dissenting.
The American Civil Liberties Union appeals from the denial of a preliminary injunction in its suit against the Cook County State’s Attorney (that is, the “D.A.” of Cook County, Illinois) to invalidate the Illinois Eavesdropping Act as a violation of freedom of speech (more precisely, freedom to publish or otherwise disseminate other people’s speech). I would affirm the district court.
The Act criminalizes “knowingly and intentionally us[ing] an eavesdropping device for the purpose of hearing or recording all or any part of any conversation” without “the consent of all of the parties to such conversation.” 720 ILCS 5/14 — 2(a)(1). My colleagues have decided to reverse, and to order the entry of a preliminary injunction against enforcement of the Eavesdropping Act. But why a preliminary injunction? The opinion gives no indication of what argument or evidence presented on remand might allow the district court again to uphold the Act.
*609The invalidation of a statute on constitutional grounds should be a rare and solemn judicial act, done with reluctance under compulsion of clear binding precedent or clear constitutional language or — in the absence of those traditional sources of guidance — compelling evidence, or an overwhelming gut feeling, that the statute has intolerable consequences. The law invalidated today is not an outdated one left on the books by legislative inertia, like many of the laws invalidated by the Supreme Court in famous cases such as Grisivold v. Connecticut, 381 U.S. 479, 85 S.Ct. 1678, 14 L.Ed.2d 510 (1965). In its present form it dates only from 1994. It is stricter than provisions found in the laws governing electronic eavesdropping in most other states because it requires both parties to consent to a recording of their conversation. Maybe it’s too strict in forbidding nonconsensual recording even when done in defense of self or others, as when the participant in a conversation records it in order to create credible evidence of blackmail, threats, other forms of extortion, or other unlawful activity, as in Glik v. Cunniffe, 655 F.3d 78 (1st Cir.2011). But that feature of the statute is irrelevant. The ACLU insists on, and the majority opinion endorses, the right to record conversations to which police officers are parties even if no party consents to the recording, as long as the officers are performing public duties (as distinct from talking with one another on a private topic) in a public place and speaking loudly enough to be heard by a person who doesn’t have special equipment for amplifying sound — in other words, a person standing nearby.
Our ruling casts a shadow over electronic privacy statutes of other states as well, to the extent that they can be interpreted to require the consent of at least one party to a conversation to record it even though the conversation takes place in a public place, if the conversation could nevertheless reasonably be thought private by the parties. The statutes of several states are so open-ended that they could easily be found invalid under the approach taken in the majority opinion. See Alaska Stat. Ann. § 42.20.310; Ark.Code Ann. § 5-60-120; CaLPenal Code § 632(c); Mich. Comp. Laws Ann. § 750.539c; N.D. Cent. Code Ann. § 12.1-15-02. The California statute is illustrative. It states that “the term ‘confidential communication’ includes any communication carried on in circumstances as may reasonably indicate that any party to the communication desires it to be confined to the parties thereto, but excludes a communication made in a public gathering or in any legislative, judicial, executive or administrative proceeding open to the public, or in any other circumstance in which the parties to the communication may reasonably expect that the communication may be overheard or recorded.” The words are clear, the meaning is clear, but the application is unclear. Should a conversation in a public place, but intended to be private, be thought a “communication that any party desires to be confined to the parties”? It is both intended to be private and remote from a communication made in a “public gathering,” a term that from its placement connotes a public meeting of some sort. But what of the exclusion of private communications that the parties “may reasonably expect ... may be overheard or recorded”? That fogs the issue of which private communications are protected. To read the statute literally would exclude all private communications, for any private communication can be overheard and recorded, even if it is a conversation in a closed room.
A number of state privacy statutes tee off from the statement in Katz v. United States, 389 U.S. 347, 351, 88 S.Ct. 507, 19 L.Ed.2d 576 (1967), that “what a person *610knowingly exposes to the public, even in his own home or office, is not a subject of Fourth Amendment protection. But what he seeks to preserve as private, even in an area accessible to the public, may be constitutionally protected.” See, e.g., Fla. Stat. § 934.02(2); Ohio Rev.Code Ann. § 2933.51(B); Texas Penal Code § 16.02(b)(1), incorporating Tex.Code Crim. P. art. 18.20 § 1(1); cf. 18 U.S.C. § 2510(2). The police in Katz had recorded the defendant’s phone call, made in a public telephone booth, by secretly fastening a microphone to the booth, and the Court held that the recording violated the Fourth Amendment because the police had no warrant. Suppose the telephone booth had had no door, or that though it had a door the booth was not soundproof and someone standing five feet away could hear the conversation. Or suppose a police officer is talking in a low voice to a crime victim on a crowded sidewalk; there are people within earshot but the conversants reasonably assume that no one is listening, though they notice someone looking at his cell phone and the recorder in the cell phone might be turned on. We can’t predict the impact of today’s decision on the laws of most other states.
