Saenz v. Playboy Enterprises, Inc.

653 F. Supp. 552, 13 Media L. Rep. (BNA) 1977, 1987 U.S. Dist. LEXIS 360

• Court: District Court, N.D. Illinois
• Decided: January 16, 1987
• Docket: 81 C 5723
• Status: Published

Opinion

MEMORANDUM AND ORDER

MORAN, District Judge.

This case consists largely of a political and ideological squabble fought out in the bog of defamation law. Plaintiff Adolph Saenz, former Secretary of the New Mexico Department of Corrections, before that an official with the United States Office of Public Safety (O.P.S.), a program of the Agency for International Development (A.I.D.) which was discontinued in 1975 under pressure from Congress, believes that defendants called him a torturer, or worse, in “Thirty-Six Hours at Santa Fe,” an article in the March 1981 Playboy. He seeks vindication not only for himself in this suit, but apparently also for his old agency. Most of the evidence he submits seems intended to show that charges made in the 1970s, that the O.P.S. was somehow implicated in incidents of torture by the police of various Latin American countries, were false.

The article’s author, defendant Roger Morris, who worked briefly for the National Security Council during the early days of Henry Kissinger’s tenure as national security advisor, but has since become a critic of American foreign policy, and publisher, defendant Playboy Enterprises, Inc. (Playboy), contend that they called Saenz no such thing. They move for summary judgment. Defendants nevertheless appear to join the battle on the ground Saenz selects. Much of the evidence they submit appears intended to show that the accusations of the 1970s against the O.P.S. were true.

This court sees no need to try the O.P.S. in absentia. Our preferred role is to resolve legal questions, not political ones. Cf Goldwater v. Carter, 444 U.S. 996, 100 S.Ct. 533, 62 L.Ed.2d 428 (1979); Mallo v. Bell, 430 U.S. 787, 798, 97 S.Ct. 1473, 1481, 52 L.Ed.2d 50 (1977). The article of which plaintiff complains was a piece of political and social criticism. We do not know whose side truth is on, but the First Amendment is on the side of the critic of government. First Amendment protection limits the extent to which defamatory meaning can be extracted from the words of the article. It also requires clear and convincing evidence that Morris and Playboy acted recklessly toward the truth before Saenz can recover from them for defamation. Because the article directly accused Saenz of almost nothing which is constitutionally defamatory, because Saenz does not make out a defamation claim on the one explicit charge, and additionally because defendants were not reckless, the motion for summary judgment is granted.

BACKGROUND

The O.P.S. was an American agency which advised the police forces of various underdeveloped countries. Plaintiff headed the O.P.S. office in Montevideo, Uruguay, between 1965 and 1969. The O.P.S. then posted Saenz first to Panama and eventually back to Washington. His successor in Montevideo, Dan Mitrione, became a hostage of the Tupamaro guerrilla movement in that country and was eventually executed when the Uruguayan government refused the guerrillas’ demands. To persons who supported American policy in Latin America, Mitrione was an honored and mourned hero. However, a character strongly resembling him had the villain’s role in a film highly critical of American policy, “State of Seige,” which accused that character and the O.P.S. generally of helping a repressive Uruagayan regime to squelch dissent. Among other things, the O.P.S. supposedly aided, abetted and even taught local police in the practice of torturing prisoners. Quite independently of the film, the O.P.S. and its International Police Academy in Washington came under criticism from others in the early 1970s, most notably Senator James Abourezk of South Dakota, who became convinced that the agency at least appeared to condone torture by the police forces of several countries which it served. He attacked it both in committees and on the Senate floor. He won over enough of his colleagues that O.P.S. funding was trimmed. A.I.D. finally eliminated the program in 1975.

In 1980, Saenz had the misfortune to be appointed New Mexico’s Corrections Secretary, the head of the state’s corrections system, two days before one of the bloodiest prison riots in American history broke out at the State Penitentiary in Santa Fe. The New Mexico Senate confirmed Governor Bruce King’s appointment of Saenz shortly after the riot. Morris, by then a writer living in Santa Fe, attacked the appointment in a series of articles which appeared in the Santa Fe Reporter, a local weekly. Morris pointed out Saenz’ former O.P.S. connections and dredged up the old accusations against the O.P.S. Governor King removed Saenz in June 1980, after he had served approximately six months.

Morris meanwhile had begun compiling materials for a book on the Santa Fe prison riot, which apparently has since appeared under the title, “The Devil’s Butcher Shop.” As authors often do on their way to a book, Morris put together an article on the same subject, which Playboy purchased and published in March 1981. That piece not only described the riot in some detail but commented on the history of the New Mexico prison system; New Mexico’s society and politics in general, and the careers of Governor King and long time Deputy Secretary of Corrections Felix Rodriguez in particular; the lack, in Morris’ opinion, of significant change in the prison system after the riot, and other more or less related matters. Tangentially to that commentary, Morris brought up the appointment of Saenz as a negative example of Governor King’s prison policy, raising the same questions he had put forward in his earlier newspaper series, albeit more briefly. Set in type, the article covered approximately eleven or twelve pages of the magazine (not counting advertisements or cartoons). Approximately twelve paragraphs of it dealt with either Saenz or the O.P.S. Some sentences were taken virtually verbatim from Morris’ Santa Fe Reporter articles.

Saenz then filed this suit for defamation. (Eventually he would also sue the Santa Fe Reporter as well, but that action in the New Mexico state courts has apparently been dismissed.) In previous memoranda this court has held that Saenz is a public figure for purposes of this suit, that New Mexico law governs, and that under that law Saenz had no claim for libel per se but had stated one for libel per quod sufficient to survive a motion to dismiss. We thereby reduced the complaint from three counts to one. We also held that the portions of the article stating that Saenz was a “drifter” and unqualified for the prisons post were not actionable because they were opinions about a public official protected under the First Amendment. However, we could not assess the rest of the claim because the complaint did not identify the other allegedly defamatory statements with enough specificity. Saenz v. Playboy Publications, Inc., No. 81 C 5723 (N.D.I11. June 14, 1984 [Available on WESTLAW, DCTU database], and May 16, 1983).

Saenz has since amended his complaint. Defendants now move for summary judgment, or in the alternative for judgment on the pleadings, on four grounds. First, they contend that under New Mexico law libel per quod cannot lie unless a plaintiff pleads extrinsic facts, and Saenz has pled no extrinsic facts. Second, libel per quod also requires special damages, and Saenz has not satisfactorily alleged special damages. Third, the First Amendment prohibits the inferences which Saenz wishes to draw from the article in order to give it defamatory meaning; as to what the article actually states, those matters are true, or at least Saenz cannot prove them false. Fourth, defendants cannot be liable as they did not knowingly publish any false statements, nor publish any such statements with reckless disregard for their truth or falsity. The first two grounds are not well taken, but the third and fourth are.

DISCUSSION

I. Pleading Requirements for Libel Per Quod

A. Special Damages

One of New Mexico’s rules for libel per quod is distinctive. To succeed, a plaintiff must plead and prove either special damages in the form of pecuniary loss, or else that the defendant knew of extrinsic facts which made his statement defamatory. Reed v. Melnick, 81 N.M. 608, 471 P.2d 178 (1970). Defendants contend that Saenz has not sufficiently pled special damages since he has not alleged a dollar amount for them.

