remedies in Light of Free Speech Doctrine
          A. Injunctions in Non-Speech Cases
          B. Permanent Injunctions
          C. Preliminary Injunctions in Piracy and Other "Easy" Cases
          D. Applying the Freedman Exception to Copyright Law?
          E. Requiring a Finding of High Probability of Success on the Merits
V. Implications for Other Intellectual Property Laws
          A. Trademarks
          B. Rights of Publicity
          C. Trade Secrets
          D. Patents
VI. Prospects for Implementation of Our Proposals
Conclusion

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 a few days or weeks.290 Copyright law today doesn't give any such assurance of rapid final adjudication on the merits, but perhaps Congress could implement some special fast-track procedure.291

We are skeptical about this proposal, in part because we are skeptical about Freedman itself. As we argue above, even relatively short delays can be significant.292 We doubt that it would be constitutional, for instance, to require that all movies, or even all movies about which libel complaints have been filed, be prescreened for possible libel, even if there were a guarantee of final adjudication on the merits within weeks. And in copyright cases, unlike libel cases, damages do seem to be a practically sufficient remedy. Especially in light of this, we believe that giving a special speech-restricting privilege to copyright owners that is denied to other victims of speech (such as people who are libeled) tends to corrode public respect for First Amendment law.293

Nonetheless, realizing that our broader proposal is quite a substantial step, we insist on at least half a loaf: at the very least, if preliminary injunctions are allowed, there must be some guarantee that the defendants can get a truly prompt final adjudication on the merits. Perhaps such a fast-track solution is politically difficult--courts after all have to adjudicate many other important cases294--but Freedman suggests that the First Amendment requires at least this much.

Time-sensitive works, be they timely news magazine articles, books, or movies, ought not be preliminarily enjoinable even under Freedman; the free speech cost here would be too great. To our knowledge, Freedman has in practice not been applied to time-sensitive works at all, and we think it should not be extended to them.

E. Requiring a Finding of High Probability of Success on the Merits

Finally, even if preliminary injunctions in copyright cases are too deeply entrenched to be dislodged by the above arguments, at the very least they should only be available when there's a truly high probability of success on the merits. Non-speech conduct may properly be preliminarily enjoined based on a finding of "sufficiently serious questions going to the merits"295 or a "better than negligible chance" of success on the merits;296 when a court is enjoining such conduct, the presumption may properly favor "preservation of the status quo."297 But a court should not enjoin speech unless it has considerably greater confidence that the speech is in fact constitutionally unprotected.298

Thus, when a court concludes that the defendant's expression is probably not substantially similar to the plaintiff's expression, or is probably a fair use, the court should never issue a preliminary injunction, even if the balance of hardships tilts in the plaintiff's favor. Likewise when the court concludes that the case is genuinely close: erroneously failing to enjoin speech is better than erroneously enjoining speech, especially when erroneous failure to enjoin speech is remediable by a damages award.299

We suggest that at the very least courts should borrow from libel law, where the Court has required that actual malice be proved by clear and convincing evidence.300 This higher quantum of proof reflects the principle that oversuppression of speech is considerably worse than underprotection of the government interest. If this judgment is true with regard to whether the plaintiff can recover at all in libel cases, it is even more appropriate when the only question is what sort of relief--a preliminary injunction, or damages plus a permanent injunction--the plaintiff can get in copyright cases. Thus, a court should not enter a preliminary injunction unless it is clearly convinced that the speech falls within the copyright exception (i.e., that it's substantially similar to the plaintiff's expression and is not a fair use).301

V.IMPLICATIONS FOR OTHER INTELLECTUAL PROPERTY LAWS

If we're right that preliminary injunctions in copyright cases restrict speech, what does that say about other types of intellectual property cases? After all, virtually all intellectual property regimes rely heavily on "property rules" and enforce injunctions--even preliminary injunctions--as a matter of course.302 Do these injunctions also violate the First Amendment?

The answer to this question is complex, and turns on the particular characteristics of each intellectual property right. We'll try to sketch the outlines of each inquiry below, considering whether laws within each of these fields restrict speech, whether the speech restriction is substantively valid, and whether the law involves prior restraints. But we stress that what follows is indeed only a sketch, much less detailed than our analysis of copyright law. We hope others will build on it to provide a more detailed treatment.

A. Trademarks

Injunctions are the most important remedy in trademark cases. This largely follows from the rationale for trademark law, which is rooted in consumer protection.303 If consumers are deceived by the defendant's use of a mark that is confusingly similar to the plaintiff's mark, awarding damages to the plaintiff won't alleviate that confusion--only preventing the defendant from using the mark at all will serve that purpose. Similarly, trademarks can rarely coexist if consumer confusion results. Thus, one side or the other will frequently have to give up the mark entirely at the conclusion of the case; injunctive relief arguably prevents wasted investment by defendants in doomed trademarks.304 Consonant with this policy goal, injunctive relief is standard in trademark cases.305

At the same time, the burden placed on trademark owners in preliminary injunction cases is somewhat higher than in copyright cases. Most courts require a significant showing of likelihood of success on the merits.306 While courts will normally presume that the plaintiff faces irreparable injury upon a showing of likely consumer confusion,307 and while the public interest in avoiding confusion often favors the plaintiff,308 courts do balance the hardships to each party, and will refuse preliminary injunctions when the equities so require.309 Further, the inquiry into the merits is a detailed one, and a greater showing is required to support an injunction that compels affirmative conduct than one that merely maintains the status quo.310

As to free speech issues, some cases have dismissed prior restraint arguments on the grounds that the trademark laws do not restrict protected speech at all, because the use of a mark is not a speech act, or is not subject to the normal protections afforded speech.311 Other courts have concluded that while trademark laws do restrict speech, they do so in a "content-neutral" manner and are therefore permissible.312 These courts have apparently confused viewpoint-neutrality with content-neutrality.313 Still other courts have simply assumed that trademarks are property interests, and that such property interests must trump the constitutional guarantee of free speech.314 The First Amendment argument cannot be disposed of so easily, however, for reasons explained above in Part II.D.1.

Most trademark cases involve "commercial speech"--that is, speech which proposes (directly or indirectly) a commercial transaction.315 Commercial speech is entitled to First Amendment protection, albeit to a lesser degree than political speech.316 Indeed, the Supreme Court has called the free flow of commercial information "indispensable" to the economy.317 Since trademarks are a form of commercial speech entitled to at least some First Amendment protection, and since a trademark injunction prevents a defendant from using a particular term or phrase, it makes little sense to argue that it doesn't restrict speech at all.318

Further, a significant minority of trademark cases do not involve commercial speech by defendants at all, but rather traditional speech entitled to full protection. Cases in this category include certain parodies of trademarks, political or socially-directed advertisements, the use of the plaintiff's marks outside of advertisements to truthfully refer to the plaintiff's product (so-called "non-trademark use"), product reviews, uses in fiction, commentary or film, and the merchandising of a trademark as a good in itself.319 In none of these cases is the trademark itself used to propose a commercial transaction; rather, it is used as the subject of speech. Still other cases may involve trade dress or product configurations in which the "trademark" is both the product itself and the advertisement for that product. In such cases, use of the trademark may be both commercial and noncommercial speech.320

There are persuasive reasons to permit political and social uses of trademarks. With the importance of brand image in today's economy, trademarks "form an important part of the public dialog on economic and social issues."321 Rochelle Dreyfuss and others have noted that modern expression frequently requires the use of trademarks in their role as social referents--whether or not the product itself is being discussed directly.322 As Judge Kozinski has noted, "[m]uch useful social and commercial discourse would be all but impossible if speakers were under threat of an infringement lawsuit every time they made reference to a person, company or product by using its trademark."323 And as the scope of trademark protection expands to all aspects of the appearance of the product itself and its packaging, more and more trademark cases actually involve noncommercial speech, at least in part.324

The strongest constitutional justification for trademark laws is that, properly construed, they prevent only commercial speech that is likely to cause consumer confusion, and that false or misleading commercial speech can be restricted.325 The Supreme Court took this view in Friedman v. Rogers.326 This may mean that trademark law is substantively constitutional, so long as it is linked to evidence that the defendant's mark is in fact misleading, and so long as the defendant's speech really is commercial.327

By itself, this characterization doesn't do away with the prior restraint problem. Just as in copyright law, the danger addressed by prior restraint doctrine is that speech will be silenced before a court can make a full determination on the merits that the speech is infringing. One certainly couldn't get a preliminary injunction against the publication of a book on the grounds that the contents of the book might eventually be demonstrated at trial to be false or even libelous and therefore constitutionally unprotected. Indeed, the highly fact-specific nature of trademark infringement analysis, which necessarily focuses on what the ordinary consumer thinks, makes it particularly important that courts not leap to the conclusion that a trademark is being infringed without a detailed consideration of the evidence.328 Unless courts can effectively conduct the final trial on the merits at the preliminary injunction hearing, we do not see how they can make this determination with sufficient certainty to satisfy the prior restraint rule.329

The Court has suggested in cautious dictum that the rule against prior restraints may not apply to commercial speech at all.330 There may be something to this distinction, though the courts have been uneven in applying it.331 If it's true that prior restraints against commercial speech are permissible, this would dispose of the prior restraint problem in some trademark cases.332 But not all trademark cases can be resolved in this way. Even if trademark law can properly reach noncommercial speech, a preliminary injunction against such speech seems to us to be an unlawful prior restraint.333

In short, trademark law is substantively constitutional, at least so long as it is limited to cases of consumer confusion. But it may suffer the same prior restraint problem as copyright law. And at the very least, some trademark preliminary injunctions--such as those restricting noncommercial speech--cannot pass constitutional muster.

B. Rights of Publicity

The right of publicity derives from the common law tort of invasion of privacy by appropriation of likeness.334 The publicity tort grants a potentially sweeping right to plaintiffs, making it unlawful to appropriate the name, likeness or identity of another for commercial advantage.335 "Identity" has been interpreted broadly; imitations of the plaintiff's voice,336 pictures of cars similar to the one the plaintiff drives,337 and even the portrayal of robots that remind people of the plaintiff338 have all been held to appropriate the plaintiff's identity. The plaintiff need not be famous, or even known at all; anyone whose likeness is used can sue under the right of publicity.339 The tort generally does not extend to the use of a plaintiff's personal attributes in traditional speech products, such as books or movies, whether factual or fictional, even when the products are sold commercially.340 Rather, it generally covers only advertisements and merchandising, such as T-shirts, coffee mugs, or busts depicting or referring to the plaintiff.341

As with trademark law, some courts and commentators have assumed that limiting the publicity right to commercial speech solves the First Amendment problem, since commercial speech is not entitled to full First Amendment protection.342 But as Judge Kozinski noted in his White v. Samsung dissent, "less protected means protected nonetheless."343 Commercial uses of someone's likeness are "speech" subject to First Amendment protection, and in many cases the speech will have clear political or social content that would be stifled by application of the right of publicity. Indeed, White itself is an example. White involved a humorous parody of the set of the game show Wheel of Fortune in 2012 A.D., in which host Vanna White had been replaced by a robot in a blonde wig. The court's conclusion that because the parody took place in an advertisement it wasn't protectable speech flies in the face of First Amendment doctrine.344

Further, the right of publicity doesn't arise only in the context of commercial speech. The diminished protection afforded commercial speech is generally limited to commercial advertising; certainly the fact that an item is sold commercially doesn't make it commercial speech, or else most newspapers, magazines, and books would fall into this less protected category.345 The courts, though, have found right of publicity violations in some noncommercial uses of a plaintiff's image. For instance, while sculptures--such as busts of Martin Luther King, Jr.--are generally seen as core protected speech, the King estate won a right of publicity case against the distributors of such sculptures.346 Paintings, photographs, and even inscriptions on T-shirts are generally seen as core protected speech,347 but they too may infringe the right of publicity.348

Indeed, from a constitutional perspective, the right of publicity is significantly more troubling than trademark or even defamation law. The goal of both trademark and defamation law is to identify and suppress false speech about a person or product which may mislead the public. By contrast, nothing in the right of publicity requires that the punished speech be false or misleading.349 No one would have thought that the robot in the Samsung ad was Vanna White, or even that she authorized the parody. Instead, White is given the right to control the circumstances in which others can portray her, discuss her, or even conjure up the idea of her,350 radical stuff in a country which prides itself on promoting robust, open debate in the press and among the public. The speech-restrictive potential of the right of publicity goes much further than that of trademark law, or even libel law, and it may mean that the doctrine as a whole is substantively unconstitutional, at least to noncommercial speech but perhaps even as to commercial speech. The First Amendment argument is especially strong in states that apply the right of publicity to merchandising as well as to ads, because merchandise itself is not commercial speech as the Court has defined the term.

In the right of publicity context, courts have sometimes been willing to acknowledge the First Amendment conflict,351 and in particular the dangers of prior restraint.352 To be sure, judicial treatment of the problem has been uneven. Some right of publicity cases have granted preliminary injunctions against imitation or appropriation without much consideration of the First Amendment.353 These cases largely focus attention on the difficulty of compensating reputational injury with damages354--the precise argument that has proven unpersuasive in the libel context.355 But other cases apply the rule against prior restraints to claims based in privacy and publicity.356 The result is a rather puzzling mix of precedents, with no clear doctrinal line separating those cases in which preliminary injunctions are granted from those in which the prior restraint rule is applied.357

For cases that apply the right of publicity to noncommercial speech, the proper outcome seems relatively clear: whether or not the right is substantively constitutional, preliminary injunctions should be considered unlawful prior restraints. For commercial speech cases, the problem is harder. If the Court concludes that the prior restraint doctrine does not apply at all to commercial speech, preliminary injunctions may continue (assuming the right of publicity is determined to be substantively constitutional). But if commercial speech is ever to be protected by the rule against prior restraints, the right of publicity seems a good case for it.

C. Trade Secrets

State (and more recently federal)358 trade secret laws protect the owners of valuable confidential information against the "misappropriation" of that information.359 The definition of "trade secret" is extremely broad, encompassing virtually every sort of information that can be both valuable and secret. But misappropriation is limited to either (1) acquisition by improper means or (2) use or disclosure in violation of a confidential relationship.360 Independent discovery, independent creation, and even "reverse engineering" a publicly available product are all permissible under trade secret law.361

The Uniform Trade Secrets Act provides that "[a]ctual or threatened misappropriation may be enjoined,"362 and preliminary injunctive relief is quite common in trade secret cases.363 In part this stems from the fragility of trade secret protection--public disclosure forever destroys the secret,364 so the harm from unauthorized disclosure is "irreparable" in a sense absent in most copyright or trademark cases.365 Nonetheless, courts in trade secret cases apply the traditional multifactor test in determining whether preliminary injunctive relief is appropriate, and (unlike in copyright cases) give some teeth to each of the requirements.366

The conflict between the First Amendment and trade secret law is less direct than that between the First Amendment and copyright law. This largely results from the limited set of things trade secret law defines as "misappropriation." Some acts of misappropriation--the mere acquisition of a secret by improper means, or the use of information without authorization (for example to design a new product)--may not involve restrictions on speech at all. Restricting disclosure of a secret does limit speech, but there the limitation could perhaps be justified on substantive constitutional grounds because the disclosure occurred in breach of an existing contractual relationship.367 Courts may enforce contracts not to speak through damage awards368 and maybe even through permanent injunctions; this should validate trade secret cases filed against people bound by confidentiality agreements.

Still, preliminary injunctive relief may potentially run afoul of prior restraint doctrine in three sets of trade secret cases. First, publication of a trade secret by a party who isn't bound by the contract might be constitutionally protected even against a damages judgment, and certainly ought to be protected against a preliminary injunction. The interest of the trade secret owner in getting such an injunction is likely to be strong, since the secret itself may be destroyed by the publication; but of course the same could have been said about the United States interest in blocking publication of the Pentagon Papers.369 Nonetheless, courts can't order newspapers not to publish materials, whether the Pentagon Papers or the formula for Coca-Cola, even if the materials were leaked by someone (such as Daniel Ellsberg) who had a duty to keep them private. Some courts have in fact rejected trade secret injunctions on free speech grounds, though the cases are not entirely consistent.370

Second, even a preliminary injunction against a party who is supposedly bound by a confidentiality contract normally would be an unconstitutional prior restraint, for the same reasons we gave in the copyright discussion. Even if one can punish or enjoin speech based on a finding that the speaker promised not to speak, the prior restraint doctrine should prevent injunctions based only on a finding that the speaker might have promised not to speak. As with copyright, there might be some exceptions similar to the ones we discussed in Part IV: Perhaps a preliminary injunction might issue when there's no serious dispute about the existence of the contract and the dispute turns only on constitutionally irrelevant questions,371 or perhaps a Freedman-like system could allow for very brief temporary injunctions with quick appellate review.372 But as a general matter, preliminary injunctions justified by the possibility that the defendant's speech breaches a contract are no more warranted than preliminary injunctions justified by the possibility that the defendant's speech is libelous.

Third, the controversial doctrine of "inevitable disclosure" in trade secret law permits injunctions against employment or prospective speech, even without direct evidence that a secret will be disclosed, where the court considers disclosure especially likely.373 Findings of inevitable disclosure are "exceptional,"374 and are generally not made in a preliminary injunction hearing. But even after a full trial on the merits, a speech-restricting injunction based not on actual or threatened disclosure of a secret, but the mere possibility of inadvertent disclosure, seems troubling from a First Amendment perspective.

Trade secret law may well be substantively constitutional. But both preliminary injunctions against the disclosure of secrets and permanent injunctions against undefined future speech raise serious constitutional issues.

D. Patents

The patent laws give those who apply for and receive a patent on an invention the exclusive right to prevent others from making, using, selling, offering for sale, or importing the invention as described in the claims of the patent.375 Inventions encompass any new, useful and nonobvious machine, process, product, or composition of matter,376 but protection does not extend to literary or artistic works that constitute mere "printed matter,"377 or anything outside the ill-defined scope of the technological arts.378 Inventions protected by patent law enjoy the benefits of a strong property rule. The patent owner gets the right to control virtually any use of the invention for the term of the patent, backed up by a presumptive entitlement to both preliminary and permanent injunctive relief. Preliminary injunctions are granted as a matter of course, on the basis of a relatively simple showing of likelihood of success on the merits. As in copyright cases, courts have found irreparable injury, the balance of hardships, and the public interest to tip in favor of the patent owner.379

Unlike copyright law, though, patent law doesn't ordinarily restrict speech. The statute prevents others from making, using, or selling machines, products, or processes. By and large, those aren't activities that involve speech, and so restricting them shouldn't raise any First Amendment concerns. But there are at least four specific circumstances in which patent law might restrict speech, and in those specific cases a conflict might arise.

First, the patent statute was amended in 1994 to make it illegal to make "offers to sell" an infringing product.380 This punishes speech--albeit commercial speech--directly, but we don't think it is especially troubling as a general matter. If a court can legitimately punish a particular transaction (say, prostitution or selling drugs), it can also sanction speech which proposes that commercial transaction.381 Were it otherwise, anti-solicitation statutes (and indeed general conspiracy statutes) might be unconstitutional. So once a court has finally determined that selling a product is patent infringement, it ought to be able to prevent offers to sell the product too. Preliminary injunctions barring offers for sale might be impermissible prior restraints on speech, but we aren't aware of any such preliminary injunctions, and we don't expect there will be any. If a court can preliminarily enjoin the sale itself, it hardly needs to enjoin the offer too.382

Second, courts have occasionally held that where practicing a method was infringement, telling someone how to practice the method was illegal because it induced infringement.383 The most obvious example is Moleculon Research Corp. v. CBS,384 where the court held that makers of the Rubik's Cube puzzle induced infringement of Moleculon's patent on a method of solving the puzzle by "dissemination of an instruction sheet teaching the method of restoring the preselected pattern with each puzzle."385 This result seems troubling, in part because the patentee himself must disclose to the world how to practice his method.386 If it really is unlawful to tell someone how to practice an invention, there is a conflict between the patent law and the First Amendment. In that limited set of cases, restricting the speech may be substantively unconstitutional; but even if it's not, preliminarily enjoining it would almost certainly be an impermissible prior restraint.387

Third, the Invention Secrecy Act of 1951388 provides that the government may effectively impound certain patent applications on national security grounds by ordering the inventor to keep the invention secret389 and imposing criminal sanctions if the order is disobeyed.390 This is clearly a substantive restriction on speech; whether it is constitutionally permissible to punish disclosure of this information will depend on whether the national security argument justifies the restriction.391 In either event, it is also a prior restraint. The statute establishes a group of agency heads, designated by the President, that must preapprove publication;392 if the group orders an invention kept secret, willful disobedience of that order is itself punishable as a crime.393 While one could perhaps characterize this as a "subsequent punishment," since a violation triggers section 186 and not a contempt proceeding, the fact that section 186 punishes disobedience of the order itself, rather than the underlying speech, seems to bring it within the scope of prior restraint doctrine.394

Finally, one district court has held that computer source code is "speech" protected by the First Amendment.395 Because software is patentable,396 taking this conclusion seriously leads to some tension between patent law and the First Amendment. If writing code is protected speech, a law which precludes the "making" of such code interferes with protected speech. Further, the substantive conflict cannot be resolved by reference to the idea-expression dichotomy or the fair use doctrine, as in copyright law, for patent law has neither.

Even if the Bernstein court's conclusion is upheld, however,397 it probably ought not extend past source code to object code.398 We think most executable software is best treated as a virtual machine rather than as protected expression.399 If we are right, patent law could constitutionally prohibit the use of infringing object code in a computer. This wouldn't totally resolve the First Amendment conflict--an effort to prevent someone from writing, teaching, or perhaps even reading400 source code might be unconstitutional, and preliminary injunctions against such conduct would almost certainly be prior restraints on speech.401 But since most of the commercial value of software comes from its functionality, the patentee's right to control the functional use of the program would be largely unaffected.402

In short, patent law is qualitatively different from copyright law because most of the acts it restricts don't involve speech at all. Thus, there is ordinarily no reason for the rule against prior restraints to come into play. But in certain narrow circumstances patent law may restrict speech. When it does, courts should give consideration to the First Amendment implications of both preliminary injunctions and permanent relief.

VI.PROSPECTS FOR IMPLEMENTATION OF OUR PROPOSALS

We readily acknowledge that our proposals are contrary to everyday intellectual property practice. Though we believe they are dictated by the Supreme Court's First Amendment case law, we realize that the traditions of the bench and bar (plus, of course, the occasional circuit court precedents) often carry more weight in court than does law review reasoning.

Nonetheless, we think our cause is far from hopeless. Quite a few judges have expressed some concern about the problem we identify;403 we hope our doctrinal and policy explanations will support the views of these judges and of others who think as they do, and will help them persuade their colleagues. More broadly, some judges--including some Supreme Court Justices--have begun to express some doubts about the practice of routinely granting injunctions in non-piracy cases.404 Such an attitude may make courts more hospitable to constitutional arguments against preliminary injunctions.

We also suspect that prior restraint arguments may have been hurt in the past by being stated too broadly--perhaps so broadly as to cover all injunctions, permanent or preliminary--and without a clear policy justification. Many judges may be reluctant to accept a total ban on injunctions, especially since banning all injunctions but tolerating all damages awards might seem like a pointless (and harmful) formalism. Our approach, on the other hand, wouldn't affect permanent injunctions or preliminary injunctions in piracy cases, which we think aren't prior restraints; it would cover only preliminary injunctions in non-piracy cases.405 This may be more palatable than a total ban on all injunctions; and we think that it's justified by our explanation of why many preliminary injunctions are indeed more constitutionally problematic than damages awards.

