Concept
Falsity, Opinion, and Parody
Freedom of the press
Falsity, Opinion, and Parody
Actual malice presupposes falsity. A publisher cannot know that a statement is false, or recklessly disregard whether it is false, unless the statement is in fact untrue, so the standard that New York Times Co. v. Sullivan imposed on suits by public officials already assumes that the defendant published something false.1 That standard also demands proof of the publisher’s state of mind by clear and convincing evidence.2 The earlier sections followed the publisher’s state of mind and the plaintiff’s status. This section turns to the statement itself and asks two questions. Who must prove that it is false? And what kind of statement can be false at all?
The common law gave the first question to the defendant. Plaintiffs did not have to prove fault, falsity, or actual damages; truth was a defense that the defendant had to plead and prove.3 Sullivan explained what is wrong with that allocation. A rule that makes critics guarantee the truth of their factual assertions leads to self-censorship, because a critic may stay silent even when the criticism is believed true and is in fact true, out of doubt that it can be proved in court or fear of the expense of trying.4 The first subsection follows that reasoning into suits by private plaintiffs.
The second question limits what the tort can reach. Defamation liability requires a false statement of fact, and a statement that cannot be shown false cannot support it.5 Free debate produces excess as well as error; Madison observed that “[s]ome degree of abuse is inseparable from the proper use of every thing; and in no instance is this more true than in that of the press.”6 The second subsection shows how courts separate assertions of fact from opinion, hyperbole, and parody.
Burden of Proving Falsity
Philadelphia Newspapers, Inc. v. Hepps decided, by a vote of 5–4, that a plaintiff who sues a newspaper for libel over a matter of public concern must prove that the statements are false.7 The case grew out of a series of articles linking the owner of a chain of stores to organized crime. The state court had put the burden of proving truth on the newspaper because the plaintiff was neither a public official nor a public figure.8 The Supreme Court rejected that allocation: “[t]o ensure that true speech on matters of public concern is not deterred, we hold that the common-law presumption that defamatory speech is false cannot stand when a plaintiff seeks damages against a media defendant for speech of public concern.”9
The rule changes less for public officials and public figures than for private plaintiffs. Officials and figures already had to prove falsity, because knowledge of falsity can be established only if the statement was in fact false, and reckless disregard for the truth also implies publication of a false statement.10 In an actual malice case, truth is not a defense; lack of truth is part of the plaintiff’s case.11 Hepps extended the same burden to the private-figure plaintiff on a matter of public concern. Against a media defendant, that plaintiff must prove falsity, and need prove the required fault only by a preponderance of the evidence.12
A burden of proof decides only the cases that the evidence leaves in doubt, and it decides them against the party who carries it. When the jury cannot tell whether a story was true, the plaintiff who bears the burden of proving falsity loses; under the common-law allocation, the newspaper would have lost on the identical record. That difference explains why the allocation matters to speech. A publisher that believes its story accurate but doubts that it can persuade a jury has reason to drop the story, and true speech on public matters is deterred. Sullivan described that mechanism, and Hepps invoked its purpose.13 Gertz v. Robert Welch, Inc. supplies the companion point: punishing error risks inducing a cautious and restrictive exercise of the freedoms of speech and press, and a rule of strict liability that compels a publisher to guarantee the accuracy of its factual assertions may lead to intolerable self-censorship.14 The price of the Hepps allocation is that a plaintiff injured by a falsehood he cannot prove false will lose.
Hepps speaks of a newspaper publishing speech of public concern, and the Court observed that when the matter is of exclusively private concern and the plaintiff is a private figure, the constitutional requirements do not necessarily demand such special protection for the media.15 Where the Constitution does not require the shift, state law governs. The constitutional elements of fault, falsity, and actual damages are minimum protections, and the states remain free to restrict tort liability further.16 Some states place the burden of proving falsity on the plaintiff even for statements that do not involve public concern, although Hepps does not require it.17 The media label counts for less than it appears to. Five Justices in Dun & Bradstreet v. Greenmoss Builders, in separate opinions, rejected any distinction between media and nonmedia defendants for purposes of defamation liability, and most cases do not distinguish between the two in applying these rules.18
Falsity carries a lighter burden than actual malice. Actual malice must be shown by clear and convincing evidence, and on summary judgment a court asks whether a jury applying that standard could reasonably find for either side.19 The burden on falsity itself is merely a preponderance of the evidence.20 A plaintiff whose proof of falsity is only slightly better than even therefore carries that element, though the plaintiff must still satisfy the higher standard on actual malice.
What the plaintiff must show false depends on what the statement communicates. A communication is defamatory only if the recipient reasonably understands it in a defamatory sense, and the context of the statement and the extrinsic facts known to the recipient count in deciding how it was understood.21 The trier of fact decides what meaning the statement conveyed and then whether that meaning is true.22 The common-law defense of truth shows how far that inquiry reaches, and it governs equally when the plaintiff carries the burden. Small discrepancies between the statement and the facts do not defeat truth if the substance of the defamatory meaning, its “sting,” is true, although a plea of truth must be as broad as the libel charged.23 A charge of a specific act is true if the plaintiff committed that act or one substantially the same, while a charge of a pattern of misconduct is not established by a single instance.24 Cases such as Air Wisconsin Airlines Corp. v. Hoeper address when an inaccuracy is serious enough to matter.
Altered quotations follow the same logic. The Court has held that even the intentional fabrication of quotations is not enough, by itself, to prove actual malice if the statements were substantially accurate in reflecting what was said.25 Exact wording matters less than whether the substance attributed to the plaintiff is false.