This court disagrees. A specific figure for the pecuniary loss is not necessary at the pleading stage. What plaintiff needs is an allegation of pecuniary loss which reasonably could be reduced to a dollar amount and proven. See Action Repair, Inc. v. American Broadcasting Co., 776 F.2d 143, 150 (7th Cir.1985). Saenz admits that he was unemployed when the Playboy article appeared, so he cannot claim that the article cost him his job. However, he had by then left New Mexico and was seeking employment in Washington. He contends that the publication of the Playboy article gave national circulation to what had until then been a merely local controversy, thus greatly extending his jobless period. The contention sounds difficult to prove, but it is not unreasonable, and it could show pecuniary loss. The complaint does not fall on that ground.

B. Extrinsic Facts

1. Inducement

Defendants also argue that libel per quod demands that a plaintiff plead and prove the extrinsic facts which have given the statement a defamatory meaning, whereas Saenz has pled no extrinsic facts. “Extrinsic facts” in this context means the “inducement” of common law pleading: facts not in the allegedly defamatory statement itself, but which were known to the readers or hearers and which, when coupled with the statement, gave it a defamatory impact. See, e.g., Ratner v. Young, 465 F.Supp. 386, 395 (D.V.I., 1979); Restatement (2d) of Torts, § 563, comment f (1977). The classic example was a newspaper’s congratulatory announcement of the birth of a child. Innocent on its face, it became defamatory when coupled with the extrinsic fact that the (childless) parents had only been married for a month. Morrison v. Ritchie & Co., [1901-02] Sess.Cas. 645 (Scot.2d Div.), 39 Scot.L.Reptr. 432 (1902).

Saenz provides no such extrinsic facts. He appears to have confused pleading an inducement with pleading malice. For example, he attempts to label statements and letters from Morris to Morris’ literary agent, and statements in internal Playboy memoranda, “extrinsic facte.” Presumably he takes this direction in the hope of shaping his claim to New Mexico’s rule for publishers who knew of extrinsic facts. He urges this court not to dismiss on this ground, to allow him to unearth more “extrinsic facts” through discovery. The effort fails. Extrinsic facte must be known to the statement’s readers or hearers. By definition, the extrinsic facts relevant to defamation could not be hidden, to be found only through discovery. Saenz has not pled any extrinsic facte.

2. “Innuendo”

However, it does not necessarily follow that he therefore fails to state a claim for libel per quod under New Mexico law. New Mexico applies an “innocent interpretation” test to distinguish between libel per se and libel per quod. Only statements that are susceptible of a single meaning which is defamatory are libel per se. If the statement can be interpreted innocently, a claim on it is only possible through libel per quod. Reed, 471 P.2d at 182; Monnin v. Wood, 86 N.M. 460, 525 P.2d 387 (N.M. App.1974). Thus any claim which depends on “insinuations [or] innuendoes” for its defamatory impact is a per quod claim. Monnin, 525 P.2d at 389. This court previously held that Saenz had no claim for libel per se, in part on the ground that the references to him in the article were not solely susceptible of a defamatory meaning.

Given that rule, however, under New Mexico law not all per quod claims would need extrinsic facte. In common law plead ing, “innuendo” had a precise and technical definition. One pled extrinsic facts in the inducement. Then, in the innuendo, one argued what the words meant in light of the inducement to those who read or heard them. Thus there could be no innuendo without extrinsic facts. See W. Prosser, Handbook of the Law of Torts 748-749 (1971). For example, the innuendo in the birth announcement case would have been a pleading that the announcement accused the wife of pregnancy before marriage. More recently, however, courts have used “innuendo” in a more general sense, to mean what a reader or hearer could infer or insinuate from the text itself. See, e.g., Sellers v. Time, Inc., 423 F.2d 887 (3d Cir.), cert, denied, 400 U.S. 830, 91 S.Ct. 64, 27 L.Ed.2d 67 (1970); Eaton, The American Law of Defamation Through Gertz v. Robert Welch, Inc. and Beyond: an Analytical Primer, 61 Va.L.Rev. 1349, 1355 n. 20 (1975). It is in this latter sense that the New Mexico courts have used “innuendo.” See Reed, 471 P.2d at 182; McGaw v. Webster, 79 N.M. 104, 44Ó P.2d 296 (1968); Monnin, 525 P.2d at 389.

Libel per quod in New Mexico thus covers two possibilities. One is like libel per quod at common law, using extrinsic facts to flesh out defamatory meaning. The other, however, arises from a statement which on its face has two or more meanings, one of them innocent, one defamatory. Mar-chiondo v. New Mexico State Tribune Co. (Marchiondo I), 98 N.M. 282, 648 P.2d 321, 327 (N.M.App.1981), overruled in part on other grounds by Marchiondo v. Brown (Marchiondo II), 98 N.M. 394, 649 P.2d 462 (1982). The latter type of libel per quod does not require proof of extrinsic facts. Rather, New Mexico law merely makes the question of which of the possible meanings the statement actually had — the question of the permissible “innuendo” in the general sense — a question of fact for the jury. Marchiondo I, 648 P.2d at 327-328; see Reed, 471 P.2d at 182.

Plaintiff argues precisely that type of claim. He does not contend that Playboy’s readers knew additional facts about him which caused a statement innocent on its face to take on a 1 defamatory meaning. Rather, he contends that although the article does not state in so many words that he was a torturer, by referring to him and the allegations against the O.P.S. in the same breath, as it were, it insinuates that he was a torturer. In other words, he contends that the article accuses him of torture by “innuendo.” Since Saenz has pled special damages, that claim appears viable under New Mexico law. McGaw, 440 P.2d at 299. The absence of extrinsic facts in the pleadings is not fatal to Saenz’s claim.

II. Constitutional Limits on Reading Criticism of a Government Agency as Defamation of an Individual

Saenz therefore argues that his claim should survive summary judgment because it raises a question of fact for a jury. The references to him in the Playboy article can be interpreted in more than one way. Only a trier of fact, he contends, may decide what meaning should be inferred from the article.

When statements appear within a work of political or social criticism, however, there are some interpretations which juries are not allowed to make, both under New Mexico law, see, e.g., Marchiondo II, 649 P.2d at 469, and as a matter of federal constitutional law. Where the statement comes as part of speech which is highly valued under the First Amendment, constitutional law imposes strict limits on the freedom of juries to interpret the statement as defamatory and so reduce its First Amendment protection. See Bose Corp. v. Consumers Union of United States, Inc., 466 U.S. 485, 501, 104 S.Ct. 1949, 1959, 80 L.Ed.2d 502 (1984); Monitor Patriot Co. v. Roy, 401 U.S. 265, 277, 91 S.Ct. 621, 628, 28 L.Ed.2d 35 (1971); Rosenblatt v. Baer, 383 U.S. 75, 82, 86 S.Ct. 669, 674, 15 L.Ed.2d 597 (1966). All the inferences which Saenz wants a trier of fact to draw from the Playboy text would run up against those limits. Some implicate the limit on reading criticism of government as defamation of an individual. Rosenblatt, 383 U.S. at 81, 86 S.Ct. at 673. Others involve the limit on inferring statements of defamatory fact out of protected expressions of opinion. Old Dominion Branch No. 496, National Association of Letter Carriers v. Austin, 418 U.S. 264, 284, 94 S.Ct. 2770, 2781, 41 L.Ed.2d 745 (1974).

A. Constitutional Limits on a Colloquium

Common law pleading rules for libel per quod also required a “colloquium.” A plaintiff who by inducement and innuendo had shown that the statement had a defamatory meaning, further had to plead and prove that the statement was made “of and concerning the plaintiff” — i.e., that the defamatory meaning attached to him personally. Restatement, § 563 comment f. A statement could be shown to be “of and concerning” him through inference as well as direct reference. Id., § 564 and comments.

