Opinion for the Court filed by Circuit Judge STARR.

Concurring opinion filed by Circuit Judge BORK, with whom Circuit Judges WILKEY, GINSBURG and Senior Circuit Judge MacKINNON join.

Concurring opinion filed by Senior Circuit Judge MacKINNON.

Opinion dissenting in part, filed by Chief Judge SPOTTSWOOD W. ROBINSON, III, with whom Circuit Judge J. SKELLY WRIGHT joins.

Opinion dissenting in part filed by Circuit Judge WALD, with whom Circuit Judges HARRY T. EDWARDS and SCALIA join.

Statement concurring in part and dissenting in part filed by Circuit Judge HARRY T. EDWARDS.

Opinion dissenting in part filed by Circuit Judge SCALIA, with whom Circuit Judges WALD and HARRY T. EDWARDS join.

This defamation action arises out of the publication of a syndicated column by Rowland Evans and Robert Novak in May 1978. The question before us is whether the allegedly defamatory statements set forth in the column are constitutionally protected expressions of opinion or, as appellant contends, actionable assertions of fact. We conclude, as did the District Court, that the challenged statements are entitled to absolute First Amendment protection as expressions of opinion.

* Rowland Evans and Robert Novak are nationally syndicated columnists whose columns appear regularly in newspapers across the country. According to the complaint in this case, which was filed by plaintiff Bertell Ollman on February 15, 1979, an Evans and Novak column appeared on or about May 4, 1978 in The Washington Post and other newspapers across the Nation. Complaint p 5. Attached to the complaint as Exhibit A was a photocopy of the column, styled "The Marxist Professor's Intentions," as it appeared in The Washington Post on May 4, 1978. A copy of that column is reproduced as an Appendix to this opinion.

The plaintiff, Bertell Ollman, is a professor of political science at New York University. The complaint averred that Mr. Ollman "is widely esteemed among his colleagues and enjoys the highest possible reputation as a scholar of integrity and a teacher." Complaint p 2. In March 1978, Mr. Ollman was nominated by a departmental search committee to head the Department of Government and Politics at the University of Maryland. The committee's recommendation "was duly approved by the Provost of the University and the Chancellor of the College Park campus." Id. p 4.

With this professional move from Washington Square to College Park, Maryland thus in the offing, the Evans and Novak article appeared. Since the years of litigation that have followed revolve entirely around this single column, we will begin by describing its contents in some detail. In our description, we will highlight the specific portions that Mr. Ollman assails as false and defamatory. The column begins as follows:

What is in danger of becoming a frivolous public debate over the appointment of a Marxist to head the University of Maryland's department of politics and government has so far ignored this unspoken concern within the academic community: the avowed desire of many political activists to use higher education for indoctrination.

The column immediately goes on to state that:

[t]he proposal to name Bertell Ollman, Professor at New York University, as department head has generated wrong-headed debate. Politicians who jumped in to oppose Ollman simply for his Marxist philosophy have received a justifiable going-over from defenders of academic freedom in the press and the university. Academic Prince Valiants seem arrayed against McCarythite [sic] know-nothings.

With these opening two paragraphs as lead-in, the authors then pose what they deemed the pivotal issue in the debate: "But neither side approaches the crucial question: not Ollman's beliefs, but his intentions. His candid writings avow his desire to use the classroom as an instrument for preparing what he calls 'the revolution.' Whether this is a form of indoctrination that could transform the real function of a university and transcend limits of academic freedom is a concern to academicians who are neither McCarthyite nor know-nothing." (Emphasis added).

The columnists thus, in the first three paragraphs, articulated a view of what should be the central question in what they viewed as a fruitless debate. The authors then go on in the next paragraph to state: "To protect academic freedom, that question should be posed not by politicians but by professors. But professors throughout the country troubled by the nomination, clearly a minority, dare not say a word in today's campus climate."

With this observation, the authors turn in the following six paragraphs to a discussion of Mr. Ollman and his writings. Evans and Novak state that "[w]hile Ollman is described in news accounts as a 'respected Marxist scholar,' he is widely viewed in his profession as a political activist. Amid the increasingly popular Marxist movement in university life, he is distinct from philosophical Marxists. Rather, he is an outspoken proponent of 'political Marxism.' " (Emphasis added).

The authors next relate Mr. Ollman's two unsuccessful efforts to win election to membership on the council of the American Political Science Association. In these elections, the column states (and appellant does not dispute) that Professor Ollman ran as a candidate of the Caucus for a New Political Science and finished last out of sixteen candidates each time. "Whether or not that represents a professional judgment by his colleagues, as some critics contend, the verdict clearly rejected his campaign pledge: 'If elected ... I shall use every means at my disposal to promote the study of Marxism and Marxist approaches to politics throughout the profession.' "

Evans and Novak then direct the four ensuing paragraphs of the column to a summary of an article by Mr. Ollman, entitled "On Teaching Marxism and Building the Movement" in the Winter 1978 issue of New Political Science. Record ("R.") 3. In this article, Mr. Ollman claims that most students conclude his political science course with a " 'Marxist outlook.' " The authors go on:

Ollman concedes that will be seen "as an admission that the purpose of my course is to convert students to socialism."

That bothers him not at all because "a correct understanding of Marxism (as indeed of any body of scientific truths) leads automatically to its acceptance." * * * The "classroom" is a place where the students' bourgeois ideology is being dismantled. "Our prior task" before the revolution, he writes, "is to make more revolutionaries."1

Moving to a brief discussion of Mr. Ollman's principal work, Alienation: Marx's Conception of Man in Capitalist Society, the authors described the work as "a ponderous tome in adoration of the master (Marxism 'is like a magnificiently rich tapestry'). Published in 1971, it does not abandon hope for the revolution forecast by Karl Marx in 1848." This brings the columnists to the last statement specifically identified in the complaint as defamatory:

Such pamphleteering is hooted at by one political scientist in a major eastern university, whose scholarship and reputation as a liberal are well known. "Ollman has no status within the profession, but is a pure and simple activist," he said. Would he say that publicly? "No chance of it. Our academic culture does not permit the raising of such questions." (Emphasis added).

Evans and Novak then bring the column to a close, indicating in the penultimate paragraph that " '[s]uch questions' would include these: What is the true measurement of Ollman's scholarship? Does he intend to use the classroom for indoctrination? Will he indeed be followed by other Marxist professors? Could the department in time be closed to non-Marxists, following the tendency at several English universities?"

In the column's final paragraph, the authors return to their opening theme that "such questions" as set forth in the previous paragraph should not be raised by politicians, even if, as the anonymous political scientist claimed, they cannot be raised within the Academy. They conclude the column by calling upon academics to address these questions:

Here are the makings of a crisis that, to protect its integrity and true academic freedom, academia itself must resolve.

On May 19, 1978, Mr. Ollman's lawyer wrote to Evans and Novak demanding retraction of the allegedly defamatory statements in the column. Letter of I. Silver to R. Evans and R. Novak (May 19, 1978). R. 1. This Evans and Novak refused to do. On May 8, however, only four days after the Evans and Novak column appeared, The Washington Post published a letter from Mr. Ollman. In this letter, Professor Ollman rejected the allegation that he used the classroom to indoctrinate students and set the column's quotations from his writings in what he viewed as their proper context. Letter from B. Ollman to the Editors of The Washington Post (May 8, 1978). R. 3.

The District Court granted Evans and Novak's motion for summary judgment, concluding that the column simply reflected the columnists' opinion and their "interpretation of [Mr. Ollman's] writings." Memorandum Opinion at 5.2 Thus, the District Court held that the opinion was absolutely protected by the First Amendment. This appeal followed.

* This case presents us with the delicate and sensitive task of accommodating the First Amendment's protection of free expression of ideas with the common law's protection of an individual's interest in reputation. It is a truism that the free flow of ideas and opinions is integral to our democratic system of government. Thomas Jefferson well expressed this principle in his First Inaugural Address, when the Nation's memory was fresh with the passage of the notorious Alien and Sedition Acts:

If there be any among us who would wish to dissolve this Union or to change its republican form, let them stand undisturbed as monuments of the safety with which error of opinion may be tolerated where reason is left free to combat it.3

At the same time, an individual's interest in his or her reputation is of the highest order. Its protection is an eloquent expression of the respect historically afforded the dignity of the individual in Anglo-American legal culture.4 A defamatory statement may destroy an individual's livelihood, wreck his standing in the community, and seriously impair his sense of dignity and self-esteem.

The judiciary's task in accommodating these competing interests is by no means new: at common law, the fair comment doctrine bestowed qualified immunity from libel actions as to certain types of opinions in order that writers could express freely their views about subjects of public interest.5 However, since Gertz v. Robert Welch, Inc., 418 U.S. 323, 94 S. Ct. 2997, 41 L. Ed. 2d 789 (1974), the nature of this accommodation has fundamentally changed. In Gertz, the Supreme Court in dicta seemed to provide absolute immunity from defamation actions for all opinions and to discern the basis for this immunity in the First Amendment. The Court began its analysis of the case by stating:

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. But there is no constitutional value in false statements of fact. Neither the intentional lie nor the careless error materially advances society's interest in "uninhibited, robust, and wide-open debate on the public issues."6

By this statement, Gertz elevated to constitutional principle the distinction between fact and opinion, which at common law had formed the basis of the doctrine of fair comment.7 Gertz's implicit command thus imposes upon both state and federal courts the duty as a matter of constitutional adjudication to distinguish facts from opinions in order to provide opinions with the requisite, absolute First Amendment protection.8 At the same time, however, the Supreme Court provided little guidance in Gertz itself as to the manner in which the distinction between fact and opinion is to be discerned. That, as we shall see, is by no means as easy a question as might appear at first blush.

Indeed, Gertz did not focus on this distinction at all. Rather, assuming without lengthy discussion that the statements in that case could be construed as statements of fact, the Court held that the plaintiff, who was a private rather than public figure, could prove that the statements at issue there were libelous upon demonstrating that they were negligently made.9 The distinction in our law between public and private figures, however, does not directly bear on the distinction between fact and opinion.10 Expressions of opinion are protected whether the subject of the comment is a private or public figure. See Lewis v. Time, Inc., 710 F.2d 549, 555 (9th Cir.1983). In a word, Gertz's reasoning immunizes an opinion, not because the opinion is asserted about a public figure, but because there is no such thing as a "false" opinion.

While Gertz is mute with respect to the method of separating fact from opinion, two Supreme Court cases do provide guidance in this respect. Old Dominion Branch No. 496, National Association of Letter Carriers v. Austin, 418 U.S. 264, 94 S. Ct. 2770, 41 L. Ed. 2d 745 (1974); Greenbelt Cooperative Publishing Association v. Bresler, 398 U.S. 6, 90 S. Ct. 1537, 26 L. Ed. 2d 6 (1970).11 In Letter Carriers, decided by the Court on the same day as Gertz, three non-union employees of the Postal Service were included on a list of names circulated by the Letter Carriers union. To the list was appended a well-known piece of trade union literature attributed to Jack London, which defined the nature of a "scab." Drawing upon Biblical references and American history, the definition ended on the following unflattering note:

The scab sells his birthright, country, his wife and his children and his fellowmen for an unfulfilled promise from his employer.

Esau was a traitor to himself; Judas was a traitor to his God; Benedict Arnold was a traitor to his country; a SCAB is a traitor to his God, his country, his family and his class.

Letter Carriers, 418 U.S. at 268, 94 S.Ct. at 2773. Holding this allegedly defamatory language to be absolutely protected, the Supreme Court reversed a libel judgment in favor of the non-union employees. While the Court grounded its decision upon federal labor laws' protection of communications in a labor dispute, rather than the First Amendment, the Court's analysis derived from Gertz's proposition that opinions cannot be false. Id. 418 U.S. at 284, 94 S. Ct. at 2781 (citing Gertz, supra, 418 U.S. at 339-340, 94 S. Ct. at 3006-3007). To demonstrate that the union's "scab" description was indeed opinion, the Court considered both its specific linguistic context and its broader social setting. The Court found, for instance, that the epithet "traitor" in the context of a well-known piece of union literature was deployed in a "loose, figurative sense" and could not be taken for an assertion that the identified employees had "committ[ed] the criminal offense of treason." Id. at 284-85, 94 S.Ct. at 2781. Moving to the social context in which the statement was made, the Court further noted that this type of "exaggerated rhetoric was commonplace in labor disputes." Thus, the Court concluded, readers would be alerted by virtue of the broad context in which the statement was made that the statement was opinion, not an imputation of actual criminal conduct. Id. at 286, 94 S.Ct. at 2782.

Letter Carriers also relied upon Greenbelt Publishing, supra, for the proposition that the allegedly libelous language must be evaluated in its broader context to assess whether a reader would have understood the allegation to be a statement of fact. Id. at 284, 94 S. Ct. at 2781. Greenbelt Publishing was, of course, a pre-Gertz case, which may be seen in retrospect as an application of the distinction between fact and opinion subsequently delineated in Gertz. In Greenbelt Publishing, a developer was attempting to secure zoning variances to construct high density housing; at the same time, the city of Greenbelt, Maryland was trying to purchase land from the developer to build a school. During the course of these negotiations, some attendees at a public meeting characterized the developer's negotiating tactics as "blackmail." The developer thereafter brought a successful libel suit against a local newspaper that printed this colorful characterization. The Supreme Court reversed the judgment, concluding that "as a matter of constitutional law, the word "blackmail" in these circumstances was ... not libel when reported." Id., 398 U.S. at 13, 90 S. Ct. at 1541. The Court noted that the Greenbelt newspaper was performing a wholly "legitimate function as a community newspaper" and that it "accurately and fully" described the developer's negotiating proposals. The Court then held that in light of the full context of the articles a reader would have understood the "blackmail" characterization as a criticism of the developer's negotiating tactics rather than as an actual criminal charge. Under the circumstances, the remark was deemed to be merely "rhetorical hyperbole." Id. at 14, 90 S.Ct. at 1542.

There is, then, limited but helpful teaching from the Supreme Court to guide us in our inquiry. With largely uncharted seas having been left in Gertz's wake, the lower federal courts and state courts have, not surprisingly, fashioned various approaches in attempting to articulate the Gertz-mandated distinction between fact and opinion. We pause here, briefly, to examine the results of the efforts of our fellow laborers in this new constitutional vineyard.

Some courts have, in effect, eschewed any effort to construct a theory and simply treated the distinction between fact and opinion as a judgment call. See, e.g., Shiver v. Apalachee Publishing Co., 425 So. 2d 1173 (Fla.Dist.Ct.App.1983). Other courts have concentrated on a single factor, such as the verifiability vel non of the allegedly defamatory statement. See, e.g., Hotchner v. Castillo-Puche, 551 F.2d 910, 913 (2d Cir.), cert. denied sub. nom. Hotchner v. Doubleday & Co., 434 U.S. 834, 98 S. Ct. 120, 54 L. Ed. 2d 95 (1977). Still others have adopted a multi-factor test, attempting to assess the allegedly defamatory proposition in the totality of the circumstances in which it appeared. See, e.g., Information Control Corp. v. Genesis One Computer Corp., 611 F.2d 781 (9th Cir.1980).12

In formulating a test to distinguish between fact and opinion, courts are admittedly faced with a dilemma. Because of the richness and diversity of language, as evidenced by the capacity of the same words to convey different meanings in different contexts, it is quite impossible to lay down a bright-line or mechanical distinction.13 Judicial decisions, however, that represent mere ad hoc judgments or which, in contrast, lay down rules of excessive complexity may deter publication of the very opinions which the Gertz-mandated distinction is designed to protect, inasmuch as potential speakers or writers would, under such regimes, be at a loss to predict what courts will ultimately deem to be opinion. While this dilemma admits of no easy resolution, we think it obliges us to state plainly the factors that guide us in distinguishing fact from opinion and to demonstrate how these factors lead to a proper accommodation between the competing interests in free expression of opinion and in an individual's reputation.

In formulating this analysis, we agree with the overwhelming weight of post-Gertz authority that the distinction between opinion and fact is a matter of law. See, e.g., Lewis v. Time, Inc., supra, 710 F.2d at 553; Rinsley v. Brandt, 700 F.2d 1304, 1309 (10th Cir.1983); Orr v. Argus-Press Co., 586 F.2d 1108, 1114 (8th Cir.), cert. denied, 440 U.S. 960, 99 S. Ct. 1502, 59 L. Ed. 2d 773 (1979). Although the Supreme Court has never directly addressed this issue, the Court has clearly ruled that questions as to other privileges derived from the First Amendment, such as the qualified privilege as to public officials and public figures, are to be decided as matters of law. See Gertz, supra, 418 U.S. at 346, 94 S. Ct. at 3010. Moreover, the predictability of decisions, which is of crucial importance in an area of law touching upon First Amendment values, is enhanced when the determination is made according to announced legal standards and when a body of public case law furnishes published examples of the manner in which these standards are to be applied.

While courts are divided in their methods of distinguishing between assertions of fact and expressions of opinion, they are universally agreed that the task is a difficult one. See, e.g., Rinsley v. Brandt, supra, 700 F.2d at 1309. To be sure, paradigm examples of statements of fact, on the one hand, and paradigm examples of expressions of opinion, on the other, can be contrasted. Clearly, in the former category are assertions that describe present or past conditions capable of being known through sense impressions. See Goodrich v. Waterbury Republican-American, Inc., 448 A.2d 1317, 1321 (Conn.1982) (citing 1 F. Harper & F. James, Torts Sec. 5.28, p. 458 n. 11, Sec. 7.8, p. 560). It is rather hard to imagine a context in which the statement, "Mr. Jones had ten drinks at his office party and sideswiped two vehicles on his way home," could be deemed to be a statement of opinion. At the other extreme are evaluative statements reflecting the author's political, moral, or aesthetic views, not the author's sense perceptions. A statement such as, "Mr. Jones is a despicable politician," is a paradigm of opinion.

It is a fitting illustration of the complexity of language and communication that many statements from which actions for defamation arise do not clearly fit into either category. These statements pose more subtle problems and are the stuff of which litigation is made. The principal difficulty arises from statements that on first analysis seem to be based upon perceptions of events, but are not themselves simply a record of those perceptions. Such statements may imply in some contexts the existence of facts not disclosed by the author.14 An example of such a statement, set forth in the Restatement (Second) of Torts, is: "Mr. Jones is an alcoholic."15 These statements obviously can be as damaging to reputation as statements which on their face describe particular historical events.

The degree to which such kinds of statements have real factual content can, of course, vary greatly. We believe, in consequence, that courts should analyze the totality of the circumstances in which the statements are made to decide whether they merit the absolute First Amendment protection enjoyed by opinion. To evaluate the totality of the circumstances of an allegedly defamatory statement, we will consider four factors in assessing whether the average reader would view the statement as fact or, conversely, opinion.16 While necessarily imperfect, these factors will, we are persuaded, assist in discerning as systematically as possible what constitutes an assertion of fact and what is, in contrast, an expression of opinion.