As for contrary circuit precedent, the mere fact that a circuit has upheld preliminary injunctions, without considering the First Amendment, isn't precedent for the proposition that such injunctions are constitutional; "cases cannot be read as foreclosing an argument that they never dealt with."406 Furthermore, a court's upholding of a permanent injunction against a First Amendment challenge should leave the question open as to preliminary injunctions, which we argue raise very different constitutional questions.407 In addition, a court's rejection of the First Amendment argument in, say, a trademark case may leave the question open for copyright and other intellectual property cases. And at least some rejections of the First Amendment argument may, on examination, prove to be dicta.408

We do not think our proposal should necessarily occasion panic among the intellectual property bar. By offering a way to distinguish object code, permanent injunctions, and clear cases of piracy from those cases presenting serious questions of copying, substantial similarity of expression, or fair use, we hope that we've restricted our approach to cases in which the copyright plaintiff's interest in a preliminary injunction is weakest and the First Amendment values are strongest.

True, copyright plaintiffs in any particular case will no doubt want a preliminary injunction any way they can get it, but copyright owners are also copyright defendants, as many movie and television studios can attest. Moreover, publishers would also do well not to trample the First Amendment in their rush to protect intellectual property; that sort of short-sightedness could come back to haunt them the next time they're accused of libel or obscenity. In those cases that do in fact present substantial constitutional questions, we think copyright law will be a stronger institution if it addresses those questions head-on, rather than trying to evade the problem.

In some circuits and for some issues, the precedent may indeed be quite clearly against us: the Second and Ninth Circuits have specifically held that preliminary injunctions, including preliminary injunctions of noncommercial speech, are permitted in trademark cases.409 The Ninth Circuit case also might be read as rejecting the prior restraint argument for copyright cases, though it's not clear whether that's dictum or holding.410 Nonetheless, in many circuits these questions remain unresolved, and even the Second and Ninth Circuits might change their minds. The cases rejecting the prior restraint argument aren't particularly well-reasoned, and generally don't confront the points we make here; these circuits might well be willing to reconsider the question en banc.

In any event, we believe that our argument is strong enough that defense lawyers should have little to lose, and perhaps quite a bit to gain, from making it. Conventional wisdom does change; tradition sometimes does yield to doctrinal and policy arguments. Traditional libel law, for instance, was firmly entrenched throughout the United States until the Court's decisions since the 1960s--not only New York Times Co. v. Sullivan,411 but also Gertz v. Robert Welch, Inc.,412 Philadelphia Newspapers, Inc. v. Hepps,413 and others--drastically restructured it.

Likewise, the formerly commonplace bans on lawyer advertising were dramatically cut back by Court's commercial speech decisions of the 1970s and 1980s.414 And in Texas Monthly, Inc. v. Bullock,415 the Court struck down formerly widespread tax exemptions for religious publications.416 With the possible exception of the libel decisions, these cases weren't prompted by sea changes in broad social attitudes or in the Court's composition. Rather, they seem to reflect the force of general constitutional doctrine applied to specific cases, and suggest that this doctrine can indeed override common law and legislative traditions.

CONCLUSION

In warning against the dangers of preliminary injunctions that restrain speech, the Supreme Court wrote:

[Allowing preliminary injunctions in obscenity cases would authorize] prior restraints of indefinite duration on the exhibition of motion pictures that have not been finally adjudicated to be obscene. . . . Such prior restraints would be more onerous and more objectionable than the threat of criminal sanctions after a film has been exhibited, since nonobscenity would be a defense to any criminal prosecution. . . . 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.417

Substitute "infringing" for "obscene" in the Court's language; the rationale remains the same, but the results reached by the courts shift radically.

We don't see the justification for this difference. Preliminary injunctions in copyright cases--and in certain other types of intellectual property cases--are at least as troublesome as preliminary injunctions in obscenity or libel cases. We believe that under the Court's prior restraint jurisprudence many intellectual property preliminary injunctions are therefore unconstitutional.

At the same time, we think that this is a good opportunity to reflect not only on preliminary injunction practice but also on prior restraint jurisprudence itself. If you find our proposal to be imprudent--if you think preliminary injunctions in copyright cases shouldn't generally be declared unconstitutional--then perhaps this suggests that prior restraint doctrine has gone too far, and might deserve some trimming for all claims. Or if you reject our broadest proposal but accept our compromise based on Freedman v. Maryland, then perhaps this shows that the much-maligned Freedman makes more sense than many might at first think.

We ultimately believe that existing prior restraint doctrine is, despite its costs, rather sensible; but we think that it does deserve reexamination, and that copyright law, a field where preliminary injunctions are routinely issued, provides a good case study for such reexamination. But regardless of whether prior restraint doctrine is reexamined, we think that courts must apply it consistently. For all its importance, intellectual property law shouldn't be entitled to a First Amendment free ride.


† Professor of Law, University of Texas School of Law; Of Counsel, Fish & Richardson P.C., Austin, Texas.

†† Acting Professor, UCLA School of Law (volokh@law.ucla.edu). Many thanks to Howard Anawalt, John Barton, Craig Bloom, Jack Brown, Dan Burk, Rochelle Cooper Dreyfuss, John Frank, Paul Geller, Mark Gergen, Paul Goldstein, Rose Hagan, Rick Hasen, Marci Hamilton, Paul Heald, Justin Hughes, Dennis Karjala, Bruce Keller, Bob Kreiss, Doug Laycock, Larry Lessig, Sandy Levinson, Jessica Litman, Dan Lowenstein, David McGowan, Peter Menell, Neil Netanel, Thu Minh Nguyen, David Nimmer, Mark Perkins, Malla Pollack, David Post, Scot Powe, David Rabban, Martin Redish, Seana Shiffrin, Lionel Sobel, Peter Swire, Ragesh Tangri, and participants in a faculty workshop at the University of Texas School of Law for their helpful comments.

1. See infra notes 109-112 and accompanying text.

2. See infra Part II.B.1.

3. See infra Part II.A.

4. See Henry P. Monaghan, First Amendment "Due Process," 83 HARV. L. REV. 518 (1970).

5. Note that we speak here of copyright claims brought based on material that's "expression" for copyright purposes--books, movies, songs, paintings, and so on. Our argument doesn't cover copyrighted software, which (at least in object code form) probably doesn't qualify as speech for First Amendment purposes. See infra Part IV.A and notes 395-402 and accompanying text.

6. An Act for the Encouragement of Learning by Vesting the Copies of Printed Books in the Authors or Purchasers of Such Copies During the Times Therein Mentioned, 1709, 8 Anne, ch. 19 (Eng.). For a discussion of analogous copyright cases prior to that time, see Howard B. Abrams, The Historic Foundation of American Copyright Law: Exploding the Myth of Common Law Copyright, 29 WAYNE L. REV. 1119 (1983); David Lange, At Play in the Fields of the Word: Copyright and the Construction of Authorship in the Post-Literate Millennium, 55 LAW & CONTEMP. PROBS. 139, 140-41 (1992).

7. See, e.g., Donaldson v. Beckett, 1 Eng. Rep. 837, 17 PARL. HIST. ENG. 953, 989 (H.L. 1774) (statement of De Grey, Lord Chief Justice) (observing that the availability of a remedy at law is no barrier to the issuance of an injunction).

8. Id.

9. Id. at 989-90.

10. Abrams, supra note 6, at 1143 (citing Eyre v. Walker (Ch. 1735), Motte v. Falkner (Ch. 1735), Walthoe v. Walker (Ch. 1736), and Tonson v. Walker (Ch. 1739), none of which produced a reported decision).

11. Donaldson, 17 PARL. HIST. ENG. at 990 (statement of De Grey, Lord Chief Justice). The practice seems to have been to grant at least some such injunctions on the papers, without a hearing. Accord Smith v. Johnson, 22 F. Cas. 579, 579 (C.C.S.D.N.Y. 1859) (No. 13,066) (discussing the practice of granting and denying preliminary injunctions on the strength of the papers).

12. Tonson v. Walker, 36 Eng. Rep. 1017, 1020 (Ch. 1752); accord Little v. Gould, 15 F. Cas. 604, 610-11 (C.C.N.D.N.Y. 1851) (No. 8,394) (discussing English precedent).

13. Wilkins v. Aikin, 34 Eng. Rep. 163, 165 (Ch. 1810).

14. See Banks v. McDivitt, 2 F. Cas. 759, 762 (C.C.S.D.N.Y. 1875) (No. 961) (noting that an injunction is not ordinarily refused unless it will work a hardship on defendant). Little suggested that refusing to grant a preliminary injunction required "serious doubt" as to the plaintiff's case after a full investigation of the facts. Little, 15 F. Cas. at 611; see also EATON SYLVESTER DRONE, A TREATISE ON THE LAW OF PROPERTY IN INTELLECTUAL PRODUCTIONS 516 (1879) (noting that where the plaintiff presents a prima facie case, "a temporary injunction will usually be granted").

15. 28 Eng. Rep. 924 (Ch. 1765).

16. See id. at 924; see also Sheriff v. Coates, 39 Eng. Rep. 61, 65 (Ch. 1830) (dissolving an ex parte injunction because of doubts about the originality of the plaintiff's fabric designs, but requiring an accounting by defendants of their sales during the interim period).

17. See Osborne, 28 Eng. Rep. at 924.

18. See also Pope v. Curl, 26 Eng. Rep. 608 (Ch. 1741) (denying defendant's motion to dissolve a preliminary injunction previously issued ex parte); Sheriff, 39 Eng. Rep. at 61 (noting an ex parte injunction that was later dissolved).

19. 34 Eng. Rep. 163 (Ch. 1810).

20. Id. at 164; see also Smith v. Chatto, 31 L.T.R. 775, 777 (V.C. 1874) (granting an injunction because it would be simpler to calculate the harm resulting from an injunction erroneously granted than the harm from one erroneously withheld).

21. See Butterworth v. Kelly, 4 T.L.R. 430 (Ch. 1888).

22. See Prince Albert v. Strange, 41 Eng. Rep. 1171, 1179 (Ch. 1849). Admittedly, injunctive relief in this case--which involved a claimed invasion of privacy resulting from the unauthorized publication of private works--appears to have been particularly appropriate. Nonetheless, the Chancellor's argument was broadly applicable to many time-sensitive copyright cases: "postponing the injunction would be equivalent to denying it altogether. The interposition of this Court in these cases does not depend upon any legal right, and to be effectual, it must be immediate." Id.

23. JOHN PAXTON NORMAN, THE LAW AND PRACTICE OF THE COPYRIGHT REGISTRATION AND PROVISIONAL REGISTRATION OF DESIGNS 57 (1851) (citing Spottiswoode v. Clarke, 41 Eng. Rep. 900, 901 (Ch. 1846)) (emphasis added).

24. See Abrams, supra note 6, at 1171-78 (discussing the origins of American copyright in English law); cf. Jane C. Ginsburg, A Tale of Two Copyrights: Literary Property in Revolutionary France and America, 64 TUL. L. REV. 991 (1990) (contrasting U.S. and French copyright history).

25. Injunctive relief was probably available according to the rules of equity from the time of the first copyright act in 1790. Act of May 31, 1790, 1 Stat. 124 (1790) (providing copyright holders with the exclusive right of publication for fourteen years). It is clearly authorized in the copyright statute passed in 1819. Act of Feb. 15, 1819, ch. 19, 3 Stat. 481, 481 (extending the circuit courts' original jurisdiction to include patent and copyright cases, and providing that they "shall have authority to grant injunctions"); see also Clayton v. Stone, 5 F. Cas. 999, 1001 n.2 (C.C.S.D.N.Y. 1829) (No. 2,872) (discussing the equity power conferred by this statute).

26. See, e.g., Wheaton v. Peters, 29 F. Cas. 862, 865-69 (C.C.E.D. Pa. 1832) (No. 17,486) (holding that equity cannot act preliminarily where the legal right is unclear), aff'd, 33 U.S. (8 Pet.) 591 (1834); Ewer v. Coxe, 8 F. Cas. 917, 920 (C.C.E.D. Pa. 1824) (No. 4,584) (refusing to grant a preliminary injunction where the court determined that the plaintiff probably did not own the copyright); Blunt v. Patten, 3 F. Cas. 763, 765-67 (C.C.S.D.N.Y. 1828) (No. 1,580) (denying a preliminary injunction where difficult factual issues required resolution but later granting a permanent injunction after trial).

Evidence of this reluctance can be found even as late as the early twentieth century. See Hoffman v. Le Traunik, 209 F. 375, 378 (N.D.N.Y. 1913) (refusing to grant a preliminary injunction where the defendant claimed that the material in question was common property); West Publ'g Co. v. Edward Thompson Co., 169 F. 833, 883-87 (E.D.N.Y. 1909) (rejecting a preliminary injunction on the ground that the copyright was of dubious validity); Savage v. Hoffmann, 159 F. 584, 586 (S.D.N.Y. 1908) (same); Howell v. Miller, 91 F. 129, 142 (6th Cir. 1898) (same); Myers v. Callaghan, 5 F. 726, 734-36 (N.D. Ill. 1881) (concluding that a preliminary injunction was properly refused, though a permanent injunction was later issued).

Of course, not all early U.S. decisions refused preliminary injunctions. See, e.g., Chicago Dollar Directory Co. v. Chicago Directory Co., 66 F. 977 (7th Cir. 1895) (per curiam) (granting injunction); Photo Drama Motion Picture Co. v. Social Uplift Film Corp., 213 F. 374, 378 (S.D.N.Y. 1914) (same); Hein v. Harris, 183 F. 107 (2d Cir. 1910) (same).

27. See, e.g., Ewer, 8 F. Cas. at 918 (criticizing English cases granting injunctions "under circumstances which rendered the title doubtful").

28. See, e.g., id. at 918-19.

29. See, e.g., Photo Drama Motion Picture Co., 213 F. at 376 (noting that disputed facts must be resolved against the plaintiff on a motion for a preliminary injunction); Hoffman, 209 F. at 378 (refusing to grant a preliminary injunction where the defendant claimed the material in question was common property); West Publ'g Co., 169 F. at 883-87 (rejecting preliminary injunction on the ground that the copyright was of dubious validity); Benton v. Van Dyke, 170 F. 203, 204 (C.C.S.D.N.Y. 1909) (holding that "the question of infringement is too much in doubt to justify the granting of a preliminary injunction" and noting that "[s]uch a writ should never issue unless the court is clearly of the opinion that the complainant will succeed at final hearing"); Savage, 159 F. at 586 (same); Littleton v. Fischer, 137 F. 684, 685 (C.C.S.D.N.Y. 1905) (holding that "the extraordinary relief of a preliminary injunction should not be granted" because "it is impossible to decide to what extent, if at all, the defendant has trespassed upon the [complainants'] rights . . . and because it does not appear that defendant is unable to respond in damages in case it shall finally appear that he has violated complainants' rights"); Howell, 91 F. at 142 (refusing to grant an injunction where infringement was not "clearly manifest from the evidence"); Scribner v. Stoddart, 21 F. Cas. 876, 877 (C.C.E.D. Pa. 1879) (No. 12,561) (noting that prelimary injunctions should "be granted with extreme caution, and only in very clear cases" because of the "summary nature of the proceeding, and the consequent liability to mistake" (citing 2 JOSEPH STORY, COMMENTARIES ON EQUITY JURISPRUDENCE 290-91 (4th ed., Charles C. Little & James Brown 1846) (1836))); Blunt, 3 F. Cas. at 766 ("All this may be susceptible of satisfactory explanation. It is at present, however, involved in too much obscurity for me to safely act . . . and I shall leave the explanation to be given on a trial at law. The application for an injunction is, accordingly, denied."). The reasoning of the court in Scribner is highly instructive:

The doubt in my mind as respects both of these cases, is such that, without more, I should feel it to be my duty to deny this motion and decline the issuing of an injunction until the questions thus involved are fully, carefully, and deliberately considered and settled. Were I to issue the process in advance of this I would incur the danger of doing serious injustice to the defendants.

Scribner 21 F. Cas. at 878; see also id. at 877 (noting that in the absence of these precautions, the preliminary injunction "instead of being an instrument to promote the public as well as private welfare, . . . will become a means of extensive and perhaps irreparable injustice").

In cases where preliminary injunctions did issue, it was generally because the court found no doubt on the merits. See, e.g., Banks v. McDivitt, 2 F. Cas. 759, 762 (C.C.S.D.N.Y. 1875) (No. 961) (enjoining the sale of part but not all of the defendant's book, noting that "[w]here an infringement is palpable, and a provisional injunction will not be attended with serious injury, it is not ordinarily refused, as to so much of the work as is a plain infringement"); see also Benton, 170 F. at 204 (noting that a preliminary injunction could issue if the court found no doubt on the merits, although the standard was not met in the present case); Littleton, 137 F. at 685 (same). But see Little v. Gould, 15 F. Cas. 604 (C.C.N.D.N.Y. 1851) (No. 8,394). In Little, the court noted that doubts as to legal rather than factual questions should not ordinarily preclude a preliminary injunction if the doubts could be resolved by detailed analysis at the preliminary injunction stage. See id. at 611. The court rejected the prior English chancery practice of referring such issues to the law courts as irrelevant in the federal courts, where law and equity were merged in a single court. See id.

30. See Scribner, 21 F. Cas. at 877 (stating that an injunction may be withheld for any equitable reason, even if the plaintiff can show a likelihood of success on the merits).

31. 209 U.S. 20 (1908).

32. Id. at 23, 24. This result was resurrected by the Court in a footnote in Campbell v. Acuff-Rose Music Co., 501 U.S. 569, 578, n.10 (1994) ("[T]he goals of the copyright law . . . are not always best served by automatically granting injunctive relief when parodists are found to have gone beyond the bounds of fair use."). For an approach that would take the Campbell dictum seriously, see Robert A. Kreiss, Access and Commercialization in Copyright Theory, 43 UCLA L. REV. 1, 71-74 (1995).

33. 185 F. 921 (C.C.N.D.N.Y. 1911).

34. Id. at 925. To some extent, of course, this reluctance reflects the narrower scope of copyright before this century. Early U.S. copyright law gave copyright owners much less protection against the preparation of derivative works. See Paul Goldstein, Derivative Rights and Derivative Works in Copyright, 30 J. COPYRIGHT SOC'Y 209 (1983). Because of this, the plaintiff's case in White was weaker than it appears today.

35. See Hansen v. Jacquard Jewelry Co., 32 F. 202, 203-04 (C.C.E.D. Mo. 1887).

36. See supra notes 19-21 and accompanying text.

37. This general equitable rule provides that preliminary injunctions should not issue if a later legal remedy would adequately compensate the plaintiff. For a discussion of the history and modern status of the rule, see DOUGLAS LAYCOCK, THE DEATH OF THE IRREPARABLE INJURY RULE 111-18 (1991) (arguing that courts have failed to apply this rule rigorously).

38. See Ladd v. Oxnard, 75 F. 703, 734 (C.C.D. Mass. 1896) (establishing the rule that a defendant could avoid a preliminary injunction by posting an adequate bond); see also Boosey v. Empire Music Co., 224 F. 646, 647 (S.D.N.Y. 1915) (ordering preliminary injunction to issue unless defendant posted bond); Sampson & Murdock Co. v. Seaver-Radford Co. 129 F. 761, 773 (C.C.D. Mass. 1904) (same); Hubbard v. Thompson, 14 F. 689 (E.D. Mo. 1882) (forecasting the rule in Ladd). But see Reed v. Holliday, 19 F. 325, 327 (W.D. Pa. 1884) ("[I]f a plaintiff shows infringement of his copyright, the court will grant an injunction without proof of actual damage."). This is exactly the opposite of the modern rule, which requires the plaintiff to post a bond as a condition of the granting of a preliminary injuction. See infra notes 78-80 and accompanying text.

39. See Trow Directory Printing Co. v. Boyd, 97 F. 586, 734 (C.C.S.D.N.Y. 1899).

40. See supra note 22 and accompanying text.

41. ARTHUR W. WEIL, AMERICAN COPYRIGHT LAW 515 (1917); see also Historical Publ'g Co. v. Jones Bros. Publ'g Co., 231 F. 638, 644 (3d Cir. 1916) (holding that a threatened infringement should have been preliminarily enjoined and reversing the lower court's failure to grant an injunction).

42. See DAVID M. RABBAN, FREE SPEECH IN ITS FORGOTTEN YEARS 169-73 (1997) (discussing the general change in injunction practice). In particular, preliminary injunctions against a wide variety of speech by labor unions were common in this era. The argument that enjoining labor speech was a prior restraint was normally dismissed out of hand, sometimes on the theory that the labor leaders were engaged in "verbal acts" rather than "real" speech. See, e.g., Gompers v. Buck's Stove & Range Co., 221 U.S. 418, 436-39 (1911). For a discussion of this now-discredited approach to injunctions restricting speech, see RABBAN, supra, at 169-73.

43. Pub. L. No. 94-553, 1976 U.S.C.C.A.N. (90 Stat.) 2541 (1976) (codified at 17 U.S.C. 101-1101 (1994)).

44. 17 U.S.C. 502(a) (1994). As Paul Goldstein notes, the apparent intent of Congress in enacting section 502(a) was to authorize temporary restraining orders as well. See 2 PAUL GOLDSTEIN, COPYRIGHT 11.1.1, at 11:5 n.5 (1995).

45. See LAYCOCK, supra note 37, at 118 (outlining the four-factor test); 11A CHARLES ALAN WRIGHT ET AL., FEDERAL PRACTICE AND PROCEDURE 131-33 (2d ed. 1995) (outlining the test and citing cases).

46. See LAYCOCK, supra note 37, at 118. One strand of legal thought has attempted to quantify the balance of the hardships, measuring the loss to both sides from wrongfully being enjoined (or wrongfully being denied an injunction), and multiplying it by the probability that the party will prevail at trial to determine a sort of "total hardship." See American Hosp. Supply Corp. v. Hosp. Prods., 780 F.2d 589, 593 (7th Cir. 1986); cf. John Leubsdorf, The Standard for Preliminary Injunctions, 91 HARV. L. REV. 525, 540-42 (1978) (suggesting a similar test). Given the uncertainty attendant upon both variables, and indeed the fact that preliminary injunctions are theoretically intended to be used only when the harm to the plaintiff cannot be measured, one might reasonably question how much precision this formulation really adds. See LAYCOCK, supra note 37, at 120; Linda S. Mullenix, Burying (With Kindness) the Felicific Calculus of Civil Procedure, 40 VAND. L. REV. 541, 569-72 (1987).

47. See LAYCOCK, supra note 37, at vii (surveying more than 1400 cases in which injunctive relief was sought and determining that courts do not apply the irreparable injury rule with any teeth).

48. See Triad Sys. Corp. v. Southeastern Express Co., 64 F.3d 1330, 1335 (9th Cir. 1995) ("In a copyright infringement action, . . . the rules are somewhat different."); Religious Tech. Ctr. v. F.A.C.T.Net, Inc., 901 F. Supp. 1519, 1523 (D. Colo. 1995) (describing the test for preliminary injunctive relief in copyright cases as "less rigorous" than the normal standard).

49. See GOLDSTEIN, supra note 44, 11.1.2.2, at 11:11-12 n.29 (collecting copyright cases that use the four-factor analysis). The exceptions are the Second and Ninth Circuits, which do not use the public interest factor, but instead only balance the three remaining factors, with a greater likelihood of success on the merits lessening the required showing of irreparable injury or balance of the hardships. See, e.g., Hasbro Bradley, Inc. v. Sparkle Toys, Inc., 780 F.2d 189, 192 (2d Cir. 1985); Apple Computer, Inc. v. Formula Int'l, Inc., 725 F.2d 521, 523 (9th Cir. 1984).