However, when allegedly defamatory statements come as part of criticism of a government agency, the question of whether they can be read to attach to an individual takes on constitutional significance. The freedom to criticize the conduct of government lies “at the very center of the constitutionally protected area of free expression.” New York Times Co. v. Sullivan, 376 U.S. 254, 292, 84 S.Ct. 710, 732, 11 L.Ed.2d 686 (1964). Damage awards from libel actions place a heavy burden on freedom of expression. Id. at 277-278, 84 S.Ct. at 724-725. It follows that criticism of government should not be transmuted into statements actionable at defamation unless we are quite sure that the publisher also attacked an individual. Id. at 288, 84 S.Ct. at 730; Rosenblatt, 383 U.S. at 81, 86 S.Ct. at 673.

In Sullivan, a newspaper advertisement harshly denounced the behavior of the Montgomery, Alabama police towards civil rights demonstrators. The city commissioner in charge of police sued the newspaper for libel. The most noted section of the opinion held that he could not collect without a showing of “actual malice” — that the newspaper not only published false statements, but either knew they were false or recklessly disregarded whether they were false or not. 376 U.S. at 280, 84 S.Ct. at 726. However, the Sullivan court also held that the degree of First Amendment protection for speech against the government imposed an additional barrier to his suit. He needed far more proof than he had shown that criticism of the police amounted to a libelous attack which attached to him. Id. at 288, 84 S.Ct. at 730. The court later expanded on this holding in Rosenblatt. A newspaper columnist had praised the financial success of the new management of a county ski area and then asked, “What happened to all the money last year? and every other year?” 383 U.S. at 78, 86 S.Ct. at 672. The previous supervisor of the ski area sued for defamation. The court found no “explicit charge” of corruption “specifically directed at the plaintiff” and held that the First Amendment did not permit the jury’s inference that the statement attacked him personally. Id. at 81-82 and n. 6, 86 S.Ct. at 673-674 and n. 6.

Plaintiff here argues that we should not grant summary judgment because a trier of fact might infer that the following passages accused him of being a torturer:

At the ardent recommendation of both the state’s chief justice and its purchasing director, the governor selected without further question a native Hispanic New Mexican and erstwhile federal official said to have much law-enforcement experience. On January 31, the date of the fateful intelligence meeting of pen officials, King announced the appointment of Adolph B. Saenz, “a person of such excellent quality and expert experience,” said the governor, that it was a “great pleasure” to have him. Summoned to the riot-torn pen that weekend, Saenz was later swiftly confirmed by the New Mexico senate to preside over rebuilding the state’s shattered facility and reputation.

What no one in the Statehouse knew, or acknowledged, was that the vaunted new corrections secretary had spent 17 years in the U.S. Office of Public Safety (O.P.S.), a C.I.A.-inspired program established in the late Fifties to advise foreign police in suppressing political dissent in Latin America and elsewhere — and then abolished by bipartisan Congressional action 20 years later amid well-documented charges of U.S. complicity in torture and political terror.

* He * * * #

By the time [Saenz] left Panama to return to teach at the International Police Academy in mid-1974, the O.P.S. was under rising condemnation in the U.S. Congress and press. When U.S. advisers were exposed in the scandal of the infamous “tiger cage” underground torture cells in South Vietnam, when then-Senator James Abourezk revealed the existence of a torture “school” in Texas for foreign police, when columnist Jack Anderson and other researchers could find theses, films and other documents at the International Police Academy dealing with torture, Congress moved in bipartisan action to abolish the O.P.S. — the only agency so eliminated in the post-war period. Without opposition from a Ford Administration fearing a full investigation, the O.P.S. was disbanded by 1975.

This court doubts that any statement of and concerning Saenz could be inferred from those passages even at common law defamation. Statements which refer to an organization ordinarily do not attach to its members. See, e.g., Provisional Government of the Republic of New Afrika v. American Broadcasting Cos., 609 F.Supp. 104, 108 (D.D.C.1985). But in any case, constitutional protection for critical assessments of government performance prevents any inference from this description of accusations against the O.P.S. as an agency to an attack on Saenz as an individual.

Saenz argues against that conclusion by attempting to distinguish Sullivan and Rosenblatt. He points out that in neither case had the allegedly defamatory text used the plaintiff’s name, while here the text expressly mentions him. However, we do not think that these holdings rest solely on the presence or absence of a plaintiff’s name somewhere in the text. See Price v. Viking Press, Inc., 625 F.Supp. 641, 645 (D.Minn.1985) (book criticizing FBI’s handling of Indian protest action mentioned name of agent). Even in cases which do not present the extra consideration of criticism of government conduct, a plaintiff cannot satisfy the requirement that something was said of and concerning him merely on the juxtaposition of his name with negative comments about others. For example, a television report did not state of and concerning an insurance agent being investigated for fraud that he had been indicted, merely by using his name and picture along with a story about two indicted agents. Lauderback v. American Broadcasting Co., 741 F.2d 193, 196 (8th Cir.1984), cert, denied, 469 U.S. 1190, 105 S.Ct. 961, 83 L.Ed.2d 967 (1985). Similarly, song lyrics which accused two witnesses of lying on the stand, did not accuse another witness of lying simply by mentioning her and substantially true facts about her in the same song. Valentine v. C.B.S., Inc., 698 F.2d 430, 431-432 (11th Cir.1983). A reference to an individual alongside criticism of others does not of itself mean that the criticism rubs off on that individual.

Thus Sullivan and Rosenblatt cannot stand merely for the proposition that a government official has a libel action if the text mentions him by name but cannot bring one if it does not. Rather, as we read them, they generally increase the burden of proving a colloquium, of proving that the statement was made “of and concerning the plaintiff,” when the plaintiff is a government official. The First Amendment protects criticism of government. Thus an official trying to prove that such criticism defamed him personally must be able to show “an explicit charge” “specifically directed” at him. Rosenblatt, 383 U.S. at 81, 86 S.Ct. at 673. Moreover, he must be able to show the attachment to him with convincing clarity. The city commissioner in Sullivan had produced seven witnesses who testified that they read the advertisement as directing charges at the commissioner. The court held that proof “constitutionally insufficient” to show a charge of and concerning the commissioner. 376 U.S. at 292, 84 S.Ct. at 732. The constitutionally sufficient standard elsewhere in the opinion is proof with “convincing clarity.” 376 U.S. at 285-286, 84 S.Ct. at 728-729. Since the testimony of seven witnesses was not enough, the court must have been applying that standard. In short, Sullivan held that when a plaintiff seeks to convert criticism of government to defamation, the First Amendment requires that he prove his colloquium with convincing clarity. We conclude that criticism of government is not actionable defamation unless a plaintiff clearly and convincingly shows an explicit charge specifically directed at him. The presence of Saenz’s name in the text is one factor in his proof, but it cannot alone provide the convincing clarity he needs.

His claim on these passages founders because they do not specifically direct any charge at him. The Rosenblatt court, in explaining why the First Amendment did not permit the jury’s finding of defamation in that case, said:

The jury was permitted to award damages upon a finding merely that respondent was one of a small group acting for an organ of government, only some of whom were implicated, but all of whom were tinged with suspicion. In effect, this permitted the jury to find liability merely on the basis of his relationship to the government agency, the operations of which were the subject of discussion.

383 U.S. at 82, 86 S.Ct. at 674. The Constitution likewise does not allow the inference Saenz argues here. At most, the passages tinge members of the O.P.S. with suspicion. Permitting the inference that they labeled Saenz a torturer would be permitting the imposition of liability merely on the basis of his relationship to a government agency.

