with whom Justice Marshall joins, dissenting.
Since this Court first hinted that the First Amendment provides some manner of protection for statements of opinion,1 notwithstanding any common-law protection, courts and commentators have struggled with the contours of this protection and its relationship to other doctrines within our First Amendment jurisprudence. Today, for the first time, the Court addresses this question directly and, to my mind, does so cogently and almost entirely correctly. I agree with the Court that under our line of cases culminating in Philadelphia Newspapers, Inc. v. Hepps, 475 U. S. 767, 777 (1986), only defamatory statements that are capable of being proved false are subject to liability under state libel law. See ante, at 16.2 I also agree with the Court that the “state*24ment” that the plaintiff must prove false under Hepps is not invariably the literal phrase published but rather what a reasonable reader would have understood the author to have said. See ante, at 16-17 (discussing Greenbelt Cooperative Publishing Assn., Inc. v. Bresler, 398 U. S. 6 (1970); Letter Carriers v. Austin, 418 U. S. 264 (1974); Hustler Magazine, Inc. v. Falwell, 485 U. S. 46 (1988)).
In other words, while the Court today dispels any mis-impression that there is a so-called opinion privilege wholly in addition to the protections we have already found to be guaranteed by the First Amendment, it determines that a protection for statements of pure opinion is dictated by existing First Amendment doctrine. As the Court explains, “full constitutional protection” extends to any statement relating to matters of public concern “that cannot ‘reasonably [be] interpreted as stating actual facts’ about an individual.” Ante, at 20. Among the circumstances to be scrutinized by a court in ascertaining whether a statement purports to state or imply “actual facts about an individual,” as shown by the Court’s analysis of the statements at issue here, see ante, at 22, and n. 9, are the same indicia that lower courts have been relying on for the past decade or so to distinguish between statements of fact and statements of opinion: the type of language used, the meaning of the statement in context, whether the statement is verifiable, and the broader social circumstances in which the statement was made. See, e. g., Potomac Valve & Fitting Inc. v. Crawford Fitting Co., 829 F. 2d 1280 (CA4 1987); Janklow v. Newsweek, Inc., 788 F. 2d 1300 (CA8 1986); Ollman v. Evans, 242 U. S. App. D. C. 301, 750 F. 2d 970 (1984), cert. denied, 471 U. S. 1127 (1985).
*25With all of the above, I am essentially in agreement. I part company with the Court at the point where it applies these general rules to the statements at issue in this case because I find that the challenged statements cannot reasonably be interpreted as either stating or implying defamatory facts about petitioner. Under the rule articulated in the majority opinion, therefore, the statements are due “full constitutional protection.” I respectfully dissent.
As the majority recognizes, the kind of language used and the context in which it is used may signal readers that an author is not purporting to state or imply actual, known facts. In such cases, this Court has rejected claims to the contrary and found that liability may not attach “as a matter of constitutional law.” Ante, at 17. See, e. g., Bresler, supra (metaphor); Letter Carriers, supra (hyperbole); Falwell, supra (parody). In Bresler, for example, we found that Bresler could not recover for being accused of “blackmail” because the readers of the article would have understood the author to mean only that Bresler was manipulative and extremely unreasonable. See ante, at 16-17. In Letter Carriers, we found that plaintiffs could not recover for being accused of being “traitor[sj” because the newsletter’s readers would have understood that the author meant that plaintiffs’ accurately reported actions were reprehensible and destructive to the social fabric, not that plaintiffs committed treason. See ante, at 17.
Statements of belief or opinion are like hyperbole, as the majority agrees, in that they are not understood as actual assertions of fact about an individual, but they may be actionable if they imply the existence of false and defamatory facts. See ante, at 18-19. The majority provides some general guidance for identifying when statements of opinion imply assertions of fact. But it is a matter worthy of further attention *26in order “to confine the perimeters of [an] unprotected category within acceptably narrow limits in an effort to ensure that protected expression will not be inhibited.” Bose Corp. v. Consumers Union of United States, Inc., 466 U. S. 485, 505 (1984). Although statements of opinion may imply an assertion of a false and defamatory fact, they do not invariably do so. Distinguishing which statements do imply an assertion of a false and defamatory fact requires the same solicitous and thorough evaluation that this Court has engaged in when determining whether particular exaggerated or satirical statements could reasonably be understood to have asserted such facts. See Bresler, supra; Letter Carriers, supra; Falwell, supra. As Justice Holmes observed long ago: “A word is not a crystal, transparent and unchanged, it is the skin of a living thought and may vary greatly in color and content according to the circumstances and the time in which it is used.” Towne v. Eisner, 245 U. S. 418, 425 (1918).