The ACLU particularly wants to record conversations to which a police officer is a party during demonstrations in public places, such as the march protesting the start of the second Iraq war that was before us in Vodak v. City of Chicago, 639 F.3d 738 (7th Cir.2011). That is its particular desire, but if its constitutional argument is correct, anyone has a constitutional right to record all such conversations, not just groups like the ACLU, and journalists, because neither the ACLU nor the press has greater First Amendment rights than other members of the public. Citizens United v. Federal Election Commission, — U.S. —, 130 S.Ct. 876, 905-06, 175 L.Ed.2d 753 (2010); Lovell v. City of Griffin, 303 U.S. 444, 452, 58 S.Ct. 666, 82 L.Ed. 949 (1938); see generally Eugene Volokh, “Freedom for the Press as an Industry, or for the Press as a Technology? From the Framing to Today,” 160 U. Pa. L.Rev. 459 (2012). Nor would the right be limited to political demonstrations; it would extend to all audible police conversations in public places, whether outdoors on sidewalks and in parks or indoors in the lobbies or other public spaces of courthouses and other government buildings.
Judges asked to affirm novel “interpretations” of the First Amendment should be mindful that the constitutional right of free speech, as construed nowadays, is nowhere to be found in the Constitution. The relevant provision of the First Amendment merely forbids Congress to abridge free speech, which as understood in the eighteenth century meant freedom only from censorship (that is, suppressing speech, rather than just punishing the speaker after the fact). A speaker could be prosecuted for seditious libel, for blasphemy, and for much other reprobated speech besides, but in a prosecution he would at least have the protection of trial by jury, which he would not have if hauled before a censorship board; and his speech or writing would not have been suppressed, which is what censorship boards do. Protection against censorship was the only protection that the amendment was understood to create. Patterson v. Colorado, 205 U.S. 454, 461-62, 27 S.Ct. 556, 51 L.Ed. 879 (1907) (Holmes, J.); Blue Canary Corp. v. City of Milwaukee, 251 F.3d 1121, 1123 (7th Cir.2001); Akhil Reed Amar, The Bill of Rights: Creation and Reconstruction 23-24 (1998); cf. 4 William Blackstone, Commentaries on the Laws of England 150-53 (1769).
The limitation of the amendment to Congress, and thus to federal restrictions on *611free speech (the First Amendment does not apply to state action), and to censorship is the original understanding. Judges have strayed so far from it that further departures should be undertaken with caution. Even today, with the right to free speech expanding in all directions, it remains a partial, a qualified, right. To make it complete would render unconstitutional defamation law, copyright law, trade secret law, and trademark law; tort liability for wiretapping, other electronic eavesdropping, and publicly depicting a person in a “false light”; laws criminalizing the publication of military secrets and the dissemination of child pornography; conspiracy law (thus including much of antitrust law); prohibitions of criminal solicitation, threats and fighting words, securities fraud, and false advertising of quack medical remedies; the regulation of marches, parades, and other demonstrations whatever their objective; limitations on free speech in prisons; laws limiting the televising of judicial proceedings; what little is left of permitted regulation of campaign expenditures; public school disciplining of inflammatory or disruptive student speech; the attorney-client, spousal, and physician-patient privileges in cases in which an attorney or spouse or physician would like to speak but is forbidden by the privilege to do so; laws making medical records confidential; and prohibitions against the public disclosure of jurors’ names in cases in which jurors might be harassed. All these legal restrictions of free speech are permitted (some because they may actually increase the amount of speech, a point I’ll come back to). The question in this case is whether a state, to protect both privacy and public safety, should be allowed in addition to forbid the recording of conversations between police officers and members of the public in a public place unless both parties to the conversation consent to being recorded for posterity.