First, we will analyze the common usage or meaning of the specific language of the challenged statement itself. Our analysis of the specific language under scrutiny will be aimed at determining whether the statement has a precise core of meaning for which a consensus of understanding exists or, conversely, whether the statement is indefinite and ambiguous. See Buckley v. Littell, 539 F.2d 882, 895 (2d Cir.1976), cert. denied, 429 U.S. 1062, 97 S. Ct. 785, 50 L. Ed. 2d 777 (1977). Readers are, in our judgment, considerably less likely to infer facts from an indefinite or ambiguous statement than one with a commonly understood meaning. Second, we will consider the statement's verifiability--is the statement capable of being objectively characterized as true or false? See, e.g., Hotchner v. Castillo-Puche, supra, 551 F.2d at 913. Insofar as a statement lacks a plausible method of verification, a reasonable reader will not believe that the statement has specific factual content. And, in the setting of litigation, the trier of fact obliged in a defamation action to assess the truth of an unverifiable statement will have considerable difficulty returning a verdict based upon anything but speculation. Third, moving from the challenged language itself, we will consider the full context of the statement--the entire article or column, for example--inasmuch as other, unchallenged language surrounding the allegedly defamatory statement will influence the average reader's readiness to infer that a particular statement has factual content. See Greenbelt Cooperative Publishing Association v. Bresler, supra, 398 U.S. at 13-14, 90 S. Ct. at 1541; cf. Restatement (Second) of Torts Sec. 563. Finally, we will consider the broader context or setting in which the statement appears. Different types of writing have, as we shall more fully see, widely varying social conventions which signal to the reader the likelihood of a statement's being either fact or opinion. See Old Dominion Branch No. 496, National Association of Letter Carriers v. Austin, supra, 418 U.S. at 286, 94 S. Ct. at 2782.

* The first factor of our inquiry is to analyze the common usage or meaning of the allegedly defamatory words themselves.17 We seek in this branch of our analysis to determine whether the allegedly defamatory statement has a precise meaning and thus is likely to give rise to clear factual implications.18 A classic example of a statement with a well-defined meaning is an accusation of a crime. To be sure, such accusations are not records of sense perceptions. Quite to the contrary, they depend for their meaning upon social normative systems. But those norms are so commonly understood that the statements are seen by the reasonable reader or hearer as implying highly damaging facts. Post-Gertz courts have therefore not hesitated to hold that accusations of criminal conduct are statements "laden with factual content" that may support an action for defamation. See, e.g., Cianci v. New Times Publishing Co., 639 F.2d 54, 63 (2d Cir.1980) (holding that an article which implied that the Mayor of Providence, R.I., had committed rape and which charged him with paying the alleged victim not to bring charges was not protected opinion). Even a somewhat less well defined accusation that a "judge is corrupt" has been held actionable. Rinaldi v. Holt, Rinehart & Winston, Inc., 42 N.Y.2d 1299, 397 N.Y.S.2d 943, 366 N.E.2d 1299, cert. denied, 434 U.S. 969, 98 S. Ct. 514, 54 L. Ed. 2d 456 (1977). "Corruption," at least in the context of public service, was deemed to imply factual allegations of bribery or other official malfeasance.

On the other hand, statements that are "loosely definable" or "variously interpretable" cannot in most contexts support an action for defamation. See Buckley v. Littell, supra, 539 F.2d at 895. In that case, a writer in his book on the political right in the United States accused columnist and author William F. Buckley, Jr., of being a "fellow traveler" of "fascists." Noting that Mr. Buckley and the author of this particular tome embraced widely different definitions of "fascism" and different views as to which journals could be described as "fascist," the court declined to develop a "correct" definition of this pivotal term.19 The Second Circuit held, rather, that the use of such expressions "cannot be regarded as having been proved to be statements of facts, among other reasons, because of the tremendous imprecision of the meaning and usage of these terms in the realm of political debate, an imprecision which is similarly echoed in the book."20 Id. at 893. Pursuing a line of analysis similar to that found in Buckley, the same court that held actionable the term "corrupt" concluded that the term "incompetent" as applied to a judge was too vague to support a claim of libel. Rinaldi v. Holt, Rinehart & Winston, Inc., supra, 397 N.Y.S. at 947, 366 N.E.2d at 1303.

The use of indefinite terms is obviously not confined to the realm of politics and public policy. In Cole v. Westinghouse Broadcasting Co., Inc., 386 Mass. 303, 435 N.E.2d 1021, cert. denied, 459 U.S. 1037, 103 S. Ct. 449, 74 L. Ed. 2d 603 (1982), the Massachusetts Supreme Judicial Court held that the statement that a reporter had engaged in "sloppy and irresponsible reporting" and had poor reporting technique was too "imprecise" to support a defamation action.21 Similarly, in Avins v. White, 627 F.2d 637 (3d Cir.), cert. denied, 449 U.S. 982, 101 S. Ct. 398, 66 L. Ed. 2d 244 (1982), the former dean of a law school claimed that his academic ability and performance had been falsely disparaged in the summary evaluation of the school's first accreditation report. The summary bluntly stated: "[T]he most important deficiency [of the law school] is an intangible one; there is an academic ennui that pervades the institution. The intellectual spark is missing in the faculty and students." Id. at 642. Emphasizing that the statement itself described its criticism as "intangible," the Avins court classified the statement as an expression of opinion.

The straightforward but important principle to be drawn from cases such as Buckley, Rinaldi, Cole and Avins is that in all types of discourse, the courts must analyze the allegedly defamatory statement to determine whether it has a sufficiently definite meaning to convey facts.

In assessing whether the challenged statements are facts, rather than opinion, courts should, secondly, consider the degree to which the statements are verifiable--is the statement objectively capable of proof or disproof? See Goodrich v. Waterbury Republican-American, supra, 448 A.2d at 1319; Hotchner v. Castillo-Puche, supra, 551 F.2d at 913.22 The reason for this inquiry is simple: a reader cannot rationally view an unverifiable statement as conveying actual facts. Moreover, insofar as a statement is unverifiable, the First Amendment is endangered when attempts are made to prove the statement true or false. Lacking a clear method of verification with which to evaluate a statement--such a labelling a well-known American author a "fascist," see Buckley v. Littell, supra--the trier of fact may improperly tend to render a decision based upon approval or disapproval of the contents of the statement, its author, or its subject.

In making this observation, we imply no criticism of a jury's ability to find facts, if facts are to be found. The rule against allowing unverifiable statements to go to the jury is, in actuality, merely one of many rules in tort law that prevent the jury from rendering a verdict based on speculation. Cf. Hobson v. Wilson, 737 F.2d 1 at 62 (D.C.Cir.1984) (permitting First Amendment interests to be compensated "if they can be conceptualized and if harm can be shown with sufficient certainty to avoid damages based ... on pure speculation"). An obvious potential for quashing or muting First Amendment activity looms large when juries attempt to assess the truth of a statement that admits of no method of verification.

Needless to say, it will often be difficult to assay whether a statement is verifiable. Statements made in written communication or discourse range over a spectrum with respect to the degree to which they can be verified rather than dividing neatly into categories of "verifiable" and "unverifiable." But even if the principle of inquiring as to verifiability provides no panacea, this approach will nonetheless aid trial judges in assessing whether a statement should have the benefit of the absolute privilege conferred upon expressions of opinion. Trial judges have rich experience in the ways and means of proof and so will be particularly well situated to determine what can be proven.

In addition to evaluating the precision-indefiniteness and verifiability-unverifiability of a challenged statement, courts should, thirdly, examine the context in which the statement occurs. Readers will inevitably be influenced by a statement's context, and the distinction between fact and opinion can therefore be made only in context. As the Supreme Court's opinions in Greenbelt and Letter Carriers suggest, the context to be considered is both narrowly linguistic and broadly social.

The degree to which a statement is "laden with factual content" or can be read to imply facts depends upon the article or column, taken as a whole, of which the statement is a part. See Information Control v. Genesis One Computer, supra, 611 F.2d at 783. The language of the entire column may signal that a specific statement which, standing alone, would appear to be factual is in actuality a statement of opinion. An example of the power of context to transform an ostensibly factual statement into one of opinion is Greenbelt Publishing. See supra I A. Because the local newspaper in that case had described the substance of the land developer's negotiating proposals, the use of the term "blackmail" to characterize those proposals was quite plainly to be seen as an expression of opinion.23

An article or column, however, plainly does not have to include a complete set of facts to make it clear that a statement is being used in a metaphorical, exaggerated or even fantastic sense. In Myers v. Boston Magazine Co., Inc., 380 Mass. 336, 403 N.E.2d 376 (1980), the court held as protected opinion a magazine's statement that a television sports reporter was "the only newscaster in town who is enrolled in a course for remedial speaking." Id., 403 N.E.2d at 377. Although the statement on its face appears quite factual, the court emphasized in its analysis that the statement appeared in an article describing the best and worst sports personalities in a series of "one-liners." Id. For instance, the court noted that another item in the article described the Boston Bruins hockey team members as looking "like a gargoyle" and that the various descriptions had corresponding cartoons. The court concluded that the average reader would have been put on notice that he or she was reading opinions, and not being showered with facts. Id., 403 N.E.2d at 379.

Another consideration in this respect, of particular relevance to the case at hand and useful in distinguishing between fact and opinion, is the inclusion of cautionary language in the text in which the statement at issue is found, see Information Control, supra, 611 F.2d at 784 (noting that the allegedly libelous statement was preceded by the phrase, "In the opinion of Genesis' management" and that this favored treating the statement which followed as opinion), or framing the statement as an interrogatory ("Is it not true that ... ?"). The rationale typically advanced for this consideration is that cautionary 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. See Pease v. Telegraph Publishing Co., 121 N.H. 62, 426 A.2d 463, 465 (1981). In a word, when the reasonable reader encounters cautionary language, he tends to "discount that which follows." See Burns v. McGraw-Hill Broadcasting Co., 659 P.2d 1351, 1360 (Colo.1983).

To be sure, there is authority against giving weight to cautionary or interrogatory language. Stating that "[i]t would be destructive of the law of libel if a writer could escape liability for accusations of crime simply by using, explicitly or implicitly, the words 'I think,' " Cianci, supra, 639 F.2d at 64, the Second Circuit in an opinion by Judge Friendly rejected the notion that cautionary language could immunize an otherwise defamatory statement. While Judge Friendly's argument is not without force, it may be overstated if applied outside the type of facts before the court in Cianci--the accusation of a crime--since cautionary language is only one of several factors to be considered in assessing an allegedly defamatory statement.24 Burns v. McGraw-Hill, supra, 659 P.2d at 1360 n. 4. When a statement is as "factually laden" as the accusation of a crime, which of course was the issue in Cianci, cautionary language is by and large unavailing to dilute the statement's factual implications. However, in statements less clearly factual, cautionary language may make a more substantial difference to the reader's understanding.

What is more, we cannot forget that the public has an interest in receiving information on issues of public importance even if the trustworthiness of the information is not absolutely certain. The First Amendment is served not only by articles and columns that purport to be definitive but by those articles that, more modestly, raise questions and prompt investigation or debate. By giving weight on the opinion side of the scale to cautionary and interrogative language, courts provide greater leeway to journalists and other writers and commentators in bringing issues of public importance to the public's attention and scrutiny.

Besides looking to the immediate context of the allegedly defamatory statement, courts should examine, finally, the broader social context into which the statement fits. Some types of writing or speech by custom or convention signal to readers or listeners that what is being read or heard is likely to be opinion, not fact.25 It is one thing to be assailed as a corrupt public official by a soapbox orator and quite another to be labelled corrupt in a research monograph detailing the causes and cures of corruption in public service. This observation reflects no novel principle. The Supreme Court has expressly recognized the importance of social context when, in finding as an expression of opinion the use of the word "traitor" as applied to an employee who crossed a picket line, the Court stated that "such exaggerated rhetoric was commonplace in labor disputes." Letter Carriers, supra, 418 U.S. at 286, 94 S. Ct. at 2782.26

Similarly, in Myers v. Boston Magazine, supra, the Massachusetts Supreme Judicial Court was even more explicit in focusing upon the reader's understanding of a particular type of writing. Emphasizing that the "magazine's statement partook of an ancient, lively tradition of criticizing, even lampooning, performers," the court concluded that the statement that a sportscaster was attending a course in remedial speaking constituted privileged opinion. Id., 403 N.E.2d at 381. In the lampooning tradition, the court emphasized, it is well understood that "a critic may resort to caricature and rhetorical license." Id. See also Pring v. Penthouse, Inc., 695 F.2d 438 (10th Cir.1982), cert. denied, --- U.S. ----, 103 S. Ct. 3112, 77 L. Ed. 2d 1367 (1983) (finding that the imputation that the plaintiff had committed sexual acts on stage at the Miss America Pageant could not support a libel action when the writing in which the statement appeared was clearly a "fantasy").

Courts have, in the same vein, considered the influence that other well established genres of writing will have on the average reader. Of particular relevance in this respect to the case before us is Loeb v. Globe Newspaper Co., 489 F. Supp. 481 (D.Mass.1980). In that case, the court observed that the article containing the alleged defamations of the publisher of the Manchester Union-Leader was situated on the Boston Globe's editorial page. The court held that, in the specific context or setting at issue there, the statement to the effect that Mr. Loeb never backed a winner in a presidential election was protected opinion. Plainly, the general understanding of the nature of the statements on the editorial page was relevant to the decision; if the statement had appeared on the front page where news is reported, it would most likely have been treated as a statement of fact. See also National Rifle Association v. Dayton Newspapers, Inc., 555 F. Supp. 1299 (S.D.Ohio 1983) (holding that the statement in an editorial that the National Rifle Association "happily encourages ... murders and robberies" was protected opinion). In short, it is well understood that editorial writers and commentators frequently "resort to the type of caustic bombast traditionally used in editorial writing to stimulate public reaction." Id. at 1309. Hence, in analyzing the distinction between fact and opinion, the court will take fully into account the different social conventions or customs inherent in different types of writing.27

After deciding that a particular statement is opinion rather than fact, courts often undertake a second mode of analysis before wrapping the statement in the mantle of the First Amendment's opinion privilege. Relying upon the Restatement (Second) of Torts Sec. 566, the courts consider whether the opinion implies the existence of undisclosed facts as the basis for the opinion.28 If the opinion implied factual assertions, courts have held that it should not receive the benefit of First Amendment protection as an opinion.

We have no quarrel with the purpose of section 566. As we have already seen, categorizing a statement as fact or opinion is a difficult task. Many statements are not simple factual statements or simple opinions, but are statements that are "ladden with factual content." Cianci, supra, 639 F.2d at 63. The Restatement is obviously designed to address the problems posed by such statements. In our view, however, the tests already articulated are a sufficient aid in determining whether a statement implies the existence of undisclosed facts. The definiteness and verifiability of a statement (factors one and two) clearly bear on the ability of a statement to carry factual implications. The linguistic and social context of the statement (factors three and four) will also influence the average reader's readiness to infer from the statement the existence of undisclosed facts. Thus, once our inquiry into whether the statement is an assertion of fact or expression of opinion has concluded, the factors militating either in favor of or against the drawing of factual implications from any statement have already been identified. A separate inquiry into whether a statement, already classified in this painstaking way as opinion, implies allegedly defamatory facts would, in our view, be superfluous. In short, we believe that the application of the four-factor analysis set forth above, and drawn from the considerable judicial teaching on the subject, will identify those statements so "factually laden" that they should not receive the benefit of the opinion privilege.29

We are fortified in this respect by section 566's potential, on occasion, to mislead. Comments to that section may be taken to imply that only the disclosure of facts which form the basis of the statement will signal to the reader that the author is not employing an opinion to imply undisclosed facts.30 To be sure, we fully agree that in some contexts statements should be submitted to the trier of fact, unless the presence of facts surrounding the statement suggests that it is merely a characterization of those facts and thus is best classified, like the characterization in Greenbelt Publishing, supra, as an opinion. For instance, in the context of a front page news story or magazine article, the presence of such facts may be the only factor that would prevent the allegedly defamatory statement from being submitted to the jury.

However, in other contexts, as we have shown above, factors besides the disclosure of facts are relevant in determining whether a statement implies factual allegations to the reasonable reader. Here, for instance, as we shall discuss more fully, that the statements challenged by Professor Ollman were found in a column on the Op-Ed page suggests, among other factors, that the statements would be understood by the reasonable reader as opinion--even in the absence of full disclosure of facts signalling to the reader that the allegedly defamatory statement was a characterization. In a word, disclosure of facts in the surrounding text is not the only signal that hard facts cannot reasonably be inferred from a statement. We think that our four-factor test takes account of the insights provided by section 566, while not rejecting the other factors that may signal that a statement is to be read as opinion.31III

Now we turn to the case at hand to apply the foregoing analysis. As we have seen, Mr. Ollman alleges various instances of defamation in the Evans and Novak column. Before analyzing each such instance, we will first examine the context (the third and fourth factors in our approach) in which the alleged defamations arise. We will then assess the manner in which this context would influence the average reader in interpreting the alleged defamations as an assertion of fact or an expression of opinion.

From the earliest days of the Republic, individuals have published and circulated short, frequently sharp and biting writings on issues of social and political interest. From the pamphleteers urging revolution to abolitionists condemning the evils of slavery, American authors have sought through pamphlets and tracts both to stimulate debate and to persuade. Today among the inheritors of this lively tradition are the columnists and opinion writers whose works appear on the editorial and Op-Ed pages of the Nation's newspapers. The column at issue here is plainly part and parcel of this tradition of social and political criticism.32

The reasonable reader who peruses an Evans and Novak 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. Readers expect that columnists will make strong statements, sometimes phrased in a polemical manner that would hardly be considered balanced or fair elsewhere in the newspaper. National Rifle Association v. Dayton Newspaper, Inc., supra, 555 F.Supp. at 1309. That proposition is inherent in the very notion of an "Op-Ed page." Because of obvious space limitations, it is also manifest that columnists or commentators will express themselves in condensed fashion without providing what might be considered the full picture. Columnists are, after all, writing a column, not a full-length scholarly article or a book. This broad understanding of the traditional function of a column like Evans and Novak will therefore predispose the average reader to regard what is found there to be opinion.33

A reader of this particular Evans and Novak column would also have been influenced by the column's express purpose. The columnists laid squarely before the reader their interest in ending what they deemed a "frivolous" debate among politicians over whether Mr. Ollman's political beliefs should bar him from becoming head of the Department of Government and Politics at the University of Maryland. Instead, the authors plainly intimated in the column's lead paragraph that they wanted to spark a more appropriate debate within academia over whether Mr. Ollman's purpose in teaching was to indoctrinate his students. Later in the column, they openly questioned the measure or method of Professor Ollman's scholarship. Evans and Novak made it clear that they were not purporting to set forth definitive conclusions, but instead meant to ventilate what in their view constituted the central questions raised by Mr. Ollman's prospective appointment. In the penultimate paragraph of the column, as we have already seen, the authors expressly posed the following "questions:"

What is the true measurement of Ollman's scholarship? Does he intend to use the classroom for indoctrination? Will he indeed be followed by other Marxist professors? Could the department in time become closed to non-Marxists, following the tendency at several English universities?

Prominently displayed in the Evans and Novak column, therefore, is interrogatory or cautionary language that militates in favor of treating statements as opinion.

* Having reviewed the context of the challenged statements, we turn next to the alleged defamation that, in our view, is most clearly opinion, namely that "[Ollman] is an outspoken proponent of political Marxism." This kind of characterization is much akin to the characterization, "fascist," found absolutely protected in Buckley v. Littell, supra. This is unmistakably a "loosely definable, variously interpretable statement[ ] of opinion ... made inextricably in the contest of political, social or philosophical debate ...." 539 F.2d at 895. It is obviously unverifiable. Since Mr. Ollman concedes that he is a Marxist, see Letter of B. Ollman to the Editors of The Washington Post (May 8, 1978), R. 3, the trier of fact in assessing the statement would have the dubious task of trying to distinguish "political Marxism" from "non-political Marxism," whatever that may be.

Nor is the statement that "[Mr. Ollman] is widely viewed in his profession as a political activist" a representation or assertion of fact. "Political activist" is a term, like "political Marxism," that is hopelessly imprecise and indefinite. It is difficult to imagine, much less construct, a means of deciding the quantum of political activity justifying the label "activist." While Mr. Ollman argues that this assertion is defamatory since it implies that he has no reputation as a scholar, we are rather skeptical of the strength of that implication, particularly in the context of this column. It does not appear the least bit evident that "scholarship" and "political activism" are generally understood to be incompatible. Moreover, Evans and Novak set out facts which signalled to the reader that this statement represents a characterization arising from the columnists' view of the facts. In the paragraph immediately following this statement, the column indicated that Mr. Ollman on no less than two occasions finished dead last among all candidates for election to the governing Council of the American Political Science Association, when he ran on the platform: "If elected ... I shall use every means at my disposal to promote the study of Marxism and Marxist approaches to politics throughout the profession." A reasonable reader would conclude that the authors' judgment that Mr. Ollman was "widely viewed as a political activist" was a characterization based upon the latter's unsuccessful electoral endeavors within his profession.