For instance, the statement that “Jones is a liar,” or the example given by the majority, “In my opinion John Jones is a liar” — standing alone — can reasonably be interpreted as implying that there are facts known to the speaker to cause him to form such an opinion. See ante, at 18-19. But a different result must obtain if the speaker’s comments had instead been as follows: “Jones’ brother once lied to me; Jones just told me he was 25; I’ve never met Jones before and I don’t actually know how old he is or anything else about him, but he looks 16; I think Jones lied about his age just now.” In the latter case, there are at least six statements, two of which may arguably be actionable. The first such statement is factual and defamatory and may support a defamation action by Jones’ brother. The second statement, however, that “I think Jones lied about his age just now,” can be reasonably interpreted in context only as a statement that the speaker infers, from the facts stated, that Jones told a particular lie. It is clear to the listener that the speaker does *27not actually know whether Jones lied and does not have any other reasons for thinking he did.3 Thus, the only fact implied by the second statement is that the speaker drew this inference. If the inference is sincere or nondefamatory, the speaker is not liable for damages.4
*28H-1 h-t
The majority does not rest its decision today on any finding that the statements at issue explicitly state a false and defamatory fact. Nor could it. Diadiun’s assumption that Milkovich must have lied at the court hearing is patently conjecture.5 The majority finds Diadiun’s statements actionable, however, because it concludes that these statements imply a factual assertion that Milkovich perjured himself at the judicial proceeding. I disagree. Diadiun not only reveals the facts upon which he is relying but he makes it clear at which point he runs out of facts and is simply guessing. Read in context, the statements cannot reasonably be interpreted as implying such an assertion as fact. See ante, at 5-7, n. 2 (reproducing the column).
Diadiun begins the column by noting that, on the day before, a Court of Common Pleas had overturned the decision by the Ohio High School Athletic Association (OHSAA) to suspend the Maple Heights wrestling team from that year’s state tournament. He adds that the reversal was based on due process grounds. Diadiun emphasizes to the audience that he was present at the wrestling meet where the brawl that led to the team’s suspension took place and that he was present at the hearing before the OHSAA. He attributes the brawl to Maple Heights coach Milkovich’s wild gestures, ranting and egging the crowd on against the competing team from Mentor. He then describes Milkovich’s testimony before the OHSAA, characterizing it as deliberate misrepresen*29tation “attempting not only to convince the board of [his] own innocence, but, incredibly, shift the blame of the affair to Mentor.” Ante, at 6, n. 2. Diadiun then quotes statements allegedly made by Milkovich to the commissioners to the effect that his wrestlers had not been involved in the fight and his gestures had been mere shrugs.
At that point in the article, the author openly begins to surmise. Diadiun says that it “seemed” that Milkovich’s and another official’s story contained enough contradictions and obvious untruths that the OHSAA board was able to see through it and that “[pjrobably” the OHSAA’s suspension of the Maple Heights team reflected displeasure as much at the testimony as at the melee. Ante, at 7, n. 2 (emphasis added). Then Diadiun guesses that by the time of the court hearing, the two officials “apparently had their version of the incident polished and reconstructed, and the judge apparently believed them.” Ibid, (emphasis added). For the first time, the column quotes a third party’s version of events. The source, an OHSAA commissioner, is described — in evident contrast to Diadiun — as having attended the proceeding. The column does not quote any testimony from the court proceeding, nor does it describe what Milkovich said in court. There is only a vague statement from the OHSAA commissioner that the testimony “sounded pretty darned unfamiliar.”6 Fbr the first time, Diadiun fails *30to claim any firsthand knowledge, after stressing that he had personally attended both the meet and the OHSAA hearing. After noting again that the judge ruled in Milkovich’s and Maple Heights’ favor, Diadiun proclaims: “Anyone who attended the meet, whether he be from Maple Heights, Mentor, or impartial observer, knows in his heart that Milkovich and Scott lied at the hearing after each having given his solemn oath to tell the truth.” Ibid.