50. See Concrete Mach. Co. v. Classic Lawn Ornaments, Inc., 843 F.2d 600, 612 (1st Cir. 1988); Dallas Cowboys Cheerleaders, Inc. v. Score Board Posters, Inc., 600 F.2d 1184, 1187 (5th Cir. 1979) ("Preliminary injunctions are a common judicial response to the imminent infringement of an apparently valid copyright."); Georgia Television Co. v. TV News Clips of Atlanta, Inc., 718 F. Supp. 939, 945 (N.D. Ga. 1989) (noting that preliminary injunctions in copyright cases are "common"); ROBERT P. MERGES ET AL., INTELLECTUAL PROPERTY IN THE NEW TECHNOLOGICAL AGE 520 (1997) (noting that "preliminary injunctions are generally granted as a matter of course"); 4 MELVILLE NIMMER & DAVID NIMMER, NIMMER ON COPYRIGHT 14.06[A], at 14-112 (1997) [hereinafter Nimmer on Copyright] (issuance of preliminary injunctions in copyright cases "is actually quite ordinary, even commonplace").

51. See Cadence Design Sys. v. Avant! Corp., 125 F.3d 824, 827 (9th Cir. 1997); Johnson Controls, Inc. v. Phoenix Control Sys., Inc., 886 F.2d 1173, 1174 (9th Cir. 1989); Creative Labs v. Cyrix Corp., 42 U.S.P.Q.2d (BNA) 1872 (N.D. Cal. 1997).

52. See, e.g., Country Kids 'N City Slicks, Inc. v. Sheen, 77 F.3d 1280, 1288 (10th Cir. 1996) (endorsing the rule, which was already in force in "the overwhelming majority of our sister circuits"); Video Trip Corp. v. Lightning Video, Inc., 866 F.2d 50, 51-52 (2d Cir. 1989); Johnson Controls, Inc. v. Phoenix Control Sys., Inc., 886 F.2d 1173, 1174 (9th Cir. 1989); West Publ'g Co. v. Mead Data Cent., Inc., 799 F.2d 1219, 1229 (8th Cir. 1986); Apple Computer, Inc. v. Formula Int'l, 725 F.2d 521, 525 (9th Cir. 1984); Apple Computer, Inc. v. Franklin Computer Corp., 714 F.2d 1240, 1254 (3d Cir. 1983); Atari, Inc. v. North Am. Philips Consumer Elec. Corp., 672 F.2d 607, 620 (7th Cir. 1982); Wainwright Sec., Inc. v. Wall Street Transcript Corp., 558 F.2d 91, 94 (2d Cir. 1977); Joshua Meier Co. v. Albany Novelty Mfg., 236 F.2d 144, 147 (2d Cir. 1956); Georgia Television Co., 718 F. Supp. at 944, 948 (presuming irreparable harm despite "speculative" nature of plaintiff's injury); see also GOLDSTEIN, supra note 44, at 11:15-18 (collecting cases); NIMMER ON COPYRIGHT, supra note 50, at 14-100 (same). This presumption is specific to copyright owners; it does not extend to accused infringers who are plaintiffs seeking a declaratory judgment of noninfringement. See Barris/Fraser Enters. v. Goodson-Todman Enters., Ltd., 638 F. Supp. 292, 295 (S.D.N.Y. 1986).

The exception to the general practice of presuming irreparable harm once likelihood of success on the merits is shown is the Fifth Circuit, which requires a showing of each of the four factors regardless of the status of the merits. See, e.g., Apple Barrel Prods., Inc. v. Beard, 730 F.2d 384, 390 (5th Cir. 1984); NIMMER ON COPYRIGHT, supra note 50, at 14-104 n.36 (collecting cases).

Despite the presumption of irreparable injury, there are occasionally decisions in these circuits in which courts deny injunctive relief because the harm to the plaintiff can be calculated precisely and awarded as damages. See, e.g., Walt Disney Prods. v. Basmajian, 600 F. Supp. 439, 442 (S.D.N.Y. 1984); American Metro Enters., Inc. v. Warner Bros. Records, Inc., 154 U.S.P.Q. 311, 311 (S.D.N.Y. 1967). These cases, though, are clearly in the minority.

53. See GOLDSTEIN, supra note 44, at 11:16.

54. See Northwestern Bell Tel. Co. v. Bedco of Minn., Inc., 501 F. Supp. 299, 303 (D. Minn. 1980).

55. See West Publ'g Co. v. Mead Data Cent., Inc., 616 F. Supp. 1571, 1581 (D. Minn. 1985), aff'd, 799 F.2d 1219 (8th Cir. 1986).

56. See Feist v. Rural Tel. Servs., 499 U.S. 340, 364 (1991) (holding that telephone white pages are not copyrightable); Matthew Bender & Co. v. West Publ'g Co., 42 U.S.P.Q.2d 1930, 1934 (S.D.N.Y. 1997) (holding that West's page numbers are not copyrightable). But see Oasis Publ'g Co. v. West Publ'g Co., 924 F. Supp. 918, 923 (D. Minn. 1996) (holding that West's page numbers are copyrightable despite Feist).

57. See Rushton v. Vitale, 218 F.2d 434, 436 (2d Cir. 1955) (considering the short-lived nature of the market as a factor in favor of the plaintiff); Houghton Mifflin Co. v. Stackpole Sons, Inc., 104 F.2d 306, 307 (2d Cir. 1939) (same); O'Neill Devs., Inc. v. Galen Kilburn, Inc., 524 F. Supp. 710, 715 (N.D. Ga. 1981) (same); Stern Elec., Inc. v. Kaufman, 523 F. Supp. 635, 638 (E.D.N.Y. 1982) (same); see also NIMMER ON COPYRIGHT, supra note 50, at 14-110 to 14-111(citing cases); GOLDSTEIN, supra note 44, at 11:17-18 n.47 (same). But see Thomas Wilson & Co. v. Irving J. Dorfman Co., 268 F. Supp. 711, 714 (S.D.N.Y. 1967) ("[S]uch an injunction would amount in substance to the ultimate relief sought in the action--a further factor weighing against the granting of this extraordinary remedy.").

58. Compare McGraw-Hill, Inc. v. Worth Publishers, Inc., 335 F. Supp. 415, 422 (S.D.N.Y. 1971) (suggesting in dicta that defendant's insolvency would favor a finding of irreparable harm), with Albert E. Price, Inc. v. Metzner, 574 F. Supp. 281, 289 (E.D. Pa. 1983) (finding irreparable harm where the defendant has sufficient financial resources, since the existence of those resources suggests that the defendant may be able to threaten the plaintiff's competitive position).

59. In both the Second and Ninth Circuits, a showing of probable success on the merits will obviate the need to prove that the balance of hardships tips in the plaintiff's favor. If the plaintiff can only show the "existence of serious questions going to the merits," however, it must also show that the balance of hardships tips in its favor. Apple Computer, Inc. v. Formula Int'l, Inc., 725 F.2d 521, 523 (9th Cir. 1984). See also Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 206-07 (2d Cir. 1979); GOLDSTEIN, supra note 44, 11.1.2.2, at 11:12. The balance in such a case need only "tip" in the plaintiff's favor in the Ninth Circuit, Apple Computer, 725 F.2d at 523, but must tip "decidedly" in the plaintiff's favor in the Second Circuit, Dallas Cowboys Cheerleaders, 604 F.2d at 207.

60. GOLDSTEIN, supra note 44, at 11:19. See, e.g., Georgia Television Co. v. TV News Clips of Atlanta, Inc., 718 F. Supp. 939, 949-50 (N.D. Ga. 1989). A particularly striking example is Worlds of Wonder, Inc. v. Veritel Learning System, Inc., 658 F. Supp. 351 (N.D. Tex. 1986), which held that the balance of hardships favored plaintiff, despite the likelihood that an injunction would put the defendant out of business entirely. See id. at 357.

The exception to this rule is, once again, the Fifth Circuit, in which the balance of hardships may favor the defendant in appropriate cases. See Plains Cotton Coop. Ass'n v. Goodpasture Computer Serv., Inc., 807 F.2d 1256, 1261 (5th Cir. 1987); Apple Barrel Prods., Inc. v. Beard, 730 F.2d 384, 390 (5th Cir. 1984); see also Religious Tech. Ctr. v. F.A.C.T.Net, Inc., 901 F. Supp. 1519, 1527 (D. Colo. 1995) (finding the balance of hardships to weigh in favor of the defendants).

61. See, e.g., Belushi v. Woodward, 598 F. Supp. 36, 37 (D.D.C. 1984).

62. See, e.g., West Publ'g Co. v. Mead Data Cent., Inc., 799 F.2d 1219, 1229 (8th Cir. 1986) (holding that a preliminary injunction was proper even though page numbers were the only allegedly infringing parts of the defendant's work).

63. Apple Computer, Inc. v. Franklin Computer Corp., 714 F.2d 1240, 1255 (3d Cir. 1983) (quoting Klitzner Indus. v. H.K. James & Co., 535 F. Supp. 1249, 1259-60 (E.D. Pa. 1982)); accord Georgia Television Co., 718 F. Supp. at 950.

64. See GOLDSTEIN, supra note 44, at 11:21 and cases cited therein. Alan Rau calls the public interest factor in copyright cases "a makeweight, serving merely to re-state the general public policy underlying the cause of action." Alan Scott Rau, Fifth Circuit Symposium: Intellectual Property, 16 TEXAS TECH L. REV. 355, 390 (1985).

65. See, e.g., West Publ'g Co. v. Mead Data Cent., Inc., 616 F. Supp. 1571, 1582 (D. Minn. 1985) (rejecting the defendant's arguments that the public interest favors dissemination), aff'd, 799 F.2d 1219 (8th Cir. 1986); Dallas Cowboys Cheerleaders, Inc. v. Scoreboard Posters, Inc., 600 F.2d 1184, 1187 (5th Cir. 1979) (same); Wainwright Sec., Inc. v. Wall Street Transcript Corp., 558 F.2d 91, 95 (2d Cir. 1977) (same). But seeF.A.C.T.Net, 901 F. Supp. at 1527 (accepting this argument for "matters of significant public controversy"); Belushi, 598 F. Supp. at 37-38 (same); Keep Thomson Governor Comm. v. Citizens for Gallen Comm., 457 F. Supp. 957, 959-960 (D.N.H. 1978) (same).

66. See 4 NIMMER ON COPYRIGHT, supra note 50, at 14-101 ("[T]he plaintiff's burden for obtaining a preliminary injunction in copyright cases collapses to showing likelihood of success on the merits.").

67. See 17 U.S.C. 106 (1994) (enumerating exclusive rights); Dr. Seuss Enters. v. Penguin Books USA, Inc., 924 F. Supp. 1559, 1563 (S.D. Cal. 1996) (defining the prima facie case).

68. See, e.g., GOLDSTEIN, supra note 44, at 11:13-14. The Copyright Act of 1976 states that "[i]n any judicial proceedings the certificate of a registration made before or within five years after first publication of the work shall constitute prima facie evidence of the validity of the copyright." 17 U.S.C. 410(c) (1994).

69. See, e.g., Georgia Television Co. v. TV News Clips of Atlanta, Inc., 718 F. Supp. 939, 945 (N.D. Ga. 1989); Association of Am. Med. Colleges v. Carey, 482 F. Supp. 1358, 1364 n.15 (N.D.N.Y. 1980), and cases cited therein. This is so despite the fact that for U.S. plaintiffs, either copyright registration or a rejection thereof is nominally a jurisdictional requirement for filing suit. See 17 U.S.C. 411(a). But see Apple Barrel Prods., Inc. v. Beard, 730 F.2d 384, 386-87 (5th Cir. 1984) ("In order to bring suit for copyright infringement, it is not necessary to prove possession of a registration certificate. One need only prove payment of the required fee, deposit of the work in question, and receipt by the Copyright Office of a registration application."). Apple Barrel is inconsistent with section 411.

70. See, e.g., Nancy H. McAleer, A Bird's Eye View of Copyright Registration, N.J. LAW., Feb. 15, 1993, at 15; Nelson R. Capes, The Software Copyright "Super Patent," COMP. LAW., June 1995, at 8.

71. See Apple Computer, Inc. v. Formula Int'l, 725 F.2d 521, 523 (9th Cir. 1984) (holding that the plaintiff's certificate of copyright registration constituted prima facie evidence of validity of copyrights and that the defendant had the burden of overcoming the presumption of validity); Original Appalachian Artworks, Inc. v. Toy Loft, Inc., 489 F. Supp. 174, 179 (N.D. Ga. 1980) ("[B]y obtaining a copyright certificate the plaintiff is entitled to a prima facie presumption of copyright validity (and the burden shifts to the defendant to prove lack of originality).").

72. See Arnstein v. Porter, 154 F.2d 464, 468 (2d Cir. 1946); GOLDSTEIN, supra note 44, at 11:14-15.

73. See, e.g., Joshua Meier Co. v. Albany Novelty Mfg. Co., 236 F.2d 144, 146-48 (2d Cir. 1956); American Direct Mktg., Inc. v. Azad Int'l, Inc., 783 F. Supp. 84, 96 (E.D.N.Y. 1992).

74. See generally Computer Assocs. v. Altai, Inc., 982 F.2d 693 (2d Cir. 1992) (noting that the defendant could rebut the plaintiff's proof of copying by demonstrating that copied elements were not themselves protectable).

75. See Kontes Glass Co. v. Lab Glass, Inc., 373 F.2d 319, 320-21 (3d Cir. 1967). But see Encyclopaedia Britannica Educ. Corp. v. Crooks, 447 F. Supp. 243, 251 (W.D.N.Y. 1978) (holding that a fair use claim would not bar a preliminary injunction and assuming for purposes of the plaintiff's motion that the plaintiff could prove its allegations of market harm caused by the defendant's work).

76. For example, in Covington Fabrics Corp. v. Artel Products, Inc., 328 F. Supp. 202 (S.D.N.Y. 1971), the court gave so much weight to the presumption of originality that accompanied the registration certificate that it rejected the testimony of the defendant's expert witness that the plaintiff's work was not in fact original. See id. at 204. The court granted a preliminary injunction, having concluded that the plaintiff's copyright registration was sufficient evidence of likelihood of success on the merits on this issue. See id. at 204-05.

77. See 4 NIMMER ON COPYRIGHT, supra note 50, at 14-104 to 14-105.

78. See id. at 14-98 to 14-99 (citing cases). For a history of bond requirements, see FELIX FRANKFURTER & NATHAN GREEN, THE LABOR INJUNCTION, 1-46 (1930).

79. See, e.g., Nintendo of Am., Inc. v. Lewis Galoob Toys, Inc., 16 F.3d 1032, 1039 (9th Cir. 1994) (upholding the award of a $15 million bond to the defendant).

80. Cf. Wales Indus., Inc. v. Hasbro Bradley, Inc., 612 F. Supp. 507, 508-09 (S.D.N.Y. 1985) (holding that the party seeking declaratory judgment of noninfringement was not entitled to a preliminary injunction since its loss from being unable to sell its product was compensable through money damages).

81. See FED. R. CIV. P. 65(b); Fort Wayne Women's Health Org. v. Brane, 734 F. Supp. 849, 850 (N.D. Ind. 1990) ("An ex parte temporary restraining order is an extraordinary remedy which will not be granted unless the movant clearly shows that such relief is warranted.").

82. See 17 U.S.C. 503(a) (1994).

83. See FED. R. CIV. P. 65 advisory committee's note; GOLDSTEIN, supra note 44, at 11:22-24 (questioning such seizures and quoting Federal Rules Advisory Committee); Marc Alexander, Discretionary Power to Impound and Destroy Infringing Articles: An Historical Perspective, 29 J. COPR. SOC'Y 479, 488-89 (1982) (asserting that the constitutionality of the rule deserves additional scrutiny); cf. Paramount Pictures Corp. v. Doe, 821 F. Supp. 82, 86-89 (E.D.N.Y. 1993) (noting Fourth Amendment problems with impoundment remedy); id. at 87 n.4 (observing that the Federal Rules Advisory Committee noted similar problems); Paul S. Owens, Impoundment Procedures Under the Copyright Act: The Constitutional Infirmities, 14 HOFSTRA L. REV. 211 (1985) (noting potential First Amendment problems with this rule).

84. Portions of the discussion in Parts II.A, II.D, and II.E are adapted from Eugene Volokh & Brett McDonnell, Freedom of Speech and Independent Judgment Review in Copyright Cases, 107 YALE L.J. 2431 (1998).

85. Cf. Spence v. Washington, 418 U.S. 405, 417 (1974) (Rehnquist, J., dissenting) (giving copyright law as an example of a speech restriction).

86. See 17 U.S.C.A. 506(a) (West 1996 & Supp. 1998) (criminalizing "willful[]" infringement that is either "for purposes of commercial advantage or private financial gain" or involves material with "a total retail value of more than $1,000").

87. See, e.g., Toksvig v. Bruce Publ'g Co., 181 F.2d 664 (7th Cir. 1950); Sheldon v. Metro-Goldwyn Pictures Corp., 81 F.2d 49 (2d Cir. 1936).

88. See, e.g., Steinberg v. Columbia Pictures Indus., 663 F. Supp. 706 (S.D.N.Y. 1987); Kisch v. Ammirati & Puris, Inc., 657 F. Supp. 380 (S.D.N.Y. 1987).

89. Tin Pan Apple Inc. v. Miller Brewing Co., 30 U.S.P.Q.2d (BNA) 1791, 1793 (S.D.N.Y. 1994).

90. See infra notes 219-220 and accompanying text (discussing such cases).

91. 471 U.S. 539 (1985).

92. See id. at 555-60.

93. Id. at 558.

94. Id. at 560.

95. Id. at 556 (citing 17 U.S.C. 102(b) (1994)).

96. 403 U.S. 713 (1971).

97. Id. at 726 n.* (Brennan, J., concurring), cited with approval in Harper & Row, 471 U.S. at 556.

98. Harper & Row, 471 U.S. at 559.

99. We mean "unprotected" in the literal sense: the speech may constitutionally be punished by copyright law, despite the First Amendment, which means that the First Amendment does not protect the speech against legal sanction. In this respect, infringing speech is just like the traditional exceptions to First Amendment protection, such as obscenity, defamation, fighting words, threats, child pornography, advocacy of unlawful conduct that's intended to and is likely to produce imminent lawlessness, publication of sailing dates of troop ships, and the like. See Simon & Schuster, Inc. v. Members of the N.Y. State Crime Victims Bd., 502 U.S. 105, 127 (1991) (Kennedy, J., concurring) (listing such exceptions).

We are not suggesting that infringing speech is "valueless." Like advocacy of unlawful conduct, or revelation of extremely sensitive government secrets, it can often make an important contribution to public debate or at least provide public entertainment. It is punishable primarily because of its perceived harm and the supposedly ample alternative avenues for expression, not because of its perceived lack of value. But whatever the reason, the First Amendment does not protect this speech against the operation of copyright law.

Of course, speech that's unprotected against copyright law might still be protected against other laws: The government may not, for instance, constitutionally apply a ban on racist speech or sexually suggestive speech even to material that's infringing, just as it may not apply a ban on racist speech even to material that constitutes fighting words. See R.A.V. v. City of St. Paul, 505 U.S. 377, 383-84 (1992).

100. See Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 562, 577 n.13 (1977) (some citations omitted):

We note that Federal District Courts have rejected First Amendment challenges to the federal copyright law on the ground that "no restraint [has been] placed on the use of an idea or concept." See also Walt Disney Productions v. Air Pirates, 345 F. Supp. 108, 115-16 (N.D. Cal. 1972) (citing [Melville B.] Nimmer, Does Copyright Abridge The First Amendment Guarantees of Free Speech and Press?, 17 UCLA L. REV. 1180 (1970) who argues that copyright law does not abridge the First Amendment because it does not restrain the communication of ideas or concepts).

See also Lee v. Runge, 404 U.S. 887, 892 (1971) (Douglas, J., dissenting from denial of certiorari) ("Serious First Amendment questions would be raised if Congress' power over copyrights were construed to include the power to grant monopolies over certain ideas.").

101. While Harper & Row suggested that the idea-expression dichotomy and at least some sort of fair use doctrine are constitutionally required, see supra text accompanying note 94, it did not say that all speech covered by current fair use doctrine is constitutionally protected. It's conceivable that Congress may to some extent constrict the fair use doctrine without transgressing the First Amendment, though it seems that it may not eliminate it altogether or constrict it too much, at least in the area of "scholarship and comment." Harper & Row, 471 U.S. at 560. On the other hand, an effort to constrict the fair use or idea-expression privileges might compel courts to expressly evaluate the constitutionality of the new rule, something that they haven't had to do in the past. On the constitutional significance of the fair use doctrine, and efforts to whittle away at it, see Jessica Litman, Reforming Information Law in Copyright's Image, 22 U. DAYTON L. REV. 587, 612 (1997) (noting that "precisely because it requires case-by-case analysis . . . fair use is a troublesome safe harbor for First Amendment rights").

102. This is not to say that the courts have always gotten it right in interpreting the fair use doctrine, or in drawing the line between idea and expression. For an argument that software blurs the latter line in disturbing ways, see Lloyd L. Weinreb, Copyright for Functional Expression, 111 HARV. L. REV. 1149 (1998); Dennis S. Karjala & Peter S. Menell, Applying Fundamental Copyright Principles to Lotus Development Corp. v. Borland International, Inc., 10 HIGH TECH. L.J. 177 (1995).

103. Many commentators who have considered the issue have come to similar substantive conclusions. See, e.g., Floyd Abrams, First Amendment and Copyright, 35 J. COPR. SOC'Y 1 (1987); Robert C. Denicola, Copyright and Free Speech: Constitutional Limitations on the Protection of Expression, 67 CAL. L. REV. 283 (1979); Charles C. Goetsch, Parody as Free Speech--The Replacement of the Fair Use Doctrine by First Amendment Protection, 3 W. NEW ENG. L. REV. 39 (1980); Paul Goldstein, Copyright and the First Amendment, 70 COLUM . L. REV . 983 (1970); Melville B. Nimmer, Does Copyright Abridge the First Amendment Guarantees of Free Speech and Press?, 17 UCLA L. REV. 1180 (1970); Hon. James L. Oakes, Copyrights and Copyremedies: Unfair Use and Injunctions, 18 HOFSTRA L. REV. 983 (1990); L. Ray Patterson, Free Speech, Copyright, and Fair Use, 40 VAND. L. REV. 1 (1987); David E. Shipley, Conflicts Between Copyright and the First Amendment After Harper & Row, Publishers v. Nation Enterprises, 1986 BYU L. REV. 983 (1986); Alfred C. Yen, A First Amendment Perspective on the Idea/Expression Dichotomy and Copyright in a Work's "Total Concept and Feel," 38 EMORY L.J. 393 (1989); Geri J. Yonover, The Precarious Balance: Moral Rights, Parody, and Fair Use, 14 CARDOZO ARTS & ENT. L.J. 79 (1996); Mel Marquis, Comment, Fair Use of the First Amendment: Parody and Its Protections, 8 SETON HALL CONST. L.J. 123 (1997); cf. Neil Weinstock Netanel, Copyright and a Democratic Civil Society, 106 YALE L.J. 283, 296-97 (1996) (offering speech-related justifications for limited copyright protection); Diane Leenheer Zimmerman, Information as Speech, Information as Goods: Some Thoughts on Marketplaces and the Bill of Rights, 33 WM. & MARY L. REV. 665, 666 (1992) (suggesting that if anything commentators understate the problems caused by the conflict between copyright law and the First Amendment).