Next we turn to Mr. Ollman's complaints about the column's quotations from and remarks about his writings, and specifically his article, "On Teaching Marxism and Building the Movement."34 We note in this respect that even before the appearance of the constitutionally based opinion privilege in Gertz, commentary on another's writing was considered a privileged occasion at common law and therefore received the benefit of the fair comment doctrine.35 When a critic is commenting about a book, the reader is on notice that the critic is engaging in interpretation, an inherently subjective enterprise, and therefore realizes that others, including the author, may utterly disagree with the critic's interpretation.36 The average reader further understands that because of limitations of space, not to mention those limitations imposed by the patience of the prospective audience, the critic as a practical matter will be able to support his opinion only by rather truncated quotations from the book or work under scrutiny. The reader is thus predisposed to view what the critic writes as opinion. In this context, courts have rightly been wary of finding statements to be defamatory, unless the statements misquote the author, put words into the author's mouth or otherwise clearly go beyond the realm of interpretation.

Evans' and Novak's statements about Mr. Ollman's article clearly do not fall into the category of misquotation or misrepresentation. First, the plaintiff complains of the following statement: "Ollman concedes that [the fact that most students have a 'Marxist outlook' after taking his course] 'will be seen as an admission that the purpose of my course is to convert students to socialism.' " Tellingly, however, the quoted words are accurately reproduced from Mr. Ollman's article. See "On Teaching Marxism and Building the Movement" at 5. To be sure, the quotation has not been printed in its complete context.37 But that is neither here nor there; the quotation of remarks without the complete context in which the remarks appeared is entirely commonplace when summarizing a written work in a brief space. We are fully aware that this practice can be highly irritating when the context does not seem fully and fairly stated. The balm for the irritation, however, cannot be a libel suit, unless triers of fact are to sit in editorial judgment.38

Professor Ollman also objects to the column's posing the question, prompted in Evans' and Novak's view by Mr. Ollman's article, of whether he intended to use the classroom for indoctrination. As we noted previously, the column in no wise affirmatively stated that Mr. Ollman was indoctrinating his students. Moreover, indoctrination is not, at least as used here in the setting of academia, a word with a well-defined meaning. To paraphrase Justice Harlan in another context, see Cohen v. California, 403 U.S. 15, 25, 91 S. Ct. 1780, 1788, 29 L. Ed. 2d 284 (1971), what is indoctrination to one person is merely the vigorous exposition of ideas to another. We therefore conclude that the column's statements concerning "indoctrination" constitute protected opinion.

Mr. Ollman also complains of the statement: "His candid writings avow his desire to use the classroom as an instrument for preparing what he calls the 'revolution'." This statement, unlike the column's remarks about indoctrination, is stated without any interrogatory language to allow the reader to discount it as opinion. However, it is clear from the context that the statement represents Evans' and Novak's interpretation of Mr. Ollman's writing. And, like the charge of indoctrination, this statement does not have a well-defined meaning or admit of a method of proof or disproof. What to one person is a patently improper use of the classroom for political purposes may represent to another no more than the imparting of ideas, in the faith that ideas have consequences.

Finally, we turn to the most troublesome statement in the column.* In the third-to-last paragraph, an anonymous political science professor is quoted as saying: "Ollman has no status within the profession but is a pure and simple activist." The District Court interpreted this remark as a statement that Mr. Ollman "lack[ed] a reputation in his field as scholar."39 Memorandum Opinion at 5.

Certainly a scholar's academic reputation among his peers is crucial to his or her career. Like the peripatetic philosophers of ancient Greece, modern scholars depend upon their reputation to enable them to pursue their chosen calling. We also acknowledge that at least one pre-Gertz case has held that the common-law privilege of fair comment does not extend protection to remarks which disparage one's status among one's peers. See Cepeda v. Cowles Magazines and Broadcasting, Inc., 328 F.2d 869 (9th Cir.1964) (holding that, inter alia, remarks that a baseball player had "doghouse status" with the San Francisco Giants' hierarchy was not protected by fair comment).40

We are of the view, however, that under the constitutionally based opinion privilege announced in Gertz, this quotation, under the circumstances before us, is protected. A confluence of factors leads us to this conclusion. First, as we have stated, inasmuch as the column appears on the Op-Ed page, the average reader will be influenced by the general understanding of the functions of such columns and read the remark to be opinion.41 The identical quotation in a newspaper article purporting to publish facts or in an academic publication which purported to rate status within a given discipline would, of course, be quite another matter. But here we deal with statements by well-known, nationally syndicated columnists on the Op-Ed page of a newspaper, the well-recognized home of opinion and comment. In addition, the thrust of the column, taken as a whole, is to raise questions about Mr. Ollman's scholarship and intentions, not to state conclusively from Evans' and Novak's first-hand knowledge that Professor Ollman is not a scholar or that his colleagues do not regard him as such.

Moreover, the anonymous professor's unflattering comment appears only after the columnists expressly state that Mr. Ollman is a professor at New York University, a highly respected academic institution, a fact which provides objective evidence of Mr. Ollman's "status." So too, the controversy itself was occasioned by Professor Ollman's nomination by the departmental search committee as chairman of an academic department at the University of Maryland, a fact stated in the column's opening paragraph which also plainly suggested to the average reader that Professor Ollman did in fact enjoy some considerable status in academia. Finally in this regard, the column expressly states that Professor Ollman's imminent ascension to the departmental chairmanship at Maryland was troubling only to a clear minority of academics. Thus, the charge of "no status" in this context would plainly appear to the average reader to be "rhetorical hyperbole" within the meaning of Greenbelt, and which in turn would lead the reader to treat the statement as one of opinion.42

We note especially in this respect that the anonymous academician quoted in the column goes on to say that he would not repeat his charge publicly, stating that: "[o]ur academic culture does not permit the raising of such questions." Thus, while Mr. Ollman's critic is asserting a proposition about Mr. Ollman, he is simultaneously implying that, in the contemporary academic environment, no evidence can publicly be adduced to support it. Whether right or wrong, this admission by the anonymous political scientist would clearly tend to make the reader treat this proposition as opinion.43

But most fundamentally, we are reminded that in the accommodation of the conflicting concerns reflected in the First Amendment and the law of defamation, the deep-seated constitutional values embodied in the Bill of Rights require that we not engage, without bearing clearly in mind the context before us, in a Talmudic parsing of a single sentence or two, as if we were occupied with a philosophical enterprise or linguistic analysis. Ours is a practical task, with elemental constitutional values of freedom looming large as we go about our work. And in that undertaking, we are reminded by Gertz itself of our duty "to assure to the freedoms of speech and press that 'breathing space' essential to their fruitful exercise." Gertz, supra, 418 U.S. at 342, 94 S. Ct. at 3008. For the contraction of liberty's "breathing space" can only mean inhibition of the scope of public discussion on matters of general interest and concern. The provision of breathing space counsels strongly against straining to squeeze factual content from a single sentence in a column that is otherwise clearly opinion.44 As the Ninth Circuit so succinctly put it, "[t]he court must consider all the words used, not merely a particular phrase or sentence." Information Control Corp. v. Genesis One Computer Corp., supra, 611 F.2d at 784.45

The judgment of the District Court is therefore

Affirmed.

What is in danger of becoming a frivolous public debate over the appointment of a Marxist to head the University of Maryland's department of politics and government has so far ignored this unspoken concern within the academic community: the avowed desire of many political activists to use higher education for indoctrination.

The proposal to name Bertell Ollman, professor at New York University, as department head has generated wrong-headed debate. Politicians who jumped in to oppose Ollman simply for his Marxist philosophy have received a justifiable going-over from defenders of academic freedom in the press and the university. Academic Prinve [sic ] Valiants seem arrayed against McCarythite [sic ] know-nothings.

But neither side approaches the central question: not Ollman's beliefs, but his intentions. His candid writings avow his desire to use the classroom as an instrument for preparing what he calls "the revolution." Whether this is a form of indoctrination that could transform the real function of a university and transcend limits of academic freedom is a concern to academicians who are neither McCarthyite nor know-nothing.

To protect academic freedom, that question should be posed not by politicians but by professors. But professors throughout the country troubled by the nomination, clearly a minority, dare not say a word in today's campus climate.

While Ollman is described in news accounts as a "respected Marxist scholar," he is widely viewed in his profession as a political activist. Amid the increasingly popular Marxist movement in university life, he is distinct from philosophical Marxists. Rather, he is an outspoken proponent of "political Marxism."

He twice sought election to the council of the American Political Science Association as a candidate of the "Caucus for a New Political Science" and finished last out of 16 candidates each time. Whether or not that represents a professional judgment by his colleagues, as some critics contend, the verdict clearly rejected his campaign pledge: "If elected ... I shall use every means at my disposal to promote the study of Marxism and Marxist approaches to politics throughout the profession."

Ollman's intentions become explicit in "On Teaching Marxism and Building the Movement," his article in the Winter 1978 issue of New Political Science. Most students, he claims, conclude his course with a "Marxist outlook." Ollman concedes that will be seen "as an admission that the purpose of my course is to convert students to socialism."

That bothers him not at all because "a correct understanding of Marxism (as indeed of any body of scientific truths) leads automatically to its acceptance." Non-Marxists students are defined as those "who do not yet understand Marxism." The "classroom" is a place where the students' "bourgeois ideology is being dismantled." "Our prior task" before the revolution, he writes, "is to make more revolutionaries. The revolution will only occur when there are enough of us to make it."

He concludes by stressing the importance to "the movement" of "radical professors." If approved for his new post, Ollman will have a major voice in filling a new professorship promised him. A leading prospect is fellow Marxist Alan Wolfe; he is notorious for his book "The Seamy Side of Democracy," whose celebration of communist China extols the beneficial nature of "brainwashing."

Ollman's principal scholarly work, "Alienation: Marx's Conception of Man in Capitalist Society," is a ponderous tome in adoration of the master (Marxism "is like a magnificently rich tapestry"). Published in 1971, it does not abandon hope for the revolution forecast by Karl Marx in 1848. "The present youth rebellion," he writes, by "helping to change the workers of tomorrow" will, along with other factors, make possible "a socialist revolution."

Such pamphleteering is hooted at by one political scientist in a major eastern university, whose scholarship and reputation as a liberal are well known. "Ollman has no status within the profession, but is a pure and simple activist," he said. Would he say that publicly? "No chance of it. Our academic culture does not permit the raising of such questions."

"Such questions" would include these: What is the true measurement of Ollman's scholarship? Does he intend to use the classroom for indoctrination? Will he indeed be followed by other Marxist professors? Could the department in time be closed to non-Marxists, following the tendency at several English universities?

Even if "such questions" cannot be raised by the faculty, they certainly should not be raised by politicians. While dissatisfaction with pragmatism by many liberal professors has renewed interest in the comprehensive dogma of the Marxists, there is little tolerance for confronting the value of that dogma. Here are the makings of a crisis that, to protect its integrity and true academic freedom, academia itself must resolve.

BORK, Circuit Judge, with whom WILKEY and GINSBURG, Circuit Judges, and MacKINNON, Senior Circuit Judge, join, concurring:

While I concur in the judgment of the court and in much of Judge Starr's scholarly exposition, I write separately because I do not think he has adequately demonstrated that all of the allegedly libelous statements at issue here can be immunized as expressions of opinion. The dissents, on the other hand, while acknowledging the importance of additional factors, seem actually premised on the idea that the law makes a clear distinction between opinions, which are not actionable as libel, and facts, which are. In my view, the law as enunciated by the Supreme Court imposes no such sharp dichotomy. Some lower courts have assumed, as do some members of this court, not only that this opinion vs. fact formula is controlling but that it is governed, at least primarily, by grammatical analysis. I think that incorrect. Any such rigid doctrinal framework is inadequate to resolve the sometimes contradictory claims of the libel laws and the freedom of the press.

This case illustrates that point. It arouses concern that a freshening stream of libel actions, which often seem as much designed to punish writers and publications as to recover damages for real injuries, may threaten the public and constitutional interest in free, and frequently rough, discussion. Those who step into areas of public dispute, who choose the pleasures and distractions of controversy, must be willing to bear criticism, disparagement, and even wounding assessments. Perhaps it would be better if disputation were conducted in measured phrases and calibrated assessments, and with strict avoidance of the ad hominem; better, that is, if the opinion and editorial pages of the public press were modeled on The Federalist Papers. But that is not the world in which we live, ever have lived, or are ever likely to know, and the law of the first amendment must not try to make public dispute safe and comfortable for all the participants. That would only stifle the debate. In our world, the kind of commentary that the columnists Rowland Evans and Robert Novak have engaged in here is the coin in which controversialists are commonly paid.

These reflections lead me to conclude that Professor Ollman cannot press a libel action. But I do not find it easy to reach that result through a blunt distinction between opinion and fact, which while sometimes useful in just that crude dichotomy, is not adequate to the task here.

This inadequacy is most apparent in dealing with what Judge Starr calls "the most troublesome statement in the column," that concerning Ollman's reputation. It will be well to place the statement more completely in its context. Toward the end of their column, Evans and Novak say this:

Ollman's principal scholarly work, "Alienation: Marx's Conception of Man in Capitalist Society," is a ponderous tome in adoration of the master (Marxism "is like a magnificently rich tapestry"). Published in 1971, it does not abandon hope for the revolution forecast by Karl Marx in 1848. "The present youth rebellion," he writes, by "helping to change the workers of tomorrow" will, along with other factors, make possible "a socialist revolution."

Such pamphleteering is hooted at by one political scientist in a major eastern university, whose scholarship and reputation as a liberal are well known. "Ollman has no status within the profession, but is a pure and simple activist," he said. Would he say that publicly? "No chance of it. Our academic culture does not permit the raising of such questions."

Judge Starr's opinion for the majority contends that, in the circumstances of this case and in the context of the column as a whole, the quoted statement that "Ollman has no status within the profession, but is a pure and simple activist" qualifies as an opinion and so is constitutionally protected. The dissents, on the other hand, suggest that an assertion about one's general reputation is an assertion of fact. If common usage were the test, and if we looked at the sentence standing alone, the dissent's characterization would certainly be correct. The challenged language is a statement that others hold a particular opinion. Whether or not they do is a question of fact, though, as I will try to show, it is a "fact" of a peculiar nature in the context of first amendment litigation. If placing the bare assertion in question into one of two compartments labelled "opinion" and "fact" were the only issue we were allowed to consider, I would join the dissent. But I do not think these simple categories, semantically defined, with their flat and barren descriptive nature, their utter lack of subtlety and resonance, are nearly sufficient to encompass the rich variety of factors that should go into analysis when there is a sense, which I certainly have here, that values meant to be protected by the first amendment are threatened.

The temptation to adhere to sharply-defined categories is understandable. Judges generalize, they articulate concepts, they enunciate such things as four-factor frameworks, three-pronged tests, and two-tiered analyses in an effort, laudable by and large, to bring order to a universe of unruly happenings and to give guidance for the future to themselves and to others. But it is certain that life will bring up cases whose facts simply cannot be handled by purely verbal formulas, or at least not handled with any sophistication and feeling for the underlying values at stake. When such a case appears and a court attempts nevertheless to force the old construct upon the new situation, the result is mechanical jurisprudence. Here we face such a case, and it seems to me better to revert to first principles than to employ categories which, in these circumstances, inadequately enforce the first amendment's design.

Viewed from that perspective, the statement challenged in this lawsuit, in terms of the policies of the first amendment, is functionally more like an "opinion" than a "fact" and should not be actionable. It thus falls within the category the Supreme Court calls "rhetorical hyperbole." See pp. 975-79, infra. I will try to set out the factors in this case that justify application of that concept.

Because Evans and Novak wrote that an anonymous political science professor said he had "no status" among political scientists, Ollman wants to ask a jury to award him $1,000,000 in compensatory damages and an additional $5,000,000 in punitive damages. In the field of journalism, these are enormous sums. They are quite capable of silencing political commentators forever. Unless the defamation was heinous and devastating, the amounts sought are entirely disproportionate. No one would think it appropriate for a state to levy such amounts as fines upon writers for statements of the sort made here. But, under current doctrine, lower courts have no way of saying that such sums may not be sought in libel actions, Gertz v. Robert Welch, Inc., 418 U.S. 323, 94 S. Ct. 2997, 41 L. Ed. 2d 789 (1974), or, indeed, of saying that damages may not be awarded as punishment or that such components of compensation as psychological anguish are inconsistent with the first amendment when the libel occurs in a public, political dispute. Time, Inc. v. Firestone, 424 U.S. 448, 460, 96 S. Ct. 958, 968, 47 L. Ed. 2d 154 (1976). Instead, unless we continue to develop doctrine to fit first amendment concerns, we are remitted to old categories which, applied woodenly, do not address modern problems.

The American press is extraordinarily free and vigorous, as it should be. It should be, not because it is free of inaccuracy, oversimplification, and bias, but because the alternative to that freedom is worse than those failings. Yet the area in which legal doctrine is currently least adequate to preserve press freedom is the area of defamation law, the area in which this action lies. We are said to have in the first amendment "a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open." New York Times Co. v. Sullivan, 376 U.S. 254, 270, 84 S. Ct. 710, 721, 11 L. Ed. 2d 686 (1964). That principle has resulted in the almost total abolition of prior restraints on publication; New York Times Co. v. United States, 403 U.S. 713, 91 S. Ct. 2140, 29 L. Ed. 2d 822 (1971); Nebraska Press Association v. Stuart, 427 U.S. 539, 96 S. Ct. 2791, 49 L. Ed. 2d 683 (1976); the curtailment of the possibility of criminal sanctions; Garrison v. Louisiana, 379 U.S. 64, 85 S. Ct. 209, 13 L. Ed. 2d 125 (1964); and, in Sullivan itself, the construction of serious obstacles to private defamation actions by government officials. The cases that came afterward deployed similar obstacles to defamation actions by "public figures," Curtis Publishing Co. v. Butts, 388 U.S. 130, 87 S. Ct. 1975, 18 L. Ed. 2d 1094 (1967); Rosenbloom v. Metromedia, Inc., 403 U.S. 29, 91 S. Ct. 1811, 29 L. Ed. 2d 296 (1971); Gertz, 418 U.S. at 345, 94 S.Ct. at 3009. Thus, we have a judicial tradition of a continuing evolution of doctrine to serve the central purpose of the first amendment.

Judge Scalia's dissent implies that the idea of evolving constitutional doctrine should be anathema to judges who adhere to a philosophy of judicial restraint. But most doctrine is merely the judge-made superstructure that implements basic constitutional principles. There is not at issue here the question of creating new constitutional rights or principles, a question which would divide members of this court along other lines than that of the division in this case. When there is a known principle to be explicated the evolution of doctrine is inevitable. Judges given stewardship of a constitutional provision--such as the first amendment--whose core is known but whose outer reach and contours are ill-defined, face the never-ending task of discerning the meaning of the provision from one case to the next. There would be little need for judges--and certainly no office for a philosophy of judging--if the boundaries of every constitutional provision were self-evident. They are not. In a case like this, it is the task of the judge in this generation to discern how the framers' values, defined in the context of the world they knew, apply to the world we know. The world changes in which unchanging values find their application. The fourth amendment was framed by men who did not foresee electronic surveillance. But that does not make it wrong for judges to apply the central value of that amendment to electronic invasions of personal privacy. The commerce power was established by men who did not foresee the scope and intricate interdependence of today's economic activities. But that does not make it wrong for judges to forbid states the power to impose burdensome regulations on the interstate movement of trailer trucks. The first amendment's guarantee of freedom of the press was written by men who had not the remotest idea of modern forms of communication. But that does not make it wrong for a judge to find the values of the first amendment relevant to radio and television broadcasting.