No reasonable reader could understand Diadiun to be impliedly asserting — as fact — that Milkovich had perjured himself. Nor could such a reader infer that Diadiun had further information about Milkovich’s court testimony on which his belief was based. It is plain from the column that Diadiun did not attend the court hearing. Diadiun also clearly had no detailed secondhand information about what Milkovich had said in court. Instead, what suffices for “detail” and “color” are quotations from the OHSAA hearing-old news compared to the court decision which prompted the column — and a vague quotation from an OHSAA commissioner. Readers could see that Diadiun was focused on the court’s reversal of the OHSAA’s decision and was angrily supposing what must have led to it.7
*31Even the insinuation that Milkovich had repeated, in court, a more plausible version of the misrepresentations he had made at the OHSAA hearing is preceded by the cautionary term “apparently” — an unmistakable sign that Diadiun did not know what Milkovich had actually said in court. “[C]au-tionary language or interrogatories of this type put the reader on notice that what is being read is opinion and thus weaken any inference that the author possesses knowledge of damaging, undisclosed facts. ... In a word, when the reasonable reader encounters cautionary language, he tends to ‘discount that which follows.’” Ollman v. Evans, 242 U. S. App. D. C., at 314, 750 F. 2d, at 983, quoting Burns v. McGraw-Hill Broadcasting Co., 659 P. 2d 1351, 1360 (Colo. 1983). See also B. Sanford, Libel and Privacy: The Prevention and Defense of Litigation 145 (1987) (explaining that many courts have found that words like “apparent” reveal “that the assertion is qualified or speculative and is not to be understood as a declaration of fact”); Information Control Corp. v. Genesis One Computer Corp., 611 F. 2d 781, 784 (CA9 1980) (explaining that a statement phrased in language of apparency “is less likely to be understood as a statement of *32fact rather than as a statement of opinion”); Gregory v. McDonnell Douglas Corp., 17 Cal. 3d 596, 603, 552 P. 2d 425, 429 (1976) (finding a letter “cautiously phrased in terms of apparency” did not imply factual assertions); Stewart v. Chicago Title Ins. Co., 151 Ill. App. 3d 888, 894, 503 N. E. 2d 580, 583 (1987) (finding a letter “couched in language of opinion rather than firsthand knowledge” did not imply factual assertions). Thus, it is evident from what Diadiun actually wrote that he had no unstated reasons for concluding that Milkovich perjured himself.
Furthermore, the tone and format of the piece notify readers to expect speculation and personal judgment. The tone is pointed, exaggerated, and heavily laden with emotional rhetoric and moral outrage. Diadiun never says, for instance, that Milkovich committed perjury. He says that “[ajnyone who attended the meet . . . knows in his heart” that Milkovich lied — obvious hyperbole as Diadiun does not purport to have researched what everyone who attended the meet knows in his heart.
The format of the piece is a signed editorial column with a photograph of the columnist and the logo “TD Says.” Even the headline on the page where the column is continued— “Diadiun says Maple told a lie,” ante, at 4 — reminds readers that they are reading one man’s commentary. While signed columns may certainly include statements of fact, they are also the “well recognized home of opinion and comment.” Mr. Chow of New York v. Ste. Jour Azur S. A., 759 F. 2d 219, 227 (CA2 1985). Certain formats — editorials, reviews, political cartoons, letters to the editor — signal the reader to anticipate a departure from what is actually known by the author as fact. See Ollman v. Evans, supra, at 317, 750 F. 2d, at 986 (“The reasonable reader who peruses [a] column on the editorial or Op-Ed page is fully aware that the statements found there are not ‘hard’ news like those printed on the front page or elsewhere in the news sections of the newspaper”); R. Smolla, Law of Defamation §6.12(4), n. 252 (1990) (col*33lecting cases); Zimmerman, Curbing the High Price of Loose Talk, 18 U. C. D. L. Rev. 359, 442 (1985) (stressing the need to take into account “the cultural common sense of the ordinary listener or reader”)-8
III
Although I agree with the majority that statements must be scrutinized for implicit factual assertions, the majority’s scrutiny in this case does not “hol[d] the balance true,” ante, at 23, between protection of individual reputation and freedom of speech. The statements complained of neither state nor imply a false assertion of fact, and, under the rule the Court reconfirms today, they should be found not libel “ ‘as a matter of constitutional law.’” Ante, at 17, quoting Bresler, 398 U. S., at 13. Readers of Diadiun’s column are signaled repeatedly that the author does not actually know what Milkovich said at the court hearing and that the author is surmising, from factual premises made explicit in the column, that Milkovich must have lied in court.9
*34Like the “imaginative expression” and the “rhetorical hyperbole” which the Court finds have “traditionally added much to the discourse of our Nation,” ante, at 18, conjecture is intrinsic to “the free flow of ideas and opinions on matters of public interest and concern” that is at “the heart of the First Amendment.” Falwell, 485 U. S., at 50. The public and press regularly examine the activities of those who affect our lives. “One of the perogatives of American citizenship is the right to criticize men and measures.” Id., at 51 (quoting Baumgartner v. United States, 322 U. S. 665, 673-674 (1944)). But often only some of the facts are known, and solely through insistent prodding — through conjecture as well as research — can important public questions be subjected to the “uninhibited, robust, and wide-open” debate to which this country is profoundly committed. New York Times Co. v. Sullivan, 376 U. S. 254, 270 (1964).