104. See Toro Co. v. R & R Prods. Co., 787 F.2d 1208, 1212 (8th Cir. 1986) (the idea-expression dichotomy is partly grounded "in the First Amendment interest in the free exchange of ideas"); Sid & Marty Krofft Television Prods., Inc. v. McDonald's Corp., 562 F.2d 1157, 1170 (9th Cir. 1977) (noting that the "idea-expression dichotomy already serves to accommodate the competing interests of copyright and the first amendment"); Walt Disney Prods. v. Air Pirates, 581 F.2d 751, 759 (9th Cir. 1978) ("'[T]he idea-expression line represents an acceptable definitional balance as between copyright and free speech interests.'"), quoting Sid & Marty Krofft, 562 F.2d at 1170.

105. We owe much of the general analysis in this section to Martin H. Redish, The Proper Role of the Prior Restraint Doctrine in First Amendment Theory, 70 VA. L. REV. 53 (1984).

106. There has been virtually no discussion of the problems we identify in this Article. The one exception is the Nimmer treatise, which observes: "Indeed, the frequency of copyright TRO's and injunctions against expressive material under the copyright rubric is surprising when juxtaposed against the reluctance of courts to engage in a prior restraint of speech under the analysis mandated by the First Amendment." NIMMER ON COPYRIGHT, supra note 50, 14.06[A].

For discussions of prior restraints in copyright law that focus on areas of conflict other than preliminary injunctions, see Henry H. Perritt, Jr., Property and Innovation in the Global Information Infrastructure, U. CHI. LEGAL F. 261, 325-29 (1996) (arguing that strict liability for Internet service providers is tantamount to a prior restraint); William L. LaFuze & Valerie K. Friedrich, Aggressive Enforcement of Rights Involving Internet Abuses, COMPUTER L. REV. & TECH. J. 1, 20-21 (1997) (noting the "surprising frequency" of TROs and injunctions despite the prior restraint doctrine, without distinguishing preliminary injunctions from permanent ones); Zimmerman, supra note 103, at 737-38 (criticizing the use of a property analysis to grant injunctive relief in light of free speech law's disfavor of prior restraints); James Thompson, Permanent Injunctions in Copyright Infringement: Moral and Economic Justifications for Balancing Individual Rights Instead of Following Harsh Rules, 7 S.CAL. INTERDISC. L.J. 477 (1998) (arguing for discretion in the grant of permanent injunctions); Tiffany D. Trunko, Note, Remedies for Copyright Infringement: Respecting the First Amendment, 89 COLUM. L. REV. 1940, 1941 (1989) (arguing that permanent injunctions "may look suspiciously like" prior restraints, yet noting that "courts routinely issue copyright injunctions"). Trunko also mentions that "preliminary injunctions can pose similar prior restraint problems." Id. at n.5. For reasons we will shortly discuss, we think that the problem is much worse for preliminary than for permanent injunctions.

107. Copyright plaintiffs who win a final judgment on the merits are generally entitled to a permanent injunction as a matter of course, though in some fairly rare cases courts have denied such injunctions. See, e.g., Abend v. MCA, Inc., 863 F.2d 1465, 1479 (9th Cir. 1988), aff'd sub nom. Stewart v. Abend, 495 U.S. 207 (1990); see also Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 578 n.10 (1994) (dictum) (suggesting this possibility for certain near-fair-use cases).

108. Pittsburgh Press Co. v. Pittsburgh Comm'n on Human Relations, 413 U.S. 376, 390 (1973) (emphasis added).

109. 354 U.S. 436 (1957).

110. 413 U.S. 376 (1973).

111. See infra Part II.B.2.

112. Pittsburgh Press, 413 U.S. at 390; accord Kingsley Books, 354 U.S. at 445 (upholding an injunction of obscenity in part because the injunction covered only speech that already had been found to be unprotected). This argument applies equally to any unprotected category of speech. See Retail Credit Co. v. Russell, 218 S.E.2d 54, 62 (Ga. 1975) (upholding an injunction against statements already found to be "false and defamatory"); O'Brien v. University Community Tenants Union, Inc., 327 N.E.2d 753, 755 (Ohio 1975) ("Once speech has judicially been found libelous, if all the requirements for injunctive relief are met, an injunction for restraint of continued publication of that same speech may be proper."); Advanced Training Sys., Inc. v. Caswell Equip. Co., 352 N.W.2d 1, 11 (Minn. 1984) (approving an injunction against libel issued after "full adversarial proceedings"); see also Kramer v. Thompson, 947 F.2d 666, 675 (3d Cir. 1991) (striking down an injunction but noting in dictum that an injunction will generally not be unconstitutional if a "jury has determined that the speech is not constitutionally protected"); Anderson v. Dean, 354 F. Supp. 639, 642 (N.D. Ga. 1973) (striking down an injunction but stating that "an injunction directed against a libel is not unconstitutional") (citation omitted); RESTATEMENT (SECOND) OF TORTS 623 special note on remedies for defamation other than damages (1977) (stating that an injunction may be available when a court has "formally determined" that the speech is unprotected); Castrol, Inc. v. Pennzoil Co., 987 F.2d 939, 949 (3d Cir. 1993) (finding, albeit in a commercial speech context, that the prior restraint doctrine did not apply "because there has been 'an adequate determination that [the expression] is unprotected by the First Amendment'" (quoting Pittsburgh Press, 413 U.S. at 390)); SEC v. Wall St. Publ'g Inst., Inc., 851 F.2d 365, 370 (D.C. Cir. 1988) (same, in a case involving securities information).

113. Compare Alexander v. United States, 509 U.S. 544, 550 (1993) ("The term prior restraint is used 'to describe administrative and judicial orders forbidding certain communications when issued in advance of the time that such communications are to occur.' Temporary restraining orders and permanent injunctions--i.e., court orders that actually forbid speech activities--are classic examples of prior restraints.") (citations omitted), with Madsen v. Women's Health Ctr., Inc. 512 U.S. 753, 764 n.2 (1994) (concluding that certain content-neutral injunctions aren't prior restraints), Pittsburgh Press, 413 U.S. at 389-90 (concluding that an injunction barring the placement of want ads in sex-segregated columns was not a prior restraint), and Kingsley Books, 354 U.S. at 441-45 (finding that a content-based injunction was not a prior restraint).

This area has been considerably confused by the Court's penchant for occasionally condemning speech restrictions as prior restraints when not even subsequent punishment of the speech would be permissible. See LAURENCE H. TRIBE, AMERICAN CONSTITUTIONAL LAW 12-34, at 1040-41 (2d ed. 1988). Two such cases include Organization for a Better Austin v. Keefe, 402 U.S. 415 (1971) (rejecting a content-based injunction against leafleting that criticized a local real estate broker as a prior restraint) and Nebraska Press Ass'n v. Stuart, 427 U.S. 539 (1976) (rejecting an injunction and noting that injunctions against press coverage of judicial proceedings may pose special dangers). It seems clear that under standard First Amendment doctrine, the leafleting at issue in Austin and the press coverage in Nebraska Press could not even have been punished after the fact, so the prior restraint inquiry added nothing useful to the analysis. We agree with Professor Tribe that "[i]n order to test the extent and strength of the prior restraint doctrine . . . one must examine expression that is at least arguably outside the ambit of substantive first amendment protection, yet inside the ban on prior restraints." TRIBE, supra, at 1040.

114. See Fort Wayne Books, Inc. v. Indiana, 489 U.S. 46, 66 (1989):

Here there was not--and has not been--any determination that the seized items were "obscene" or that a RICO violation has occurred . . . . [T]he petition for seizure and the hearing thereon were aimed at establishing no more than probable cause to believe that a RICO violation had occurred, and the order for seizure recited no more than probable cause in that respect . . . . [O]ur cases firmly hold that mere probable cause to believe a legal violation has transpired is not adequate to remove books or films from circulation.

See also M.I.C., Ltd. v. Bedford Township, 463 U.S. 1341, 1343 (1983) (Brennan, Circuit Justice) (issuing an emergency stay and noting that a stay is constitutionally mandated when "the trial court's broad proscription will bar, in advance of any final judicial determination that the suppressed films are obscene, the exhibition of any film that might offend the court's ban"); Vance v. Universal Amusement Co., 445 U.S. 308, 316-17 (1980) (holding that a preliminary injunction was a prior restraint).

115. See, e.g., Vance, 445 U.S. at 312; Blount v. Rizzi, 400 U.S. 410, 420 (1971).

116. See, e.g., Philip Hamburger, The Development of the Law of Seditious Libel and the Control of the Press, 37 STAN. L. REV. 661, 671-73 (1984).

117. See, e.g., Freedman v. Maryland, 380 U.S. 51 (1965).

118. 283 U.S. 697 (1931).

119. See id. at 706.

120. The one possible exception here is commercial speech, which might be subjected to various prescreening systems that are forbidden for noncommercial speech. See infra notes 330-331 and accompanying text (discussing the uncertainty in the case law on this point). But even if this exception is sound, it fits well into our general framework. Schemes that restrain speech before a court makes a final adjudication that the speech is unprotected risk restraining protected speech as well as unprotected speech. When that risk is borne by speech that is fully protected--such as nonobscene movies or nonlibelous newspaper articles--the prior restraint cases hold that this is too grave a restraint on free speech. But when the risk is borne by commercial speech, which even when protected is considered to be of lower value than noncommercial speech, see infra notes 316-318 and accompanying text, then the First Amendment cost of erroneous temporary restraint is considerably less.

121. 445 U.S. 308 (1980).

122. Id. at 312.

123. Id. at 316; see also Blount v. Rizzi, 400 U.S. 410, 420 (1971) (holding that a determination by a judge of "probable cause" that speech is obscene is insufficient to justify a restriction); State v. Book-Cellar, Inc., 679 P.2d 548, 553-55 (Ariz. Ct. App. 1984) (upholding a statute that authorized preliminary injunctions against the distribution of obscenity by requiring "that a final judicial determination [be] made by the end of 60 days from the issuance of a preliminary injunction," a safeguard compelled by Freedman v. Maryland, 380 U.S. 51 (1965)); Paradise Hills Assocs. v. Procel, 1 Cal. Rptr. 2d 514, 519 (Cal. Ct. App. 1991) ("A preliminary injunction is a prior restraint."); Cohen v. Advanced Med. Group, 496 S.E.2d 710, 710-11 (Ga. 1998) (striking down a preliminary injunction of libel on the grounds that the injunction was not "'entered subsequent to a verdict in which a jury found that statements made by [defendant] were false and defamatory'" (quoting High Country Fashions, Inc. v. Marlenne Fashions, Inc., 357 S.E.2d 576, 577 (Ga. 1987))); City of Cadillac v. Cadillac News & Video, Inc., 562 N.W.2d 267, 270 (Mich. Ct. App. 1996) (striking down a preliminary injunction of obscenity on the grounds that the injunction would permit "removal of allegedly obscene materials from circulation before a judicial determination whether the material is obscene, with none of the safeguards" established in Freedman). Freedman is discussed further infra Part II.B.3.

124. Vance, 445 U.S. at 317.

125. See, e.g., Auburn Police Union v. Carpenter, 8 F.3d 886, 903 (1st Cir. 1993) (stressing that an injunction of charitable solicitation was permitted only "after a final adjudication on the merits that the speech is unprotected"); St. Margaret Mercy Healthcare Ctrs., Inc. v. Ho, 663 N.E.2d 1220, 1223-24 (Ind. Ct. App. 1996) (holding that a preliminary injunction of an alleged libel was an unconstitutional prior restraint); Gilbert v. National Enquirer, Inc., 51 Cal. Rptr. 2d 91, 96-99 (Cal. Ct. App. 1996) (same); Pirmantgen v. Feminelli, 745 S.W.2d 576, 577-78 (Tex. Ct. App. 1988) (same); cf. San Antonio Community Hosp. v. Southern Cal. Dist. Council of Carpenters, 125 F.3d 1230, 1240 (9th Cir. 1997) (Kozinski, J., dissenting) (criticizing the majority, and arguing that "this . . . is the first ever [case] (so far as I am aware) to uphold [on appeal] a preliminary injunction against speech covered by [New York Times Co. v. Sullivan]"); San Antonio Community Hosp. v. Southern Cal. Dist. Council of Carpenters, 137 F.3d 1090, 1092-93 (9th Cir. 1998) (Reinhardt, J., dissenting from the denial of rehearing en banc) (stating that "until today, no court has ever upheld [on appeal] an injunction prohibiting the type of speech at issue here"). But see San Antonio Community Hosp., 125 F.3d at 1233-39 (concluding that a preliminary injunction in a labor union libel case was not a prior restraint because the statements were so misleading as to be fraudulent, and "[t]he First Amendment does not protect fraud"); Vondran v. McLinn, No. C 95-20296 RPA, 1995 WL 415153, at *4-*6 (N.D. Cal. July 5, 1995) (acknowledging that speech may be enjoined only when the statements have actually been found to be libelous, but nonetheless preliminarily enjoining libelous statements based on a mere finding of likelihood of success on the merits); Mercy Health Servs. v. 1199 Health & Human Serv. Employees Union, 888 F. Supp. 828 (W.D. Mich. 1995) (allowing a preliminary injunction in a labor union libel case without even considering the prior restraint question); Bingham v. Struve, 591 N.Y.S.2d 156, 158-59 (Sup. Ct. App. Div. 1992) (ordering a preliminary injunction against a libel on a matter of private concern, concluding that the libel was constitutionally unprotected but not considering the prior restraint dimensions).

126. Pittsburgh Press Co. v. Pittsburgh Comm'n on Human Relations, 413 U.S. 376, 390 (1973).

127. See Richard Feiner & Co. v. Turner Entertainment Co., 98 F.3d 33, 34 (2d Cir. 1996).

128. Id.

129. Publications Int'l, Ltd. v. Meredith Corp., 88 F.3d 473, 480 (7th Cir. 1996) (citing Roland Machinery v. Dresser Indus., Inc., 749 F.2d 380, 387 (7th Cir. 1984)); see also Fashion Victim, Ltd. v. Sunrise Turquoise, Inc., 785 F. Supp. 1302, 1307 (N.D. Ill. 1992) (denying a preliminary injunction because the plaintiff failed to demonstrate a "'better than negligible'" chance that it would succeed on the merits (quoting Roland Machinery, 749 F.2d at 387)); Jackson v. MPI Home Video, 694 F. Supp. 483, 492 (N.D. Ill. 1988) (finding a "greater than negligible chance" that the plaintiff could gain injunctive relief).

130. New Line Cinema Corp. v. Bertlesman Music Group, Inc., 693 F. Supp. 1517, 1531 (S.D.N.Y. 1988).

131. See Walker v. City of Birmingham, 388 U.S. 307 (1967).

132. See supra Part I.B. Professor Doug Laycock suggests that courts are effectively applying the irreparable injury standard for permanent injunctions to preliminary injunction cases. See LAYCOCK, supra note 37, at 37-48.

133. Cheever v. Academy Chicago, Ltd., 690 F. Supp. 281, 289 (S.D.N.Y. 1988).

134. See, e.g., material from Alexander v. United States, quoted supra note 113.

135. See infra Part II.B.3.

136. See, e.g., Fort Wayne Books, Inc. v. Indiana, 489 U.S. 46 (1989).

137. See John Calvin Jeffries, Jr., Rethinking Prior Restraint, 92 YALE L.J. 409, 427-29 (1983) (arguing that subsequent criminal punishments usually chill speech more than injunctions); Stephen R. Barnett, The Puzzle of Prior Restraint, 29 STAN. L. REV. 539, 550-51 (1977) (same).

138. See University of Texas v. Camenisch, 451 U.S. 390, 395 (1981) ("A party . . . is not required to prove his case in full at a preliminary-injunction hearing."); Compuserve, Inc. v. Cyberpromotions, Inc., 962 F. Supp. 1015, 1019 (S.D. Ohio 1997) ("A preliminary injunction is customarily granted on the basis of procedures that are less formal and evidence that is less complete than in a full trial on the merits.").

139. We'll discuss in Part III whether this judgment is sound. But, sound or not, it is quite clearly part of free speech jurisprudence. See, e.g., supra note 114 and accompanying text.

140. Vincent Blasi, Toward a Theory of Prior Restraint: The Central Linkage, 66 MINN. L. REV. 11, 92-93 (1981).

141. See Kingsley Books, Inc. v. Brown, 354 U.S. 436, 447-48 (1957) (Brennan, J., dissenting). We are somewhat skeptical that juries will in practice prove more protective of speech than judges, but such a discussion is beyond the scope of this Article.

142. See Redish, supra note 105. To briefly summarize--and thus necessarily vastly oversimplify--Professor Redish's points: (1) Permanent injunctions against speech found by a court to be unprotected are in practice not especially likely (particularly when compared with criminal punishments) to deter or restrict speech that should be protected. (2) Permanent injunctions are not especially likely to be overused, because, like civil and criminal punishments, they can only be entered after a full trial on the merits. (3) The main supposed procedural advantage of civil and criminal trials--the presence of the jury--is in fact not especially speech-protective. (4) The danger of adjudication in the abstract based only on speculation about future harm is often as present in criminal prosecutions, because many of the free speech exceptions (e.g., threats, fighting words, incitement, solicitation) allow the punishment of speech based solely on a sufficient risk that it will cause harm. (5) While audiences may be understandably skeptical about the credibility of speech that has been precleared by censors, they may be equally skeptical about speech that has been published with an eye towards avoiding subsequent punishment. See id. at 59-75.

For an economic defense of Redish's theory, see Christine Jolls et al., A Behavioral Approach to Law and Economics, 50 STAN. L. REV. 1471, 1516-17 (1998).

143. Redish, supra note 105, at 88-89.

144. Id. at 55 (emphasis in original). See generally Part III (explaining why we think the Vance view is correct, especially in copyright cases).

145. 388 U.S. 130 (1967).

146. Id. at 149 (citation omitted). As an example of one of the "strong arguments," the Court cited Roscoe Pound, Equitable Relief Against Defamation and Injuries to Personality, 29 HARV. L. REV. 640 (1916)--a leading article urging that libels be enjoinable. The only citation the Court gave for the proposition that it had "rejected all manner of prior restraint on publication" was Near v. Minnesota, 283 U.S. 697 (1931), which rejected an injunction against speech that had not yet been adjudicated to be constitutionally unprotected.

More recently, in Organization for a Better Austin v. Keefe, 402 U.S. 415 (1971), the Court cited Near v. Minnesota to similar effect:

It is elementary, of course, that in a case of this kind the courts do not concern themselves with the truth or validity of the publication. Under Near v. Minnesota, the injunction, so far as it imposes prior restraint on speech and publication, constitutes an impermissible restraint on First Amendment rights.

Id. at 418 (citation omitted). Again, however, the Court doesn't explain how Near applied to libel cases generally, nor did it distinguish Kingsley Books, which held that some speech may be enjoined once it's found to be unprotected. See Kingsley Books, 354 U.S. at 445. In addition, the statement in Keefe, like that in Curtis Publishing, is in dictum and in passing: Keefe was not a libel case, and the speech at issue in Keefe would have been protected against subsequent punishment as well as against prior restraint. See Frisby v. Schultz, 487 U.S. 474, 485 (1988) (stating that speech in residential areas that's aimed at reaching the public--the sort of speech involved in Keefe--may not be banned).

147. 8 Paige Ch. 24, 26 (N.Y. Ch. 1839):

It is very evident that this court cannot assume jurisdiction of the case presented by the complainant's bill, or of any other case of the like nature, without infringing upon the liberty of the press, and attempting to exercise a power of preventive justice which, as the legislature has decided, cannot safely be entrusted to any tribunal consistently with the principles of a free government.

Accord New York Juvenile Guardian Soc'y v. Roosevelt, 7 Daly 188, 191 (C.P.N.Y. 1877); State ex rel. Liversey v. Judge Civ. Dist. Ct., 34 La. Ann. 741, 746 (1882).

148. See, e.g., Wilson v. Superior Ct., 532 P.2d 116, 120 (Cal. 1975) (purporting to apply both the United States and California constitutions); Gilbert v. National Enquirer, Inc., 51 Cal. Rptr. 2d 91, 96 (Cal. Ct. App. 1996) (same); Guste v. Connick, 515 So. 2d 436, 438 (La. 1987); Willing v. Mazzocone, 393 A.2d 1155, 1157 (Pa. 1978); see also Kramer v. Thompson, 947 F.2d 666 (3d Cir. 1991) (discussing at length Pennsylvania's constitutional prohibition against prior restraint of libelous speech).

149. See supra note 112.

150. See United States v. Thirty-Seven Photographs, 402 U.S. 363, 367-75 (1971); 11126 Baltimore Blvd., Inc. v. Prince George's County, 58 F.3d 988, 1001-02 (4th Cir. 1995) (holding that an eight-month delay for a final judicial determination could not be considered the brief and prompt review "that is required to guard against the abridgment of protected speech"); State v. Book-Cellar, Inc., 679 P.2d 548, 555 (Ariz. App. 1984) (stating that for prior restraints to "pass constitutional muster," adequate procedural safeguards, including a prompt and final judicial determination, must exist).

151. 380 U.S. 51, 59 (1965).

152. See, e.g., MARTIN H. REDISH, FREEDOM OF EXPRESSION 151 (1984).

153. But see Times Film Corp. v. City of Chicago, 365 U.S. 43 (1961) (upholding such a system aimed at screening out obscene movies).

154. But cf. Blount v. Rizzi, 400 U.S. 410, 417 (1971) (suggesting that such a system might be constitutional when applied to material that's supposedly obscene if the system contained the procedural protections required by Freedman).

155. See National Socialist Party of Am. v. Village of Skokie, 432 U.S. 43, 44 (1977); cf. New York State Ass'n of Career Sch. v. State Educ. Dep't, 823 F. Supp. 1096, 1106 (S.D.N.Y. 1993) (holding that a scheme requiring preapproval of adult education curricula was a prior restraint, but concluding that such a restraint was valid because it passed strict scrutiny and incorporated procedural safeguards like those present in Freedman).

156. See Freedman, 380 U.S. at 58-59.

157. 393 U.S. 175 (1968).

158. See id. at 180 (stating that because the TRO was obtained "without any effort, however informal, to invite or permit [the opposing party's] participation in the proceedings," it must be set aside).

159. See id.

160. Id. at 183; cf. NIMMER ON COPYRIGHT, supra note 50, 14.07, at 14-128 to 14-130 (discussing this point in the related context of ex parte impoundment of allegedly infringing articles under the Copyright Act).

161. See Carroll, 393 U.S. at 184-85 ("We need not here decide that it is impossible for circumstances to arise in which the issuance of an ex parte restraining order for a minimum period could be justified because of the unavailability of the adverse parties or their counsel, or perhaps for other reasons.").