So it is with defamation actions. We know very little of the precise intentions of the framers and ratifiers of the speech and press clauses of the first amendment. But we do know that they gave into our keeping the value of preserving free expression and, in particular, the preservation of political expression, which is commonly conceded to be the value at the core of those clauses. Perhaps the framers did not envision libel actions as a major threat to that freedom. I may grant that, for the sake of the point to be made. But if, over time, the libel action becomes a threat to the central meaning of the first amendment, why should not judges adapt their doctrines? Why is it different to refine and evolve doctrine here, so long as one is faithful to the basic meaning of the amendment, than it is to adapt the fourth amendment to take account of electronic surveillance, the commerce clause to adjust to interstate motor carriage, or the first amendment to encompass the electronic media? I do not believe there is a difference. To say that such matters must be left to the legislature is to say that changes in circumstances must be permitted to render constitutional guarantees meaningless. It is to say that not merely the particular rules but the entire enterprise of the Supreme Court in New York Times v. Sullivan was illegitimate.

We must never hesitate to apply old values to new circumstances, whether those circumstances are changes in technology or changes in the impact of traditional common law actions. Sullivan was an instance of the Supreme Court doing precisely this, as Brown v. Board of Education, 347 U.S. 483, 492-95, 74 S. Ct. 686, 690-92, 98 L. Ed. 843 (1954), was more generally an example of the Court applying an old principle according to a new understanding of a social situation. The important thing, the ultimate consideration, is the constitutional freedom that is given into our keeping. A judge who refuses to see new threats to an established constitutional value, and hence provides a crabbed interpretation that robs a provision of its full, fair and reasonable meaning, fails in his judicial duty. That duty, I repeat, is to ensure that the powers and freedoms the framers specified are made effective in today's circumstances. The evolution of doctrine to accomplish that end contravenes no postulate of judicial restraint. The evolution I suggest does not constitute a major change in doctrine but is, as will be shown, entirely consistent with the implications of Supreme Court precedents.

We now face a need similar to that which courts have met in the past. Sullivan, for reasons that need not detain us here, seems not to have provided in full measure the protection for the marketplace of ideas that it was designed to do. Instead, in the past few years a remarkable upsurge in libel actions, accompanied by a startling inflation of damage awards, has threatened to impose a self-censorship on the press which can as effectively inhibit debate and criticism as would overt governmental regulation that the first amendment most certainly would not permit. See Lewis, New York Times v. Sullivan Reconsidered: Time to Return to "The Central Meaning of the First Amendment," 83 Colum.L.Rev. 603 (1983).1 It is not merely the size of damage awards but an entire shift in the application of libel laws that raises problems for press freedom. See Smolla, Let the Author Beware: The Rejuvenation of the American Law of Libel, 132 U.Pa.L.Rev. 1 (1983).2 Taking such matters into account is not, as one dissent suggests, to engage in sociological jurisprudence, at least not in any improper sense. Doing what I suggest here does not require courts to take account of social conditions or practical considerations to any greater extent than the Supreme Court has routinely done in such cases as Sullivan. Nor does analysis here even approach the degree to which the Supreme Court quite properly took such matters into account in Brown, 347 U.S. at 492-95, 74 S.Ct. at 690-92. Matters such as the relaxation of legal rules about permissible recovery, the changes in tort law to favor compensation, and the existence of doctrinal confusion, see Smolla,supra, are matters that courts know well. Indeed, courts are responsible for these developments.

The only solution to the problem libel actions pose would appear to be close judicial scrutiny to ensure that cases about types of speech and writing essential to a vigorous first amendment do not reach the jury.3 See Bose Corp. v. Consumers Union of United States, Inc., --- U.S. ----, 104 S. Ct. 1949, 1965, 80 L. Ed. 2d 502 (1984). This requires a consideration of the totality of the circumstances that provide the context in which the statement occurs and which determine both its meaning and the extent to which making it actionable would burden freedom of speech or press. That, it must be confessed, is a balancing test and risks admitting into the law an element of judicial subjectivity. To that objection there are various answers. A balancing test is better than no protection at all. Given the appellate process, moreover, the subjective judgment of no single judge will be controlling. Over time, as reasons are given, the element of subjectivity will be reduced. There is, in any event, at this stage of the law's evolution, no satisfactory alternative. Hard categories and sharply-defined principles are admirable, if they are available, but usually, in the world in which we live, they share the problem of absolutes, of which they are a subgenre: they do not stand up when put to the test of hard cases. In the process of "balancing," I will state my reasons fully so that it may be judged whether they are rooted adequately in central first amendment concerns and so that guidance may be given as to how I think cases should be decided in the future.

Two general considerations lead me to conclude that Professor Ollman should not be allowed to try his case to a jury. First, the state of doctrine in this area, if not precisely embryonic, is certainly still developing. Nothing in case law that is binding upon this court requires us to ignore context and the purposes of the first amendment and, instead, to apply a rigid opinion-fact dichotomy and to define the compartments of that dichotomy by semantic analysis. Indeed, the Supreme Court has indicated that we are not to do that. See pp. 975-79, infra. We are required, therefore, to continue the evolution of the law in accordance with the deepest rationale of the first amendment. Second, the central concerns of the first amendment are implicated in this case so that a damage award would have a heavily inhibiting effect upon the journalism of opinion. On the other hand, the statement challenged, in practical impact, is more like an expression of opinion than it is like an assertion of fact. It is the kind of hyperbole that must be accepted in the rough and tumble of political argument.

It is plain, I think, that the opinion-fact dichotomy is not as rigid as the various dissents suppose. There is no need to become caught up in a debate about the true nature of the allegedly libelous statement in terms of that dichotomy. The formalistic distinction between the two would be binding on us, sitting as an en banc court, only if the Supreme Court had required it. The thought that the Supreme Court has required it rests upon what I believe to be a misapprehension of dicta in Gertz v. Robert Welch, Inc., 418 U.S. 323, 94 S. Ct. 2997, 41 L. Ed. 2d 789. The facts of that case are important, if only by contrast with other cases, to an understanding of still-evolving doctrine in this area. Plaintiff Gertz was a lawyer who represented the family of a youth killed by a policeman in civil litigation against the policeman. In his capacity as counsel, Gertz attended the coroner's inquest but otherwise did nothing more than press the civil suit. The defendant, which published a monthly magazine, ran an article that portrayed Gertz as "an architect of the 'frame-up' " against the police officer, implied that Gertz had a lengthy criminal record, called him a "Leninist" and a "Communist-fronter," and identified him as an official of an organization that advocated violent seizure of the government. 418 U.S. at 326, 94 S.Ct. at 3000. None of this was true. The Court introduced its discussion of the governing considerations with an observation that was not necessary to the decision:

We begin with the common ground. 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. But there is no constitutional value in false statements of fact. Neither the intentional lie nor the careless error materially advances society's interest in "uninhibited, robust, and wide-open" debate on public issues. New York Times Co. v. Sullivan, 376 U.S., at 270 [84 S.Ct. at 721]. They belong to that category of utterances which "are no essential part of any exposition of ideas, and are of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality." Chaplinsky v. New Hampshire, 315 U.S. 568, 572 [62 S. Ct. 766, 769, 86 L. Ed. 1031] (1942).

Id., 418 U.S. at 339-40, 94 S. Ct. at 3007 (footnote omitted).

In Gertz, it was obvious that most of the assertions that were the subject of the action purported to be flat statements of fact. The two statements that might arguably have been statements of opinion were that Gertz was a "Leninist" and a "Communist-fronter." 418 U.S. at 326, 94 S. Ct. at 3000. The Court did not discuss their proper categorization. But as Judge Friendly said in Cianci v. New Times Publishing Co., 639 F.2d 54, 61 (2d Cir.1980), these assertions must have been "deemed sufficiently 'factual' to support an action for defamation," since the Supreme Court remanded the case for jury trial.

For this reason, it is instructive to compare the Court's treatment of an even more clearly "factual" assertion in Greenbelt Cooperative Publishing Association v. Bresler, 398 U.S. 6, 90 S. Ct. 1537, 26 L. Ed. 2d 6 (1970). Plaintiff Bresler, a real estate developer and builder, engaged in negotiations with the City Council of Greenbelt, Maryland, for zoning variances so that he could build high-density housing on land he owned. Simultaneously, the city was trying to acquire another tract of land from Bresler to construct a high school. The concurrent negotiations gave each side bargaining leverage. Bresler, of course, could vary the price for the tract depending on the city's attitude toward the variances. A newspaper accurately reported the public debate at city council meetings at which Bresler's negotiating demands were denounced as "blackmail." Bresler sued, alleging that the articles imputed a crime to him. The Court held that this denunciation was a constitutionally protected statement since here the word "blackmail" was no more than "rhetorical hyperbole, a vigorous epithet used by those who considered Bresler's negotiating position extremely unreasonable." Id., 398 U.S. at 14, 90 S. Ct. at 1542. The context in which the words appeared was such that no reader could have thought that Bresler was charged with a crime.

The analytical approach of Bresler was reaffirmed in Old Dominion Branch No. 496, National Association of Letter Carriers v. Austin, 418 U.S. 264, 285-86, 94 S. Ct. 2770, 2781-82, 41 L. Ed. 2d 745 (1974), a case argued and handed down on the same days as Gertz. In Letter Carriers, a union newsletter, Carrier's Corner, published the names of those, including plaintiffs, who had not joined the union under the heading "List of Scabs." Just above the list the newsletter printed a particularly derogatory definition of the term "scab" attributed to Jack London which included the statement that a scab was "a traitor to his God, his country, his family and his class."4 The Court quoted the reasoning of Bresler about the meaning imparted by context and then said:

It is similarly impossible to believe that any reader of the Carrier's Corner would have understood the newsletter to be charging the appellees with committing the criminal offense of treason. As in Bresler, Jack London's "definition of a scab" is merely rhetorical hyperbole, a lusty and imaginative expression of the contempt felt by union members towards those who refuse to join.

418 U.S. at 285-86, 94 S. Ct. at 2782 (footnote omitted).

A comparison of Gertz, on the one hand, with Bresler and Letter Carriers, on the other, indicates the actual state of the law. The fact that the epithets "Leninist" and "Communist-fronter" were deemed actionable, while the epithets "blackmail," "scab," and "traitor" were not, demonstrates that, when it comes to first amendment analysis, the Supreme Court does not employ a simplistic opinion-fact dichotomy. A statement that, on its face and standing alone, sounds like an assertion of fact may not be actionable. Context is crucial and can turn what, out of context, appears to be a statement of fact into "rhetorical hyperbole," which is not actionable. Thus, it is clear that the Supreme Court, in the service of the first amendment, employs a test which requires consideration of the totality of the circumstances in which a statement appears.5

Courts other than the Supreme Court agree that context may make non-actionable statements that are facially assertions of fact. Thus, the Ninth Circuit has said that "even apparent statements of fact may assume the character of statements of opinion, and thus be privileged, when made in public debate, heated labor dispute, or other circumstances in which an 'audience may anticipate efforts by the parties to persuade others to their positions by use of epithets, fiery rhetoric or hyperbole.' " Information Control Corp. v. Genesis One Computer Corp., 611 F.2d 781, 784 (9th Cir.1980). Moreover, "the test to be applied in determining whether an allegedly defamatory statement constitutes an actionable statement of fact requires that the court examine the statement in its totality in the context in which it was uttered or published." Id. at 784. It is not unusual to protect false statements of fact where, because of the context, they would have been understood as part of a satire or fiction. In Myers v. Boston Magazine Co., 380 Mass. 336, 403 N.E.2d 376 (1980), a magazine called the plaintiff the "worst" sports announcer in Boston and stated that he was "enrolled in a course for remedial speaking." Holding that the distinction between opinion and fact is a question of law, the court said the statement, in context, was one of opinion and would reasonably be understood to suggest that the plaintiff should have been enrolled in such a course. 403 N.E.2d at 379. The remarks about plaintiff appeared in a series of categorizations of various people as the best and worst in their fields. As the court noted, the "pervasive mood" was one of "rough humor." Id., 403 N.E.2d at 377. See Pring v. Penthouse International, Ltd., 695 F.2d 438, 443 (10th Cir.), cert. denied, --- U.S. ---- 103 S. Ct. 3112, 77 L. Ed. 2d 1367 (1983) (in fictional account false statement of facts constitutionally protected as "obviously a complete fantasy").6

I trust I have said enough to demonstrate that in Supreme Court decisions and the decisions of other courts there is no mechanistic rule that requires us to employ hard categories of "opinion" and "fact"--defined by the semantic nature of the individual assertion--in deciding a libel case that touches upon first amendment values.7 We must turn instead to the totality of the circumstances of the case to determine whether a statement may be actionable.

There are several factors that convince me Ollman cannot maintain this action. These considerations are of the type that the Supreme Court and other courts have deemed important: the danger to first amendment freedoms and the functional meaning of the challenged statement as shown by its context and its qualities as recognizable rhetorical hyperbole. The factors here are: Ollman, by his own actions, entered a political arena in which heated discourse was to be expected and must be protected; the "fact" proposed to be tried is in truth wholly unsuitable for trial, which further imperils free discussion; the statement is not of the kind that would usually be accepted as one of hard fact and appeared in a context that further indicated it was rhetorical hyperbole.

Plaintiff Ollman, as will be shown, placed himself in the political arena and became the subject of heated political debate. That fact has significance in two ways. The first, and more conventional, point is that the existence of a political controversy is part of the total context that gives meaning to statements made about Ollman. When we read charges and countercharges about a person in the midst of such controversy we read them as hyperbolic, as part of the combat, and not as factual allegations whose truth we may assume. It will be seen, as the events are recounted, how true that is in Ollman's case.

My second point is less conventional, though by no means ruled out by case law as a next step in the evolution of doctrine in this troubling field. It is this: in order to protect a vigorous marketplace in political ideas and contentions, we ought to accept the proposition that those who place themselves in a political arena must accept a degree of derogation that others need not. Because this would represent a further development of the law I have argued it more fully than the first point. But it is not necessary to accept this proposition in order to accept the first point, that political controversy is part of the context that tends to show that some apparently factual assertions should be treated as rhetorical hyperbole and hence as opinions.

It is common ground that the core function of the first amendment is the preservation of that freedom to think and speak as one pleases which is the "means indispensable to the discovery and spread of political truth." Whitney v. California, 274 U.S. 357, 375, 47 S. Ct. 641, 648, 71 L. Ed. 1095 (1927) (Brandeis, J., concurring). Necessary to the preservation of that freedom, of course, is the willingness of those who would speak to be spoken to and, as in this case, to be spoken about. This is not always a pleasant or painless experience, but it cannot be avoided if the political arena is to remain as vigorous and robust as the first amendment and the nature of our polity require.

In deciding a case like this, therefore, one of the most important considerations is whether the person alleging defamation has in some real sense placed himself in an arena where he should expect to be jostled and bumped in a way that a private person need not expect. Where politics and ideas about politics contend, there is a first amendment arena. The individual who deliberately enters that arena must expect that the debate will sometimes be rough and personal. This would not be true of a political scientist who confined himself to academic pursuits and eschewed political proselytizing. Such a person might legitimately expect that, should columnists for some reason become interested in him, any criticism levelled would stick close to his work, and that, if assertions were made about his reputation, they would be actionable if false.

But Ollman has, as is his undoubted right, gone well beyond the role of the cloistered scholar, and he did so before Evans and Novak wrote about him. As the column recounts, and its literal accuracy in these respects is not challenged, Professor Ollman was an active proponent not just of Marxist scholarship but of Marxist politics. He wrote an article called "On Teaching Marxism and Building the Movement," which asserted that his classroom was a place where the students' "bourgeois ideology is being dismantled," that his endeavor was to "make more revolutionaries," and that "radical professors" are important to "the movement." His book approved the "youth rebellion" as helping make possible "a socialist revolution." Twice he put himself forward for election to the council of the American Political Science Association, campaigning on the promise that, "If elected ... I shall use every means at my disposal to promote the study of Marxism and Marxist approaches to politics throughout the profession." It was plain that Ollman was a political activist and that he saw his academic post as, among other things, a means of advancing his political goals. This is controversial behavior for an academic, no matter what political creed he espoused, and was bound to raise for debate the question whether he used his position as a teacher to indoctrinate the young with his political beliefs.

It was thus inevitable that when Ollman, who was a political figure, put himself forward as a candidate for the chairmanship of the department of politics and government at the University of Maryland there would be a public political controversy. But more took place, both upon Ollman's initiative and the initiatives of others, that confirmed his status as a figure in a political arena before the Evans and Novak column appeared.

A hot public controversy erupted the day after Ollman's nomination for the chairmanship of the department was disclosed. Among the participants in the dispute, which was extensively covered by the news media, were the Republican Acting Governor of Maryland, two members of the university's board of regents, a state senator, a member of the Prince George's County council, the associate general secretary of the American Association of University Professors, the Washington Post columnist Richard Cohen, and the three Democratic candidates for governor.8 Ollman's nomination thus became an issue in the 1978 Maryland gubernatorial race. The debate about his nomination and politics received nationwide press coverage.

In the midst of this controversy, Ollman announced that he had begun to market a new board game called "Class Struggle," which he said he had been working on for seven years. He said, "This game will give our people [a] view of how our society works, and for whom." Players representing workers moved a little hammer around the board; those representing capitalists moved a little top hat. Players moved to the final confrontation--revolution. " 'Not a violent overthrow,' Ollman emphasized, 'but a structural change.' " The Washington Post, Apr. 28, 1978. The Evans and Novak column appeared on May 4.

The president of the university rejected Ollman's appointment, and The Washington Post, in an editorial generally critical of the decision, said: "A teacher's politics may be his own business, but it becomes a legitimate criterion by which to judge his appointment when it calls into question his classroom intentions. In recent weeks, Mr. Ollman's public statements have not made his case more appealing. To many, his remarks have suggested that he is in fact more interested in polemics than in political science." The Washington Post, July 23, 1978, at C 6, col. 1.

The important point about all of this is that Ollman was not simply a scholar who was suddenly singled out by the press or by Evans and Novak. Whatever the merits of his scholarship, he was also a political man who publicly tried to forward his political goals. He had entered the political arena before he put himself forward for the department chairmanship. That candidacy merely widened the area within which he was known and raised for debate a topic of legitimate political concern, a debate which his further actions fueled. That being so, he must accept the banging and jostling of political debate, in ways that a private person need not, in order to keep the political arena free and vital.

Ollman may not be required to accept the same degree of buffeting that a candidate for a major office must, but when he chose to become a spokesman for Marxism to be implemented politically, when he stated that his teaching effectively converted students to Marxism, when he stated that he wanted to spread Marxist approaches to politics throughout a profession of teachers and writers, when he stated that he favored revolution by structural change, when he marketed a game designed to teach the general public about class struggle, and when he stood for an office that would extend his influence over teaching and writing, and hence over the development of the political views of the young--when Professor Ollman chose that path he became a figure in whom the public might legitimately be interested, and about whose intentions and professional status public questions might legitimately be raised. In a word, when he did those things, Ollman entered a first amendment arena and had to accept the rough treatment that arena affords.

The concept of the public, political arena that I have employed has at least some of the same functional characteristics as the concept of a person who is a public figure for limited purposes. That similarity may prompt the objection that the public figure concept applies only to distinguish between negligence and actual malice for purposes of liability. That is, of course, an accurate statement of current doctrine, but I know of no case holding that the concept may not be put to the use proposed, to assist in deciding how much public bumping a person must accept as a risk of the controversies he chooses to engage in.