Did NASA officials ignore sound warnings that the Challenger Space Shuttle would explode? Did Cuban-American *35leaders arrange for John Fitzgerald Kennedy’s assassination? Was Kurt Waldheim a Nazi officer? Such questions are matters of public concern long before all the facts are unearthed, if they ever are. Conjecture is a means of fueling a national discourse on such questions and stimulating public pressure for answers from those who know more. “ ‘The maintenance of the opportunity for free political discussion to the end that government may be responsive to the will of the people and that changes may be obtained by lawful means, an opportunity essential to the security of the Republic, is a fundamental principle of our constitutional system.’” Id,., at 269 (quoting Stromberg v. California, 283 U. S. 359, 369 (1931)).
What may be more disturbing to some about Diadiun’s conjecture than, say, an editorial in 1960 speculating that Francis Gary Powers was in fact a spy, despite the Government’s initial assurances that he was not, is the naiveté of Diadiun’s conclusion. The basis of the court decision that is the subject of Diadiun’s column was that Maple Heights had been denied its right to due process by the OHSAA. Diadiun, as it happens, not only knew this but included it in his column. But to anyone who knows what “due process” means, it does not follow that the court must have believed some lie about what happened at the wrestling meet, because what happened at the meet would not have been germane to the questions at issue. There may have been testimony about what happened, and that testimony may have been perjured, but to anyone who understands the patois of the legal profession there is no reason to assume — from the court’s decision — that such testimony must have been given.
Diadiun, therefore, is guilty. He is guilty of jumping to conclusions, of benightedly assuming that court decisions are always based on the merits, and of looking foolish to lawyers. He is not, however, liable for defamation. Ignorance, without more, has never served to defeat freedom of speech. “The constitutional protection does not turn upon ‘the truth, popularity, or social utility of the ideas and beliefs which are *36offered.’” New York Times, supra, at 271 (quoting NAACP v. Button, 371 U. S. 415, 445 (1963)).
I appreciate this Court’s concern with redressing injuries to an individual’s reputation. But as long as it is clear to the reader that he is being offered conjecture and not solid information, the danger to reputation is one we have chosen to tolerate in pursuit of “ ‘individual liberty [and] the common quest for truth and the vitality of society as a whole.’” Falwell, supra, at 50-51 (quoting Bose Corp., 466 U. S., at 503-504). Readers are as capable of independently evaluating the merits of such speculative conclusions as they are of evaluating the merits of pure opprobrium. Punishing such conjecture protects reputation only at the cost of expunging a genuinely useful mechanism for public debate. “In a society which takes seriously the principle that government rests upon the consent of the governed, freedom of the press must be the most cherished tenet.” Edwards v. National Audubon Society, Inc., 556 F. 2d 113, 115 (CA2), cert. denied sub nom. Edwards v. New York Times Co., 434 U. S. 1002 (1977).
It is, therefore, imperative that we take the most particular care where freedom of speech is at risk, not only in articulating the rules mandated by the First Amendment, but also in applying them. “ ‘Whatever is added to the field of libel is taken from the field of free debate.’” New York Times, supra, at 272 (quoting Sweeney v. Patterson, 76 U. S. App. D. C. 23, 24, 128 F. 2d 457, 458, cert. denied, 317 U. S. 678 (1942)). Because I would affirm the Ohio Court of Appeals’ grant of summary judgment to respondents, albeit on somewhat different reasoning, I respectfully dissent.