B. Constitutional Limits on “Innuendo”

The policy goals of Sullivan and the language of Rosenblatt also set constitutional limits on the defamatory “innuendo” which may be drawn from criticism of government, just as there are limits on the colloquium. The Sullivan court applied First Amendment values to defamation law and concluded that statements on political issues or about public officials deserve more protection from defamation actions than other kinds of statements. It also held that some of that protection functioned at the stage of construction of the allegedly defamatory text — specifically, in that instance, a restriction on the colloquium which could be extracted from criticism of government. Rosenblatt then held that a public official attempting to transform criticism of government into defamation must be able to show “an explicit charge.” 383 U.S. at 81, 86 S.Ct. at 673. The Court’s choice of language indicates a similar principle of construction applicable to the “innuendo” (in its non-technical sense) a public official may assert. If the accusation must be explicit, it follows that defamatory meaning cannot rest on inference and insinuation. In other words, Rosenblatt strongly suggests that when the allegedly defamatory text involves criticism of governmental conduct, a public official cannot show defamatory meaning by “innuendo” from the text.

We know of no case which spells out such a principle of construction in so many words. However, the principle is consistent not only with the results of Sullivan and Rosenblatt but also with later decisions involving criticism of governmental officials. If criticism of public officials is to have the “breathing space” which Sullivan, 376 U.S. at 272, 84 S.Ct. at 721, mandates for it, then the question of whether that criticism should be protected from the threat of a libel suit must be defined largely by the statement’s function and its practical impact. If it functions to criticize governmental performance and the speaker intends it to improve government, even though the improvement the speaker has in mind is the removal from office of a particular official or that he never be considered for office again, then the statement ought to enjoy substantial First Amendment protection. The official should not be able to use inference and implication to transform it into defamation of him as a private individual. See generally Oilman v. Evans, 750 F.2d 970, 979-981 (D.C.Cir.1984) (en banc), cert, denied, 471 U.S. 1127, 105 S.Ct. 2662, 86 L.Ed.2d 278 (1985); id. at 994-1005 (Bork, J., joined by three other judges, concurring) (reviewing cases); Lewis, New York Times v. Sullivan Reconsidered: Time to Return to “The Central Meaning of the First Amendment,” 83 Colum.L.Rev. 603, 620-622 (1983).

Thus, for example, the Eighth Circuit, sitting en banc, refused to permit a public official to rest an action for defamation merely on the “innuendoes” in a statement criticizing him. Janklow v. Newsweek, Inc., 788 F.2d 1300 (8th Cir.1986). A news magazine had discussed charges against a current governor made some nine years previously when he had been state attorney general. In its history of the events, it said, “Eight months later [the governor] was prosecuting [the author of the charges].” In fact the prosecution was under way before the charges had ever been made, and was merely continuing eight months later. The governor found in that statement a defamatory implication that he had initiated the prosecution solely for revenge, thus abusing his office. The court held that the statement was protected by the First Amendment. It relied for its result on the distinction between fact and opinion (discussed infra, section III) rather than a restriction on permissible “innuendo.” But it read the statement narrowly in part because it “involve[d] criticism of the motives and intentions of a public official,” and under those circumstances refused to extract defamatory meaning from an “imprecise” accusation. 788 F.2d at 1305. In short, the court could equally have based its result on the lack of an explicit charge.

Similarly, in Davis v. Costa-Gavras, 619 F.Supp. 1372, 1384 (S.D.N.Y.1985), the court refused to insinuate a defamatory charge from a text criticizing governmental conduct, though it did not cite Rosenblatt and it used the word “concrete” rather than “explicit.” Three American officials who served in Chile during the period recreated in the film, “Missing,” claimed that the film defamed them by accusing them of complicity in the murder of an American citizen during the military coup which overthrew Chilean President Salvador Allende. Undisputedly, the film strongly charged American involvement in the coup, had three characters easily identified with these individuals, and portrayed the individuals unfavorably. Two of them, however, had no claim. They could not show a concrete accusation that they were personally involved in the murder. 619 F.Supp. at 1383-1386. See also Oilman, 750 F.2d at 980 (majority) (emphasizing that accusation must have “definiteness” to be defamatory); id. at 1004-1005 (Bork, J., concurring) (reading Supreme Court cases as indicating different standards of construction when the plaintiff has voluntarily “entered the political arena” and the statement dealt with “a topic of legitimate political concern”).

We conclude that our reading of Sullivan and Rosenblatt, while perhaps novel, is at most a different route of reaching the result which other courts have found that the First Amendment demands. The defamation claims of government officials arising from a text which criticizes governmental conduct must be based on statements which make an explicit charge, not on insinuations of defamatory meaning from possible implications in the text. To avoid chilling criticism of government, an official cannot constitutionally argue a defamatory “innuendo” from a text which merely may have implied something negative about him.

Saenz contends that defendants again called him a torturer in the following graphic and unpleasant passage:

The whistle was blown on O.P.S. after the body of U.S. police adviser Daniel Mitrione was found crumpled in an old Buick convertible on a barrio side street of Montivedeo, Uruguay, on August 10, 1970. Kidnapped by the Tupamaro urban guerrillas, and then killed when the Uruguayan government refused to ransom him by freeing political prisoners, Mitrione was flown back to the States to a martyr’s funeral by the Nixon Administration. While the murder aroused anti-terrorist sentiments, it also stirred a mounting controversy over U.S.-supported repression in Latin America. Scores of Latin journalists, clergy and others told of grisly police torture of political prisoners in Uruguay, Brazil and elsewhere. Stripped, beaten, sexually abused, tortured under water and on racks, burned with electric needles under fingernails, shocked with electrical wires on the breasts of women and the testes of men, the victims described their agonies in accounts that repeatedly implicated the O.P.S. U.S. advisers were said to have supplied the torture devices, instructed their Latin police clients in the latest techniques, in some cases even been present at or participated in the sessions.

Mitrione, Washington’s official martyr, became a prominent figure in much of the emerging scandal. But in the accumulating evidence about police torture in Uruguay, there were also numerous reports of atrocity for sometime prior to Mitrione’s arrival in Montevideo in the late summer of 1969. The evidence poured in not only from political dissidents and victims but from a multi-party inquiry by the Uruguayan senate, from the Nobel Prize-winning Amnesty International, from a former Uruguayan police commissioner who resigned in revulsion, from another police official who was tortured himself as a suspected Tu-pamaro spy, from Catholic priests and then from the U.S. Catholic Conference of Bishops. Torture in Uruguay, said the array of authorities, had been “common,” “normal,” “habitual” before 1969. And the U.S. adviser who had been Mi-trione’s predecessor for four years, whose office was on the first floor of the Montevideo jefatura, where torture reportedly took place and the screams of victims reverberated, who by his own account had intimate and influential relations with the Uruguayan police, was Adolph Saenz.

From Montevideo, allegations of torture by his police clients would follow Saenz through subsequent assignments in Colombia and Panama____

This passage, without doubt, makes an accusation against Saenz personally as well as against the O.P.S. generally. However, though it portrays him unfavorably, as in Davis the portrayal cannot constitutionally support the inference Saenz wants a trier of fact to make. See 619 F.Supp. at 1384, 1386. Since the passage comes in the context of criticism of a government agency, any charge must be explicit. This text does not explicitly accuse Saenz of actively participating in torture. Indeed, it quite specifically says that any actual torture was carried on by the “police clients,” the local police of Uruguay, and perhaps of Colombia and Panama. What it says of Saenz, at most, is that he knew or should have known of torture and perhaps condoned it, which is a quite different charge. Cf Davis, 619 F.Supp. at 1385-1386. Defendants do not contest that interpretation of the passage. Saenz cannot show from these passages a concrete accusation that he or any other individual O.P.S. officer personally participated in torture.