Two of the dissenting opinions (Wald and Scalia, JJ.) maintain that commentary about public figures is already adequately protected by the actual malice requirement of New York Times v. Sullivan. According to this view, there is no reason to go beyond Sullivan and accord greater first amendment protection to some false political statements made knowingly and with actual malice. But the Supreme Court has already placed the law in precisely the posture to which the dissent objects. Gertz, of course, means that a statement characterized as an opinion cannot be actionable even if made with actual malice and even if it severely damages the person discussed. In such circumstances, society must depend upon the competition of ideas to correct pernicious opinions rather than on "the conscience of judges and juries."

Bresler and Letter Carriers make the point even clearer. In both, apparent factual assertions--in Bresler that plaintiff engaged in "blackmail"; in Letter Carriers that plaintiffs were "scabs" and "traitors"--were held not actionable because, in context, the reader would take them not as assertions of fact but as vigorous hyperbole. In neither case did the Court inquire about actual malice. It assumed that even if these statements were made with actual malice, they were protected because the context in which they appeared alerted the reader that the statements were not to be read as factual allegations. Thus, the Supreme Court has obviously recognized that the actual malice requirement of Sullivan does not always provide adequate protection and the Court has provided the additional protection that the first amendment requires.

In this respect, I am doing no more than following Supreme Court precedent. As I said at the outset of this subsection, part of the context here is the existence of a vigorous political controversy that Ollman himself fueled and which conditions the way a reader understands the kind of charge that Evans and Novak related.

Judge Wald's dissent objects that making the distinction between a person who has stepped into the political arena and one who has not is a task too baffling for judges. The answer is that this is exactly the task that judges must perform in deciding whether a person has become a public figure.

But I have suggested, though it is not essential to my result, that the law consider the existence of political controversy and the concept of a political arena in an additional way. That concept could be used to set a kind of de minimis level for rough statements about persons who enter a first amendment arena and become, in essence, public figures for limited purposes. This is a different spectrum from that of the actual malice-negligence distinction but surely one to which the concept of a public figure or a political individual is relevant. Indeed, though the law has not yet had occasion to consider this point, Americans have a kind of common understanding or social usage that runs along these lines. The United States has just been through an intense political campaign. In this highly charged atmosphere, many cruel and damaging things were said about various candidates for major political offices. Some of the statements made may well meet the law's standards for actual malice--reckless disregard for the truth of the matter asserted. Examples will no doubt spring to mind. Yet if the statement is of the sort that we recognize as rhetorical hyperbole, we would be astonished and highly disapproving if the defamed candidate brought an action for libel. We expect people who engage in controversy to accept that kind of statement as their lot. We think the first amendment demands a hide that tough. As I have said, Ollman may not be required to be as thick-skinned as a candidate for major political office but, as a political man, he shares some of the same responsibility. I do not say that this point alone is sufficient to decide the case, but it weighs, and, I think, weighs heavily, on the side of holding the statement not actionable.

But, in any event, it is indisputable that this swirling public debate provided a strong context in which charges and countercharges should be assessed. In my view, that context made it much less likely that what Evans and Novak said would be regarded as an assertion of plain fact rather than as part of the judgments expressed by each side on the merits of the proposed appointment.

Particularly troubling in a first amendment context is the kind of fact that is proposed for trial and, on either side's demand, jury determination. Here it is well to recur to one of the functions of the rough division between opinions and facts. It is relatively easy to litigate a false statement of fact; it may be impossible to prove or disprove an opinion. Courts of law may reasonably limit their dockets to questions which they are competent to resolve. Accordingly, the opinion-fact division serves a purpose by confining the category of actionable statements to those which lend themselves to competent judicial resolution of the truthfulness of their content. Viewed from that juridical perspective, the statement in question here is qualitatively more like an opinion than a fact. It is simply not fit for jury determination.

The evidence is mounting that juries do not give adequate attention to limits imposed by the first amendment and are much more likely than judges to find for the plaintiff in a defamation case. It is appropriate for judges, therefore, to take cases from juries when they are convinced that a statement ought to be protected because, among other reasons, the issue it presents is inherently unsusceptible to accurate resolution by a jury. As the Supreme Court said in Bose Corp. v. Consumers Union of United States, Inc., --- U.S. ----, 104 S. Ct. 1949, 80 L. Ed. 2d 502, appellate courts must independently examine the record in first amendment cases to ensure that constitutional values are not endangered. "The requirement of independent appellate review ... reflects a deeply held conviction that judges--and particularly members of [the Supreme] Court--must exercise such review in order to preserve the precious liberties established and ordained by the Constitution." Id., 104 S.Ct. at 1965. The underlying principle, it seems to me, requires judges to decide when allowing a case to go to a jury would, in the totality of the circumstances, endanger first amendment freedoms. That danger is overwhelming when the issue is of the sort presented here.

The issue the dissents would have tried--the political science academic community's opinion of Professor Ollman's stature as a political scientist--is inherently incapable of being adjudicated with any expectation of accuracy. One dissent (Wald, J.) suggests that "[o]ne could, for instance, devise a poll of American Political Science Association members as to their opinion, on a scale of one to ten, of the scholarly value of Ollman's work. Testimony of prominent political scientists or other measures of reputation would also serve to verify or refute the statement about Ollman's reputation without sending the jury onto a sea of speculation." But this suggestion is itself abstract speculation. Some element of realism is necessary in these matters. Let us try to imagine the nature of the trial and what the jury could make of such evidence.

As every presidential campaign reminds us, there is a great spread in the results of public opinion polls, even in the results of polls taken at the same time by a number of reputable and experienced polling organizations. There are scientific and professional disputes about polling methodology, about the representativeness of the sample or of those who respond to the questionnaire, since it is often true that those who respond have markedly different views from those who do not respond. (The problems of sampling, as will be seen, are very much present with a group whose members are as disparate as political scientists.) There are disputes about the phrasing and the order of the questions put, and whether such matters skewed the results. Indeed, if the column's assertion about his status among academics harmed Ollman's status among academics, the poll would be seriously biased. (If there was no such harm, of course, Ollman would not have much of a case.) All of these disputes would occur about the poll suggested by the dissent, and would be tried with experts in statistics, psychology, and perhaps other disciplines offering the jury conflicting scientific arguments. Perhaps both the plaintiff and defendants would devise and send out questionnaires so that the jury, weighing scientific arguments about which experts cannot agree, would have to decide which poll was the more methodologically sound. I do not think the results of a trial on issues like these could be anything but random and, whatever we might be willing of necessity to allow in a different kind of trial, I would be utterly unwilling to let first amendment freedoms ride upon an outcome determined by chance.

Let us suppose, however, that the jury chooses one poll as methodologically more acceptable than the other. And let us suppose that the results show that most of the scores awarded Ollman range between 2 and 7, with a scattering of 1's and 10's, and a mean of 3.5 and a median of 4. What on earth is a jury to make of that? That Ollman has high status?, that he has low status but not "no status"? If low status, is that close enough to "no status" to afford the statement of "no status" protection as permissible hyperbole? It is not at all clear what the term "no status" connotes. The term is so vague as to suggest little more than general, but not necessarily universal, disapproval. Thus, if the profession were sharply divided so that a fifth of those responding ranked Ollman at 8 and the remainder ranked him at 1, would the jury be permitted to find that, in effect, showed "no status" or would it be instructed that any favorable opinion showed "some status" so that the column's statement was one of false fact?

How is the jury, or an appellate court, to know whether knowledge that the poll was for use in a lawsuit skewed the results? The controversy and this case are widely known, especially among academic political scientists. But the professors who fill out the questionnaires will not be available for examination. Indeed, in order to avoid one kind of bias, they would have to be promised anonymity. How are we to know whether the political stance of the combatants--that Ollman is a Marxist and Evans and Novak are generally regarded as conservatives--skewed the results? Indeed, must not the ideological coloration of the entire political science academic profession become an issue for the jury in evaluating the poll? If that community is conservative, would they rank Ollman lower for purposes of a lawsuit against Evans and Novak than their real estimate of his professional qualities? If that community leans to the left, would its members, for similar political reasons, rank him higher? Would not the investigation into opinions about Ollman necessarily include an investigation of the political opinions of the relevant academic community?

Matters are really worse than this, however. Academic political scientists number in the tens of thousands. With the exception of a few very prominent persons, the quality of no one's work is known throughout the profession. The profession is fragmented and contains many subsets. Knowledge of a professor's work is likely to be confined to one or a few such subsets. Thus, political scientists who view themselves as devoted to value-free empirical studies are unlikely to have any informed estimate of the work done by most persons working in political philosophy. More than this, we are not talking about opinions concerning the professional credentials of a faculty member in the school of engineering or medicine, fields in which ideology plays little or no part in estimations of status. We are talking about an academic specialty which, as anyone remotely familiar with it knows, is politically highly charged and riven. Political outlook may color professional estimation. In this field there are varieties of liberals, conservatives, libertarians, Marxists, and Straussians. Suppose, to put a not wholly unreasonable hypothetical, that on the questionnaire the dissent proposes, Ollman received 9's and 10's from Marxists and 1's and 2's from Straussians. It may be doubted that either set of numbers has any significance that a jury should be entitled to consider. If views of professional status are colored or determined by political or philosophical agreement or disagreement, is that the "status" we are interested in? Presumably, if Ollman has been defamed, it is in relation to a more objective, or less political, status. At least, he puts the matter that way. See p. 1010, infra.

The suggestion that reputation could be verified by the testimony of prominent political scientists cures none of this. If prominent political scientists could be induced to testify, and if those who could be induced represented a fair cross-section of the academic community, both heroic assumptions, the jury would be left with contradictory opinions about opinions. I do not know how the jury could reach any informed judgment unless it were told that any opinion favorable to Ollman meant that the allegation of "no status" was false.

The problem of trying academic reputation to a jury is very similar to the problem a faculty faces when it tries to determine whether to vote to award tenure to a candidate. Judge Winter, himself a veteran of tenure debates, described the situation in Zahorik v. Cornell University, 729 F.2d 85 (2d Cir.1984):

[T]enure decisions are a source of unusually great disagreement. Because the stakes are high, the number of relevant variables is great and there is no common unit of measure by which to judge scholarship, the dispersion of strongly held views is greater in the case of tenure decisions than with employment decisions generally.... [A]rguments pro and con are framed in largely conclusory terms which lend themselves to exaggeration, particularly since the stauncher advocates on each side may anticipate and match an expected escalation of rhetoric by their opponents. Moreover, disagreements as to individuals may reflect long standing and heated disputes as to the merits of contending schools of thought or as to the needs of a particular department.... [A] file composed of irreconcilable evaluations is not unusual.

Id. at 93.

I can testify that this description is accurate, though perhaps understated. The faculty member who has not read the candidate's publications himself and formed his own judgment is helpless before the impressive, well-documented but diametrically opposed arguments of others. The jury would certainly be in a far worse position to judge.9

Academic reputation, in short, seems to me peculiarly unsuited to a trial at law unless the person in question is one of the few universally acknowledged throughout the profession to be a major figure. Ollman is not claimed to be that. This concern may or may not be weighty enough by itself to deny Ollman access to the jury. I tend to think it may be. But I need not decide that because the points I am making are intended to be cumulative and this point certainly goes to the question of the degree of risk we are willing to impose upon the exercise of political comment.

The statement of "no status" is very unlikely to be read as a flat statement of fact. Rather, it strikes the reader primarily as an exaggerated expression of the anonymous professor's own view of Ollman's academic credentials. It is wrong to speak as though there is always a sharp distinction between opinion and fact. There certainly is at the extremes an obvious difference in kind. The assertion that "Jones stole $100 from the church poor box last Friday night," cannot be tortured into an opinion, just as the assertion that "I think Jones is the kind of man who would steal from the church poor box" is obviously only a statement of the speaker's opinion of Jones' character. But the statement that "Half the people in this town think Jones is the kind of man who would steal from the poor box" is not quite like either of the first two. It is less harmful than the first and perhaps more damaging than the second. I say "perhaps" because the assertion of what others think always has a ring of hyperbole about it. The hearer knows that what he is being told is, in fact, one man's opinion about others' opinions. It can be called an assertion of fact, which in a sense it is, but it is also the kind of criticism that we are used to hearing and about which we regularly suspend judgment. Told by Smith that Jones actually stole the money, we think that Smith would not dare say such a thing if it were not so. There is a hard quality to the statement: it is capable of proof or disproof and it describes a physical action that did or did not take place. Told by Smith that half the town thinks Jones is the kind of fellow who would steal the money, we instantly discount it as an expression of Smith's antipathy to Jones. We think it may or may not be so and we realize that there is very little chance of verifying the truth of the assertion as made.

So it is here with the statement that Ollman has no status within the profession of political scientists. It is one man's impression or opinion relayed by Evans and Novak. The reader does not accept it as a concrete fact. He understands that the speaker thinks poorly of Ollman. He gathers that Ollman is a controversial figure within the profession, which certainly appears to be true. Indeed, the column contains information from which the reader might draw the same conclusion even if Evans and Novak had not made it explicit. Earlier than the passage under discussion, the column stated:

He [Ollman] twice sought election to the council of the American Political Science Association as a candidate of the "Caucus for a New Political Science" and finished last out of 16 candidates each time. Whether or not that represents a professional judgment by his colleagues, as some critics contend, the verdict clearly rejected his campaign pledge: "If elected ... I shall use every means at my disposal to promote the study of Marxism and Marxist approaches to politics throughout the profession."

The results of these two elections would certainly appear to be a rejection of Ollman's campaign pledge, and the fact that he made the pledge coupled with the results of the two elections certainly give grounds for supposing that Ollman is an "activist" and that his stature in the profession, or in important segments of the profession, might well be low. Indeed, the column contains accurate quotations from Ollman's writings that would strongly suggest such an assessment, by some members of the profession, might be likely. I have already rehearsed these in connection with Ollman's status as a political actor.

This raises the question of what academic reputation or status is. Men and women engaged in academic life are judged by colleagues on various scales of values. That fact might prove troublesome at trial. But Ollman, interestingly enough, advances a quite conventional standard by which status should be judged: "Plaintiff's occupation is that of scholar and teacher. It is commonly expected that a person in that position will be open-minded and fair-minded, will not attempt to indoctrinate students, and will seek the truth through research and testing and will communicate the results of his search by means of publications which adhere to certain objective canons of scholarship." Brief for Appellant at 6. If the ideal of the scholar seeking truth dispassionately is the standard, as most lay readers of newspapers undoubtedly believe that it is, then the column's quotations from his writings and from his electioneering statements, as well as his own public statements about, and the marketing of, his board game, Class Struggle, indicate that he has upon more than one occasion significantly departed from it. Thus, the anonymous professor's remark that Ollman had "no status" would be taken as a comment upon what the column and the news stories had already revealed.

When we come to the context in which this statement occurred, it becomes even more apparent that few people were likely to perceive it as a direct assertion of fact, to be taken at face value. That context was one of controversy and opinion, and it is known to be such by readers. It is significant, in the first place, that the column appeared on the Op-Ed pages of newspapers. These are pages reserved for the expression of opinion, much of it highly controversial opinion. That does not convert every assertion of fact on the Op-Ed pages into an expression of opinion merely by its placement there. It does alert the reader that he is in the context of controversy and politics, and that what he reads does not even purport to be as balanced, objective, and fair-minded as he has a right to hope to be the case with what is contained in the news columns of the paper. The Op-Ed pages are known to be a forum for controversy, often heated controversy, analogous in many respects to the context of a labor dispute. The latter, of course, was found to impart corrective meaning to the very unpleasant assertions challenged in Letter Carriers.

In this case, moreover, the column was identified as written by Evans and Novak, men who are widely known, and certainly known to readers of the Op-Ed pages, as purveyors of opinion who are frequently controversial. More than this, before the reader comes to the passage in question, he will have discovered many times over that Evans and Novak are, to say the least of it, suspicious of Ollman's intentions and that they regard him as a remarkably wayward academic. All of that impression is conveyed in language and expressions of opinion that no one on this court finds actionable. By the time the reader comes to the assertion of an anonymous professor's statement of academic opinion about Ollman, he is, I think, likely to read the remark as more of the same. He is most unlikely to regard that assertion as to be trusted automatically. It is an assertion of a kind of fact, it is true, but a hyperbolic "fact" so thoroughly embedded in opinion and tendentiousness that it takes on their qualities.

It is important to be clear about this. It is the totality of these circumstances that show the statement to be rhetorical hyperbole. If the statement were that a person is known by his friends to be an alcoholic or that a professor's written works were plagiarized, then it would be a very different kind of factual assertion from that involved here, one taken more seriously by readers, and not mitigated by context.10

I have attempted the kind of contextual inquiry that I think the Supreme Court's cases indicate and the rationale of the first amendment mandates. I am persuaded that Ollman may not rest a libel action on the statement contained in the Evans and Novak column.

MacKINNON, Senior Circuit Judge (concurring):

Bertell Ollman is a political scientist who, inter alia, was an associate professor teaching Marxism at New York University. His attitudes toward Marxism and his manner of teaching are fully described in the accompanying opinions by Judges Bork and Starr, so they will not be repeated here. Suffice it to say that Ollman's expressed attitude toward teaching Marxism to college students raised considerable controversy among many teachers of political science and others throughout the nation. In 1978 Ollman applied to the University of Maryland for the post of Chairman of the department of politics and government.

The University of Maryland is established as a public university and constitutes a "state agency." Md.Educ.Code Ann. Sec. 13-101 (1978). The government of the University is vested in a Board of Regents consisting of 15 members. Except for the State's Secretary of Agriculture, its Regents are all appointed by the Governor with the advice and consent of the Maryland Senate. Id. Sec. 13-102. The University is tax-supported, receiving very substantial sums of money for its maintenance and operation in the form of appropriations from the Maryland Legislature.

Ollman's application was approved by a search committee of the department of government and politics at the University, and that committee nominated him to head that department. His nomination was approved by the Provost of the University and the Chancellor of the main campus of the University. When these facts became public, an intense public controversy arose over the propriety of appointing Ollman to head the political science department of a public university supported by tax payers' money. This discussion raised many issues that can only be described as political in nature. The most obvious of these was whether members of the public, which necessarily included tax paying citizens of Maryland whose sons and daughters attended the University, should support the appointment of a professor who taught Marxism as Ollman did to lead the political science department that was supposed to instruct their children in government. The faculty and president of the University may have had other questions about Ollman's competence. The public, however, was concerned about his teaching objectives and methods.

At this juncture, the Evans and Novak column appeared. The article is set forth as an appendix to Judge Starr's opinion so one can judge it in its entirety. For those interested in knowing the outcome of this controversy, it is a fact of wide public knowledge that the president of the University ultimately refused to approve Ollman's appointment.1

In due time Ollman sued Evans and Novak for libel. The complaint alleges that five statements about him in the Evans and Novak column of May 4, 1978, were false and defamatory (J.A. 7). It is clear to me that four of the five statements are constitutionally protected as expressions of opinion for reasons generally agreed upon by the court as stated in our other opinions herewith. The discussion in this opinion is thus limited to the fifth statement in the article which stated that "Ollman has no status within the profession but is a pure and simple activist."2 When this case was before the original panel, I dissented from the majority opinion which held that this statement constituted actionable defamation. My stated reasons are set forth generally in the margin hereof.3 This opinion sets forth reasons in addition to those in my panel dissent as to why the statement in question must be regarded as an expression of opinion.