See, e. g., New York Times Co. v. Sullivan, 376 U. S. 254, 292, n. 30 (1964) (“Since the Fourteenth Amendment requires recognition of the conditional privilege for honest misstatements of fact, it follows that a defense of fair comment must be afforded for honest expression of opinion based upon privileged, as well as true, statements of fact”); Gertz v. Robert Welch, Inc., 418 U. S. 323, 339-340 (1974) (“Under the First Amendment there is no such thing as a false idea. However pernicious an opinion may seem, we depend for its correction not on the conscience of judges and juries but on the competition of other ideas”).
The defendant in the Hepps case was a major daily newspaper and, as the majority notes, see ante, at 16, the Court declined to decide whether the rule it applied to the newspaper would also apply to a nonmedia defendant. See 475 U. S., at 779, n. 4. I continue to believe that “such a distinction is ‘irreconcilable with the fundamental First Amendment principle *24that “[t]he inherent worth of . . . speech in terms of its capacity for informing the public does not depend upon the identity of the source, whether corporation, association, union, or individual.’”” Id., at 780 (Brennan, J., concurring) (citations omitted).
The Restatement (Second) of Torts § 566, Comment c (1977), makes a similar observation. It explains that a statement that “I think C must be an alcoholic” is potentially libelous because a jury might find that it implies the speaker knew undisclosed facts to justify the statement. In contrast, it finds that the following statement could not be found to imply any defamatory facts:
“A writes to B about his neighbor C: ‘He moved in six months ago. He works downtown, and I have seen him during that time only twice, in his backyard around 5:30 seated in a deck chair with a portable radio listening to a news broadcast, and with a drink in his hand. I think he must be an alcoholic.’ ”
Yet even though clear disclosure of a comment’s factual predicate precludes a finding that the comment implies other defamatory facts, this does not signify that a statement, preceded by only a partial factual predicate or none at all, necessarily implies other facts. The operative question remains whether reasonable readers would have actually interpreted the statement as implying defamatory facts. See ante, at 20, n. 7; see generally Note, 13 Wm. Mitchell L. Rev. 545 (1987); Comment, 74 Calif. L. Rev. 1001 (1986); Zimmerman, Curbing the High Price of Loose Talk, 18 U. C. D. L. Rev. 359 (1985).
See ante, at 20, n. 7 (noting that under Philadelphia Newspapers, Inc. v. Hepps, 475 U. S. 767 (1986), “the issue of falsity relates to the defamatory facts implied by a statement” (emphasis changed)). Hepps mandates protection for speech that does not actually state or imply false and defamatory facts — independently of the Bresler-Letter Carriers-Falwell line of cases. Implicit in the constitutional rule that a plaintiff must prove a statement false to recover damages is a requirement to determine first what statement was actually made. The proof that Hepps requires from the plaintiff hinges on what the statement can reasonably be interpreted to mean. For instance, if Riley tells his friends that Smith cheats at cards and Smith then proves that he did not rob a convenience store, Smith cannot recover damages for libel on that basis because he has proved the wrong assertion false. Likewise, in the example in text, Jones cannot recover for defamation for the statement “I think Jones lied about his age just now” by producing proof that he did not lie about his age because, like Smith, he would have proved the wrong assertion false. The assertion *28Jones must prove false is that the speaker had, in fact, drawn the inference that Jones lied.
Conjecture, when recognizable as such, alerts the audience that the statement is one of belief, not fact. The audience understands that the speaker is merely putting forward a hypothesis. Although the hypothesis involves a factual question, it is understood as the author’s “best guess.” Of course, if the speculative conclusion is preceded by stated factual premises, and one or more of them is false and defamatory, an action for libel may lie as to them. But the speculative conclusion itself is actionable only if it implies the existence of another false and defamatory fact.
The commissioner is quoted as having said: “T can say that some of the stories told to the judge sounded pretty darned unfamiliar .... It certainly sounded different from what they told us.’” Ante, at 7, n. 2. This quotation might also be regarded as a stated factual premise on which Diadiun’s speculation is based. However, Milkovich did not complain of the quotation in his pleadings. In any event, it is unlikely that it would be found defamatory. Diadiun had already characterized the testimony of the two officials before the OHSAA as “obvious untruths.” Thus, the commissioner’s alleged assertion that the testimony in court was different is quite nebulous. It might indicate that the officials told the truth in court, in contrast to the version given to the commissioners, or that the *30officials discussed entirely different issues, rather than that they told a new lie.