162. See, e.g., Procter & Gamble Co. v. Bankers Trust Co., 78 F.3d 219, 225-26 (6th Cir. 1996) (setting aside an ex parte TRO against publication of confidential court documents, even though the district court concluded that it needed to issue the TRO to have time to consider the evidence); Howard Gault Co. v. Texas Rural Legal Aid, Inc., 848 F.2d 544, 561 (5th Cir. 1988) (setting aside an ex parte TRO against union picketing); Anderson v. Dean, 354 F. Supp. 639, 642 (N.D. Ga. 1973) (holding that an ex parte TRO against making libelous statements was unconstitutional unless adequate safeguards were present); United Farm Workers v. Quincy Corp., 681 So. 2d 773, 777 (Fla. Dist. Ct. App. 1996) (setting aside an ex parte TRO against union demonstrations and picketing); State v. Alston, 887 P.2d 681, 688 (Kan. 1994) (setting aside sua sponte a gag order prohibiting the publication of information that would have allegedly interfered with the fairness of a criminal trial). But see Falwell v. Penthouse Int'l, Ltd., 521 F. Supp. 1204, 1206 (W.D. Va. 1981) (mentioning an ex parte TRO that enjoined the distribution of an issue of Penthouse on the grounds that it might have constituted "false light" invasion of privacy and infringement of common law copyright, but not discussing the First Amendment).

163. Compare Dallas Cowboys Cheerleaders, Inc. v. Scoreboard Posters, Inc., 600 F.2d 1184, 1188 (5th Cir. 1979) (concluding that the prior restraint doctrine doesn't apply to copyright and trademark cases because "[t]he first amendment is not a license to trammel on legally recognized rights in intellectual property"), with L.L. Bean, Inc. v. Drake Publishers, Inc., 811 F.2d 26, 28-29 (1st Cir. 1987) (disagreeing with this view in the trademark and parody context, and noting that "the constitutional issue raised here cannot be dispensed with by simply asserting that Bean's property right need not yield to the exercise of first amendment rights"). See also Wendy J. Gordon, A Property Right in Self-Expression: Equality and Individualism in the Natural Law of Intellectual Property, 102 YALE L.J. 1533, 1537 (1993) (expressing concern that "[t]he incantation 'property' seems sufficient to render free speech issues invisible").

164. See, e.g., Marrero v. City of Hialeah, 625 F.2d 499, 514 (5th Cir. 1980) (noting that Florida law recognizes business reputation as a property interest); Nossen v. Hoy, 750 F. Supp. 740, 743 (E.D. Va. 1990) (holding that "an individual holds a . . . property interest in his or her reputation" for purposes of Washington and Virginia conversion law). Given the economic value of reputation--something any businessperson would attest to--such an approach isn't surprising. To most businesses and businesspeople, one's reputation is considerably more valuable than the copyrights one may own.

Reputation is generally not a property interest for purposes of the federal Due Process Clause, so the government may defame people without incurring constitutional liability. See Paul v. Davis, 424 U.S. 693, 711-12 (1976). But this merely means that the government doesn't have to consider reputation to be a property interest; the government may still do so if it wants to.

165. H.R. 3883, 104th Cong. (1996) (introduced by Rep. Torricelli).

166. See Texas v. Johnson, 491 U.S. 397, 406 (1989) (holding that flag burning is protected speech); United States v. Eichman, 496 U.S. 310, 315 (1990) (declining to reverse Johnson).

167. The property argument has appeared in other contexts as well. In Federal Election Commission v. International Funding Institute, Inc., 969 F.2d 1110 (D.C. Cir. 1992) (en banc), for example, the Court of Appeals for the D.C. Circuit upheld a restriction on the use of political contributor names collected by the Federal Election Commission. The FEC got the names from political committees (which are required to give them to the FEC) and then revealed them to any member of the public on the condition that they not be used for fundraising purposes. See id. at 1112-13. Judges Buckley and Randolph would have upheld this condition simply on the grounds that "treating the lists as property provides reason enough to find [the condition] constitutional." Id. at 1120 (Buckley, J., concurring); see also id. at 1121 (Randolph, J., concurring). We don't agree with this logic, and instead prefer the majority's reasoning, which conceded that the condition restricted speech, but upheld the condition on the grounds that the government may attach such strings when it releases information that it has no obligation to release. See id. at 1114 (following Seattle Times Co. v. Rhinehart, 467 U.S. 20 (1984)).

168. See generally San Francisco Arts & Athletics, Inc. v. United States Olympic Comm., 483 U.S. 522 (1987) (quasi-trademark); Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539 (1985) (copyright); Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 562 (1977) (right of publicity).

169. See Hudgens v. NLRB, 424 U.S. 507 (1976).

170. Cf. Madsen v. Women's Health Ctr., Inc., 512 U.S. 753, 764 n.2 (1994) (concluding that certain content-neutral injunctions that may incidentally affect expression are not prior restraints).

171. See, e.g., Rogers v. Grimaldi, 875 F.2d 994, 999 (2d Cir. 1989) (rejecting the trespass law analogy in a trademark case on the grounds that "a restriction on the location of a speech is different from a restriction on the words the speaker may use"); L.L. Bean, Inc. v. Drake Publishers, Inc., 811 F.2d 26, 29 (1st Cir. 1987) ("The first amendment issues involved in this case cannot be disposed of by equating the rights of a trademark owner with the rights of an owner of real property . . . ."). But see Dallas Cowboys Cheerleaders v. Pussycat Cinema, Ltd., 604 F.2d 200, 206 (2d Cir. 1979) ("Plaintiff's trademark is in the nature of a property right, and as such it need not 'yield to the exercise of First Amendment rights under circumstances where adequate alternative avenues of communication exist.'" (quoting Lloyd Corp. v. Tanner, 407 U.S. 551 (1972) (involving trespass law) (citations omitted))).

172. See supra notes 91-101 and accompanying text.

173. Among many other discussions of this "public good" characteristic of intellectual property, see Peter S. Menell, An Analysis of the Scope of Copyright Protection for Application Programs, 41 STAN. L. REV. 1045 (1989).

174. Compare College Sav. Bank v. Florida Prepaid Postsecondary Educ. Expense Bd., 148 F.3d 1343, 1347 (Fed. Cir. 1998) cert. granted, 1999 WL 5331 (U.S. Jan. 8, 1999) (No. 98-531) (holding that patents are property for Fourteenth Amendment purposes), with Chavez v. Arte Publico Press, 139 F.3d 504, 510 (5th Cir. 1998) (holding that copyrights are not property for Fourteenth Amendment purposes).

175. Cf. Lloyd Corp., 407 U.S. at 567 (holding that there is no First Amendment right to access private property to distribute handbills).

176. See, e.g., New York Times Co. v. United States, 403 U.S. 713, 731 n.1 (1971) (White, J., concurring) (suggesting that private judicial enforcement of copyrights might not raise the same First Amendment risks as government suits seeking to enjoin the press from publishing information about government affairs).

177. Criminal libel law may still be constitutional, see Garrison v. Louisiana, 379 U.S. 64 (1964), but such prosecutions are very rare.

178. See libel cases cited supra notes 123 and 125.

179. See New York Times Co. v. Sullivan, 376 U.S. 254, 265 (1964) ("Although [a libel suit] is a civil lawsuit between private parties, the [state] courts have applied a state rule of law . . . . It matters not that that law has been applied in a civil action . . . ."); cf. Lugar v. Edmondson Oil Co., 457 U.S. 922, 940-41 (1982) (holding that judicial enforcement of a civil writ against the defendant's property is state action).

There's no doubt that a court's enforcement of copyright law to restrict private speech constitutes state action. Cf. Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 562, 574-78 (1977) (assuming that private enforcement of a government-granted intellectual property right--here the right of publicity--constitutes state action); Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539, 558-60 (1985) (discussing First Amendment defense to a copyright claim without even suggesting that there's any state action difficulty); Cardtoons, L.C. v. Major League Baseball Players Ass'n, 95 F.3d 959, 968 (10th Cir. 1996) (stating that private enforcement of the right of publicity constitutes state action). See also Julie E. Cohen, A Right to Read Anonymously: A Closer Look at "Copyright Management" in Cyberspace, 28 CONN. L. REV. 981, 1020-21 (1996) (arguing that a finding of state action is implicit in the fact that copyright protection is limited by the First Amendment); Robert C. Denicola, Trademarks as Speech: Constitutional Implications of the Emerging Rationales for the Protection of Trade Symbols, 1982 WIS. L. REV. 158, 193 n.146 ("The impact of the first amendment on the exclusive rights granted to copyright owners remains unclear. . . . [However, t]he applicability of the constitutional limitations has not been challenged.") (citations omitted); Marquis, supra note 103, at 136 (noting that the Supreme Court has determined that the First Amendment is "implicated where a private party seeks enforcement of a government-created right of action that limits free speech") (citation omitted).

180. Privately enforced laws might sometimes prove to be more restrictive than government-enforced ones. See Hughes Aircraft Co. v. United States, 117 S. Ct. 1871, 1877-78 (1997) (suggesting that private, self-interested enforcement of legal rules may be more zealous and more thorough than direct government enforcement).

181. But see Dr. Seuss Enters. v. Penguin Books USA, Inc., 109 F.3d 1394, 1403 n.11 (9th Cir. 1997) (citing Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 206 (2d Cir. 1979) ("The prohibition of the Lanham Act is content neutral, and therefore does not arouse the fears that trigger the application of constitutional 'prior restraint' principles." (citations omitted in original)).

182. See, e.g., Burson v. Freeman, 504 U.S. 191, 198 (1992) (plurality opinion) (holding that content-based restrictions, even viewpoint-neutral ones, are constitutionally suspect); Boos v. Barry, 485 U.S. 312, 321 (1988) (same); Arkansas Writers' Project, Inc. v. Ragland, 481 U.S. 221, 230 (1987) (same); Consolidated Edison Co. v. Public Serv. Comm'n, 447 U.S. 530, 537-38 (1980) (same); Carey v. Brown, 447 U.S. 455, 462 n.6 (1980) (same).

183. See Twentieth Century-Fox Film Corp. v. MCA, Inc., 715 F.2d 1327, 1329 (9th Cir. 1983).

184. See, e.g., Winters v. New York, 333 U.S. 507, 510 (1948) (holding that lowbrow entertainment is "as much entitled to the protection of free speech as the best of literature"); Southeastern Promotions, Ltd. v. Conrad, 420 U.S. 546 (1975) (holding that the play Hair, and theatrical productions generally, were constitutionally protected); Ward v. Rock Against Racism, 491 U.S. 781 (1989) (holding that music is fully constitutionally protected); Hurley v. Irish-American Gay, Lesbian, and Bisexual Group of Boston, 515 U.S. 557 (1995) (holding that a nonpolitical parade, even one that lacks a particularly clear and articulate message, is fully constitutionally protected). Commentators have likewise warned against testing First Amendment protection by reference to the supposed political value of the defendant's work. See Denicola, supra note 103, at 313-15; Marci A. Hamilton, Art Speech, 49 VAND. L. REV. 73, 103-06 (1996).

185. 445 U.S. 308 (1980).

186. See Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539, 558 (1985) (involving excerpts from President Ford's memoirs); see also Religious Tech. Ctr. v. Netcom On-Line Communication Servs., Inc., 907 F. Supp. 1361 (N.D. Cal. 1995) (involving religious scriptures).

187. See Time, Inc. v. Bernard Geis Assoc., 293 F. Supp. 130, 141-44 (S.D.N.Y. 1968).

188. See Los Angeles News Serv. v. Tullo, 24 U.S.P.Q.2d 1026 (9th Cir. 1992).

189. See, e.g., Time, Inc. v. Firestone, 424 U.S. 448 (1976); cf. Cantrell v. Forest City Publ'g Co., 419 U.S. 245 (1974) (false light invasion of privacy case involving entertainment); Time, Inc. v. Hill, 385 U.S. 374 (1967) (same).

190. Technically, Monica Lewinsky's statements on the notorious audiotapes are probably not protected by copyright, because the tapes were not created "by or under [her] authority," see 17 U.S.C. 101, 102 (1994), but this doesn't affect our broader point.

191. Harper & Row, 471 U.S. at 558.

192. The Copyright and Patent Clause, U.S. CONST. art. I, 8, cl. 8, is discussed in detail infra at Part II.D.6.

193. As Jessica Litman notes, "the harmony between the First Amendment and copyright doesn't inhere in their essential nature; rather, it derives from accommodations and restrictions we have built into copyright to enhance its role as an engine of free expression." Litman, supra note 101, at 619. See also David L. Lange, The Intellectual Property Clause in Contemporary Trademark Law: An Appreciation of Two Recent Essays and Some Thoughts About Why We Ought to Care, 59 LAW & CONTEMP. PROBS. 213, 240 (1996) (echoing Litman's view as to the way in which intellectual property laws enhance free expression principles); Kreiss, supra note 32 at 6-10 (noting the tension between protecting copyrighted works and encouraging the dissemination of information and speech). For a stronger argument challenging the core compatibility of copyright and the First Amendment, see Lange, supra note 6, at 140-44 (arguing that copyright grew out of government censorship efforts: "what comes naturally to copyright . . . is the deliberate, if selective, suppression and advancement of speech").

Justin Hughes has offered a First Amendment defense of copyright based on the listener's as well as the speaker's need for integrity, arguing that "freedom of expression is meaningless without assurances that the expression will remain unadulterated." Justin Hughes, The Philosophy of Intellectual Property, 77 GEO. L.J. 287, 359 (1988). Hughes takes as his paradigm cases instances in which defendants have altered an author's work in ways the author finds objectionable, and then attributed the adulterated work to the original author. See id. at 360-65. The cases he discusses are troubling, but primarily because they involve misrepresentations as to the source of various parts of the work. Such deceptive speech may in some situations constitute a "false designation of origin" actionable under the trademark laws; whether or not it does, we would distinguish misrepresentations about the source of speech from additions, deletions or changes to the speech itself.

194. Gertz v. Robert Welch, Inc., 418 U.S. 323, 400 (1974) (White, J., dissenting) (endorsing the view that "fascists' effective use of defamatory attacks on their opponents" suggests that "'the law of libel . . . [may be] important for modern democratic survival'" (quoting David Riesman, Democracy and Defamation: Fair Game and Fair Comment I, 42 COLUM. L. REV. 1085, 1088 (1942))).

195. See, e.g., ANDREA DWORKIN & CATHARINE A. MACKINNON, PORNOGRAPHY AND CIVIL RIGHTS: A NEW DAY FOR WOMEN'S EQUALITY 47-48 (1988); see also Paul E. McGreal, Constitutional Illiteracy, 30 IND. L. REV. 693, 697 n.30 (1997) (discussing this argument).

196. See, e.g., Mari J. Matsuda, Public Response to Racist Speech: Considering the Victim's Story, 87 MICH. L. REV. 2320, 2337 & n.86 (1989). But see David F. McGowan & Ragesh K. Tangri, A Libertarian Critique of University Restrictions of Offensive Speech, 79 CAL. L. REV. 825 (1991) (critiquing Matsuda's argument in detail).

197. See Cass R. Sunstein, Free Speech Now, in THE BILL OF RIGHTS IN THE MODERN STATE 255, 270-71 (Geoffrey R. Stone et al. eds., 1992); Owen M. Fiss, Free Speech and Social Structure, 71 IOWA L. REV. 1405, 1410 (1986). But see Bradley A. Smith, Money Talks: Speech, Corruption, Equality, and Campaign Finance, 86 GEO. L.J. 45, 64-76 (1997) (agreeing that one goal of the First Amendment was to ensure open and robust public discussion, but arguing that "this was to be achieved by protecting individual liberty interests against government interference, not through an activist government role in political debate").

198. See generally Eugene Volokh, Freedom of Speech and the Constitutional Tension Method, 3 U. CHI. L. SCH. ROUNDTABLE 223 (1996) (discussing and criticizing arguments based on the supposed tension between the Free Speech Clause and "values" flowing from other constitutional provisions).

199. U.S. CONST. art. I, 8, cl. 8.

200. See Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539, 558 (1985); see also Cable/Home Communication Corp. v. Network Prods., Inc., 902 F.2d 829, 849 (11th Cir. 1990) ("Since the Copyright Act is the congressional implementation of a constitutional directive to encourage inventors by protecting their exclusive rights in their discoveries, copyright interests also must be guarded under the Constitution, and injunctive relief is a common judicial response to infringement of a valid copyright."); 1 RODNEY A. SMOLLA, SMOLLA AND NIMMER ON FREEDOM OF SPEECH 15:58, at 15-87 to -88 (arguing that since "copyright protection is itself a value of constitutional dimension, vindicating the directive of the Constitution's Copyright Clause," injunctions should be permitted).

201. See, e.g., Lamont v. Postmaster Gen., 381 U.S. 301, 307 (1965) (striking down a restriction on mailing of communist advocacy); see also United States v. Grace, 461 U.S. 171 (1983) (striking down on First Amendment grounds a law, enacted pursuant to the Federal District Clause power, prohibiting the public display of any flag, banner, or device on the grounds of the Supreme Court); Alex Kozinski & Eugene Volokh, A Penumbra Too Far, 106 HARV. L. REV. 1639, 1649 (1993).

202. See, e.g., Time Warner Entertainment Co. v. Does, 876 F. Supp. 407 (E.D.N.Y. 1994) (holding that an ex parte seizure of supposedly infringing materials, though authorized by the Copyright Act, was impermissible under the Fourth Amendment).

203. See Patton v. United States, 281 U.S. 276, 298 (1930) ("The first ten amendments and the original Constitution were substantially contemporaneous and should be construed in pari materia.").

204. Harper & Row, 471 U.S. at 558.

205. But see Religious Tech. Ctr. v. Netcom On-Line Communication Servs., Inc., 923 F. Supp. 1231, 1258 (N.D. Cal. 1995) (concluding that Harper & Row implicitly validated all the provisions of the Copyright Act, including those providing for preliminary injunction).

206. See, e.g., Cadence Design Sys., Inc. v. Avant! Corp., 125 F.3d 824, 826-27 (9th Cir. 1997).

207. We are skeptical that this is necessarily true, however, and might therefore question the Cadence court's conclusion that the presumption of irreparable harm can never be rebutted in a copyright case by a showing that monetary damages would be adequate. See id. at 829.

208. See, e.g., Twentieth Century Music Corp. v. Aiken, 422 U.S. 151, 156 (1975). We stress the "primarily"--many scholars, and some countries, have argued for the primacy of noneconomic motivations for protecting copyright owners. Still, under United States law, the predominant justification for copyright is the incentive it creates. For further discussion, and reference to numerous authorities, see Mark A. Lemley, Romantic Authorship and the Rhetoric of Property, 75 TEX. L. REV. 873, 888-90 (1997).

In any case, even if copyright law is also seen as a means of protecting an author's feelings or reputation, this would only make clearer the analogy to libel law, where preliminary injunctions to protect feelings or reputation are forbidden.

209. See Willing v. Mazzocone, 393 A.2d 1155, 1158 (Pa. 1978) (holding that a preliminary injunction against libel was improper even though the defendant was indigent and could not pay damages).

210. See supra Part II.B.1.

211. See In re Capital Cities/ABC, Inc., 918 F.2d 140, 144 (11th Cir. 1990) (dictum) (arguing that an injunction requiring defendant to remove infringing material "does not 'give directives as to the content of expression,' only the manner of expression" (quoting Goldblum v. National Broad. Corp., 584 F.2d 904, 907 (9th Cir. 1978))); Marquis, supra note 103, at 136-37. This approach apparently derives from Lloyd Corp. v. Tanner, 407 U.S. 551 (1972), where the Court held that the owner of a private shopping center could exclude speakers from his property, so long as adequate public fora existed that the speakers could use. See id. at 567. It is not at all clear to us that the "private property" analogy works well here, where the "property" at issue is intellectual rather than real (and is itself the speech at issue). See supra Part II.D.1. Nor will it necessarily always be the case that defendants can make the same point without using the plaintiff's copyrighted work.

212. See supra notes 151-156 and accompanying text.

213. 403 U.S. 15 (1971).

214. Id. at 26.

215. See infra Part III.C.

216. Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 562, 573 (1977).

217. Zacchini, from where the "only question is who gets to do the publishing" quote is drawn, involved literal copying. Id. at 570-74.

218. See, e.g., Sheldon v. Metro-Goldwyn Pictures Corp., 81 F.2d 49, 56 (2d Cir. 1936) ("True, much of the [defendant's work] owes nothing to the [plaintiff's work]; . . . but that is entirely immaterial; it is enough that substantial parts were lifted; no plagiarist can excuse the wrong by showing how much of his work he did not pirate.").

219. See Dr. Seuss Enters. v. Penguin Books USA, Inc., 109 F.3d 1394 (9th Cir. 1997); Tyler T. Ochoa, Dr. Seuss, the Juice, and Fair Use: How the Grinch Silenced a Parody, 46 J. COPR. SOC'Y (forthcoming 1999) (criticizing this result).

220. See Woods v. Universal City Studios, Inc., 920 F. Supp. 62 (S.D.N.Y. 1996). For similar cases, see Ringgold v. Black Entertainment Television, Inc., 126 F.3d 70 (2d Cir. 1997) (concluding that a preliminary injunction may be appropriate against a television show which portrayed portions of plaintiff's copyrighted poster in the background for a total of 26 seconds); Aielleen Fajardo, Holy Case of Copyright Infringement, Batman!, 4 UCLA ENT. L. REV. 263 (1996) (discussing a sculptor's case against the makers of Batman Forever for showing a brief glimpse of his sculpture, displayed on a street in Los Angeles). But see Chase-Riboud v. Dreamworks Inc., 45 U.S.P.Q.2d 1259 (C.D. Cal. 1997) (refusing to enjoin the movie Amistad days before release, both because of doubts about the plaintiff's claim and because the balance of hardships tipped in the defendant's favor); Sandoval v. New Line Cinema Corp., 973 F. Supp. 409 (S.D.N.Y. 1997) (holding that a fleeting use of copyrighted photographs as background in the movie Seven was a fair use).

221. See, e.g., Rosemont Enters. v. Random House, Inc., 366 F.2d 303, 306-07 (2d Cir. 1966) (describing the trial cout's reliance on copyright law to suppress an unauthorized biography of Howard Hughes); Religious Tech. Ctr. v. Netcom On-Line Communication Servs., Inc., 923 F. Supp. 1231 (N.D. Cal. 1995) (holding that the Church of Scientology could enjoin the defendant in order to keep its works from being further distributed to the public); New Era Publications Int'l v. Carol Publ'g Group, 904 F.2d 152 (2d Cir. 1990) (same); Salinger v. Random House, Inc., 811 F.2d 90 (2d Cir. 1987) (holding that J.D. Salinger could enjoin the defendant from distributing a book that contained excerpts from Salinger's private letters); Francis X. Clines, Creator of Religious Art Prevails on 'Devil' Film, N.Y. TIMES, Feb. 14, 1998, at A5 (discussing the successful copyright suit by a sculptor against the makers of the film Devil's Advocate, who used a similar sculpture in ways the artist found offensive); Richard Hurowitz, Surviving Copyright Infringement: Fair Use of Protected Works in "Biopics," 22 COLUM.-VLA J.L. & ARTS 247, 247 n.2 (1998) (noting that Picasso's heirs used copyright to prevent his works from appearing in the film Surviving Picasso "because they did not like the film's portrayal of the artist"); cf. 17 U.S.C. 106A (1990) (adding new protections, albeit limited ones, that allow artists to restrict certain behavior that they believe is prejudicial to their "honor or reputation").

222. Cf. Zacchini, 433 U.S. at 573-74 (stressing that Zacchini was asking only for damages, not an injunction).

223. See, e.g., United States v. 37 Photographs, 402 U.S. 363, 374 (1971); Blount v. Rizzi, 400 U.S. 410, 421 (1971); Freedman v. Maryland, 380 U.S. 51, 59 (1965). But see infra Part IV.D for some of our criticisms of such systems.

224. See Freedman, 380 U.S. at 59 ("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.").

225. 5 U.S. (1 Cranch) 137, 176-78 (1803).