It is crystal clear that Evans' and Novak's article was directed at the public political discussion that surrounded Ollman's nomination and that it must be judged under the rule early announced by Judge (later Justice) Lurton and adopted in Washington Post Co. v. Chaloner, 250 U.S. 290, 293, 39 S. Ct. 448, 63 L. Ed. 987 (1819), that "a publication claimed to be defamatory must be read and construed in the sense in which the readers to whom it is addressed would ordinarily understand it. So the whole item ... should be read and construed together, and its meaning and significance thus determined." (Emphasis added.) A California appellate court elaborated on this principle in Desert Sun Pub. Co. v. Superior Court, Etc., 158 Cal. Rptr. 519, 521, 97 Cal. App. 3d 49 (1979): "A political publication may not be dissected and judged word for word or phrase by phrase. The entire publication must be examined." The statement in question here was directed to and uttered in a public controversy that went far beyond Ollman's merely academic credentials and related to matters of great public political interest.

Here we must focus on the sense in which the controversy ignited by Ollman's nomination was political. Here is a professor who has committed himself publicly to using the university as a means to promote the Marxist agenda. Here also is a university, an arm of a constitutional government, which is dedicated, among other things, to educating its citizens in the principles of politics. Ollman's nomination thus must inevitably have raised in the public's mind questions about the mission of a public university, the scope of academic freedom, and the responsibilities, if any, of public universities and the political science profession in a society like ours dedicated not only to free debate, but to preserving the institutions that make free debate possible. Little wonder that some must have thought, like Florence Nightingale did of hospitals and disease, that while a political science department might want to study Marxism, it should not promote it. Others equally must have held that a public university must tolerate even the advocacy of a political science opposed to public values. It was in the midst of a fray that must have inspired many political arguments that Evans and Novak made a small contribution: They reported that an unidentified liberal political science professor had stated Ollman had "no status" in his profession. How would those participating in or following the debate described above interpret this report? This is essentially the test Washington Post Co. instructs us to employ.

There can be no doubt that in the context of heated political debate, Evans' and Novak's article would have been taken by the readers to whom it was addressed as a statement of political opinion. To understand why this is so, we need to appreciate the special nature of political discussion, and how it differs from disputes that can be settled by mere reference to a survey or an almanac.

Discussions about fundamental political issues, such as academic freedom and public universities, take one into the realm of contested concepts. Here opposing sides may disagree not only about what policy they would prefer to see enacted; they disagree even about the meanings of words. Words such as "freedom," "liberty," "education," and others are used in arguments about basic political issues in a manner very different from the way that they are used in less controversial contexts. "Status," the word on whose meaning this case turns, is just the sort of word that in a political controversy like that over Ollman may mean something quite different to each faction or person who uses it. In these politically controversial settings, the average reader would treat statements that rely on words like "status" to convey their sense as statements of opinion.

In fact, words like "status" are used to express approval in a way that gives mere opinion a superficial air of scientific truth. Yet efforts to measure such magnitudes quantitatively are as misconceived as trying to measure class, clout or charisma. If one says someone has "class," for example, it sounds like one means someone has some measurable quantity of an enviable personal property. It is impossible of course to attach a precise meaning to "no class," or "no status" (or little or much) unless we first determine what "class" or "status" is. But if most people say Mr. X has no class, does that mean he is a boor, or that most people have bad taste? In fact, such statements merely express one's admiration or contempt for something about the person. As to what that something is, everyone would, if pressed, produce their own definition. "Class," like "status," is one of those properties that is ascribed to a person in a statement that sounds like it asserts a fact when it only asserts an opinion. This is partly because such statements are unverifiable, and the case law on the fact-opinion dichotomy in libel law emphasizes this central requirement that to be libelous, a statement must be verifiable. Hotchner v. Castillo-Puche, 551 F.2d 910, 913 (2d Cir.), cert. denied, 434 U.S. 834, 98 S. Ct. 120, 54 L. Ed. 2d 95 (1977).4 This point, however, is only half of the story. What is central here is why a claim that someone has little or no "status" is unverifiable in a politically controversial context. It is unverifiable because whatever method one chooses to measure "status" and other "quantities" of its ilk, one must commit oneself to some politically controversial view about what that so-called quantity really consists of. Choosing a method, a survey for example, settles nothing; it merely shifts the political debate from whether a person has "status," or some other such quality, to how best to measure it.

This essential contestability of the meanings of some words used in political controversy is not a problem that can be solved by simply taking a survey, or some other ad hoc means of definition. Even if a libel plaintiff took such a survey, a court would still have to instruct a jury on how to interpret that evidence. It would therefore have to have a theory about what status really is, or what the word "status" really means. One opinion here suggests such a theory: a political scientist has no or little status, it states, if a decisive majority of his peers respond in a scientifically conducted survey (poll) that the political scientist in question is not a good political scientist. But is that really what it means to have or not to have status? Suppose a very small, secretive minority of the professor's peers revere him as a pioneer and admires him all the more because the majority, unable to appreciate his genius, holds him in contempt? By the test the suggested poll would impose, that professor, it seems, would have little status. But suppose that small minority is called the Faculty Appointments Committee of the Department of Government, Harvard University? Does the opinion of those three persons suddenly give our professor "status," thus trumping the view of the decisive majority? Does the same conclusion apply if the minority is the equivalent committee at the University of Paris, the University of Moscow, or West Point? It seems wrong to say authoritatively that such a man has "no status," or "low to moderate status," though that is precisely the result our hypothetical survey would yield. Indeed, it would be especially ironic to use a test that equated status with popularity among one's peers in a profession whose universally acknowledged founder lost a similar popularity contest, and probably would have lost it much worse than he did if only his fellow political theorists had been polled. See Plato, The Trial and Death of Socrates (G.M. Grube, trans. 1975). While one could suggest weighting votes for intensity of admiration or contempt, or for the status of the voter, to rehabilitate the survey, it is easy to see that efforts like these would only succeed in begging the question. One cannot know how to reform the survey unless one knows what it is the survey is supposed to measure. Unless, of course, this definition of "status" as survey results is meant to be self-evidently valid. If that is the suggestion, it falls far short of convincing.

The suggested poll presupposes that out of the aggregation of the opinions of the country's political scientists, one can get a verifiable fact, an objective measure of Ollman's status. But if each political scientist polled is just expressing his opinion, why should we expect the aggregation of all this opinion to be transformed somehow into a fact? The resultant numbers constitute a fact, but the result does not determine "status" because of the unverifiable and essentially contestable interpretations that those voting in the poll, and those who might rely upon it, might give to the term "status." The methodology of the poll suggests, however, that we would not even inquire into the reasons the various respondents to the survey might have for answering as they might. Their evaluations might be stupid, but in the survey method of defining "status," this is supposed not to bother us. The reason for this could only be that we have already decided that status just is the result of a poll like that recommended by the opinion survey method. But this answer shares the circular fallacy of Professor Binet's legendary reply to the question "What is intelligence?" in the so-called IQ controversy: "It is what my test measures." In fact, of course, there is no agreement as to just what standards should be used to evaluate the professional status of an educator, or a lawyer, doctor or any other professional for that matter.

As the Second Circuit recently noted in an opinion by Judge Winter, "the number of relevant variables [in evaluating teachers] is great and there is no common unit of measure by which to judge scholarship." Zahorik v. Cornell University, 729 F.2d 85, 93 (2d Cir.1984) (judicial review of university tenure decisions). Disagreements even among members of the same faculty as to the qualifications of individuals "may reflect long standing and heated disagreement as to the merits of contending schools of thought." Id. This is all the more true where the internecine feuds of political scientists converge with the wider political debates in which the public takes part. To resort to a survey of a profession ridden with political strife in the hopes of settling the meaning of an ambiguous term itself used in the course of another heated political debate evinces an almost quixotic faith in the power of the techniques of social science. Evaluations of personal or professional status, however, are so inherently subjective that I do not believe that they can in general be considered statements of fact. A social scientist might plausibly design a survey to measure what he or she decides to call "status" for the purposes of a particular study. But to suggest that these poll results are what the reader takes the world "status" to mean when used in a political debate is unreasonable.

This perspective, it should be noted, is not at all at odds with the traditional trust the law of libel places in the jury to make determinations about such matters as a plaintiff's reputation in his or her community. See Partial Dissent of Judge Wald at 1-2. The way a word such as "reputation" is used in a court of law may be quite different from how it is used in the heat of a political debate. Courts and legislatures are in the business of giving words special and, one hopes, precise definitions and creating rules of evidence that in part determine what certain legal words mean. These words are used for particular purposes that are often quite different from the use a layman would make of them, especially during a political argument. That juries, with the help of judges and lawyers, apply legal concepts to particular facts in the special context of a trial may, from a logical point of view, have nothing to do with what people mean in ordinary affairs when they use some word or term. To compare evidence a jury could receive about reputation and the suggested survey confuses the competence of juries to use some specially defined term once it is given to them with the appropriateness of giving them a particular definition in the first place, here of what "status" meant in a political argument.

My view does not suggest that all terms and concepts become meaningless in the context of a political debate. There is indeed no suggestion here that "status" is meaningless. Rather, the argument here is that its meaning is variable, unverifiable, controversial, a matter of opinion, whom you listen to, and whose side you are on, among other things. The word does not have a "precise core of meaning." Political debate is rife with such richly if unquantifiably meaningful terms, as a moment's reflection will show. This is to be expected. Facts have a status that opinions do not, and dressing the latter up as the former may serve legitimate political purposes, just as hyperbole may.

There is another reason the approach of the suggested poll is mistaken. A rule of law which took the vast category of words often used in political debate to make seemingly factual but in fact value-laden statements and translated them into the verifiable quantities of social science would be a pernicious step in a dangerous direction. It would not only inhibit the rough and tumble of political debate, as Judge Bork so persuasively argues. It would do so in a very unsettling way. If the proposed poll is taken as illustrative, under such a rule judges would presumably arm juries with some of the more dubious techniques of modern social science and instruct them to translate essentially contested political terms into measurable quantities so they could decide whether a controversial remark was defamatory. How could participants in political debate respond to this rule except by trying to cleanse their talk of those terms, like "status," that suggest some factual content (though no one can say authoritatively what that content is) but assert political values as well? In place of those terms, those who enter the political arena would be well advised to substitute a paler lingo that tries to keep facts and opinions discrete, as does the vocabulary invented by the social scientists. Such a development would tend to impoverish perhaps the most important prerequisite of free political discussion: the language of ideas in which we conduct it.

The analysis above argues that statements which rely for their meaning on essentially contested concepts, as those used in political debates frequently do, cannot state the sort of facts necessary to support an action at libel. Judge Starr's scholarly opinion sets out a four factor test to distinguish facts and opinion sets out a four factor test to distinguish facts and opinions that emphasizes the common usage or meaning of the challenged statement, whether the statement is verifiable, the context in which the statement is made, and the broader social context into which the statement fits. Judge Starr then applies this test to Ollman's five claims, including the prickly fifth claim, in a manner in which we concur. These factors are undoubtedly the most important in determining whether a statement is an actionable libel. They are consistent with the approach of this opinion, but it is important these factors not be taken mechanically, but as a way to analyze the "totality of the circumstances." Here Judge Bork's skillful employment of "the concept of a public, political arena" is crucial to a proper understanding of the analysis Judge Starr elucidates. With the qualification that Judge Bork's stress throughout his opinion on the special nature of the "first amendment arena" is necessary to a proper understanding of Judge Starr's analysis, I concur generally with the opinions of Judges Bork and Starr.

SPOTTSWOOD W. ROBINSON, III, Chief Judge, with whom J. SKELLY WRIGHT, Circuit Judge, joins, dissenting in part:

Slightly more than a year ago, I undertook to set forth, both carefully and fully, my views on the exceedingly difficult problem presented in this case.1 Today, even with the benefit of additional argument, my colleagues' excellent presentations, and intense restudy and reconsideration of my own, those views remain essentially unchanged. This opinion, then, is largely the earlier version, revised modestly in the interest of clarity.

The one major issue presented by this appeal is whether the allegedly defamatory statements of which Ollman complains are representations of fact capable of supporting an action for libel or, instead, assertions of opinion unconditionally protected by the First Amendment. The District Court, ruling that they are no more than opinion, granted summary judgment for the defense,2 and a majority of this court now affirms. I think there may well be material issues of fact affecting the availability of the opinion privilege for several of the statements.

* The First Amendment embodies a special solicitude for unfettered expression of opinion. A decade ago, in Gertz v. Robert Welch, Inc.,3 the Supreme Court stated:

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.4

This passage was the first clear verbalization by the Court of the degree to which the Constitution preempts local defamation law in the area of opinion.

Previously, however, the Court had hinted at limitations on governmental power to impose civil or criminal liability for statements of belief, judgment or sentiment. In New York Times Co. v. Sullivan,5 the landmark decision explicating the interplay between the constitutional guaranties of free speech and press and the common law sanctions for defamation, the Court observed: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. Both defenses are of course defeasible if the public official proves actual malice....6

In Garrison v. Louisiana,7 a prosecution for criminal libel, the Court again adverted to fair comment, finding it unnecessary to decide in the context of that case "whether appellant's statement was factual or merely comment, or whether a State may provide any remedy, civil or criminal, if defamatory comment alone, however vituperative, is directed at public officials."8

The opinion privilege articulated in Gertz thus was foreshadowed, to some extent at least, in earlier pronouncements, although the degree of constitutional protection to be afforded opinion was not fully apparent. But while Gertz confirms the existence of an absolute privilege for expressions of opinion, neither that nor any other Supreme Court decision has provided much guidance for identifying statements that are opinion for First Amendment purposes.

New York Times involved misstatements, obviously factual in nature, concerning the handling of incidents of racial unrest by police,9 and its reference to fair comment appears only in a footnote at the end of the opinion. The allegedly libelous statements in Garrison--comments disparaging the judicial conduct of certain judges--were more problematic, but the Court disposed of the case on the ground that the criminal statute at issue unconstitutionally penalized both spitefully-motivated, though truthful, criticism of public officials and negligently-made misstatements about them.10 Since the Court invalidated the statutory basis for the prosecution, it did not find it necessary to classify the remarks as fact or opinion.11

Greenbelt Cooperative Publishing Association v. Bresler,12 a pre-Gertz decision, has subsequently been treated by the Court as an opinion case,13 though it did not expressly refer to the opinion-fact dichotomy. There the defendant newspaper had published two articles reporting the proceedings of city council meetings at which Bresler's request for zoning variances for certain land and the city's effort to acquire other land owned by him were subjects of heated debate, and had quoted several speakers' characterization of Bresler's position as "blackmail."14 The Court found that the jury had been improperly instructed on malice, and went on to hold that the "blackmail" spoken of could not support a libel verdict even with a proper instruction.15 Noting that the reports of the meetings were accurate and full,16 the Court concluded:

It is simply impossible to believe that a reader who reached the word "blackmail" in either article would not have understood exactly what was meant: it was Bresler's public and wholly legal negotiating proposals that were being criticized. No reader could have thought that either the speakers at the meetings or the newspaper articles reporting their words were charging Bresler with the commission of a criminal offense. On the contrary, even the most careless reader must have perceived that the word was no more than rhetorical hyperbole, a vigorous epithet used by those who considered Bresler's negotiating position extremely unreasonable. Indeed, the record is completely devoid of evidence that anyone in the city of Greenbelt or anywhere else thought Bresler had been charged with a crime.

To permit the infliction of financial liability upon the [newspaper] for publishing these two news articles would subvert the most fundamental meaning of a free press, protected by the First and Fourteenth Amendments.17

Old Dominion Branch 496, National Association of Letter Carriers v. Austin,18 a companion case to Gertz, presented a somewhat similar set of facts. The defendant union had listed the plaintiffs as "scabs" in its monthly newsletter, and had quoted a "well-known piece of trade union literature, generally attributed to author Jack London,"19 defining "scab" as, among other things, "a traitor to his God, his country, his family and his class."20 Once again the Court found defective the malice instruction given to the jury;21 it then held the critical statements nonlibelous as a matter of federal law.22 Repeating Gertz' observation that "there is no such thing as a false idea,"23 the Court reasoned that, in the context of the labor dispute which gave rise to the newsletter statements, "use of words like 'traitor' cannot be construed as representations of fact.... Such words were obviously used here in a loose, figurative sense to demonstrate the union's strong disagreement with the views of those workers who oppose unionization."24 And, recalling Greenbelt, the Court found it "similarly impossible to believe that any reader ... would have understood the newsletter to be charging the appellees with committing the criminal offense of treason.... Jack London's 'definition of a scab' is merely rhetorical hyperbole, a lusty and imaginative expression of the contempt felt by union members toward those who refuse to join."25

Gertz involved accusations in the defendant's magazine that the plaintiff had a large police file, had been an official of an organization advocating the violent seizure of government, and was a "Leninist" and a "Communist-frontier."26 Although the first two observations seem clearly factual, the latter two could be regarded as expressions of the author's opinion. The Court did not, however, discuss the applicability of the opinion privilege to these statements;27 rather, it noted that "Leninist" and "Communist-frontier" are "generally considered defamatory."28 Holding that Gertz, as a private figure, did not have to meet the demanding New York Times standard of proof on the issue of malice,29 the Court remanded the case for a new trial.30 There was no suggestion that any of the accusations were incapable, as a matter of constitutional law, of being deemed libelous.

In a more recent decision on defamation, the Court reversed a defendant's summary judgment in a case featuring statements arguably classifiable as opinion. In Hutchinson v. Proxmire,31 a United States Senator and his legislative assistant publicized a description of the plaintiff's governmentally-funded animal research as "nonsense" and "transparent worthlessness;"32 branding this use of public monies "outrageous," they had added that "[i]n fact, the good doctor has made a fortune from his monkeys and in the process made a monkey out of the American taxpayer."33 The Court ruled that these observations were not immunized by the Speech and Debate Clause,34 determined that the plaintiff was not a public figure,35 and sent the case back with no intimation that any of the allegedly defamatory remarks were within the opinion privilege.

Rounding out the Supreme Court cases having some bearing on the meaning of "opinion" are two decisions involving articles summarizing documents susceptible of differing interpretations.36 In Time, Inc. v. Pape,37 a pre-Gertz case involving a public-figure plaintiff, Time, a weekly magazine, in the course of recapitulating a report of the Civil Rights Commission, quoted from the report's summary of the allegations of a civil rights complaint without indicating that charges of police brutality against Pape were not independent findings by the Commission.38 The issue before the Court thus was whether Time had "engaged in a 'falsification' [of the report] sufficient in itself to sustain a jury finding of 'actual malice'."39 Looking at the report as a whole, the Court saw "a document that bristled with ambiguities,"40 and characterized Time's summary of what the document "said" as a statement of interpretation rather than one of historic fact.41 Deeming Time's reading of the report a plausible one, the Court held that no libel recovery was possible in such circumstances.42 It is unclear from the Pape opinion, however, whether the Court reasoned that the challenged statement was constitutionally immune to post hoc evaluation of truth or falsity because it was actually an expression of opinion,43 or rather that the statement was capable of being adjudged erroneous but incapable of being labeled malicious because of the reasonableness of Time's interpretation.44

This uncertainty about the rationale of Pape appears to have been resolved by Time, Inc. v. Firestone,45 a post-Gertz case in which a private-figure plaintiff charged Time with erroneously reporting that her husband had been granted a divorce on grounds of "extreme cruelty and adultery."46 The divorce decree prompting the news item was hardly a model of clarity,47 and Time, citing Pape, argued that a "rational interpretation of an ambiguous document is constitutionally protected."48 The Court disagreed, explaining that Pape was an application of the actual-malice standard,49 thus foreclosing the possibility that Pape was really an "opinion" case. The Court made clear that Time's interpretation of the divorce decree could be found factually incorrect, and liability could be imposed if Time bore the blame.50

Gertz' pronouncement that the First Amendment confers an absolute privilege on expressions of opinion stands as one of the cardinal principles of free speech and press. Yet, as my brief review of pertinent Supreme Court defamation cases illustrates, it is a principle whose implementation depends entirely upon a precedent determination that the allegedly defamatory statement is actually one of opinion rather than fact. Lacking guidance from the Supreme Court in such an endeavor, federal courts of appeals,51 the highest courts of several states,52 the American Law Institute,53 and various commentators54 have attempted to ascertain just what kinds of statements are protected opinion. Cognizant of the difficulty of the undertaking, and resigned to the fact that any effort to superimpose a categorical ordering on this infinitely variable area entails some oversimplification, I nevertheless believe that helpful guidelines can be fashioned to inform the disposition of cases presenting the problem.