A person who is talking with a police officer on duty may be a suspect whom the officer wants to question; he may be a bystander whom the police are shooing away from the scene of a crime or an accident; he may be an injured person seeking help; he may be a crime victim seeking police intervention; he may be asking for directions; he may be arguing with a police officer over a parking ticket; he may be reporting a traffic accident. In many of these encounters the person conversing with the police officer may be very averse to the conversation’s being broadcast on the evening news or blogged throughout the world. In some instances such publicity would violate the tort right of privacy, a conventional exception to freedom of speech as I have noted. Restatement (Second) of Torts §§ 652A, 652D (1977) (“unreasonable publicity given to [another person’s] private life”); Wolfe v. Schaefer, 619 F.3d 782, 784 (7th Cir.2010); Reuber v. Food Chemical News, Inc., 925 F.2d 703, 718-19 (4th Cir.1991) (en banc) (“publicizing] private facts in a highly offensive manner about an issue not of public concern”); Miller v. Motorola, Inc., 202 Ill.App.3d 976, 148 Ill.Dec. 303, 560 N.E.2d 900 (1990). This body of law is endangered by today’s ruling.
Privacy is a social value. And so, of course, is public safety. The constitutional right that the majority creates is likely to impair the ability of police both to extract information relevant to police duties and to communicate effectively with persons whom they speak with in the line of duty. An officer may freeze if he sees a journalist recording a conversation between the officer and a crime suspect, crime victim, or dissatisfied member of the public. He may be concerned when any stranger moves into earshot, or when he sees a recording device (even a cell phone, for modern cell phones are digital audio re*612corders) in the stranger’s hand. To distract police during tense encounters with citizens endangers public safety and undermines effective law enforcement.
The majority opinion disclaims any intention of “immunizing] behavior that obstructs or interferes with effective law enforcement.” I am not reassured. A fine line separates “mere” recording of a police-citizen encounter (whether friendly or hostile) from obstructing police operations by distracting the officers and upsetting the citizens they are speaking with. Today’s ruling may cause state and federal judicial dockets in Illinois to swell because it will unwittingly encourage police officers to shoo away bystanders, on the authority of cases like Colten v. Kentucky, 407 U.S. 104, 109-10, 92 S.Ct. 1953, 32 L.Ed.2d 584 (1972); cf. City of Houston v. Hill, 482 U.S. 451, 462 n. 11, 107 S.Ct. 2502, 96 L.Ed.2d 398 (1987); King v. Ambs, 519 F.3d 607, 613-15 (6th Cir.2008), when the officer wants to have a private conversation in a public place.
That the Eavesdropping Act, despite its name, does not punish the bystander who overhears a conversation without recording it does not have the significance that the majority opinion gives it. There is an important difference, well articulated in Justice Harlan’s dissent in United States v. White, 401 U.S. 745, 787-89, 91 S.Ct. 1122, 28 L.Ed.2d 453 (1971) (footnotes omitted), between human and mechanical eavesdropping:
The impact of the practice of third-party bugging, must, I think, be considered such as to undermine that confidence and sense of security in dealing with one another that is characteristic of individual relationships between citizens in a free society. It goes beyond the impact on privacy occasioned by the ordinary type of “informer” investigation .... The argument of the plurality opinion, to the effect that it is irrelevant whether secrets are revealed by the mere tattletale or the transistor, ignores the differences occasioned by third-party monitoring and recording which insures full and accurate disclosure of all that is said, free of the possibility of error and oversight that inheres in human reporting.
Authority is hardly required to support the proposition that words would be measured a good deal more carefully and communication inhibited if one suspected his conversations were being transmitted and transcribed. Were third-party bugging a prevalent practice, it might well smother that spontaneity — • reflected in frivolous, impetuous, sacrilegious, and defiant discourse — that liberates daily life. Much off-hand exchange is easily forgotten and one may count on the obscurity of his remarks, protected by the very fact of a limited audience, and the likelihood that the listener will either overlook or forget what is said, as well as the listener’s inability to reformulate a conversation without having to contend with a documented record. All these values are sacrificed by a rule of law that permits official monitoring of private discourse limited only by the need to locate a willing assistant.
The distinction that Justice Harlan drew between an overheard private conversation recalled from memory and one that is recorded is something everyone feels — and feels more acutely in the electronic age than 41 years ago. Walter Kirn, “Little Brother Is Watching,” New York Times Magazine (Oct. 17, 2010); William Saletan, “Bugged Naked: Webcams, Sex, and the Death of Privacy,” Slate (Oct. 1, 2010); William Safire, “To Stop the Eavesdrop,” New York Times (Dec. 20, 1999). Americans face new challenges to privacy because of the amount of personal informa*613tion stored and publicly accessible online and the ubiquity of recording devices. Lizette Alvarez, “Spring Break Gets Tamer as World Watches Online,” New York Times (March 16, 2012); Jeffrey Rosen, “The Web Means the End of Forgetting,” New York Times (July 25, 2010); Jonathan Zittrain, “Privacy 2.0,” 2008 U. Chi Legal Forum 65, 81-91. Lacking relevant expertise, lacking evidence, forced back on intuition, judges should hesitate to invalidate legislative attempts to solve these problems.