Saenz maintains that the text must at least be read to say that he taught torture, supplied torture devices and watched torture sessions. However, the passage does not say that of Saenz. The paragraph in which those comments occur deals generally with the O.P.S. and the charges which emerged after 1970 about O.P.S. operations “in Uruguay, Brazil and elsewhere.” Stretching a point, it may possibly make those accusations of Mitrione, who is of course dead and cannot be defamed, though we doubt that even that reading is constitutionally permissible. But Saenz appears only in the next paragraph, after words which signal a shift in subject matter both as to time and place. It says that while most of the charges about Uruguay focused on Mitrione’s period and later, there were allegations of torture there, and testimony from victims, which went to the period from before 1969 when Saenz was head of the O.P.S. in Montevideo. Explicitly, concretely, it says no more of Saenz than that he was in a position to know about torture by Uruguayans and perhaps also Colombians and Panamanians, leaving the reader to conclude that Saenz either knew or should have known of it.

III. Constitutional Protection for Opinion on Political and Social Issues and About Public Officials

Any discussion of the explicitness of a charge in an allegedly defamatory text, flows naturally into a discussion of another obstacle to Saenz’s suit: the strong First Amendment protection for opinion in works of political and social criticism. In such a context the question of what a particular passage means is a matter of constitutional law as well as tort law, regardless of whether the text criticizes a government agency or not. Buckley v. Littell, 539 F.2d 882, 888 and n. 3 (2d Cir.1976), cert, denied, 429 U.S. 1062, 97 S.Ct. 785, 50 L.Ed.2d 111 (1977) (“fascist” and “fellow traveler” held expressions of opinion). Cf Austin, 418 U.S. at 284, 94 S.Ct. at 2781 (labor union publication’s opinion that a “scab” is a “traitor” held not actionable). A trier of fact may not infer a defamatory factual statement from a protected expression of opinion. Austin, 418 U.S. at 285 n, 16, 94 S.Ct. at 2782 n. 16. Whether or not the statement is constitutionally protected is a question of law for the court. Janklow, 788 F.2d at 1305 n. 7; Oilman, 750 F.2d at 978. 1

A. The Oilman Factors

Courts have struggled with various definitions of opinion, but recently most courts have looked to the en banc decision of the District of Columbia Circuit in Oilman for guidance. See, e.g., Janklow, 788 F.2d at 1302; Woods v. Evansville Press Co., 791 F.2d 480, 487 (7th Cir.1986). In Oilman, a nationally syndicated columnist denounced a Marxist professor up for an appointment as department chairman of a state university. The court found that the entire article was opinion, including a quote from an anonymous colleague that the professor had “no status” among scholars in his discipline. In so holding, that court “analyze[d] the totality of the circumstances in which the statements are made,” using several factors which would tend to affect “whether the average reader would read the statement as fact or, conversely, opinion.” Oilman, 750 F.2d at 979. In this approach no single factor is necessarily decisive, not least because the question is ultimately one of the appropriate constitu tional protection for the statement rather than the average reader’s reaction to it. Id. at 979 n. 16; see also id. at 1004-1005 (Bork, J., concurring); Janklow, 788 F.2d at 1302. 2 But the factors sharpen the focus of the analysis and we adopt them here.

1. type of publication

One factor is the type of publication or broadcast in which the statements are found — whether a reader is more likely to expect fact or opinion given the nature of the publication. Oilman, 750 F.2d at 983-984. For example, a reader reasonably anticipates a greater proportion of opinion in a weekly national magazine such as Newsweek or Time than in a local daily newspaper where the traditions of straight factual reporting are stronger. Janklow, 788 F.2d at 1304 (implication about attitude of Western governor toward Indians held opinion); Lewis v. Time, Inc., 710 F.2d 549, 553 (9th Cir.1983) (identifying lawyer as one of the shadier practitioners in his profession held opinion). In the instant case, the article appeared in Playboy. It is fair to say that this type of publication has not won its circulation on the strength of its straight news reporting. Indeed, a reader of this type of publication is at least as likely to look to it for fantasy as for fact. Pring v. Penthouse International, Ltd., 695 F.2d 438 (10th Cir.1982), cert, denied, 462 U.S. 1132, 103 S.Ct. 3112, 77 L.Ed.2d 1367 (1983). The reader of this publication was fairly on notice to expect something other than fact from the outset, indeed from the front cover.

2. nature, tone and style of the work

Another factor is the degree to which the statements appear to be made in what one court has called a “public context.” Jank-low, 788 F.2d at 1304; cf. Oilman, 750 F.2d at 983. Certain types of work are obviously intended as contributions to public debate on significant issues: political and social criticism, speech about government and its officers. The nature of the work itself signals to the reader a strong likelihood of the presence of opinion. Janklow, 788 F.2d at 1304-1305; Oilman, 750 F.2d at 988 and n. 35; id. at 1004 (Bork, J., concurring). Such commentary, of course, is also the type of work entitled to maximum First Amendment protection. FCC v. League of Women Voters of California, 468 U.S. 364, 382, 104 S.Ct. 3106, 3119, 82 L.Ed.2d 278 (1984).

A closely-related factor, at least when the subject matter is political and social criticism, is the tone and style of the piece. Certain types of work signal by their style, or the traditions with which they identify, that they are vehicles for criticism and so for opinion. Oilman, 750 F.2d at 982; cf. Greenbelt Cooperative Publishing Ass’n v. Bresler, 398 U.S. 6, 14, 90 S.Ct. 1537, 1541, 26 L.Ed.2d 6 (1970) (context of political debate affects meaning of language). For example, the editorial cartoon of a newspaper is instantly recognizable as an expression of opinion. Keller v. Miami Herald Publishing Co., 778 F.2d 711 (11th Cir.1985). Readers also expect opinion from a column with a byline located on the editorial page. Oilman, 750 F.2d at 986. Even without such obvious signals, however, the tone, style and choice of language of a work can put the reader on notice that the purpose of the piece is to persuade, to spark debate, to criticize — in other words, to offer opinion. Id. at 987; Lewis, 710 F.2d at 553; Price, 625 F.Supp. at 647. The writer makes it clear that he is speaking from a particular point of view and the reader expects expression of that point of view.

“Thirty-Six Hours at Santa Fe” appeared under the byline of Roger Morris. In tone and style, from the outset, it signals to the reader that it is a piece of political and social criticism, made in a public context, intended to persuade and spark debate. Its text begins:

Built a quarter century ago with the usual grudging appropriation and local graft, the New Mexico state penitentiary stands 11 miles outside fashionable Santa Fe....

Long before any mention of Saenz is made, Morris’ choice of language and subject matter made it clear that he spoke from a point of view highly critical of New Mexico politics, the prison system and the King administration. For example:

Still worse, along with the swelling inmate population came the squalid sociology of New Mexico’s corrections bureaucracy. The state in the Fifties left the custody and reform of its felons largely to a poor, meagerly educated underclass of native Hispanics. The arrangement only mirrored local society beyond the prison walls, in which most of New Mexico’s Hispanic majority — a people of proud cultural parochialism and ancestry in this country before the Pilgrims landed at Plymouth Rock — lived in economic and social subordination to their own small aristocracy and to the growing Anglo minority whose money and men controlled the state.

******

In the politics of New Mexico’s poverty and racism, the penitentiary promptly became a center of bureaucractic nepotism and corruption, fostering the inevitable clique of administrators and guards tied by family, complicity and a shared incompetence in jobs none could afford to lose.

******

Finally, the circle closed with politicians who found no reason to clean out a state sinecure that was a virtual commissary of votes and graft.