I start with candid recognition that the universe of statements cannot be neatly divided, by some logically discernible equator, into hemispheres of fact and opinion. Fact is the germ of opinion, and the transition from assertion of fact to expression of opinion is a progression along a continuum. A reviewing court's charge is to determine, in light of the considerations inspiring First Amendment jurisprudence and the surviving policies underlying common law protection of reputation, the point at which we should draw the line marking off the portion of speech to be accorded the absolute constitutional protection of opinion55 rather than the conditional privilege afforded representations of fact.56

At one end of the continuum are statements that may appropriately be called "pure" opinion. These are expressions which commonly are regarded as incapable of being adjudged true or false in any objective sense of those terms. Matters of personal taste, aesthetics, literary criticism, religious beliefs, moral convictions, political views and social theories would fall within this category.57 These are statements which by nature "could be corrected by discussion,"58 and accordingly statements whose survival in our society's discourse should be committed to "the competition of the market" in ideas.59

Also near the pure-opinion end of the continuum, I think, are those "loosely definable, variously interpretable"60 derogatory remarks that frequently are flung about in colloquial argument and debate.61 The hallmark of these statements is not that they are innocuous or impotent, but rather that they are so far in the realm of vernacular epithet as to become expressions of generalized criticism or dislike, without any specific factual moorings. Evaluating such statements as true or false is problematic largely because of the difficulty of arriving at a consensus on precisely what evidence would be relevant and sufficient to justify their use.62

Finally, metaphorical language is also allied to pure opinion. When context makes it apparent that a word is being used figuratively or imaginatively without any intention to rely on its literal meaning,63 the labels "true" and "false" are inapposite.

All of these types of statements seem clearly to fall within the ambit of the constitutional opinion privilege. They would be recognized by most listeners and readers64 simply as expressions of personal taste or conviction, or as rhetorical outlets for venting anger or contempt without imputing any specific wrongdoing, or as colorful and hyperbolic applications of language. This is not to say that publication of such statements will never be damaging to the reputation of their targets. At this end of the continuum, however, our First Amendment commitment to free circulation of ideas and beliefs--no matter how unfair, unreasonable, or unseemly they may appear to be--bars the law of defamation from assessing, according to some standard of orthodoxy, the propriety of or the justification for such statements.

Expressions at or near the pure-opinion end of the continuum probably constitute only a relatively small portion of the statements that become subjects of defamation lawsuits. Perhaps far more common, and certainly more perplexing, are statements that reflect the author's deductions or evaluations but are "laden with factual content."65 The apparent proportions of opinion and fact in these "hybrid" statement varies considerably. For example, a statement that "Jones is incompetent to handle that job" suggests some factual underpinning but, on the whole, imports a fairly high degree of subjective judgment. By contrast, a statement that "Smith is a murderer" appears much closer to an assertion of objective fact. Analytically, however, the accusation of murder could be regarded as an opinion, for it, like the charge of incompetence, reflects a conclusion ultimately reached by the author on the basis of an amalgamation and interpretation of underlying facts.

Hybrid statements differ from pure opinion in that most people would regard them as capable of denomination as true or false, depending upon what the background facts are revealed to be. At the same time, they generally are not propositions that a scientist or logician would regard as provable facts. The hard question is whether these kinds of statements, which both express the author's judgment and indicate the existence of specific facts warranting that judgment are within the absolute privilege for opinion.

When the proponent of a hybrid statement discloses to the reader the pertinent background facts with reasonable completeness and accuracy, there is a strong argument for including the statement within the realm of absolute privilege.66 In these circumstances, the reader can easily recognize the statement as the author's synthesis and, placing it beside the predicate facts, can make up his own mind about how much weight and credence to give to the author's conclusion. In effect, the underlying facts transform the hybrid statement in the eyes of the reader from a judgment suffused with assumed but unspecified facts into a simple statement of opinion drawn from specific factual premises. Having supplied an accurate account of those facts, the author cannot be said to have misled or deceived the reader about the matter discussed, even if the author's ultimate conclusion--the hybrid statement--may in some sense be erroneous. And although the author's derogatory judgment may carry some power to damage reputation simply by virtue of his stature in the reader's estimation, it seems safe to posit a fairly high correlation between whatever damage is inflicted and the reader's personal appraisal of how reasonable that judgment is in light of the facts set forth.

In these circumstances, hybrid statements would seem to pose little or no threat to the reputation interest safeguarded by defamation law because they could be subjected to rigorous and fair evaluation by fully-informed readers.67 At the same time, their claim to First Amendment protection would be great because they would share the primary immunizing characteristic of pure opinion, for the presence of the background data would ensure that the only really active element of the statement is its judgmental or interpretive component.

This balancing of First Amendment and defamation implications of hybrid statements works flawlessly where the critical background facts are accurately set forth. A significant imbalance results, however, when a hybrid statement appears without any recitation of the underlying facts, or when those facts are stated incompletely or erroneously.68 When that is the case, the reader is unable to place the author's judgment in perspective, because he either is completely unaware of the predicate facts or is in some degree misled as to what they are. False hybrid statements obviously can wreak considerable damage to reputation.69 A reader supplied with no background at all may well assume that there are facts which support the derogatory conclusion, particularly if it is announced by the author with apparent assurance.70 A reader given materially incorrect or incomplete facts, mistakenly supposing that the pertinent data are accurately assembled before him, might give even more credence to the author's conclusion.71 Hybrid statements unaccompanied by any predicate facts, or attended by defective recitals of the underlying facts, thus should occupy a very different position in the concerns of libel law, for their claim to First Amendment protection is far less compelling. If the background data reaching the reader are deficient, the hybrid statement is as much a representation of the facts it implies as it is a judgment or interpretation of the communicated data.72

The Gertz Court immediately after proclaiming immunity for expressions of opinion, warned that

there is no constitutional value in false statements of fact. Neither the intentional lie nor the careless error materially advances society's interest in "uninhibited, robust, and wide-open" debate on public issues.... They belong to that category of utterances which "are no essential part of any exposition of ideas, and are of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality."73

I would hold that a hybrid statement is absolutely privileged as opinion when it is accompanied by a reasonably full and accurate narration of the facts pertinent to the author's conclusion. I would further hold that hybrid statements not so accompanied are not entitled to that degree of protection unless those facts are already known to the author's listeners or readers.74 I do not mean that the author must supply every little detail that conceivably might have some bearing. What I do mean is that the author's presentation must be reasonable--enough to enable the audience to fairly judge the conclusion stated.

Nor do I suggest that errors or omissions in recitals of the predicate facts automatically disentitle authors from asserting the opinion privilege for hybrid statements, for First Amendment jurisprudence requires the court to take into account the author's culpability, if any, for deficiencies in the factual presentation.

As the Gertz Court acknowledged, "[a]lthough the erroneous statement of fact is not worthy of constitutional protection, it is nevertheless inevitable in free debate,"75 and "punishment of error runs the risk of inducing a cautious and restrictive exercise of the constitutionally guaranteed freedoms of speech and press."76 On the other hand, as the Court pointed out, "absolute protection for the communications media requires a total sacrifice of the competing value served by the law of defamation."77 "The New York Times standard,"78 the Court said, "defines the level of constitutional protection appropriate to the context of defamation of a public person,"79 and, "so long as they do not impose liability without fault, the States may define for themselves the appropriate standard of liability for a publisher or broadcaster of defamatory falsehood injurious to a private individual."80 "This approach," the Court added,

provides a more equitable boundary between the competing concerns involved here. It recognizes the strength of the legitimate state interest in compensating private individuals for wrongful injury to reputation, yet shields the press and broadcast media from the rigors of strict liability for defamation. At least this conclusion obtains where, as here, the substance of the defamatory statement "makes substantial danger to reputation apparent."81

The considerations underlying the standards of care developed in New York Times and its progeny for measuring liability for defamatory falsehood apply equally to situations wherein an author states the background facts incompletely or incorrectly. I would conclude, then, that if critical background data are omitted or are erroneous,82 the absolute opinion privilege is still available when the infirmity is not traceable, in the case of a public official or public figure, to actual malice or reckless disregard of truth or falsity, or, in the case of a private figure, to culpability great enough to incur liability under relevant state defamation law.83

This approach is dictated by the need to account adequately for the factual load carried by the hybrid statement. As the Supreme Court recently recognized,

[some] sort of inaccuracy ... is commonplace in the forum of robust debate.... "Realistically, ... some error is inevitable; and the difficulties of separating fact from fiction convinced the Court in New York Times,84 Butts,85 Gertz86 and similar cases to limit liability to instances where some degree of culpability is present in order to eliminate the risk of undue self-censorship and the suppression of truthful material." "[E]rroneous statement is inevitable in free debate, and ... must be protected if the freedoms of expression are to have the 'breathing space' that they 'need ... to survive.' "87

By denying the opinion privilege only in those instances in which authors of hybrid statements are guilty of culpable error or omission, those who satisfy the applicable standard of care--the same standard which they inevitably must meet to keep their defamatory factual assertions from ripening into monetary judgments--would be free of potential liability in expressing their conclusions and interpretations, no matter how unreasonable or intemperate they may be. No author need fear that an insignificant or inadvertent error in factual presentation would transform his otherwise absolutely-protected statement into an actionable claim. Under my mode of analysis, only those indulging in culpable behavior could be deterred from expressing their ideas, and I see no constitutional imperative for extending absolute protection to authors who have misled their readers by refusing or culpably failing to provide reasonably full and accurate background data.88 Even in these latter instances, the hybrid statement, though forfeiting all right to absolute privilege, will be afforded that quantum of unqualified protection accorded a purely factual misstatement under the circumstances.89

A requirement of substantial disclosure of material background facts is hardly a novel step. The Restatement (Second) of Torts takes the position that "[a] defamatory communication may consist of a statement in the form of an opinion, but a statement of this nature is actionable only if it implies the allegation of undisclosed defamatory facts as the basis for the opinion."90 This formulation envisions a situation where the derogatory opinion seemingly is based on facts unstated by the author, and known to or assumed by the parties to the communication91--as for example, an unelucidated statement that a named person is a thief or an alcoholic.92 The rationale is that such a statement gives rise to the inference that there are undisclosed facts justifying the opinion and that these facts are defamatory.93 So, while the opinion component of the communication cannot subject the author to liability, the implied factual component might.94

I subscribe to this position, and I believe the reasoning underlying it extends further. While the Restatement's rule does not itself address the derogatory hybrid statement accompanied by an incomplete disclosure of background facts, the propensity for implication underlying the rule may similarly affect such a statement. The author's recountal of some of the background facts normally creates the inference that there are no other facts pertinent to the opinion expressed; absent some contrary indication, recipients of the communication would naturally derive that understanding. If, then, the undisclosed background facts strip away the justification the disclosed facts proffered for the disparaging remark, the communication cannot automatically be deemed a mere expression of absolutely-protected opinion, for it incorporates a falsehood by inference. The communication is really a false and defamatory representation that, squarely on the basis of such facts as were disclosed, the subject of the comment is guilty of the defamatory behavior charged.95

To recapitulate, I think the absolute First Amendment opinion privilege proclaimed in Gertz should be held to shield four categories of statements. The first includes expressions of personal taste, sentiment and values that are inherently or essentially subjective in nature. In the second group are those general derogatory epithets and "undefined slogans"96 flung about in the course of political, economic and social debate that express contempt or extreme disagreement without connoting any particular factual basis. Third is language which, from its context, obviously is used in the figurative or hyperbolic sense. These three types are characterized by the absence of any suggestion that they are grounded upon any specific factual predicate, and I would locate them near the pure-opinion end of the continuum. The fourth category embraces statements, which I have termed "hybrids," that both intimate the existence of specific facts and convey the author's judgment on or interpretation of those facts, provided that such statements are accompanied by a reasonably full and accurate account of the material background facts, or that incompleteness or inaccuracy of the predicate data is nonculpable according to the applicable standard of care.97

I do not mean to imply that I perceive my delineation of the scope of the opinion privilege to be clearly mandated by any of the Supreme Court's defamation decisions. I believe, however, that it is responsive to, and certainly not inconsistent with, what little can be gleaned from them. For example, Greenbelt98 and Letter Carriers99 are, by their own terms, instances of protection accorded language used figuratively or hyperbolically.100 In Hutchinson,101 where the challenged press release called the plaintiff's research "transparent worthlessness" and remarked that he was personally profiting from a pointless expenditure of tax monies, the plaintiff claimed that the release "contained an inaccurate and incomplete summary of his research,"102 Hence, if the allegedly libelous comments in that case are viewed as hybrids, issue squarely had been joined on whether a full and fair account of the factual predicate had been provided. The report of the content of the divorce decree in Firestone103 was not the kind of literary criticism that would be absolutely privileged as a subjective expression of pure opinion, and if in the realm of opinion at all the report was at best a hybrid presented with no background data whatever.104 Thus, the dispositions in those cases are perfectly harmonious with my view on the scope of the opinion privilege.

I now turn my attention to the passages of the syndicated column which are subjects of complaint in the case at bar. I agree with the District Court that Evans' and Novak's characterization of Ollman as an "outspoken proponent of 'political Marxism' "105 is absolutely privileged. It falls well within the class of "loosely definable, variously interpretable statements of opinion ... made inextricably in the contest of political, social or philosophical debate...."106 "Political Marxism" is much too ambiguous a slogan to permit a court to determine whether it is really defamatory, much less to ascertain whether the claim that Ollman propounds it is actually false.107

Presenting a different problem, however, is the column's observation that

[w]hile Ollman is described in news accounts as a "respected Marxist scholar," he is widely viewed in his profession as a political activist.108

Like "political Marxism," the term "political activist" would not normally be deemed defamatory.109 As it appears here, however, it reasonably could be read as implying the antithesis of scholarship.110 This interpretation is reinforced by the authors' later comments:

Such pamphleteering is hooted at by one political scientist in a major eastern university, whose scholarship and reputation as a liberal are well known. "Ollman has no status within the profession, but is a pure and simple activist," he said.111

The District Court characterized these statements as the authors' submission "that [Ollman] lacks a reputation in his field as a scholar."112 I agree that a jury reasonably could find113 the overall import of the remarks to be that, although Ollman has been described in the press as a respected scholar, his professional colleagues actually do not regard him as such. Without doubt, scholarship is the quintessential attribute of professorial competence. To say that a professor's academic peers, who presumably are those most capable of evaluating the real merit of his work, do not rate him highly as a scholar is to impugn his professional reputation severely.

A statement that Ollman's peers do not respect him as a scholar stands, I submit, on quite different footing from a statement that Evans and Novak do not themselves rank him as one. The latter might well fall into the category of pure opinion, as a subjective appraisal of the value of Ollman's writings.114 The former, however, if not actually a representation of fact, certainly rises no higher than a hybrid statement.115 It may convey the authors' ultimate assessment of what the political science profession thinks of Ollman, but it also implies the existence of facts inducing that conclusion, such as evaluations of Ollman's work by a sampling of academicians, critical reviews of his articles, or a poll taken of members of the profession. Our attention is directed to a passage describing how Ollman came in last in two American Political Science Association elections,116 but just what this fact has to do with Ollman's scholarly reputation in the profession is not immediately apparent; indeed, the article itself professes some uncertainty about "[w]hether or not [the election results] represent[ ] a professional judgment by his colleagues."117 Ollman, on the other hand, points to a 1978 published survey in which, so he claims, "a poll of 317 leading and representative political scientists" ranked him "10th in the entire field of all political scientists in terms of occupational prestige."118 I thus think that, although the matter is not wholly free from doubt, the paucity of supporting facts in the column, coupled with the survey Ollman proffers, raises a genuine issue as to whether there was a culpable error or omission in the background facts presented to the reader.119

I come finally to a set of statements relating to Ollman's writings and to what assertedly they reveal about his objectives as an instructor:

His candid writings avow his desire to use the classroom as an instrument for preparing what he calls "the revolution."

* * *

* * *

Ollman's intentions become explicit in "On Teaching Marxism and Building the Movement," his article in the Winter 1978 issue of New Political Science. Most students, he claims, conclude his course with a "Marxist outlook." Ollman concedes that will be seen "as an admission that the purpose of my course is to convert students to socialism."

That bothers him not at all because "a correct understanding of Marxism (as indeed of any body of scientific truths) leads automatically to its acceptance." Non-Marxist students are defined as those "who do not yet understand Marxism." The "classroom" is a place where the students' "bourgeois ideology is being dismantled." "Our prior task" before the revolution, he writes, "is to make more revolutionaries. The revolution will only occur when there are enough of us to make it."120

I would note initially that these excerpts do not represent literary, scholarly or ideological criticism. They do not advance the authors' personal views of such attributes as Ollman's writing style, the quality of his analysis, or the value or correctness of the ideas he advances. While comments of this type, I assume, would be the kind of pure opinion that lies at the core of the opinion privilege,121 the quoted passages purport to describe the substantive content of Ollman's article. A jury reasonably could read these passages as saying that Ollman, in his writings, openly admits that he wishes to use the classroom to indoctrinate his students and transform them into Marxists.122 To be sure, whenever an author undertakes to encapsulate and describe the contents of another's lengthy work, the product is apt to reflect some amount of the author's own interpretation and judgment. Here, however, a significant component of factual representation also comes through, particularly in such strong and apparently unequivocal phrases as "[h]is candid writings avow," "Ollman's intentions become explicit" and "Ollman concedes."123 I therefore think these passages should properly be regarded as hybrid statements of what Ollman's writings say about his intentions in the classroom.124

A fair amount of background material on this point is provided in the column under attack, largely in the form of direct quotations from Ollman's writings. There is some question, however, as to the completeness and accuracy with which these predicate facts are set out. The District Court, after review of the article, found that "[w]hile [Evans and Novak] refer to [Ollman's] writings and speeches, Ollman's statements are selected to reflect [their] opinion. Portions contrary to Evan's [sic] and Novak's viewpoint are carefully omitted."125 The court also suggested that "this may be thought of as biased journalism,"126 and an examination of the full text of the sources quoted could lead one to believe that this appellation may not be undeserved.127

I conclude, then, that these passages also present a genuine issue whether the absolute privilege for opinion has been forfeited by culpable omissions or errors in the supporting facts which the article offered its readers.

WALD, Circuit Judge, with whom Circuit Judges EDWARDS and SCALIA join, dissenting in part:

I basically agree with the plurality's outline of the appropriate strategy for identifying absolutely privileged opinion and its judgment that most of the statements made by Evans and Novak about the plaintiff are non-actionable statements of opinion. However, in my mind the columnists' statement that "Ollman has no status within the profession, but is a pure and simple activist" is an assertion of fact for which its authors can be made to answer, consistent with the requirements of the first amendment, in a suit for libel.

In many areas of the law, the factual nature of statements about reputation is recognized and indeed taken for granted. Lay witnesses are generally allowed to testify as to someone's reputation in the community for veracity or violence, for example, although they cannot give their personal opinion as to those matters. See McCormick on Evidence Sec. 44 (Cleary ed. 1984). Expert witnesses are often asked in the course of their testimony whether other authors, scholars or practitioners are generally regarded as authorities in the field, see 7 Wigmore on Evidence Sec. 1984 (Chadbourne rev. 1974), and their own qualifications may be established or attacked on the basis of professional reputation, see 5 id. Sec. 1621.