226. New York Times Co. v. Sullivan, 376 U.S. 254, 269 (1964) (footnotes omitted):

[W]e are compelled by neither precedent nor policy to give any more weight to the epithet 'libel' than we have to other 'mere labels' of state law. Like insurrection, contempt, advocacy of unlawful acts, breach of the peace, obscenity, solicitation of legal business, and the various other formulae for the repression of expression that have been challenged in this Court, libel can claim no talismanic immunity from constitutional limitations. It must be measured by standards that satisfy the First Amendment.

227. Beauharnais v. Illinois, 343 U.S. 250, 255-56 (1952) ("'There are certain well-defined and narrowly limited classes of speech, the prevention and punishment of which have never been thought to raise any Constitutional problem. These include the lewd and obscene, the profane, the libelous, and the insulting or "fighting" words . . . .'" (quoting Chaplinsky v. New Hampshire, 315 U.S. 568, 571-72 (1942))), dismissed as overbroad dictum by New York Times Co. v. Sullivan, 376 U.S. 254, 268 (1964).

228. See, e.g., Roth v. United States, 354 U.S. 476, 488-89 (1957) (describing the constitutional boundaries of unprotected obscenity, which were narrower than had historically been applied); New York Times Co. v. Sullivan, 376 U.S. 254 (1964) (same, as to libel).

229. As late as 1908, the Supreme Court refused a permanent injunction to a copyright plaintiff on the ground that the defendant's work contained substantial new expression in addition to the copied material. See Dun v. Lumbermen's Credit Ass'n, 209 U.S. 20, 23-24 (1908). See generally Lemley, supra note 208, at 886-87.

230. See Richard Delgado & Jean Stefancic, Ten Arguments Against Hate-Speech Regulation: How Valid?, 23 N. KY. L. REV. 475, 484 (1996):

Powerful actors like government agencies, the writers' lobby, industries, and so on have always been successful at coining free speech 'exceptions' to suit their interest--copyright, false advertising, words of threat, defamation, libel, plagiarism, words of monopoly, and many others. But the strength of the interest behind these exceptions seems no less than that of a black undergraduate subjected to vicious abuse while walking late at night on campus.

See also Martin E. Lee, The Price We Pay: The Case Against Racist Speech, Hate Propaganda and Pornography, NAT'L CATHOLIC REP., Oct. 4, 1996, at 17 (book review):

Noting routine exceptions to free speech absolutism (copyright, trademark and such) that hew to business interests, the essays cite studies that document the heavy toll inflicted by the multibillion dollar porn industry, as it profits from a kind of hate speech that degrades women and children. . . . This book provides a sober rejoinder to cliche-ridden thinking by highlighting the profound power imbalance and social inequities that dim the luster of the First Amendment.

See also Richard Delgado & David H. Yun, Pressure Valves and Bloodied Chickens: An Analysis of Paternalistic Objections to Hate Speech Regulation, 82 CAL. L. REV. 871, 892 (1994) ("Perhaps . . . in twenty or fifty years we will look upon hate speech rules with the same equanimity with which we now view defamation, forgery, obscenity, copyright, and dozens of other exceptions to the free speech principle, and wonder why in the late twentieth century we resisted them so strongly.").

231. We are indebted to Doug Laycock for this point.

232. For challenges to the orthodoxy of prior restraint law, see LUCAS A. POWE, JR., THE FOURTH ESTATE AND THE CONSTITUTION 140-69 (1991); Frederick Schauer, Uncoupling Free Speech, 92 COLUM. L. REV. 1321, 1353 n.98 (1992); Frederick Schauer, Fear, Risk and the First Amendment: Unraveling the "Chilling Effect," 58 B.U. L. REV. 685, 725-30 (1978).

233. See supra note 131 and accompanying text.

234. Scot Powe notes that the threat of criminal prosecution did not stop the New York Times from publishing the Pentagon Papers, but that a lower court injunction did prevent publication while the injunction was in effect. See POWE, supra note 232, at 153.

235. See, e.g., JAMES POOLEY, TRADE SECRETS 7.02[2][a], at 7-7 (1997):

[S]ecuring an injunction early in a case often leads to prompt resolution of the entire matter. The defendant, constrained by the risk of contempt and looking down a long and expensive road toward a trial that he's likely to lose, will feel strongly motivated to settle, sometimes accepting business terms that would not be available to the plaintiff from a court.

236. Richard Feiner & Co. v. Turner Entertainment Co., 98 F.3d 33, 34 (2d Cir. 1996).

237. Publications Int'l, Ltd. v. Meredith Corp., 88 F.3d 473, 480 (7th Cir. 1996). At least one court has applied this standard to issue a preliminary injunction even though the plaintiff seemed unlikely to prevail on the merits. See Art Line, Inc. v. Universal Design Collections, Inc., 966 F. Supp. 737 (N.D. Ill. 1997).

238. See supra notes 59-66 and accompanying text.

239. There is a large literature on the importance not only of granting incentives to original authors, but also of encouraging the dissemination of new ideas and expression and allowing change and improvement to existing works. See, e.g., Kreiss, supra note 193 (discussing the copyright value of access and dissemination); Mark A. Lemley, The Economics of Improvement in Intellectual Property Law, 75 TEX. L. REV. 989, 993-1000 (1997) (emphasizing the need for others to build on the works of original authors, and citing sources).

240. See FED. R. CIV. P. 65(b).

241. Ideally, we would empirically study the degree to which decisions at the preliminary injunction stage match the ultimate decision on the merits, but we know no way to do this reliably and cost-effectively, especially given the number of preliminary injunction decisions that are unreported and the number of cases that settle shortly after the preliminary injunction.

242. Freedman v. Maryland, 380 U.S. 51, 59 (1965); see also CMM Cable Rep., Inc. v. Ocean Coast Properties, Inc., 48 F.3d 618, 620 (1st Cir. 1995) (applying this standard to trademark infringement); Religious Tech. Ctr. v. Netcom On-Line Communication Servs., Inc., 907 F. Supp. 1361, 1382 (N.D. Cal. 1995) (applying this standard to copyright infringement); Value Group, Inc. v. Mendham Lake Estates, 800 F. Supp. 1228, 1231 (D.N.J. 1992) (same); Cheever v. Academy Chicago, Ltd., 690 F. Supp. 281, 289 (S.D.N.Y. 1988) (same).

243. Religious Tech. Ctr. v. F.A.C.T.Net, Inc., 901 F. Supp. 1519, 1523 (D. Colo. 1995); cf. Financial Control Assocs., Inc. v. Equity Builders, Inc., 799 F. Supp. 1103, 1113 (D. Kan. 1992) ("The movant has the burden to establish by clear proof its right to a preliminary injunction."). This mantra, however, is more often breached than observed. See NIMMER ON COPYRIGHT, supra note 50, 14.06[A], at 14-112 to -113 ("[P]reliminary injunctions are sometimes reflexively labeled an 'extraordinary remedy'; in actual practice, nonetheless, their issuance is actually quite ordinary, even commonplace.") (footnotes omitted).

244. See, e.g., City of Lakewood v. Plain Dealer Publ'g Co., 486 U.S. 750, 755-56 (1988); Grayned v. City of Rockford, 408 U.S. 104, 108 (1972).

245. See City of Lakewood, 486 U.S. at 758; Grayned, 408 U.S. 104 at 108-09.

246. Cf. Amy B. Cohen, Copyright Law and the Myth of Objectivity: The Idea-Expression Dichotomy and the Inevitability of Artistic Value Judgments, 66 IND. L.J. 175 (1990) (arguing that the idea-expression dichotomy is indeterminate); Netanel, supra note 103, at 303-04 ("[T]he copyright law safeguards that have made First Amendment defenses seem overly intrusive and unnecessary have in fact been only sporadically effective in protecting First Amendment values," in part because "while the idea/expression dichotomy makes sense in principle, it is notoriously malleable and indeterminate"); id. at 381 (describing how "prevailing uncertainties" in copyright law interfere with the creation of certain kinds of new works); Robert H. Rotstein, Beyond Metaphor: Copyright Infringement and the Fiction of the Work, 68 CHI.-KENT L. REV. 725, 777-78 (1993) (citing sources); Alfred C. Yen, A First Amendment Perspective on the Idea/Expression Dichotomy and Copyright in a Work's "Total Concept and Feel," 38 EMORY L.J. 393 (1989) (explaining how the vagueness of the idea-expression dichotomy can deter constitutionally protected speech).

247. See MELVILLE B. NIMMER, NIMMER ON FREEDOM OF SPEECH: A TREATISE ON THE FIRST AMENDMENT 2.02, at 2-10 (1984).

248. What government employees say on matters of private concern is entirely unprotected against government-as-employer retaliation. See Connick v. Myers, 461 U.S. 138, 147-49 (1983). This part of the government employee speech test is a rule, not a balancing test; regardless of whether it's a correct rule, our focus in the text is on the balancing test used for public-concern speech.

249. Pickering v. Board of Educ., 391 U.S. 563, 568 (1968); accord Rankin v. McPherson, 483 U.S. 378, 388 (1987).

250. Connick, 461 U.S. at 150.

251. See Waters v. Churchill, 511 U.S. 661, 671-75 (1994) (plurality opinion) (discussing differences in the First Amendment tests applied when the government is acting as sovereign and when the government is acting as employer). For elaboration of this distinction, see Robert C. Post, Between Governance and Management: The History and Theory of the Public Forum, 34 UCLA L. REV. 1713 (1987).

252. See supra Part II.B.1 for a discussion of the Court's cases, and supra Part II.B.2 for a rough justification of their soundness.

253. See Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539, 558 (1985).

254. Cf., e.g., Rutan v. Republican Party, 497 U.S. 62, 74 (1990) (stressing that, as a general rule, content-based substantive speech restrictions are unconstitutional when there are less restrictive alternatives available).

255. Courts may be erroneously borrowing the "irreparable harm" concept from the permanent injunction context--where irreparable harm is definitionally assumed when property is involved--into the preliminary injunction context, where a higher quantum of injury had theoretically been required. See LAYCOCK, supra note 37, at 111.

256. See MERGES ET AL., supra note 50, at 515 ("[Intellectual property rights] are exceedingly difficult to value . . . . partly because, by statutory definition, each of these rights is in some sense unique.").

257. See LAYCOCK, supra note 37, at 111.

258. See, e.g., Nintendo of Am., Inc. v. Lewis Galoob Toys, Inc., 16 F.3d 1032 (9th Cir. 1994) (reviewing a dispute over the amount that the plaintiff had been forced to post as a preliminary injunction bond to compensate the defendant after the injunction proved to have been wrongly issued).

259. See supra Part II.D.1.

260. See infra Part III.C.

261. For example, in Salinger v. Random House, Inc., 811 F.2d 90 (2d Cir. 1987), J.D. Salinger objected to the publication of a book containing excerpts from his private letters. Similarly, in Religious Tech. Ctr. v. Netcom On-Line Communication Servs., Inc., 907 F. Supp. 1361 (N.D. Cal. 1995), the copyright holders objected to the Internet publication of L. Ron Hubbard's unpublished works. In Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994), the copyright holders of Roy Orbison's song Oh Pretty Woman objected to 2 Live Crew's adaptation of the song. And in Walt Disney Prods. v. Air Pirates, 581 F.2d 751 (9th Cir. 1978), Disney objected to the use of its characters in counterculture comic books that dealt with sex and drugs. We refer here primarily to traditional copyright law's ability to protect an artist's moral rights, not to the specific and narrow moral rights provisions of 17 U.S.C. 106A (1994).

262. See Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539, 558 (1985). For a further discussion of Harper & Row, see supra notes 91-98 and accompanying text.

263. The interest may be stronger where the adaptation involves a misrepresentation--that is, where the defendant passes off an adulterated work as the original. See Hughes, The Philosophy of Intellectual Property, supra note 193, at 358-65; Justin Hughes, "Recoding" Intellectual Property and Overlooked Audience Interests, 77 TEX. L. REV. (Forthcoming February 1999). But the moral rights go beyond misrepresentations, and prohibit alteration even if it is clear who has written what.

264. See, e.g., Milkovich v. Lorain Journal Co., 497 U.S. 1, 19-20 (1990) (holding that statements of opinion that don't contain implicit false factual assertions are protected); Hustler Magazine, Inc. v. Falwell, 485 U.S. 46, 50-52, 56 (1988) (holding that emotionally distressing statements are constitutionally protected unless they fall into an established First Amendment exception, at least when the statements are about public figures).

265. See 17 U.S.C. 106A (1994) (providing moral rights protection only to a limited class of paintings, drawings, prints, sculptures, and photographs, and only against destruction, intentional distortion, physical mutilation, and misattribution).

266. See Campbell v. Acuff-Rose Music Inc., 510 U.S. 569 (1994) (allowing 2 Live Crew's parody of Roy Orbison's song Oh Pretty Woman).

267. See, e.g., id. at 591-92 ("[W]hen a lethal parody, like a scathing theater review, kills demand for the original, it does not produce a harm cognizable under the Copyright Act.") (citing Fisher v. Dees, 794 F.2d 432, 438 (9th Cir. 1986) ("Biting criticism suppresses demand; copyright infringement usurps it.")).

268. See Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974) (false statements of fact); Miller v. California, 413 U.S. 15 (1973) (obscenity).

269. See Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539 (1985)

270. See supra note 220 and accompanying text. The "balance of hardships" prong may sometimes prevent such injunctions, but many courts generally ignore the balance of hardships once the plaintiff shows a likelihood of success on the merits. See supra notes 59-62 and accompanying text.

271. 467 F. Supp. 990 (W.D. Wis.), appeal dismissed, 610 F.2d 819 (7th Cir. 1979).

272. But see L.A. Powe, Jr., The H-Bomb Injunction, 61 U. COLO. L. REV. 55 (1990) (describing the case and suggesting that the preliminary injunction was less appropriate than many might think).

273. See supra Part II.C.

274. But see infra Part IV.C (discussing the exception for cases involving literal copying).

275. This assumes that copying and distribution of executable object code, as opposed to source code, is not "speech." If we're wrong in assuming this, then even the reproduction of object code would be problematic. Cf. infra notes 395-402 and accompanying text (discussing an analogous issue in the patent context).

Likewise, creating a new building based on a protected architectural work is in our view probably sufficiently nonexpressive that a preliminary injunction against this conduct wouldn't be a prior restraint.

276. See supra Part II.B.1.

277. See supra notes 145-149 and accompanying text.

278. See Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539, 555-60 (1985).

279. Of course, the injunction should not enjoin speech that isn't infringing. For instance, an injunction barring any use of plaintiff's work, even use only of ideas and not expression, or uses that are fair, would be wrong as a constitutional matter as well as wrong under substantive copyright law.

280. See supra text following note 255. On the economic propriety of the choice between the two, see Guido Calabresi & A. Douglas Melamed, Property Rules, Liability Rules, and Inalienability: One View of the Cathedral, 85 HARV. L. REV. 1089, 1106-10 (1972) (arguing that transaction costs and valuation difficulties influence the choice between property and liability rules); Ian Ayres & Eric Talley, Solomonic Bargaining: Dividing a Legal Entitlement to Facilitate Coasean Trade, 104 YALE L.J. 1027, 1092-94 (1995) (arguing that a divided entitlement rule may facilitate preference revelation during bargaining); Michael A. Heller, The Tragedy of the Anticommons: Property in the Transition From Marx to Markets, 111 HARV. L. REV. 621, 631 (1998) (arguing that overdivided property entitlements may inhibit the efficient use of property). On the implications of this work for intellectual property, see Robert P. Merges, Contracting into Liability Rules: Intellectual Property Rights and Collective Rights Organizations, 84 CAL. L. REV. 1293, 1391-93 (1996) (arguing that legal property rules can support the private creation of liability rules where efficient); Lemley, supra note 239, at 1061-65 (arguing that where efficient licensing does not occur, property rules must be allocated so as to achieve optimal improvement).

281. See id.

282. Unless, of course, the restriction is improperly content- or viewpoint-based, such as a special copyright rule that provides greater (or lesser) copyright protection but that applies only to, say, sexually suggestive speech, religious speech, racist speech, or communist speech. See R.A.V. v. City of St. Paul, 505 U.S. 377, 387-88 (1992) (holding that content-based regulations are presumptively invalid).

283. See, e.g., 17 U.S.C. 104A (1994).

284. There might be other constitutional objections, however. See Peter A. Jaszi, Goodbye to All That--A Reluctant (and Perhaps Premature) Adieu to a Constitutionally-Grounded Discourse of Public Interest in Copyright Law, 29 VAND. J. TRANSNAT'L L. 595, 606-08 (1996) (arguing that section 104A is inconsistent with the grant of power in the Copyright and Patent Clause).

285. See 17 U.S.C. 502(a) (1994).

286. See Duncan v. Louisiana, 391 U.S. 145, 159 (1968).

287. See, e.g., Melancon v. McKeithen, 345 F. Supp. 1025 (E.D. La.), aff'd without opinion sub nom. Hill v. McKeithen, 409 U.S. 943 (1972).

288. Summary judgment for the plaintiff in a libel case is rather rare, especially if the plaintiff is a public figure and the question therefore focuses on the defendant's mental state, but it is theoretically possible. See, e.g., Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (1986); Orlowski v. Koroleski, 651 N.Y.S.2d 137 (App. Div. 1996).

289. 380 U.S. 51 (1965).

290. See id. at 59 ("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."). For a more detailed discussion of the Freedman exception, see supra notes 150-156 and accompanying text.

291. Courts have discretion to order the trial of the action on the merits to be advanced and consolidated with the hearing of the application for a preliminary injunction. See FED. R. CIV. P. 65(a)(2); see also Surgidev Corp. v. Eye Tech., Inc., 648 F. Supp. 661, 668 (D. Minn. 1986) (applying the rule). Combining the preliminary injunction hearing with a full trial on the merits should satisfy the requirements of the First Amendment, though it may present Seventh Amendment problems if one of the parties requests a jury trial. See Feltner v. Columbia Pictures Television, Inc., 118 S. Ct. 1279 (1998) (holding that the Seventh Amendment requires right to jury trial in copyright cases whenever damages are sought).

292. See supra Part III.B.

293. See supra Part II.E.

294. Congress has not generally set up prior restraint schemes with fast-track review in other areas of the law, despite the Court's conclusion in Freedman that some such schemes would be valid. See Freedman, 380 U.S. at 59.

295. Richard Feiner & Co. v. Turner Entertainment Co., 98 F.3d 33, 34 (2d Cir. 1996).

296. Publications Int'l, Ltd. v. Meredith Corp., 88 F.3d 473, 480 (7th Cir. 1996).

297. Freedman, 380 U.S. at 59.

298. Cf. TRIBE, supra note 113, 12-36, at 1047 (suggesting that prior restraints may be permissible where the standard of proof on the merits is raised sufficiently).

299. We use "error" here in the sense of a determination at the preliminary injunction stage that is different from the final determination on the merits. Technically, this need not be an error at all; under the traditional preliminary injunction balance, a court may correctly enter a preliminary injunction, because of the likelihood of success on the merits and the balance of the hardships, but then correctly dissolve it when the final decision on the merits is reached. But the speech will have been suppressed based on a legal claim which turned out to be erroneous.

300. See Gertz v. Robert Welch, Inc., 418 U.S. 323, 342 (1974) (holding that clear and convincing evidence is required in public figure libel cases); cf. People v. Mitchell Bros.' Santa Ana Theater, 180 Cal. Rptr. 728, 730 (Ct. App. 1982) (holding that clear and convincing evidence is required in civil obscenity injunction cases).

301. Cf. Oakes, supra note 103, at 992-97 (suggesting that courts should pay more attention to traditional equitable factors when deciding whether to grant copyright injunctions, rather than simply presuming irreparable harm).

302. See MERGES ET AL., supra note 50, at 297-98.

303. Commentators and courts have increasingly articulated a "property" rationale for trademarks. See Stephen L. Carter, Does It Matter Whether Intellectual Property is Property?, 68 CHI.-KENT L. REV. 715, 715 (1993). But trademarks do not bear many of the indicia of real property, or even other forms of intellectual property. Rather, they have traditionally been thought of as a means of preventing consumers from being deceived. See MERGES ET AL., supra note 50, at 524-26.

304. See George Washington Mint, Inc. v. Washington Mint, Inc., 349 F. Supp. 255, 263 (S.D.N.Y. 1972) (contending that a preliminary injunction can be "an act of kindness" to the defendant, if issued before the defendant can invest substantial resources in an infringing mark).

305. Some have suggested that the proper goal of trademark law is not protecting consumers at all, but rather protecting investments in brand quality and reputation. See, e.g., MERGES ET AL., supra note 50, at 527-530 (discussing the economic rationales for trademark protection). To the extent that this is true, injunctive relief should be less important, since incentives to invest in brand quality might be preserved by damage awards rather than injunctive relief.

306. See 5 J. THOMAS MCCARTHY, MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION 30:32 (4th ed. 1998) [hereinafter MCCARTHY ON TRADEMARKS] (collecting cases from different circuits).

307. See id. 30:47 (collecting cases); LeSportsac, Inc. v. K Mart Corp., 754 F.2d 71, 79 (2d Cir. 1985) ("Likelihood of confusion is itself strong evidence that in the absence of an injunction [the plaintiff] might face irreparable harm."); Omega Importing Corp. v. Petri-Kine Camera Co., Inc., 451 F.2d 1190, 1195 (2d Cir. 1971) ("Where there is . . . high probability of confusion, injury irreparable in the sense that it may not be fully compensable in damages almost inevitably follows.").

308. See, e.g., Opticians Ass'n of Am. v. Independant Opticians of Am., 920 F.2d 187, 197-98 (3d Cir. 1990). But see Calvin Klein Cosmetics Corp. v. Lenox Lab., Inc., 815 F.2d 500, 505 (8th Cir. 1987) (observing that the public's interest in competition and low prices weighed against an injunction). See generally 5 MCCARTHY ON TRADEMARKS, supra note 306, 30:52.

309. See 5 MCCARTHY ON TRADEMARKS, supra note 306, 30:51 (noting that an injunction may be inappropriate where the resulting financial loss to the defendant "significantly outweighs" any damage to the plaintiff).

310. See id. 30:50.

311. See Planned Parenthood Fed'n of Am., Inc. v. Bucci, 42 U.S.P.Q.2d 1430, 1440 (S.D.N.Y. 1997) (holding that the defendant's use of the "Planned
Parenthood" mark could be prevented by preliminary injunction because such use was "not part of a communicative message"). Even the McCarthy treatise says that the rule against prior restraints will not normally apply to trademark cases, though it doesn't explain why. See 5 MCCARTHY ON TRADEMARKS, supra note 306, 31:142. We discuss one possible explanation--that the prior restraint rule doesn't apply to commercial speech--below. See infra notes 330-332 and accompanying text.

312. See Dr. Seuss Enters. v. Penguin Books USA, Inc., 109 F.3d 1394, 1403 n.11 (9th Cir. 1997) ("reject[ing] outright" the prior restraint argument, and citing cases rejecting the proposition that trademark laws violate the First Amendment); Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 206 (2d Cir. 1979) (asserting that a preliminary injunction is "content neutral" and "therefore does not arouse the fears that trigger the application of constitutional 'prior restraint' principles"). The Dr. Seuss opinion didn't even consider whether the preliminary injunction it granted on the copyright claim in the same case might be an unconstitutional prior restraint. See Dr. Seuss Enters., 109 F.3d at 1403 n.11.