Both state and federal courts have found that audiences can recognize conjecture that neither states nor implies any assertions of fact, just as they can recognize hyperbole. For example, in Potomac Valve & Fitting Inc. v. Crawford Fitting Co., 829 F. 2d 1280, 1290 (CA4 1987), the court found that a disparaging statement about a product test in an industry newsletter, set forth following a list of seven observations about the test’s methodology, “readily appears to be nothing more than the author’s personal inference from the test results. The premises are explicit, and the reader is by no means required to share [the author’s] conclusion.” For the same reason, the court in Dunlap v. Wayne, 105 Wash. 2d 529, 540, 716 P. 2d 842, 849 (1986), concluded: “Arguments for actionability disappear when the audience members know the facts underlying an assertion and can judge the truthfulness of the allegedly defamatory statement themselves.” See also National Assn. of Government Employees, Inc. v. Cen*31tral Broadcasting Corp., 379 Mass. 220, 226, 396 N. E. 2d 996, 1000 (1979) (finding that, as listeners were told the facts upon which a radio talk show host based her conclusion, they “could make up their own minds and generate their own opinions or ideas which might or might not accord with [the host’s]”).
The common-law doctrine of fair comment was also premised on such an observation. Where the reader knew or was told the factual foundation for a comment and could therefore independently judge whether the comment was reasonable, a defendant’s unreasonable comment was held to defame “ ‘himself rather than the subject of his remarks. ’ ” Hill, Defamation and Privacy Under the First Amendment, 76 Colum. L. Rev. 1205, 1229 (1976) (quoting Popham v. Pickbimi, 7 H. & N. 891, 898, 158 Eng. Rep. 730, 733 (Ex. 1862) (Wilde, B.)). “As Thomas Jefferson observed in his first Inaugural Address . . . error of opinion need not and ought not be corrected by the courts ‘where reason is left free to combat it.'" Potomac, supra, at 1288-1289, quoting Thomas Jefferson’s first Inaugural Address (The Complete Jefferson 385 (S. Padover ed. 1943)).
The readers of Diadiun’s column would also have been alerted to regard any implicit claim of impartiality by Diadiun with skepticism because Diadiun’s newspaper is published in the county in which Mentor High School — home to the team that was allegedly mauled at the wrestling meet — is located. Where readers know that an author represents one side in a controversy, they are properly warned to expect that the opinions expressed may rest on passion rather than factual foundation. See, e. g., Potomac Valve & Fitting Inc. v. Crawford Fitting Co., 829 F. 2d, at 1290 (explaining that the contents of a company’s newsletter would be understood as reflecting the professional interests of the company rather than as “a dispassionate and impartial assessment” of a test of a competitor’s product); Information Control Corp. v. Genesis One Computer Corp., 611 F. 2d 781, 784 (CA9 1980) (recognizing that statements in the early weeks of litigation by one side about the other were likely to include unsubstantiated charges, but that these “are highly unlikely to be understood by their audience as statements of fact”).
Milkovich does not challenge the accuracy of any of Diadiun’s stated premises. Nor does he complain or proffer proof that Diadiun had not, in fact, concluded from the stated premises that Milkovich must have lied in court. There is, therefore, no call to consider under what circumstances *34an insincere speculation would constitute a false and defamatory statement under Philadelphia Newspapers, Inc. v. Hepps, 475 U. S. 767 (1986). However, I would think that documentary or eyewitness testimony that the speaker did not believe his own professed opinion would be required before a court would be permitted to decide that there was sufficient evidence to find that the statement was false and submit the question to a jury. Without such objective evidence, a jury’s judgment might be too influenced by its view of what was said. As we have long recognized, a jury “is unlikely to be neutral with respect to the content of speech and holds a real danger of becoming an instrument for the suppression of those ‘vehement, caustic, and sometimes unpleasantly sharp attacks,’. . . which must be protected if the guarantees of the First and Fourteenth Amendments are to prevail.” Monitor Patriot Co. v. Roy, 401 U. S. 265, 277 (1971) (quoting New York Times, 376 U. S., at 270). See also Bose Corp. v. Consumers Union of United States, Inc., 466 U. S. 485, 510-511, and n. 29 (1984) (discussing the risks of submitting various questions to juries where freedom of speech is at stake); Gertz, 418 U. S., at 349 (expressing concern about juries punishing unpopular opinion rather than compensating individuals for injuries sustained by the publication of a false fact); R. Smolla, Law of Defamation §§ 6.05(3)(a)-(c) (1990); Zimmerman, 18 U. C. D. L. Rev., at 430.