A stem that calls a statement false and defamatory invites the traditional reflex that the defendant must prove it true, so the analysis should begin with the setting. If a newspaper has published speech on a public matter, the plaintiff must prove falsity, whether the plaintiff is a senator or a shopkeeper, and a plaintiff who proves fault but leaves falsity in equipoise still loses. If the stem involves a private speaker on a private matter, the sound answer is that the Constitution does not decide the burden and the jurisdiction’s own rule controls. Falsity and fault are separate elements, and a plaintiff can prove one and fail on the other.
Opinion and Parody
Milkovich v. Lorain Journal Co. refused to treat the word “opinion” as a password. The judgment under review rested in part on the ground that the article was an “opinion” protected from state defamation law by the First Amendment, and the Supreme Court held that the First Amendment does not prohibit applying Ohio’s libel laws to the alleged defamations in the article.27 The Court said that the language in Gertz was not “intended to create a wholesale defamation exemption for anything that might be labeled ‘opinion.’”28 It tested the content of the statement instead: “a statement on matters of public concern must be provable as false before there can be liability under state defamation law,” and “a statement of opinion relating to matters of public concern which does not contain a provably false factual connotation will receive full constitutional protection.”29
The inquiry therefore turns from the label to the content. The focus is not on whether a statement is opinion but on whether it contains, directly or by clear implication, factual statements, and false statements of fact are a prerequisite for defamation liability.30 The reason lies in how recipients take a statement. A communication is defamatory if the recipient reasonably understands it in a defamatory sense, and the context counts in deciding how it was understood.31 A statement may assert the fact of bribery outright or reasonably imply it, and either can support liability.32 A listener who hears that someone took a bribe, or that in the speaker’s opinion someone took one, comes away believing the same thing about what happened. If the label controlled, every defamer would simply add it.
An opinion is safest when the audience can see the facts behind it. If the speaker discloses true facts and then draws a conclusion, readers can weigh the facts and reject the conclusion, and the conclusion asserts nothing further that a record could prove false. The risk grows when the opinion depends on unstated facts. If those facts are false or incomplete, or the speaker’s assessment of them is erroneous, the statement may imply a false assertion of fact. A statement in the form of an opinion is also defamatory if it expressly relies on another defamatory statement that turns out to be untrue, even if the speaker reasonably thought it true.33
The accusation and the insult may sound equally harsh. What separates them is whether a reader can test the statement against anything: audit records can answer the pension charge, and nothing can prove that a vote shows someone to be a fool. A plaintiff who clears that hurdle has not won. The plaintiff must still prove that the implied fact is false and must meet the fault standard for the plaintiff’s status, which for a public official or public figure is actual malice.34
Hyperbole and parody receive similar treatment. Where the actual malice standard applies, both genres are absolutely protected unless the publication can reasonably be said to imply a false statement of facts.35 That leaves the key question of when statements are rhetorical hyperbole that the First Amendment protects and when they are factual statements that can support a defamation action.36 Hustler Magazine v. Falwell tested the protection against a far more offensive publication. Parodying a liquor advertisement, Hustler carried an invented interview in which Rev. Jerry Falwell described his “first time” as a drunken episode with his mother in an outhouse, and the page said that it was an ad parody not to be taken seriously.37 Falwell’s libel claim failed because the jury found that the parody did not describe actual facts about him, but his claim for emotional distress succeeded in the courts below.38
The Court of Appeals had treated the absence of facts as irrelevant to the distress claim, asking only whether publication was sufficiently outrageous.39 Falwell asked the Supreme Court to hold that a state’s interest in protecting public figures from emotional distress justifies denying First Amendment protection to speech that is patently offensive and intended to inflict emotional injury, even when the speech could not reasonably be read as stating actual facts about the public figure.40 The Court declined.41 It held that recovery for intentional infliction of emotional distress had to meet the New York Times standards.42
The speaker’s intent to inflict distress did not change that result. The law does not give much solicitude to an intent to inflict emotional distress, and most jurisdictions make it civilly culpable where the conduct is sufficiently “outrageous.” But in debate about public affairs many things done with motives less than admirable are protected. Garrison v. Louisiana held that expression was protected even when the speaker was motivated by hatred or ill will, and the First Amendment prohibits treating bad motive as controlling in public debate about public figures.43 Outrageousness and intent to injure measure offensiveness and motive. The New York Times standards measure something else, a false statement of fact. A labeled parody that no reasonable reader would take as fact contains no such statement, so the claim fails at the first requirement.44
Two limits keep parody from becoming a label-based immunity. A parody that a reasonable reader would take as a real report implies a false statement of fact, and the protection does not apply to it.45 The Falwell decision is also limited to public figures. Whether its rationale reaches speech by a private figure about public affairs remains a question, since the reasoning would still apply, but one can ask why the First Amendment should protect outrageous speech intended to do harm.46 Subject matter counts as well. In Snyder v. Phelps, protesters at a military funeral displayed hateful and outrageously offensive signs attacking the dead soldier and his family, and because the speech was a matter of public concern, the family could not recover for intentional infliction of emotional distress.47
A stem that quotes a speaker saying “in my opinion” invites the conclusion that the statement is protected, and a stem that describes a savagely offensive parody invites the conclusion that an emotional-distress claim succeeds because the content is outrageous. Both conclusions start from the wrong question. Milkovich shows that the opinion label protects nothing when the sentence implies a provable fact, and Hustler shows that outrageousness cannot carry a public figure’s tort claim when no false statement of fact is present. In each case the question is what a reasonable reader would take the speaker to be asserting, and the answer decides whether the analysis goes on to falsity and fault or ends at once.