313. See supra Part II.D.3.

314. See, e.g., Dallas Cowboys Cheerleaders, Inc. v. Scoreboard Posters, Inc., 600 F.2d 1184, 1188 (5th Cir. 1979) ("The first amendment is not a license to trammel on legally recognized rights in intellectual property."); Walt Disney Co. v. Powell, 698 F. Supp. 10, 12 n.2 (D.D.C. 1988) (ruling that the First Amendment "gives no right to steal another's statutorily protected intellectual property"); Fund for Community Progress v. United Way of Southeastern New England, 695 A.2d 517, 523 (R.I. 1997) (holding that "there is no speech interest involved in the use of another's private property for purposes of solicitation").

The Supreme Court seemed to take this approach in San Francisco Arts & Athletics, Inc. v. United States Olympic Committee, 483 U.S. 522, 532 (1987) ("[W]hen a word acquires value as the result of organization and the expenditure of labor, skill and money by an entity, that entity constitutionally may obtain a limited property right in the word.") (internal quotation marks omitted), though only after stressing that the law involved in that case--a quasi-trademark statute--applied predominantly to use of a quasi-trademark in commercial speech, and would not be "applied to infringe significantly on noncommercial speech rights." Id. at 535-36 & n.15. For criticism of the Court's conclusion in San Francisco Arts & Athletics, see JAMES BOYLE, SHAMANS, SOFTWARE, AND SPLEENS: LAW AND THE CONSTRUCTION OF THE INFORMATION SOCIETY 145-48 (1996); Gordon, supra note 163, at 1583-91. Boyle quite reasonably points out that the constitutional protection given to flag burning in Texas v. Johnson, 491 U.S. 397 (1989), seems at odds with the result in San Francisco Arts & Athletics. See BOYLE, supra, at 147-48. While the Supreme Court is right to suggest that trademark law is substantively constitutional, at least within boundaries, see San Francisco Arts & Athletics, 483 U.S. at 536, it has not addressed what those boundaries are, and therefore invocation of the Court's language is not terribly helpful in resolving the conflict.

315. See Board of Trustees v. Fox, 492 U.S. 469, 473-74 (1989). For discussions of the characteristic elements of commercial speech, see Bolger v. Youngs Drug Products Corp., 463 U.S. 60, 67 n.14 (1983) (declining to answer whether reference to a particular product is a necessary element of commercial speech); 5 MCCARTHY ON TRADEMARKS, supra note 306, 31:139, at 31-197 ("[A] firm's trademark is the most important element of commercial speech . . . ."). On the difficulty of defining commercial speech, see Alex Kozinski & Stuart Banner, Who's Afraid of Commercial Speech?, 76 VA. L. REV. 627, 638-48 (1990); David F. McGowan, Comment, A Critical Analysis of Commercial Speech, 78 CAL. L. REV. 357, 381-410 (1990). Note that however commercial speech is defined, it cannot be so broad as to encompass any speech which is made for an ultimate purpose that is commercial or for-profit. See Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 761 (1976). The content of a book, for example, is not "commercial speech" regardless of how much money the publisher will make from it.

316. See Virginia State Bd. of Pharmacy, 425 U.S. at 771-72 & n.24 (observing that, while commercial speech is protected, the differences between commercial speech and "other forms" suggest that "a different degree of protection is necessary"); Central Hudson Gas & Elec. Corp. v. Public Serv. Comm'n, 447 U.S. 557, 562-63 (1980) (same).

317. Virginia State Bd. of Pharmacy, 425 U.S. at 765.

318. Cf. Consumers Union of the United States, Inc. v. General Signal Corp., 724 F.2d 1044, 1049 (2d Cir. 1983) (holding in a copyright case that the defendant's advertisements conveyed "useful information which is protected by the First Amendment").

319. For a discussion of the use of trademarks in political and social speech, see Mark A. Lemley, The Modern Lanham Act and the Death of Common Sense, 108 YALE L.J. (forthcoming May 1999).

320. Think of the picture of a building and skyline at issue in Rock & Roll Hall of Fame & Museum, Inc. v. Gentile Productions, 134 F.3d 749 (6th Cir. 1998). A postcard, photograph, or poster that's displayed in a store is both a speech product (akin to a book or a newspaper), which is noncommercial speech, and an advertisement for itself, which is commercial speech.

321. 5 MCCARTHY ON TRADEMARKS, supra note 306, 31:146; see also L.L. Bean, Inc. v. Drake Publishers, Inc., 811 F.2d 26, 30-31 (1st Cir. 1987) ("[T]rademarks offer a particularly powerful means of conjuring up the image of their owners and thus become an important, perhaps at times indispensable, part of the public vocabulary." (quoting Denicola, supra note 179, at 195-96)) (internal quotation marks omitted). See generally Lemley, supra note 319.

322. See Rochelle Cooper Dreyfuss, Expressive Genericity: Trademarks as Language in the Pepsi Generation, 65 NOTRE DAME L. REV. 397, 424 (1990) (observing that "Betty Crocker has replaced Hestia in the public consciousness"); see also Keith Aoki, How the World Dreams Itself to be American: Reflections on the Relationship Between the Expanding Scope of Trademark Protection and Free Speech Norms, 17 LOYOLA L.A. ENT. L.J. 523, 528-31 (1997) (arguing that trademarks and insignia, instead of identifying the source or origin of goods, are increasingly being used as status symbols); Robert C. Denicola, Institutional Publicity Rights: An Analysis of the Merchandising of Famous Trade Symbols, 62 N.C. L. REV. 603 (1984) (discussing the increased use of logos and trademarks on consumer goods); Mira Wilkins, The Neglected Intangible Asset: The Influence of the Trademark on the Rise of the Modern Corporation, 34 BUS. HIST. 66, 87-88 (1992) (discussing the impact of trademark law on economic growth); Tara J. Goldsmith, Note, What's Wrong with this Picture? When the Lanham Act Clashes with Artistic Expression, 7 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. 821, 858-80 (1997) (arguing that works of art that contain trademarks or brand names should be protected as artistic expression under the First Amendment).

323. New Kids on the Block v. News Am. Publ'g, Inc., 971 F.2d 302, 307 (9th Cir. 1992); accord Yankee Publ'g, Inc. v. News Am. Publ'g, Inc., 809 F. Supp. 267, 275-76 (S.D.N.Y. 1992) (holding that the First Amendment is implicated by expressive, rather than commercial, uses of a trademark).

324. On this expansion, see Graeme B. Dinwoodie, Reconceptualizing the Inherent Distinctiveness of Product Design Trade Dress, 75 N.C. L. REV. 471 (1997). An obvious example is Romm Art Creations, Ltd. v. Simcha International, Inc., 786 F. Supp. 1126 (E.D.N.Y. 1992), where the court held that a distributor could be prevented under trademark law from selling paintings resembling the famous Women and Cafes series by Itzchak Tarkay. See id. at 1134-35.

325. A number of courts have taken this approach. See, e.g., Anheuser-Busch, Inc. v. Balducci Publications, 28 F.3d 769, 776 (8th Cir. 1994) (rejecting First Amendment parody defense where the court thought the parody was likely to confuse consumers); E. & J. Gallo Winery v. Gallo Cattle Co., 967 F.2d 1280, 1297 (9th Cir. 1992) (holding that misleading and confusing commercial speech can be restricted under the First Amendment); see also Denicola, supra note 179, at 165 ("Protection is normally limited to instances in which there is a threatened appropriation or injury to goodwill arising through consumer confusion."); Alex Kozinski, Trademarks Unplugged, 68 N.Y.U. L. REV. 960, 973 (1993) (suggesting that trademark law should be limited to the prevention of confusion in the marketplace). Some courts, however, have proven a little too willing to find that a parody is likely to confuse consumers. See, e.g., Elvis Presley Enters., Inc. v. Capece, 141 F.3d 188, 198-200, 204 (5th Cir. 1998) (holding that the use of the name "Velvet Elvis" by a bar was likely to confuse consumers as to sponsorship, and rejecting a parody defense); Anheuser-Busch, 28 F.3d at 776 (concluding that consumers would think "Michelob Oily" parody actually involved a new kind of beer).

It is worth noting that the argument that false or misleading speech is unprotected--the one most commonly used to defend trademark laws against First Amendment attack--offers no support for dilution statutes such as the Federal Trademark Dilution Act of 1995, 15 U.S.C. 1125(c) (Supp. 1996) (providing that the owner of a mark shall be entitled to an injunction against an infringing user "if such use begins after the mark has become famous and causes dilution of the distinctive quality of the mark"). The substantive constitutionality of dilution statutes is beyond the scope of this Article. For discussions of the issue, see Megan E. Gray, Defending Against a Dilution Claim: A Practitioner's Guide, 4 TEX. INTELL. PROP. L.J. 205, 225-27 (1996); Malla Pollack, Time to Dilute the Dilution Statute and What Not to Do When Opposing Legislation, 78 J. PAT. & TRADEMARK OFFICE SOC'Y 519, 540-41 (1996); Kenneth L. Port, The "Unnatural" Expansion of Trademark Rights: Is a Federal Dilution Statute Necessary?, 18 SETON HALL L EGIS. J. 433, 446 (1994); Marla J. Kaplan, Note, Antidilution Statutes and the First Amendment, 21 SW. U. L. REV. 1139 (1992).

326. 440 U.S. 1, 15 (1979) (holding that a state's interest in protecting the public from confusion is sufficient to withstand a First Amendment challenge).

327. In practice, some trademark parody cases involving social and political messages seem to turn on whether the parody was successful--that is, on whether consumers were actually confused, or whether they understood that the use of the plaintiff's trademark was a parody. See, e.g., Hormel Foods Corp. v. Jim Henson Prods., Inc., 73 F.3d 497, 505 (2d Cir. 1996) (holding that the use of character named "Spa'am" didn't infringe the Spam trademark because confusion was unlikely); L.L. Bean, Inc. v. Drake Publishers, Inc., 811 F.2d 26, 33 (1st Cir. 1987) (holding that a "noncommercial parody" challenged under antidilution statute was subject to First Amendment protection); cf. LightHawk, the Environmental Air Force v. Robertson, 812 F. Supp. 1095, 1102 (W.D. Wash. 1993) (holding, in a non-trademark case, that use of Smokey the Bear in a political satire was not likely to cause confusion and was subject to First Amendment protection). But see Anheuser-Busch, 28 F.3d at 776 (suggesting in dictum that the First Amendment might require courts to tolerate confusion if necessary to convey the political message of a parody). By and large, we think this approach is reasonable. But see Steven M. Perez, Comment, Confronting Biased Treatment of Trademark Parody Under the Lanham Act, 44 EMORY L.J. 1451 (1995) (arguing that likelihood-of-confusion analysis in trademark parody cases is too unpredictable to be consistent with the First Amendment).

328. Some courts have acknowledged this prior restraint problem. See Cliffs Notes, Inc. v. Bantam Doubleday Dell Publ'g Group, Inc., 886 F.2d 490, 497 (2d Cir. 1989) (holding that the risk of confusion "is outweighed by the public interest in free expression" and that the district court "unjustifiably impose[d] the drastic remedy of a prepublication injunction upon the cover of a literary parody"); Pocket Books, Inc. v. Dell Publ'g Co., Inc., 267 N.Y.S.2d 269, 273-74 (Sup. Ct. 1966) (reaching the same result in an unfair competition action).

329. See Nike, Inc. v. "Just Did It" Enters., 6 F.3d 1225, 1232 (7th Cir. 1993) ("Neither we nor the district court can conclude as a matter of law that [the] parody does or does not confuse the purchasing public. Too many disputed facts require a trial for resolution."). While not all cases will involve factual disputes about consumer confusion (counterfeiting cases may be a notable exception), many will.

In San Francisco Arts & Athletics, Inc. v. United States Olympic Committee, 483 U.S. 522 (1987), the Court upheld a preliminary injunction in a quasi-trademark case without serious consideration of the prior restraint problem. But see Waters v. Churchill, 511 U.S. 661, 678 (1994) (plurality opinion) (cautioning that "cases cannot be read as foreclosing an argument that they never dealt with"); Miller v. California Pac. Med. Ctr., 991 F.2d 536, 541 (9th Cir. 1993) (invoking the "venerable principle" that a court is not bound by a prior decision that failed to address an issue the later court finds persuasive).

330. See Friedman v. Rogers, 440 U.S. 1, 10 (1979) (noting that different circumstances "may also make inapplicable the prohibition against prior restraints" in commercial speech cases (quoting Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 772 n.24 (1976))); see also Kleiner v. First Nat'l Bank of Atlanta, 751 F.2d 1193, 1204 (11th Cir. 1985) (interpreting these decisions as meaning that "commercial speech seldom implicates the traditional concerns underlying the prior restraint doctrine"). In Central Hudson Gas & Electric Corp. v. Public Service Commission, 447 U.S. 557 (1980), the Court suggested that the Commission might impose a "system of previewing advertising campaigns to insure that they will not defeat conservation policy." Id. at 571 n.13. Whether this prescreening policy would have to be accompanied by the sorts of procedural safeguards that the rule against prior restraints otherwise requires was not addressed by the Court.

331. Administrative agencies impose injunctions against commercial speech in various circumstances without courts raising prior restraint concerns. The Federal Trade Commission and the Securities and Exchange Commission enjoin deceptive speech, see FTC v. Brown & Williamson Tobacco Corp., 778 F.2d 35, 44 (D.C. Cir. 1985) (affirming an injunction based on past deceptive statements); SEC v. Lowe, 725 F.2d 892, 901-02 (2d Cir. 1984) (holding that the SEC can revoke an individual's registration as an investment adviser and prohibit him from publishing an investment newsletter), rev'd on other grounds, 472 U.S. 181 (1985); Sears, Roebuck & Co. v. FTC, 676 F.2d 385, 391-92 (9th Cir. 1982) (holding that the FTC may require prior substantiation of product performance claims after finding a past claim to have been false), though not all of these restrictions occur before a final determination on the merits. State bar associations regulate lawyer advertising by requiring ads to be submitted in advance for approval, a classic form of prior restraint. See Bates v. State Bar of Ariz., 433 U.S. 350, 382 (1977) (recognizing that the regulation of attorney advertising was subject to the First Amendment); cf. Fargo Women's Health Org., Inc. v. Larson, 381 N.W.2d 176, 182 (N.D. 1986) (upholding preliminary injunction against deceptive advertising by medical clinic).

On the other hand, some cases have flatly rejected a distinction between commercial and noncommercial speech for purposes of the rule against prior restraints. See New York Magazine v. Metropolitan Transp. Auth., 136 F.3d 123, 131 (2d Cir. 1998):

[W]e see no reason why the requirement of procedural safeguards should be relaxed whether speech is commercial or not. . . . [T]he difficulty of the question alone convinces us that the requirement of procedural safeguards in a system of prior restraints should not be loosened even in the context of commercial speech.

See also In re Search of Kitty's East, 905 F.2d 1367, 1371 n.4 (10th Cir. 1990) (noting that the Supreme Court has not distinguished between political and commercial speech in judicial review of prior restraint cases); Sears, Roebuck & Co., 676 F.2d at 399 n.31 ("The doctrinal question whether prior restraint analysis is properly applicable to any commercial speech question remains open.") (citations omitted).

A number of trademark cases involving commercial speech have applied the rule against prior restraints. See, e.g., Consumers Union v. General Signal Corp., 724 F.2d 1044, 1052-53 (2d Cir. 1983) (striking down preliminary injunction); Better Bus. Bureau v. Medical Dirs., Inc., 681 F.2d 397, 404-05 (5th Cir. 1982) (modifying overly broad injunction); see also William W. Van Alstyne, Remembering Melville Nimmer: Some Cautionary Notes on Commercial Speech, 43 UCLA L. REV. 1635 (1996) (discussing the First Amendment, commercial speech and trademark law).

332. Alternatively, some courts have suggested that prior restraints against commercial speech are still troubling, but should be subject to less exacting scrutiny than restraints on core protected speech, just as substantive evaluation of the two types of speech differs. See Beneficial Corp. v. FTC, 542 F.2d 611, 619 (3d Cir. 1976) (concluding that a prior restraint must be "reasonably necessary" to achieve the remedial aim, and tailored to that aim).

333. Where speech involving trademarks is both commercial and noncommercial speech, at the very least a preliminary injunction would have to be tailored to restrict only the advertising or marketing effects of the speech, and not the underlying speech itself. Thus, we can imagine a preliminary injunction against a photograph like the one in Gentile that restricted the display or packaging of the photograph, but not one that prevented it from being sold at all. See Rock & Roll Hall of Fame & Museum v. Gentile, 134 F.3d 749 (6th Cir. 1998).

334. On the origins of the tort, see W. PAGE KEETON ET AL., PROSSER AND KEETON ON THE LAW OF TORTS 117 (5th ed. 1984).

335. See RESTATEMENT (SECOND) OF TORTS 652C (1977); see also Midler v. Ford Motor Co., 849 F.2d 460, 463 (9th Cir. 1988) ("[W]hen a distinctive voice of a professional singer is widely known and is deliberately imitated in order to sell a product, the sellers . . . have committed a tort in California."). For a discussion of the diverse rationales behind the right of publicity, see Michael Madow, Private Ownership of Public Image: Popular Culture and Publicity Rights, 81 CAL. L. REV. 125 (1993).

336. See Waits v. Frito-Lay, Inc., 978 F.2d 1093, 1100 (9th Cir. 1992); Midler, 849 F.2d at 463.

337. See Motschenbacher v. R.J. Reynolds Tobacco Co., 498 F.2d 821, 827 (9th Cir. 1974).

338. See White v. Samsung Elecs. Am., Inc., 971 F.2d 1395, 1399 (9th Cir. 1992).

339. See RESTATEMENT (THIRD) OF UNFAIR COMPETITION 46 cmt. b (1995).

340. See id. 47 & cmt. c; see also Hicks v. Casablanca Records, 464 F. Supp. 426, 433 (S.D.N.Y. 1978).

341. See RESTATEMENT (THIRD) OF UNFAIR COMPETITION 47 & cmt. b (1995). We speak here only of advertisements for non-speech products, such as cars or video recorders. If a speech product uses someone's name or likeness, advertisements for that speech product can generally likewise use that person's name or likeness. See Montana v. San Jose Mercury News, Inc., 40 Cal. Rptr. 2d 639, 643 (Ct. App. 1995) (holding that a newspaper could feature its own article on Joe Montana in advertisements without violating Montana's right of publicity); RESTATEMENT (THIRD) OF UNFAIR COMPETITION 47 cmt. a (1995).

342. See 2 J. THOMAS MCCARTHY, THE RIGHTS OF PUBLICITY AND PRIVACY 11.6[C] (1998) (listing and discussing cases).

343. White v. Samsung Elecs. Am., Inc., 989 F.2d 1512, 1519 (9th Cir. 1993) (Kozinski, J., dissenting from denial of rehearing en banc).

344. See White, 989 F.2d at 1519-20 (Kozinski, J., dissenting from denial of rehearing en banc) (criticizing the panel's conclusion as inconsistent with the First Amendment). For other examples of uses for "commercial advantage" that nonetheless convey a political or social message, see Cardtoons, L.C. v. Major League Baseball Players' Ass'n, 95 F.3d 959, 969 (10th Cir. 1996) (holding that parody baseball cards are protected speech); Matthews v. Wozencraft, 15 F.3d 432, 440 (5th Cir. 1994) (holding that a novel about a fictional character recognizable as the plaintiff was protected speech).

345. See supra note 315.

346. See Martin Luther King, Jr., Ctr. for Social Change, Inc. v. American Heritage Prods., Inc., 296 S.E.2d 697 (Ga. 1982).

347. See Ayres v. City of Chicago, 125 F.3d 1010, 1014 (7th Cir. 1997) ("[T]here is no question that the T-shirts are a medium of expression prima facie protected by the free-speech clause of the First Amendment, and they do not lose their protection by being sold rather than given away."); Gaudiya Vaishnava Soc'y v. San Francisco, 952 F.2d 1059, 1064 (9th Cir. 1990) (holding that merchandise which exhibits protected expression is protected speech); cf. Lakewood v. Plain Dealer Publ'g Co., 486 U.S. 750, 756 n.5 & 768 (1988) (holding that the act of selling protected works is itself protected speech).

348. See RESTATEMENT (THIRD) OF UNFAIR COMPETITION 47 cmt. b (1995) ("An unauthorized appropriation of another's name or likeness for use on posters, buttons, or other memorabilia is . . . ordinarily actionable as an infringement of the right of publicity."); Wendt v. Host Int'l, Inc. 125 F.3d 806, 811-12 (9th Cir. 1997) (holding that licensees of "Cheers" television show could violate the right of publicity by using robots that looked like characters from the show in "Cheers" theme bars); Factors Etc., Inc. v. Pro Arts, Inc., 579 F.2d 215, 221 (2d Cir. 1978); (holding that a poster depicting Elvis Presley infringed the Presley estate's right of publicity) Winterland Concessions Co. v. Sileo, 528 F. Supp. 1201, 1214 (N.D. Ill. 1981) (holding that T-shirts depicting band logos violated the right of publicity), aff'd on other grounds, 735 F.2d 257 (7th Cir. 1984). In our view, the sort of "decor speech" at issue in Wendt doesn't fit within the traditional definition of commercial speech, because it doesn't propose a commercial transaction.

349. To be sure, some appropriations of the plaintiff's name or likeness may mislead the public, as where Tom Waits or Bette Midler are wrongly thought to be endorsing commercial products. See Waits v. Frito-Lay, Inc., 978 F.2d 1093, 1104 (9th Cir. 1992); Midler v. Ford Motor Co., 849 F.2d 460, 462 (9th Cir. 1988). But if such confusion in fact results, trademark and unfair competition law provide a sufficient remedy. And courts have applied the right of publicity even where the statements at issue are entirely true, and indeed helpful to consumers. See, e.g., Nature's Way Prods., Inc. v. Nature-Pharma, Inc., 736 F. Supp. 245 (D. Utah 1990) (holding that the right of publicity precluded truthful advertisement that the plaintiff had developed the defendant's recipe, without considering the First Amendment implications).

350. Cf. Waits, 978 F.2d at 1100 (characterizing the right of publicity as a plaintiff's "personal property right . . . to control the use of his identity").

351. The general First Amendment issue was discussed in Zacchini v. Scripps-Howard Broadcasting Co., 433 U.S. 562 (1977), albeit only with respect to a particular strand of the right of publicity (the right to exclusive control over broadcasts of a performer's entire act). For detailed discussions of the First Amendment problems raised by the right of publicity that do not focus on prior restraints, see Barbara A. Burnett, The Property Right of Publicity and the First Amendment: Popular Culture and the Commercial Persona, 3 HOFSTRA PROP. L.J. 171 (1990); Ira J. Kaplan, They Can't Take That Away From Me: Protecting Free Trade in Public Images from Right of Publicity Claims, 18 LOYOLA L.A. ENT. L.J. 37 (1997); Roberta Rosenthal Kwall, The Right of Publicity vs. the First Amendment: A Property and Liability Rule Analysis, 70 IND. L.J. 47 (1994); Pamela Samuelson, Reviving Zacchini: Analyzing First Amendment Defenses in Right of Publicity and Copyright Cases, 57 TUL. L. REV. 836 (1983); Edward C. Wilde, The Scope of Liability Under California's Right of Publicity Statutes: Civil Code Sections 990 and 3344, 5 UCLA ENT. L. REV. 167 (1998); Edgar Sargent, Comment, Right of Publicity Tarnishment and the First Amendment, 73 WASH. L. REV. 223 (1998).