******

[Governor Bruce] King’s hallmark as a politician would be a penchant for expensive cowboy boots and the good-ol’-boy twang and manner of the great panhandle plains that lap over into the state from the east, depositing grassland, money, political reaction and ethnic bigotry.

******

The carnage at the New Mexico pen, and the sequel of official blunders and evasions through 1980, were a natural outgrowth of the state’s shabby prison politics over the preceding years. Throughout the Seventies, incompetence or corruption, or perhaps both, ruled the local correction system with brazen impunity, the misgovemment of the penitentiary a mocking fulfillment of the New Mexico state motto, “crescit eundo,” “It grows as it goes.”

We do not think it possible that a reader could confuse “Thirty-Six Hours” with purely objective, neutral, straight factual reporting. Its nature, tone and style all indicate opinion.

3. difficulty of inferring factual statements when the context is opinion

That a piece is easily identifiable as a work of political criticism and opinion does not, of course, give the writer license to make whopping factual misstatements blatantly injurious to reputations. See, e.g., Cianci v. New Times Publishing Co., 639 F.2d 54, 64 (2d Cir.1980) (accusation that mayor once bought his way out of a rape charge); Brown & Williamson Tobacco Corp. v. Jacobson, 644 F.Supp. 1240, 1256-1258 (N.D.I11.1986) (“they’re liars” held factual although made during identifiable “opinion” portion of news broadcast). But it does affect whether statements can fairly be read to convey facts when seen in context. Oilman, 750 F.2d at 980 and n. 18, 991; see also 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). When an ambiguous statement is surrounded by opinion, it is far more likely to be seen as an expression of opinion than as an assertion of fact. Oilman, 750 F.2d at 981, 991; see also Janklow, 788 F.2d at 1304; Buckley, 539 F.2d at 892-894. A reader of criticism expects rhetorical hyperbole and vivid metaphor, so the use of lively language is understood as hyperbole and metaphor, not as fact. Austin, 418 U.S. at 272, 283, 94 S.Ct. at 2780; Greenbelt, 398 U.S. at 14, 90 S.Ct. at 1541; Keller, 778 F.2d at 716; see also Mr. Chow of New York v. Ste. Jour Azur, S.A., 759 F.2d 219, 228 (2d Cir.1985) (restaurant reviewer not liable for describing Chinese dumplings as resembling “bad Italian ravioli”). New Mexico law in particular has shown a high degree of tolerance for fiery political invective and sarcasm, shrugging off such epithets as “rabid” and “violent” as mere opinion. Marchiondo I, 648 P.2d at 331-332; Kutz v. Independent Publishing Co., 97 N.M. 243, 638 P.2d 1088 (N.M.App.1981). And in an opinionated piece, the mere fact that the account is selective, omitting some material, thus producing a treatment which is not entirely fair, is no surprise. The omission and slanting does not convert opinion into a statement of fact unless it leaves a serious misimpression about the facts. Janklow, 788 F.2d at 1306.

B. Application to a Public Official

First Amendment protection for robust debate on important political issues requires us to find that the Constitution bars an action on nearly everything which Saenz alleges is defamatory. As the Supreme Court has recently observed, “bringing critical judgment to bear on public affairs” is part of the historic responsibility of a free press. FCC, ' 468 U.S. at 382, 104 S.Ct. at 3119. To protect that historic role, what comments about political issues and public officials mean for the purpose of defamation law must be determined in light of the constitutional imperatives which encourage such comment. Buckley, 539 F.2d at 888-889. As Judge Bork of the District of Columbia Circuit has put it,

... 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 ... he should expect to be jostled and bumped in a way that a private person need not expect ... the debate will sometimes be rough and personal.

Oilman, 750 F.2d at 1002 (Bork, J., concurring). The debate may fairly include dredging up old accusations out of his distant past, see Monitor, 401 U.S. at 275, 91 S.Ct. at 627, as well as very harsh language about his present conduct. Greenbelt, 398 U.S. at 13, 90 S.Ct. at 1541. A court must keep that context and the strength of constitutional protection for it in mind when deciding whether a given statement constitutes fact or opinion. Austin, 418 U.S. at 284, 94 S.Ct. at 2781; Greenbelt, 398 U.S. at 14, 90 S.Ct. at 1541; Oilman, 750 F.2d at 1004-1005 (Bork, J., concurring).

1. Saenz as an O.P.S. official

All the statements about Saenz’s O.P.S. career fall into the category of protected opinion. Morris could properly express opinion about Saenz’s past in order to discuss whether or not he was the best person available to head the prison system in 1980. See Monitor, 401 U.S. at 275, 91 S.Ct. at 627; Ocala Star-Banner Co. v. Damron, 401 U.S. 295, 91 S.Ct. 632, 28 L.Ed.2d 57 (1971). Such an opinion can include a materially accurate report of past allegations levied at a public official without stepping over the line into fact, as long as the publisher does not himself espouse and endorse the charges. Compare Cian-ci, 639 F.2d 54, with Edwards v. National Audubon Society, Inc., 556 F.2d 113 (2d Cir.), cert, denied, 434 Ü.S. 1002, 98 S.Ct. 647, 54 L.Ed.2d 498 (1977) (newspaper report that society called scientists “liars” protected). Though there are minor errors, and some disputed matters relating to Saenz personally, in Morris’ account — for example, the O.P.S. was an agency of A.I.D. rather than the C.I.A., and Saenz disputes the location of his office — essentially the account recapitulates the Charges made about O.P.S. in the 1970s. Saenz contends that Morris inaccurately dates the charges, since the first published reports apparently did not appear until 1970, but those reports referred to incidents which occurred in 1969 and earlier. Saenz also thinks that Morris should have included O.P.S. denials of the charges, but omission of a denial does not convert such an account of the past into a presently actionable assertion of fact. Janklow v. Newsweek, Inc., 759 F.2d 644, 648 (8th Cir.1985), reh’g en banc on other grounds, Janklow, 788 F.2d 1300.

The protective umbrella of opinion even covers the charge that defendants admit they did make, that Saenz knew or should have known of torture going on around him. It is true that the question of whether Saenz knew of torture in Uruguay or not would, as a state-of-mind question, be called a “question of fact” if it was an element of a tort and disputed on a summary judgment motion. But a question of fact for purposes of Federal Rule of Civil Procedure 56 is not the same thing as a statement of fact for purposes of the distinction between fact and opinion in defamation law. Janklow, 788 F.2d at 1302. In a discussion of a public official, questions about his state of mind — what he knew or didn’t know, his intent or motives — are often topics of public debate on which an expression of opinion is appropriate. Garrison v. Louisiana, 379 U.S. 64, 77, 85 S.Ct. 209, 217, 13 L.Ed.2d 125 (1964); Oilman, 750 F.2d at 1004 (Bork, J., concurring).

This court concludes that the statement is protected as a matter of both constitutional and New Mexico law. See Janklow, 788 F.2d at 1305; Marchiondo II, 649 P.2d at 468 (charge of intent to use improper influence). The text does not suggest that the opinion is based on undisclosed defamatory facts. Cf. Marchiondo I, 648 P.2d at 232; Kutz, 638 P.2d at 1090. Indeed, the basis for the opinion, the accusations of others that the Uruguayan police tortured their prisoners and Saenz’s own assertion of close relations with the police, is set out in the article itself. The charge that Saenz knew or should have known of torture is protected by the First Amendment.