Police may have no right to privacy in carrying out official duties in public. But the civilians they interact with do. The majority opinion “acknowledge[s] the difference in accuracy and immediacy that an audio recording provides as compared to notes or even silent videos or transcripts” but says that “in terms of the privacy interests at stake, the difference is not sufficient to justify criminalizing this particular method of preserving and publishing the public communications of these public officials” (emphasis in original). The assertion lacks a supporting argument, and by describing the recording as a “method of preserving and publishing the public communications of these public officials” neglects the fact that the recording will publish and preserve what the civilians with whom the police are conversing say, not just what the police say. The further statement that these “are not conversations that carry privacy expectations even though uttered in public places” implies that anything said outdoors is ipso facto public. Yet people often say things in public that they don’t expect others around them to be listening to, let alone recording for later broadcasting, and we are given no reason to think that this is never the case when someone complains to a police officer, or otherwise speaks with one, “in public” in the sense of being in a place in which there are other people about.
Suppose a police detective meets an informant in a park and they sit down on a park bench to talk. A crime reporter sidles up, sits down next to them, takes out his iPhone, and turns on the recorder. The detective and the informant move to the next park bench to continue their conversation in private. The reporter follows them. Is this what the Constitution privileges?
It is small consolation to be told by the majority that “the ACLU plans to record openly, thus giving the police and others notice that they are being recorded” (emphasis in original). All the ACLU means is that it won’t try to hide its recorder from the conversants whom it wants to record, though since the typical recorder nowadays is a cell phone it will be hidden in plain view. A person who doesn’t want his conversations to be recorded will have to keep a sharp eye out for anyone nearby holding a cell phone, which in many urban settings is almost everyone. The ubiquity of recording devices will increase security concerns by distracting the police.
There is more on the state’s side of this case than privacy of communications and the effectiveness of law enforcement — and the more is the same First Amendment interest that the ACLU says it wants to promote. The majority opinion concedes that “conversational privacy” “serves First Amendment interests,” but thinks there can be no conversational privacy when the conversation takes place in a public place; it says that “this case has nothing to do with private conversations.” But private talk in public places is common, indeed ubiquitous, because most people spend a lot of their time in public places; because they rely on their anonymity and on the limited memory of others to minimize the risk of publication; because public places are (paradoxically) often more private than *614private places (imagine if detectives could meet with their informants only in police stations); and because eavesdropping on strangers is actually rather uncommon because it is so difficult in most cases to understand a conversation between strangers. “Anyone who’s overheard conversations on the street or in a restaurant knows that conversations between strangers are often unintelligible. There is the public language we employ when talking to strangers and the elliptical private language that we use when talking to people whom we know. Strangers need an interpreter .... ” United States v. Curescu, 674 F.3d 735, 740 (7th Cir.2012).
I disagree with the majority that “anyone who wishes to speak to police officers in confidence can do so,” and “police discussions about matters of national and local security do not take place in public where bystanders are within earshot.” Forget national security; the people who most need police assistance and who most want their conversations kept private are often the people least able to delay their conversation until they reach a private place. If a person has been shot or raped or mugged or badly injured in a car accident or has witnessed any of these things happening to someone else, and seeks out a police officer for aid, what sense would it make to tell him he’s welcome to trot off to the nearest police station for a cozy private conversation, but that otherwise the First Amendment gives passersby the right to memorialize and publish (on Facebook, on Twitter, on YouTube, on a blog) his agonized plea for help? And as in our informant example, many of the persons whom police want to talk to do not want to be seen visiting police stations.
Accuracy is a social value, and a recording of a conversation provides a more accurate record of the conversation than the recollection of the conversants: more accurate, and also more truthful, since a party to a conversation, including a police officer, may lie about what he heard or said. But on the other side of the balance are the inhibiting effect of nonconsensual recording of conversations on the number and candor of conversations (and hence on values that the First Amendment protects); the baleful effect on privacy; the negative effect on law enforcement; and the litigation likely to be engendered by police officers’ shooing away intruders on their private conversations with citizens. These are significant social costs, and the majority opinion offers no basis in fact or history, in theory or practice, in constitutional text or judicial precedent, for weighting them less heavily than the social value of recorded eavesdropping.