352. See, e.g., Cardtoons, L.C. v. Major League Baseball Players Ass'n, 95 F.3d 959, 972-73 (10th Cir. 1996) ("[E]levating the right of publicity above the right to free expression would likely prevent distribution of the parody trading cards. This would . . . also have a chilling effect upon future celebrity parodies."); Page v. Something Weird Video, 40 U.S.P.Q.2d 1196 (C.D. Cal. 1996) (finding that the defendant's use of the plaintiff's likeness was protected by the First Amendment).

353. See Factors Etc., Inc. v. Pro Arts, Inc., 579 F.2d 215 (2d Cir. 1978); Bi-Rite Enters. v. Bruce Miner Poster Co., 616 F. Supp. 71 (D. Mass. 1984), aff'd, 757 F.2d 440 (1st Cir. 1985); Ali v. Playgirl, Inc., 447 F. Supp. 723 (S.D.N.Y. 1978); Aguilar v. Ragusa, 572 N.Y.S.2d 164 (App. Div. 1991); Reilly v. Rapperswill Corp., 377 N.Y.S.2d 488 (App. Div. 1975); Rosemont Enters. v. Choppy Prods., 347 N.Y.S.2d 83, 85 (Sup. Ct. 1972).

354. See, e.g., Ali, 447 F. Supp. at 729 ("Such difficulty in determining monetary damages . . . satisfactorily establishes irreparable injury for . . . preliminary equitable relief.").

355. See supra Part II.D.8.

356. See New York Magazine v. Metropolitan Transp. Auth., 136 F.3d 123, 131-32 (2d Cir. 1998); In re Providence Journal Co., 820 F.2d 1342, 1353 (1st Cir. 1986); Jones v. Turner, 23 Media L. Rep. 1122 (S.D.N.Y. 1994); Taylor v. National Broad. Co., 22 Media L. Rep. 2433, 2436-39 (Cal. Super. 1994); Rosemont Enters. v. McGraw-Hill Book Co., 380 N.Y.S.2d 839, 843 (Sup. Ct. 1975). For arguments by commentators, see Darren F. Farrington, Note, Should the First Amendment Protect Against Right of Publicity Infringement Actions Where the Media is the Merchandiser? Say It Ain't So, Joe, 7 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. 779, 785-89 (1997).

357. Roberta Kwall describes the interaction between the right of publicity and the First Amendment as an area of "massive confusion." Kwall, supra note 351, at 48.

358. See Economic Espionage Act of 1996, Pub. L. No. 104-294, 1996 U.S.C.C.A.N. (110 Stat.) 3488 (codified at 18 U.S.C. 1831-1839 (Supp. II 1996)).

359. See UNIF. TRADE SECRETS ACT 1 (1985). The Uniform Act has been enacted in 41 jurisdictions as of this writing. See MERGES ET AL., supra note 50, at 32.

360. See UNIF. TRADE SECRETS ACT 1.

361. See generally MERGES ET AL., supra note 50, at 75.

362. UNIF. TRADE SECRETS ACT 2.

363. See, e.g., SI Handling Sys., Inc. v. Heisley, 753 F.2d 1244, 1263-64 (3d Cir. 1985); Merck & Co. v. Lyon, 941 F. Supp. 1443, 1455-62 (M.D.N.C. 1996); KLM Royal Dutch Airlines, N.V. v. deWit, 415 N.Y.S.2d 190, 191 (Sup. Ct. 1979).

364. See, e.g., Religious Tech. Ctr. v. Lerma, 908 F. Supp. 1362, 1368 (E.D. Va. 1995) (holding that information posted on the Internet becomes part of the public domain).

365. See CPG Prods. Corp. v. Mego Corp., 502 F. Supp. 42, 45 (S.D. Ohio 1980) (granting a temporary restraining order to prevent imminent disclosure).

366. See, e.g., Continental Group, Inc. v. Amoco Chems. Corp., 614 F.2d 351, 357-58 (3d Cir. 1980) (noting importance of detailed rather than conclusory analysis of the public interest factor); Uncle B's Bakery, Inc. v. O'Rourke, 920 F. Supp. 1405, 1422-38 (N.D. Iowa 1996) (giving a detailed analysis of all four factors); see also Restatement (Third) of Unfair Competition 44(2) (1995) (detailing additional factors to consider). Professor Milgrim, in his treatise on trade secret law, notes:

A trade secret preliminary injunction is typically issued only if there is a reasonable certainty that plaintiff will prevail in the ultimate disposition of the controversy, and that in the absence of the preliminary injunction plaintiff risks irreparable injury. . . . In those circuits where the public interest has to be evaluated, it must be closely analyzed and may not be simply resolved on a conclusory basis.

3 Roger M. Milgrim, Milgrim on Trade Secrets 14.01[1], at 14-2 to 14-17 (1996) (citations omitted).

367. In fact, the relationship need only be "confidential," Unif. Trade Secrets Act 1 (amended 1985), 14 U.L.A. 433, 437 (1990), but virtually all confidential relationships in trade secret cases are based either on the existence of an actual written contract, or on an agreement implied by the court from the circumstances. See Restatement (Third) of Unfair Competition 41 at 471-73 (1995); Pooley, supra note 235, 6.03[2], at 6-21 to 6-25.

368. See, e.g., Cohen v. Cowles Media Co., 501 U.S. 663 (1991) (holding that a journalist's agreement to keep a source secret was enforceable by suit for damages against the journalist). For an argument that enforcing a contract not to disclose a trade secret does not violate public policy, see Alan E. Garfield, Promises of Silence: Contract Law and Freedom of Speech, 83 Cornell L. Rev. 261, 300-03 (1998).

369. Likewise, as we discuss in Part II.D.8 supra, even though the publication of a libel can irremediably hurt someone's reputation, libels can't be preliminarily enjoined.

370. In CBS, Inc. v. Davis, 510 U.S. 1315 (Blackmun, Circuit Justice 1994), the plaintiff sought to enjoin the broadcast of a videotape of its meat packing operations on a number of theories. One of its arguments was that showing the videotape would disclose its trade secrets. See id. at 1316. Justice Blackmun stayed a preliminary injunction, noting that "[s]ubsequent civil or criminal proceedings, rather than prior restraints, ordinarily are the appropriate sanction for calculated defamation or other misdeeds in the First Amendment context." Id. at 1318. In another case, the Oregon Supreme Court struck down as a prior restraint a secrecy order imposed by the trial court to protect the trade secret once the trial began. See Oregon ex rel. Sports Management News, Inc. v. Nachtigal, 921 P.2d 1304, 1309-10 (Or. 1996); see also Religious Tech. Ctr. v. Lerma, 897 F. Supp. 260, 262-63 (E.D. Va. 1995) (refusing a preliminary injunction in a trade secret and copyright case on prior restraint grounds); Note, Trade Secrets in Discovery: From First Amendment Disclosure to Fifth Amendment Protection, 104 Harv. L. Rev. 1330, 1333-34 (1991) (discussing judicial presumption in favor of disclosing trade secrets based on a First Amendment claim of public access to pretrial discovery materials). A similar problem is raised in cases under the Economic Espionage Act (EEA), though it is complicated by the right to a public trial in criminal cases. See United States v. Hsu, 982 F. Supp. 1022, 1023-24 (E.D. Pa. 1997) (denying the government's motion for a protective order to protect trade secrets stolen in violation of the EEA), rev'd, 155 F.3d 189 (3d Cir. 1998); James H.A. Pooley et al., Understanding the Economic Espionage Act of 1996, 5 Tex. Intell. Prop. L.J. 177, 206-09 (1997) (addressing the risk of disclosing trade secrets when seeking to enforce violations of the EEA).

On the other hand, courts do sometimes issue injunctions in these cases. See Garth v. Staktek Corp., 876 S.W.2d 545, 550 (Tex. Ct. App. 1994) (recognizing that preliminary injunction against revelation of a trade secret was a prior restraint, but concluding that it was permissible because it was "necessary to provide meaningful legal protection to the owners of intellectual property"); Loveall v. American Honda Motor Co., 694 S.W.2d 937, 939-40 (Tenn. 1985) (noting the "serious" irreparable injury associated with disclosure of trade secrets obtained in discovery).

371. As with copyright, we see trade secret law as having a constitutionally mandated core--for instance, the requirement that the defendant actually agreed not to speak--coupled with some elements that are constitutionally irrelevant, such as the requirements that the trade secret have economic value and not be generally known. See Unif. Trade Secrets Act 1(4)(i), 14 U.L.A. 438 (1990). If the dispute is about whether the constitutionally mandated requirements are satisfied, then a preliminary injunction based on a mere likelihood that they are satisfied would be an improper prior restraint. But if it's clear that these requirements are met, then a preliminary injunction would be permissible even if it's not certain whether the other conditions are present.

372. See supra note 156 and accompanying text.

373. See Pepsico, Inc. v. Redmond, 54 F.3d 1262, 1269-72 (7th Cir. 1995) (enjoining employee from immediate employment at competitor corporation in order to prevent "inevitable disclosure" of trade secrets). But see Campbell Soup Co. v. Giles, 47 F.3d 467, 471-72 (1st Cir. 1995) (refusing to grant injunction because employer failed to show that former employee would inevitably disclose trade secrets); cf. Merck & Co. v. Lyon, 941 F. Supp. 1443, 1464-65 (M.D.N.C. 1996) (granting limited injunction prohibiting plaintiff's former employee from discussing certain subjects with new employer).

374. Pooley, supra note 235, 7.02[2][b][ii], at 7-17.

375. See 35 U.S.C. 271(a) (Supp. II 1996).

376. See 35 U.S.C. 101 (1994). The Supreme Court has recognized this scope as expansive, including "anything under the sun that is made by man." Diamond v. Chakrabarty, 447 U.S. 303, 309 (1980) (quoting legislative history).

377. 1 Donald S. Chisum, Patents 1.02[4], at 1-17 (1998); see also In re Schrader, 22 F.3d 290, 297-99 (Fed. Cir. 1994) (Newman, J., dissenting) (discussing and criticizing the history of the printed matter and "method of doing business" doctrines); Hotel Security Checking Co. v. Lorraine Co., 160 F. 467, 469-72 (2d Cir. 1908) (holding a claim which merely set down on paper a non-novel improvement to the art of cash-registering and account-checking unpatentable); Ex parte Gwinn, 112 U.S.P.Q. 439, 446 (Pat. Off. Bd. App. 1955) (stating that claims are patentable only where they "define either new features of structure or new relations of printed matter to structure, or both, . . . [and] give rise to some new and useful function or effect"). The Federal Circuit's recent decision in State Street Bank & Trust Co. v. Signature Financial Services, Inc., 149 F.3d 1368 (Fed. Cir. 1998), abolished the business methods exception, and therefore cast some doubt on the continued vitality of the printed matter exception as well.

378. For the technological arts language used in "mental steps" cases, see In re Musgrave, 431 F.2d 882, 890-93 (C.C.P.A. 1970); see also Donald S. Chisum, The Future of Software Protection: The Patentability of Algorithms, 47 U. Pitt. L. Rev. 959, 967-71 (1986) (discussing the doctrines of patentability developed by cases such as Musgrave). While the technological arts formulation has been criticized, and its continued applicability is unclear, certainly something more than the literal language of 35 U.S.C. 101 distinguishes the useful arts within the subject of patent law from the liberal arts covered instead by copyright. For a discussion of this thorny issue, see Dennis S. Karjala, A Coherent Theory for the Copyright Protection of Computer Software and Recent Judicial Interpretations, 66 U. Cinn. L. Rev. 53 (1997).

379. See, e.g., H.H. Robertson Co. v. United Steel Deck, Inc., 820 F.2d 384, 390-91 (Fed. Cir. 1987) (upholding an injunction against patent infringers after balancing the hardships); Smith Int'l, Inc. v. Hughes Tool Co., 718 F.2d 1573, 1577-78 (Fed. Cir. 1983) (analyzing the irreparable harm and public interest factors and granting an injunction against patent infringers); Merges et al., supra note 50, at 298-303. One study found that the majority of recent reported patent decisions granted preliminary injunctions. See Steven E. Shapiro, Preliminary Injunction Motions in Patent Litigation, 33 IDEA 323, 325 (1993).

380. See Act of Dec. 8, 1994, Pub. L. No. 103-465, 533(a), 108 Stat. 4988 (1994) (amending 35 U.S.C. 271).

381. See Pittsburgh Press Co. v. Pittsburgh Comm'n on Human Relations, 413 U.S. 376, 388 (1973) ("We have no doubt that a newspaper constitutionally could be forbidden to publish a want ad proposing a sale of narcotics or soliciting prostitutes."); Levi Strauss & Co. v. Shilon, 43 U.S.P.Q.2d 1616, 1618-19 (9th Cir. 1997) (rejecting a First Amendment challenge to the Lanham Act on the grounds that an offer to sell goods is not protected where the sale itself would be unlawful).

382. The only circumstance in which we can imagine such an injunction being sought is where the defendant is advertising on behalf of a third party, and is not itself making or selling the goods. Even there, the plaintiff would seem better off suing the direct infringer than an entity whose only connection to the infringement was publishing an advertisement. The offer for sale provision may have more use near the end of the patent term, because it allows patentees to sue for infringement based on an offer that occurred before the patent expired, even though the sale itself took place afterwards. But an injunction seems unlikely in these circumstances, because these suits by their nature will generally be filed after the patent has already expired, and when the patentee is entitled only to damages. There is a plausible argument that enjoining offers to sell in this situation violates the First Amendment, because the underlying sale is lawful. See Robert Ryan Morishita, Note, Patent Infringement After GATT: What Is an Offer to Sell?, 1997 Utah L. Rev. 905, 920-32.

383. See 35 U.S.C. 271(b) (1994) (providing that one who "actively induces infringement" is "liable as an infringer").

384. 594 F. Supp. 1420 (D. Del. 1985), aff'd in part, rev'd in part, and remanded, 793 F.2d 1261 (Fed. Cir. 1986).

385. Moleculon, 793 F.2d at 1272.

386. See 35 U.S.C. 112 (1994) (requiring patent applications to contain sufficient instructions on how to make and use the product). Presumably CBS would not have been guilty of inducing infringement merely for enclosing a copy of the Moleculon patent itself, even though the patent necessarily teaches the claimed method.

387. Patent law does allow preliminary injunctions against inducers and contributory infringers.

388. Ch. 4, 66 Stat. 3 (1952) (codified as amended at 35 U.S.C. 181 (1994)).

389. See 35 U.S.C. 181 (1994) (providing that if a government agency determines that a patent application might "be detrimental to the national security," the PTO "shall order that the invention be kept secret").

390. See 35 U.S.C. 186 (1994) (providing for criminal punishment for willful disobedience of a secrecy order).

391. We don't want to delve into the national security case law here. For a discussion in the context of the Invention Secrecy Act, see Lee Ann Gilbert, Patent Secrecy Orders: The Unconstitutionality of Interference in Civilian Cryptography Under Present Procedures, 22 Santa Clara L. Rev. 325 (1982); Sabing H. Lee, Protecting the Private Inventor Under the Peacetime Provisions of the Invention Secrecy Act, 12 Berkeley Tech. L.J. 345 (1997).

392. See 35 U.S.C. 181 (1994).

393. See 35 U.S.C. 186 (1994).

394. See Lee, supra note 391, at 387-89 (characterizing secrecy orders as prior restraints because "the primary purpose . . . is to suppress information"). Lee notes, however, that some prior restraints may be justified on national security grounds. See id. at 390-95.

395. See Bernstein v. United States Dep't of State, 945 F. Supp. 1279, 1287-88 (N.D. Cal. 1996). The case is not a patent case, but instead involves the constitutionality of U.S. restrictions on the export of strong encryption. For a more detailed analysis, see Thinh Nguyen, Note, Cryptography, Export Controls, and the First Amendment in Bernstein v. United States Department of State, 10 Harv. J.L. & Tech. 667 (1997).

396. For the contours of software patent law, see Mark A. Lemley & David W. O'Brien, Encouraging Software Reuse, 49 Stan. L. Rev. 255 (1997).

397. Two other district courts have come to the opposite conclusion. See Junger v. Daley, 8 F. Supp. 2d 708 (N.D. Ohio 1998); Karn v. United States Dep't of State, 925 F. Supp. 1 (D.D.C. 1996).

398. Source code written by programmers is converted into object code, a series of binary electrical signals represented by 1s and 0s that computers can read.

399. On this distinction, see Weinreb, supra note 102, at 1167-71; Karjala & Menell, supra note 102, at 186-92. There is no question that this distinction is at least somewhat artificial; certain programs (such as those written in LISP) might be thought of as combining source and object code. But on the whole, we think there is a practical difference for speech purposes between machine-readable instructions and human-readable source code.

400. On the existence of a First Amendment right to read, see Julie E. Cohen, A Right to Read Anonymously: A Closer Look at "Copyright Management" in Cyberspace, 28 Conn. L. Rev. 981 (1996).

401. Cf. Bernstein v. United States Dep't of State, 945 F. Supp. 1279, 1287 (N.D. Cal. 1996) (invoking prior restraint doctrine to invalidate government regulations prohibiting export).

402. This is not the only conceivable way in which patents on speech could run afoul of the First Amendment. Certainly, if the PTO began granting patents on information or writings simply because they were encoded on a computer-readable medium, the First Amendment would come into play. For suggestions of movement in this direction, see State Street Bank & Trust Co. v. Signature Fin. Group, Inc., 149 F.3d 1368, 1373 (Fed. Cir. 1998) (holding that "the transformation of [computer-readable] data, representing discrete dollar amounts, by a machine [such as a computer] through a series of mathematical calculations into a final share price," constitutes a patentable application of a mathematical logarithm); In re Beauregard, 53 F.3d 1583, 1584 (Fed. Cir. 1995) (noting that the PTO agrees that software on a floppy disk is patentable subject matter); U.S. Patent & Trademark Office, Examination Guidelines for Computer-Related Inventions, 61 Fed. Reg. 7478 (1996) (drawing the line of patentability between "functional" and "nonfunctional" descriptive material embedded in a computer-readable medium).

403. See, e.g., Cliffs Notes, Inc. v. Bantam Doubleday Dell Publ'g Group, Inc., 886 F.2d 490, 497 (2d Cir. 1989) (reversing a preliminary injunction in a trademark case on the grounds that "the drastic remedy of a prepublication injunction" was unjustified given "the public interest in free expression"); Mutual of Omaha Ins. Co. v. Novak, 836 F.2d 397, 405 (8th Cir. 1987) (Heaney, J., dissenting) (arguing that a permanent injunction in a trademark case was an unconstitutional prior restraint); Religious Tech. Ctr. v. Netcom On-Line Communication Servs., Inc., 907 F. Supp. 1361, 1383 (N.D. Cal. 1995) (refusing to issue a preliminary injunction in a copyright case, in part based on the concern that it would constitute a prior restraint); Religious Tech. Ctr. v. F.A.C.T.Net, 901 F. Supp. 1519, 1527 (D. Colo. 1995) (refusing to issue a preliminary injunction in a copyright case, in part because such an injunction would inhibit "dialogue on matters of public concern"); Religious Tech. Ctr. v. Lerma, 897 F. Supp. 260, 262-63 (E.D. Va. 1995) (refusing to issue a preliminary injunction in a copyright and trade secret case, in part based on the concern that it would constitute a prior restraint); supra notes 352 and 356 (discussing right of publicity cases which applied the prior restraint rule); supra note 370 (discussing trade secret cases which applied the prior restraint rule, including Justice Blackmun's in chambers opinion in CBS, Inc. v. Davis, 510 U.S. 1315 (1994) (Blackmun, Circuit Justice)).

404. See Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 578 n.10 (1994) (stating that "courts may also wish to bear in mind that the goals of the copyright law, 'to stimulate the creation and publication of edifying matter,' are not always best served by automatically granting injunctive relief when parodists are found to have gone beyond the bounds of fair use" (quoting Pierre N. Leval, Toward a Fair Use Standard, 103 Harv. L. Rev. 1105, 1134 (1990))); Abend v. MCA, Inc., 863 F.2d 1465, 1479 (9th Cir. 1988) (refusing to issue a permanent injunction against a movie based on plaintiff's short story), aff'd sub nom. Stewart v. Abend, 495 U.S. 207 (1990), cited with approval in Campbell, 510 U.S. at 578 n.10; Leval, supra at 1132 (stating that in non-piracy cases, especially those "raising reasonable contentions of fair use," "there may be a strong public interest in the publication of the secondary work [and] the copyright owner's interest may be adequately protected by an award of damages for whatever infringement is found"), cited with approval in Campbell, 510 U.S. at 578 n.10.

405. See supra Parts IV.B-C.

406. Waters v. Churchill, 511 U.S. 661, 678 (1994) (plurality opinion) (citing United States v. L.A. Tucker Truck Lines, Inc., 344 U.S. 33, 38 (1952)); see also Miller v. California Pac. Med. Ctr., 991 F.2d 536, 541 (9th Cir. 1993) ("It is a venerable principle that a court isn't bound by a prior decision that failed to consider an argument or issue the later court finds persuasive.").

407. See, e.g., Anheuser-Busch, Inc. v. Balducci Publications, 28 F.3d 769, 778 (8th Cir. 1994) (remanding for entry of permanent injunction, based on a final finding on the merits of a likelihood of confusion, and specifically noting the prior restraint dangers of broader injunctions); Cable/Home Communication Corp. v. Network Prods., Inc., 902 F.2d 829, 849-50 (11th Cir. 1990) (upholding permanent injunction); Mutual of Omaha, 836 F.2d at 403 (upholding permanent injunction, based on a final finding on the merits of a likelihood of confusion).

408. See In re Capital Cities/ABC, Inc., 918 F.2d 140, 144-45 (11th Cir. 1990) (holding only that the prior restraint doctrine doesn't bar a court from ordering a defendant to turn over its script for discovery purposes; stating in dictum that some sort of injunction against copyright infringement would be permissible, without indicating whether that would be a preliminary or a permanent injunction); Cable/Home Communication Corp., 902 F.2d at 849-51 (discussing copyright law, but in the context of upholding a permanent injunction entered under federal satellite telecommunications law, and specifically relying on the fact that the speech involved was commercial speech).

409. See Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 206 (2d Cir. 1979); Dr. Seuss Enters. v. Penguin Books USA, Inc., 109 F.3d 1394, 1403 n.11 (9th Cir. 1997).

410. The court said "[w]e reject outright Penguin and Dove's claim that the injunction in this case constitutes a prior restraint in violation of free speech guaranteed by the United States Constitution," Dr. Seuss Enters., 109 F.3d at 1403 n.11, but it said this only in the trademark section of its discussion, and cited only trademark cases in support of its conclusion. See id.

411. 376 U.S. 254, 283 (1964).

412. 418 U.S. 323, 347-50 (1974) (imposing negligence requirement for the formerly strict liability regime of private figure libel).

413. 475 U.S. 767, 776 (1986) (holding that private figure plaintiffs have the burden of proving falsehood, contrary to the common law rule).

414. See, e.g., Bates v. State Bar, 433 U.S. 350 (1977). But see id. at 391 (Powell, J., concurring in part and dissenting in part) (arguing that "[i]t has long been thought that price advertising of legal services inevitably will be misleading" and is thus constitutionally unprotected).

415. 489 U.S. 1 (1989).

416. See id. at 30-31 (Scalia, J., dissenting) (listing many such exemption statutes).

417. Vance v. Universal Amusement Co., 445 U.S. 308, 316-17 (1980) (citations omitted).