In so holding, we do not imply that the First Amendment would necessarily protect as opinion the same charge levied at a private citizen. Cf Oilman, 750 F.2d at 1004-1005 (Bork, J., concurring). Because the context, political criticism of public officials, implicates the First Amendment, a conclusion that the statement is protected becomes more likely for a public official than for a private citizen in at least two ways. First, precisely because we have freedom of speech, and precisely because political expression lies at the core of that freedom, the average reader expects a freer flow of opinionated invective in political criticism than in works which have no political purpose. Thus he is more prepared to take a statement made in such a work as opinion. Secondly, because criticism of government is often impossible without criticism of its officials, a court charged with balancing First Amendment values against private interests in reputation often will, and should, reach a different balance when the plaintiff is a public official, protecting the statement as opinion. See Jankow, 788 F.2d at 1303; Oilman, 750 F.2d at 997-1000, 1004-1005 (Bork, J., concurring). Morris’ criticism of Saenz’s attitudes in Uruguay is protected in large part because the First Amendment protection of opinion covers a wider range of critical comments about public officials than about private citizens.

2. Saenz as a New Mexico prisons official

Constitutional protection also extends to most of the other matters of which Saenz complains. He argues that the following passage is defamatory because a trier of fact could infer a factual statement from it that in New Mexico he set up a corrections system based on torture, and himself tortured prisoners:

Within weeks, Saenz transferred Griffin and brought Rodriguez back as acting warden. Although his own O.P.S. history had by then been exposed, he also named as departmental deputy an old friend and former O.P.S. officer in Colombia and Guatemala. Over the next four months, the habits of the jefatura governed the corrections system. Pen and department were closed to the press. Budgets were juggled, employees threatened with lie detector tests, critical journalists smeared as “Marxists” and liars. The pen was left to its old unaccountable masters, while King was conducted on periodic tours to assure the public of suitable progress. Reformers accompanied state-supreme-court justices on a spring tour of the pen, and when they insisted on the justices’ seeing the strip cells, when the tour then found fresh blood on the floor of one of the strip cells and the attending guard furiously thrust the keys at a woman reformer, saying, “Here, you run this god-damned place,” the state chief justice (and Saenz’s erstwhile patron) merely laughed.

By mid-June, Saenz had obliterated what morale remained in his beauracra-cy, hired one nephew to a key job and intervened to prevent the firing of another, and was at last under widespread fire in the press and even among out-of-session legislators. King then went to Lake Tahoe for a governors’ conference, taking Rodriguez with him. When he returned, he fired Saenz and named an aide to be acting corrections secretary.

Again the text cannot constitutionally be stretched to carry the meaning Saenz wants to ascribe to it. Frankly, we are not sure what the passage means, but that is all the more reason to find it constitutionally-protected opinion. Imprecise language is language from which a specific factual charge cannot be parsed, and without a specific factual charge the language is probably expressing opinion. Janklow, 788 F.2d at 1304; Oilman, 750 F.2d at 980; Buckley, 539 F.2d at 893; cf. Austin, 418 U.S. at 284, 94 S.Ct. at 2781.

The phrase, “the habits of the jefatura governed the corrections system,” comes the closest to making the allusion Saenz claims is there. The jefatura was the Montevideo police headquarters (although the article does not identify it as such) “where torture reportedly took place.” But to make that connection the American reader would have to first remember the single reference to. the jefatura from three pages before, then read the phrase as a factual statement rather than the metaphor it obviously is, then decide that one of the “habits” referred to was torture — all this despite the fact that the passage appears to go on to explain what the metaphor means, namely barriers to press access, budget juggling, attacks on the press and so forth. We do not think the average reader would go through all these gymnastics. Indeed, most readers probably found the phrase, with its foreign language word, incomprehensible, ignored it and went on. The statement is not fact, but rather opinion.

First Amendment protection also prevents the other inferences Saenz wants us to draw. Saenz does not deny that he hired a former O.P.S. officer as his deputy, nor does he deny the essentials of the reformers’ tour of the penitentiary — rather, he contends that recounting these facts accuses him of torture, a load which the language cannot carry. He does deny that he obliterated morale. The context and choice of language, however, signals a broadly-brushed stroke of hyperbole which states opinion. See Keller, 778 F.2d at 716; Lauderback, 741 F.2d at 197. Moreover, a reference to morale is a statement about an intangible, not subject to easy verification, which is another indicator of opinion. Oilman, 750 F.2d at 990-991 and n. 43; Avins v. White, 627 F.2d 637, 642-643 (3d Cir.), cert, denied, 449 U.S. 982,101 S.Ct. 398, 66 L.Ed.2d 244 (1980) (evaluation of atmosphere and spirit at law school). The statement is not actionable.

The First Amendment protections for criticism of government and political debate thus have pruned the complaint until nearly nothing is left of it. However, though it receives little attention in his brief, Saenz also denies that he either hired a nephew or tried to stop the firing of another. Apparently the kernel of truth at the center of this allegation is that his wife’s nephew worked with another New Mexico agency — a job he held before Saenz was appointed. Saenz denies, while defendants affirmatively assert, that he intervened on the nephew’s behalf, but in any case there is no evidence of two nephews. We are reluctant to find a factual statement in “a single sentence in a [work] that is otherwise so clearly opinion.” Oilman, 750 F.2d at 991. However, the statement is about Saenz personally. It accuses him of nepotism and it cannot be dismissed as metaphor or hyperbole. Cf. Mr. Chow, 759 F.2d at 226-227 (from a long complaint about restaurant review, only the statement about the number of dishes used to serve Peking duck was actionable). We are equally reluctant to characterize the difference between one and two nephews, or between hiring and intervention to protect and no action at all, as exaggeration for effect. It is criticism of a public official, but criticism which makes an explicit, specifically factual charge. This remnant of the complaint survives First Amendment analysis.

That conclusion, however, will not save Saenz’s claim. Under New Mexico law his action for libel per quod still needs either extrinsic facts of which defendant knew, or special damages. Reed, 471 P.2d at 181. He has pled no extrinsic facts. The special damages he has pled have no causative nexus with this defamatory statement. He does not allege that the duration of his unemployment was extended because of a charge that he used his influence to help his nephews. Since the allegedly defamatory statement did not cause the alleged special damage, his complaint still does not state a claim. Defendants are entitled to judgment as a matter of law.

IV. Absence of Clear and Convincing Proof of Malice

As an additional ground for our judgment, we hold that Saenz has not submitted sufficient evidence of malice to survive defendants’ summary judgment motion. Plaintiff, as a public official, cannot succeed merely by proving any defamatory statements false. He must also prove the defendants published them either knowing they were false or with reckless disregard for their truth or falsity — the constitutional standard which the Supreme Court calls “actual malice.” Sullivan, 376 U.S. at 280, 84 S.Ct. at 726. Moreover, he must prove this “actual malice” by clear and convincing evidence. Philadelphia Newspapers, Inc. v. Hepps, 475 U.S.-,-, 106 S.Ct. 1558, 1562, 89 L.Ed.2d 783 (1986). If on defendants’ motion for summary judgment it is obvious that plaintiff does not and will not have this clear and convincing evidence, the summary judgment motion should be granted. Anderson v. Liberty Lobby, Inc., 477 U.S.-,-, 106 S.Ct. 2505, 2513, 91 L.Ed.2d 202 (1986). Saenz appears to have no evidence of actual malice at all.