Similarly, as the plurality concedes, the law of libel has long recognized the basically factual nature of attacks on reputation. I do not dispute the plurality's assertion that the first amendment often demands modifications of the common law of libel so as to limit the chilling effect of potential civil liability on an "uninhibited, robust, and wide-open debate on public issues." New York Times v. Sullivan, 376 U.S. 254, 270, 84 S. Ct. 710, 721, 11 L. Ed. 2d 686 (1964). In particular, the first amendment demands that we evaluate the allegedly libelous statement in the context in which it appeared to determine whether it can claim the constitutional privilege for statements of opinion. Yet I find that a fair application of both the plurality's test and the approach suggested by Judge Bork in his concurrence indicates that the statement before us is more a statement of fact than of opinion.

The plurality would ask four questions about the particular statement at issue: (1) do the words have a "precise core of meaning"; (2) is the statement verifiable; (3) how do the immediate context--in this case the article--and (4) the broader context affect the likelihood that the statement will be read as an assertion of fact? Although Judge Bork calls for a more flexible, ad hoc balancing approach to the fact-opinion distinction, his analysis of this case strikes me as conceptually indistinguishable from the plurality's approach. I fully agree that the distinction between fact and opinion is rarely self-evident or exact and that we should not attempt to impose any mechanical set of categories on the complexities of libel litigation. Although the task may not always be an easy one, however, we are surely obliged to articulate some set of principles to guide the district court in determining which types of statements can give rise to a libel action.

Indeed, despite the plea for a case-by-case consideration of the "totality of circumstances," Judge Bork apparently recognizes precisely this obligation. After purporting to engage in an open-textured balancing of first amendment values, Judge Bork relies on three factors of his own in order to immunize libel defendants from suit. He reasons that (1) Ollman should be expected to endure the challenged statement because he placed himself in a public, political debate, (2) the factual nature of the "no status" statement is inherently unsuitable for jury determination, and (3) the functional meaning and general context of the statement indicate its rhetorical purpose. The first of these factors represents an unprecedented extension into the fact-opinion doctrine of the distinction between public and private officials for the purposes of defamation suits. The second two considerations merely restate the plurality's test. The challenged statement is surely capable of adjudication if it admits of a stable core of meaning and if Ollman's professional reputation is in fact verifiable. Similarly, the functional meaning or practical impact of the "no status" assertion can only be determined in light of the factual and social context surrounding the appellees' column.

In any event, I believe that the challenged statement is properly characterized as a factual assertion rather than a rhetorical hyperbole under either the majority's or Judge Bork's approach. The statement that Ollman has no status within his profession undoubtedly admits of a sufficiently ascertainable and stable core of meaning: a decisive plurality of his fellow political scientists do not regard him as a good scholar. That one might find a wide diversity of views among political scientists about Ollman's work and about what constitutes scholarly excellence in no way undermines the commonly understood meaning of a statement like this about reputation. The statement says to the ordinary reader that, however each individual scholar evaluates excellence, there is an overwhelming consensus that Ollman does not have it.

Furthermore, Ollman's scholarly reputation is adequately verifiable. One could, for instance, devise a poll of American Political Science Association members as to their opinion, on a scale of one to ten, of the scholarly value of Ollman's work. Testimony of prominent political scientists or other measures of reputation would also serve to verify or refute the statement about Ollman's reputation without sending the jury into a sea of speculation.

As both Judge Bork and Judge MacKinnon point out, neither a poll nor the testimony of his peers will, in all likelihood, conclusively establish Ollman's professional reputation in the eyes of the jury. Nonetheless, juries traditionally are called on to resolve conflicting opinions in libel cases, and the uncertainties endemic to determining a person's reputation do not, in themselves, render the issue "inherently unsusceptible to accurate resolution by a jury." Op. of Bork, J., at p. 1005. Whatever their limits as truth finding devices, expert testimony or a poll could surely establish whether Ollman enjoys some reputation as an academic scholar as opposed to a mere activist--whether that scholarly reputation is supported by consensus or sharp disagreement among his colleagues. Given appropriate instruction by the trial judge, a jury is as well equipped to determine whether an individual has or has not established professional reputation in this context as it is in a host of others. Although I share Judge Bork's concern that juries may, in some defamation cases, tend to underemphasize the limits imposed by the first amendment, I cannot subscribe to his astonishing view that "[t]he only solution to the problem libel actions pose would appear to be close judicial scrutiny to ensure that cases about the types of speech and writing essential to a vigorous first amendment do not reach the jury." Id. at p. 997 (emphasis added). Instead, I believe that any such problems should be remedied through careful supervision by the trial judge and vigorous appellate review, not through stripping the jury of its historic function merely because qualities such as "professional reputation" are difficult to adjudicate.1

The plurality cites the statement that "[o]ur academic culture does not permit the raising of such questions" as a concession of non-verifiability by Evans and Novak and their source that should warn the reader not to accept the foregoing statement about reputation as one of fact. Op. of Starr, J., at p. 991. But to me--and I believe to the ordinary reader as well--the liberal professor's refusal to be cited publicly means simply that Ollman's writings are not openly attacked in the academic community as mere polemics. Moreover, the majority's implication that Ollman has no verifiable reputation--that there is no way of evaluating the conglomeration of his colleagues' opinions, public or private, of his work--is belied by the characterization of the political scientist quoted as one "whose scholarship and reputation as a liberal are well known," as well as by the complex procedures for hiring, evaluation and tenure decisions set up by academic institutions throughout the nation. As judges we are familiar as well with how prominently academic reputation and stature figures in judicial nominations, evaluations and confirmation proceedings.

The plurality readily concedes that a statement about one's professional reputation, even the very statement before us, might be deemed a factual assertion in a different context. Yet the majority concludes that the facts, noted in the article, that Ollman was at the time a professor at New York University and was the top candidate for the position of chairman of the political science department at the University of Maryland would undermine a reader's belief in the factual accuracy of the statement. See Op. of Starr, J., at p. 990 & n. 42. But as I read the article, these "facts" could as well be understood as an assertion that Ollman's prominence is due solely to vociferousness and is entirely out of proportion to his poor reputation as a scholar among his peers. Indeed, the article as a whole, while it purports merely to raise questions about Ollman's qualifications, promotes itself as a call to sanity and objectivity and away from mere polemics. Thus, the immediate context in which this statement was made does little to warn a reader to regard with skepticism what might otherwise appear to be an assertion of fact.

In his concurrence, Judge Bork advances the further argument that the "no status" statement is, in its "practical impact," in the nature of opinion or rhetorical hyperbole because it is attributed to an anonymous source who is reporting the opinions of others. In the context of Evans' and Novak's column, however, the attribution of Ollman's utter lack of professional status to a political scientist "whose scholarship and reputation as a liberal are well known" gives the statement more rather than less of a factual and verifiable quality.

Under either the plurality's or Judge Bork's analysis, then, we are left with the bareboned fact that this article was written by Evans and Novak, well known political columnists, and appeared on the op-ed page. I agree wholeheartedly with both the plurality and Judge Bork that editorial pieces such as this one are commonly filled with "rhetorical hyperbole" and are often read with a degree of skepticism as to their factual content. The first amendment demands extraordinary caution in subjecting to the burdens of a lawsuit isolated statements in this kind of writing. The very statement before us, if adjudged to be a factual assertion, would have to cross numerous sturdily-constructed constitutional hurdles. In particular, Ollman would have to persuade a jury that the statement was false--that he indeed enjoyed a reputation as an academic scholar--and, if he were ruled a public figure subject to the standards of New York Times v. Sullivan, 376 U.S. 254, 84 S. Ct. 710, 11 L. Ed. 2d 686 (1964), that Evans and Novak made the statement with malice or reckless disregard for its truth or falsity.2 Furthermore, a jury verdict on these issues is subject to more searching appellate review than under the "clearly erroneous" standard. Bose Corp. v. Consumers Union, --- U.S. ----, 104 S. Ct. 1949, 80 L. Ed. 2d 502 (1984). Our decision today, however, means that, even assuming that the statement was utterly false, that it was made with knowledge of its falsity, and that it precipitated Ollman's loss of an important academic position and a decline in his professional standing, the statement's authors cannot be made to answer in a suit for libel. I do not believe that the first amendment requires this result, and I therefore respectfully dissent.

HARRY T. EDWARDS, Circuit Judge, concurring in part and dissenting in part:

For the most part, I thoroughly agree with and I am happy to concur in Judge Starr's thoughtful and well-reasoned opinion. Unfortunately, I cannot fully subscribe to the result reached.

After agonizing over this case, I have finally concluded that it is untenable even to suggest that the statement "Ollman has no status within the profession, but is a pure and simple activist " is an absolutely privileged "opinion." Indeed, as a former member of the academic community, I am somewhat taken aback by the notion that one's reputation within the profession (which is easily verifiable) may be so freely and glibly libelled. I can find no meaningful case authority to convince me that the First Amendment is designed to condone such loose muckraking.

Had Evans and Novak said that, in their view, Ollman "appeared to be a person without real status within the profession," this might be a different case. But they went much further and cited another "well known" scholar to support a verifiable claim that Ollman in fact had "no status within the profession." I agree with Judge Wald that "the statement says to the ordinary reader [and to the sophisticated reader as well] that, however each individual scholar evaluates excellence, there is an overwhelming consensus that Ollman does not have it." This is not a privileged opinion.

Having reached this conclusion, I concur in part in Judge Starr's opinion and concur in full in Judge Wald's and Judge Scalia's partial dissents.

SCALIA, Circuit Judge, with whom Circuit Judges WALD and HARRY T. EDWARDS, join, dissenting in part.

More plaintiffs should bear in mind that it is a normal human reaction, after painstakingly examining and rejecting thirty invalid and almost absurd contentions, to reject the thirty-first contention as well, and make a clean sweep of the matter. I have no other explanation for the majority's affirmance of summary judgment dismissing what seems to me a classic and cooly crafted libel, Evans and Novak's disparagement of Ollman's professional reputation. Judge Wald's opinion has fully responded to the straightforward contention of the majority opinion that this disparagement should be regarded as a mere nonactionable statement of opinion. I write separately to survey in somewhat greater detail the concurrence's more scenic route to what turns out to be the same destination.

It seems to me that the concurrence embarks upon an exercise of, as it puts it, constitutional "evolution," with very little reason and with very uncertain effect upon the species. Existing doctrine provides ample protection against the entire list of horribles supposedly confronting the defenseless modern publicist:

--The need to give special scope to political rhetoric is already met by recognition that hyperbole is an expected form of expression in that context. If Evans and Novak had chosen to call Ollman a traitor to our nation, fair enough. No reasonable person would believe, in that context, that they really meant a violation of 18 U.S.C. Sec. 2381 (1982). See Old Dominion Branch No. 496, National Association of Letter Carriers v. Austin, 418 U.S. 264, 285-86, 94 S. Ct. 2770, 2781-82, 41 L. Ed. 2d 745 (1974). The concurrence correctly claims the defense of this doctrine for the "no status" assertion. Surely it did not mean that Ollman had no status--only that his regard in the profession was not high. But to say, as the concurrence does, that hyperbole excuses not merely the exaggeration but the fact sought to be vividly conveyed by the exaggeration is to mistake a freedom to enliven discourse for a freedom to destroy reputation. The libel that "Smith is an incompetent carpenter" is not converted into harmless and nonactionable word-play by merely embellishing it into the statement that "Smith is the worst carpenter this side of the Mississippi."

--The expectation that one who enters the "public, political arena," Bork op. at 1004, must be prepared to take a certain amount of "public bumping," id., is already fulsomely assured by the New York Times Co. v. Sullivan, 376 U.S. 254, 84 S. Ct. 710, 11 L. Ed. 2d 686 (1964), requirement of actual malice in the defamation of public figures. One would think, from the concurrence's lugubrious description of the plight of the modern political publicist, that Evans and Novak were to be held to the truth of what they said--whereas in fact, in order to find them liable a public-figure plaintiff (such as the concurrence's argument assumes Ollman to be, see Waldbaum v. Fairchild Publications, Inc., 627 F.2d 1287, 1292 (D.C.Cir.), cert. denied, 449 U.S. 898, 101 S. Ct. 266, 66 L. Ed. 2d 128 (1980)), must establish not only that their allegation was false, but also that they knew it to be so, or acted with reckless disregard of its falsity, and must establish that by "clear and convincing proof." Gertz v. Robert Welch, Inc., 418 U.S. 323, 342, 94 S. Ct. 2997, 3008, 41 L. Ed. 2d 789 (1974) (emphasis added). This is a formidable task, and in the present case it is likely that the defendants would have to do no more to defeat it than to establish that a "political scientist in a major eastern university, whose scholarship and reputation as a liberal are well known" did indeed tell them what they printed. See St. Amant v. Thompson, 390 U.S. 727, 731, 88 S. Ct. 1323, 1325, 20 L. Ed. 2d 262 (1968).

--The difficulty of proving academic reputation, which the concurrence dwells upon at some length, Bork op. at 1005-08, is fully accounted for under current law by the fact that any failure of proof harms the plaintiff's rather than the defendant's case--and harms it in particularly devastating fashion when the "clear and convincing evidence" standard applicable to public figures governs. If the statistical evidence were indeed as inconclusive as the concurrence portrays, the result would be precisely what the concurrence desires, a dismissal of the suit.

--The problem that "juries ... are much more likely than judges to find for the plaintiff in a defamation case," id. at 1006, surely a reprehensible failing, has been met by the Supreme Court's holding that "[j]udges ... must independently decide whether the evidence in the record is sufficient to cross the constitutional threshold that bars the entry of any judgment that is not supported by clear and convincing proof of 'actual malice.' " Bose Corp. v. Consumers Union, Inc., --- U.S. ----, 104 S. Ct. 1949, 1965, 80 L. Ed. 2d 502 (1984).

It is difficult to see what valid concern remains that has not already been addressed by first amendment doctrine and that therefore requires some constitutional evolving--unless it be, quite plainly, the concern that political publicists, even with full knowledge of the falsity or recklessness of what they say, should be able to destroy private reputations at will.

When its lengthy "balancing" of the "totality of the circumstances" is complete, the concurrence ends up straddling two propositions: First, that the reasonable meaning of this statement is not that Ollman is poorly regarded within his profession. Second, that such an unquestionable libel is permitted in the course of political polemics. The first of these propositions distorts reality. I do not contest the principle that a politically disputatious context, like any element of context, can have some effect upon the properly understood meaning of a statement. If, for example, in the course of a diatribe against Marxist political thought Evans and Novak had written that "Ollman is an incompetent political scientist," the reader might understand that this was merely a corollary of their opinion that Marxism is spinach. But here they did not say he was incompetent. They said that his professional peers regarded him as incompetent--and there is no way that conclusion can be understood to be a product of their econo-political opinions. In fact, they went even further out of their way to dissociate this factual statement from their opinions: they put it in the mouth of one whom they describe as (1) an expert on the subject of status in the political science profession, and (2) a political liberal, i.e., one whose view of Ollman would not be distorted on the basis of greatly differing political opinion. They were saying, in effect, "This is not merely our prejudiced view; it is the conclusion of an impartial and indeed sympathetic expert." Try as they may, however, to convey to the world the fact that Ollman is poorly regarded in his profession, the concurrence insists upon calling it an opinion. It will not do.

Hence the second thread of argument which is subtly woven through the concurring opinion: In the field of political polemics, even statements that are fact rather than opinion must be excused because the reader "is most unlikely to regard [them] as to be trusted automatically." Bork op. at 1010 (emphasis added). Once the reader is "alert[ed] ... that he is in the context of controversy and politics, and that what he reads does not even purport to be as balanced, objective, and fair-minded as ... what is contained in ... news columns," id., he can expect libelous factual statements to be "more of the same," id. And since he would be a fool to believe them, they are not actionable. I am not prepared to accept this novel view that since political debate is always discounted, a decent amount of defamation in that context is protected by the first amendment. Besides the fact that it is unprecedented,1 it is impracticable. Whereas there are some rational limits (if only vague ones) upon what sorts of statements can be considered opinion and hence nondefamatory--limits which are plainly exceeded here--there is really no mechanism to gauge how much defamation is a decent amount.

It is this "risk of judicial subjectivity," id. at 997, rather than that which inheres in the unavoidable need in all libel cases to balance the "totality of the circumstances," id. which troubles me. Beyond that, I may add, I distrust the more general risk of judicial subjectivity presented by the concurrence's creative approach to first amendment jurisprudence. It is an approach which embraces "a continuing evolution of doctrine," id. at 995, not merely as a consequence of thoughtful perception that old cases were decided wrongly at the time they were rendered (see, e.g., Brown v. Board of Education, 347 U.S. 483, 74 S. Ct. 686, 98 L. Ed. 873 (1954)); and not even in response to a demonstrable, authoritatively expressed development of public values (see, e.g., Roberts v. Louisiana, 428 U.S. 325, 336 (1976) (plurality opinion)); but rather in reaction to judicially perceived "modern problems," Bork op. at 995, which require "evolution of the law in accordance with the deepest rationale of the first amendment," id. at 998.2 It seems to me that the identification of "modern problems" to be remedied is quintessentially legislative rather than judicial business--largely because it is such a subjective judgment; and that the remedies are to be sought through democratic change rather than through judicial pronouncement that the Constitution now prohibits what it did not prohibit before. The concurrence perceives a "modern problem" consisting of "a freshening stream of libel actions, which ... may threaten the public and constitutional interest in free, and frequently rough, discussion," id. at 993, and of claims for damages that are "quite capable of silencing political commentators forever," id. at 995. Perhaps that perception is correct, though it is hard to square with the explosion of communications in general, and political commentary in particular, in this "Media Age." But then again, perhaps those are right who discern a distressing tendency for our political commentary to descend from discussion of public issues to destruction of private reputations; who believe that, by putting some brake upon that tendency, defamation liability under existing standards not only does not impair but fosters the type of discussion the first amendment is most concerned to protect; and who view high libel judgments as no more than an accurate reflection of the vastly expanded damage that can be caused by media that are capable of holding individuals up to public obloquy from coast to coast and that reap financial rewards commensurate with that power. I do not know the answers to these questions, but I do know that it is frightening to think that the existence or nonexistence of a constitutional rule (the willfully false disparagement of professional reputation in the context of political commentary cannot be actionable) is to depend upon our ongoing personal assessments of such sociological factors. And not only is our cloistered capacity to identify "modern problems" suspect, but our ability to provide condign solutions through the rude means of constitutional prohibition is nonexistent. What a strange notion that the problem of excessive libel awards should be solved by permitting, in political debate, intentional destruction of reputation--rather than by placing a legislative limit upon the amount of libel recovery. It has not often been thought, by the way, that the press is among the least effective of legislative lobbyists.

In recent years, the Supreme Court confronted a similar assertion of a "modern problem" that required a new first amendment mutant. The omnipresence of the modern press, the popularity of "investigative reportage," and the eagerness of many dissident groups actively to seek out press coverage, have with increasing frequency caused members of the press to be in possession of information regarding unlawful activity, necessary for the detection or prevention of crime. The Court was asked, as the concurrence asks us here, not to take a "wooden" or "mechanical" view of the first amendment, and to proclaim that in modern circumstances it prevents the subpoena of such information. Of course the Court declined. Branzburg v. Hayes, 408 U.S. 665, 92 S. Ct. 2646, 33 L. Ed. 2d 626 (1972). And of course the problem has not gone unaddressed. Many states have enacted "press shield" laws, see In re Roche, 381 Mass. 624, 411 N.E.2d 466, 474 n. 13 (1980), and the federal Justice Department has promulgated regulations, 28 C.F.R. Sec. 50.10 (1983), which approach the issue in a much more calibrated fashion than judicial prohibition could achieve.

For the foregoing reasons, I join Judge Wald's dissent on the professional status point.