Defendants have submitted a thick pile of source materials which Morris used for his article. They argue that the sources prove the charges true; certainly they meet defendants’ burden of showing that Morris and Playboy did not recklessly disregard the truth. The materials Saenz offers in response, ostensibly to create an issue of fact on malice, all deal with the O.P.S. portion of the controversy. One way to show malice is to show that the statements were a complete fabrication — in other words, that defendant made them up. St. Amant v. Thompson, 390 U.S. 727, 732, 88 S.Ct. 1323, 1326, 20 L.Ed.2d 262 (1968). However, defendants did not fabricate the allegations against the O.P.S. There were charges that its agents at least made no effort to discourage torture. There were “tiger cages” in South Vietnam. Some policeman students at the International Police Academy apparently did write theses which took torture as their subject matter (defendants have submitted excerpts), although it was not a “torture” school in the sense that it taught or advocated torture. Finally, Congress did take action against the O.P.S. Whether Congress acted fairly, or moved too quickly on too little information, is not a question to be answered in this court. Political truth is an elusive concept, and the First Amendment most assuredly does not assign the determination of it to judges and juries. Time, Inc. v. Hill, 385 U.S. 374, 406, 87 S.Ct. 534, 551, 17 L.Ed.2d 456 (1967) (Harlan, J., concurring and dissenting); Sullivan, 376 U.S. at 271, 84 S.Ct. at 721.

The allegations about the O.P.S., of course, are not of and concerning Saenz. The allegation that does touch him is the accusation that he knew of torture. We have held that it is protected opinion. Even if it were not, on this record it seems unlikely that plaintiff could prove that charge false, let alone published with malice. The First Amendment imposes the burden of proof on the plaintiff, which means that when the truth of the matter is unknowable, a defamation plaintiff cannot recover. Hepps, 475 U.S. at-, 106 S.Ct. at 1564. Whether the Montevideo police actually tortured their prisoners before 1969, and if so whether Saenz knew or should have known of it, are technically questions of fact. But except for Saenz’s own testimony, the relevant witnesses and other evidence are mostly in Uruguay, either geographically or politically inaccessible to this court. Cf. Oilman, 750 F.2d at 990-991 n. 42; id. at 1006-1008 (Bork, J., concurring) (impossibility of trying the question of an individual’s status in a scholarly discipline). Any trial on these issues would probably degenerate into an inconclusive swearing contest.

However, we need not reach that question, because plaintiff has offered nothing relevant to a showing of actual malice. Saenz’s argument for proof of actual malice rests first of all on his construction of what the article said. He contends that Morris and Playboy labeled him a torturer both in Uruguay and New Mexico. That charge he expects to prove false, and contends that he can also prove that Morris knew it was false. But actual malice is not defined with reference to the statements a plaintiff thinks wére made. Woods, 791 F.2d at 487. As we have determined, the text will not support that construction. Even if it could, there would still be no issue of actual malice unless there was some reason to think that defendants intended or at least knew of the possibility of such construction. A publisher is not required to “guarantee the truth of all the inferences a reader might reasonably draw from such a publication.” Id. Even less must he guard against unreasonable inferences.

Plaintiffs also point to correspondence between Morris and his literary agent which, he contends, shows that Morris intended to damage Saenz’s reputation. However, the evidence misses the point. It is the publisher’s attitude toward truth, not toward the plaintiff — his intent to publish something false, not any evil motives he may have harbored — which matters for proof of constitutional malice. Henry v. Collins, 380 U.S. 356, 85 S.Ct. 992, 13 L.Ed.2d 892 (1965). This court, having had this lawsuit on its docket for some time, has no difficulty believing that these parties do not like each other. But we cannot constitutionally infer recklessness toward the truth from mere evidence of hostility between the parties. Rebozo v. Washington Post Co., 637 F.2d 375, 380 (5th Cir.), cert, denied, 454 U.S. 964, 102 S.Ct. 504, 70 L.Ed.2d 379 (1981). The correspondence is not evidence of actual malice.

Saenz also contends that the article shows bias: that its choice of language and shading tilts against him, and that it omitted his denials and the denials of those in O.P.S. before him. His contentions are well-grounded in fact but, again, beside the point. Just as hostility toward the plaintiff is not recklessness toward the truth, so the absence of perfect fairness to the plaintiff is not the equivalent of knowingly perpetrating a falsehood. Slanting or shading, particularly in an interpretive piece, does not show constitutional malice unless the underlying evidence which the piece is discussing is clear and unambiguous, and the line defendant took simply was not a possible rational interpretation of that evidence. Bose, 466 U.S. at 512-513, 104 S.Ct. at 1966; Time, Inc. v. Pape, 401 U.S. 279, 290, 91 S.Ct. 633, 639, 28 L.Ed.2d 45 (1971). We do not find the facts on torture and plaintiffs relation to it free of ambiguity, nor can we say that Morris’ critical interpretation of them goes beyond the bounds of rationality.

Saenz also apparently thinks he can somehow make a case for malice from the fact that many of Morris’ sources for his account of O.P.S. are persons on the left side of the political spectrum. Saenz points out, for example, that Morris drew material from Philip Agee, the C.I.A. agent who resigned and now calls himself a socialist; from the protests of Uruguayan leftists, including guerrillas; from the memoirs of a Cuban C.I.A. contact who turned out to be a double agent working for the Cuban Communist Party; from a book by a New York Times reporter who spoke with Cuban Communists, among others; and from stories in the Greek newspaper “Avghe.” Saenz brings to our attention that the latter calls itself “a morning newspaper of the left.” This court does not quite know what to make of this argument. We are not aware of any rule of defamation law that requires a presumption of falsity for any statements which originate with the New York Times or anyone to the left of it. We reiterate that the Constitution has not committed the finding of political truth to the federal judiciary. In any case, Morris merely condensed and reported what these persons said, as well as what Senator Abourezk, Amnesty International, and others said. When reporting charges made by others, failure to give the other side of the controversy is not of itself evidence of malice. Edwards, 556 F.2d at 120-121.

Saenz separately attempts to attach malice to Playboy for what he considers inadequate checking of the facts. The magazine’s editors admit that for the most part they checked the portions of Morris’ article about Saenz only against the Santa Fe reporter series on Saenz, which of course Morris also wrote. As a magazine, publishing well after the fact, Playboy was not under the deadline pressures of “hot” news. Saenz contends that the editors knew that Morris bore ill will toward him, thanks to certain communications which he offers in evidence. Since they knew of Morris’ ill feeling, he argues, they should not have checked Morris only against Morris.

Again, however, Saenz is off the mark. We are not as sure as Saenz that the internal communications show Morris’ hostility to him. But even if they do, ill will toward Saenz still does not signify recklessness toward the truth. The evidence of Playboy’s fact checking on this record puts it well short of the finest examples of a journalistic search for accuracy. But for purposes of constitutional malice, Playboy’s editors were under no obligation to check Morris’ facts at all, unless something blatant put them on notice that he was reckless about the truth. St. Amant, 390 U.S. at 731-732, 88 S.Ct. at 1325-1326; Woods, 791 F.2d at 485. Serious factual improbabilities or inconsistencies in the article itself, or resting it solely on an inherently unreliable source such as an anonymous telephone call, would be sufficiently blatant. Id. However, knowledge of an author’s ill will toward his subject does not constitute such notice. Hotchner, 551 F.2d at 914. Similarly, absent strong indicators of falsity or unreliability, reliance on a single source does not amount to malice. Woods, 791 F.2d at 488. Moreover, any distinction between “hot” and “cold” news has no relation to the constitutional standard. Barger v. Playboy Enterprises, Inc., 564 F.Supp. 1151, 1156 (N.D.Calif. 1983), affd mem., 732 F.2d 163 (9th Cir.), cert, denied, 469 U.S. 853, 105 S.Ct. 175, 83 L.Ed.2d 110 (1984). Playboy’s reliance on Morris as its principal source is not evidence of actual malice.

Finally, Saenz contends that even if none of these facts standing alone constitutes evidence of actual malice, taken together they constitute negligence, which raises an issue of malice for a trier of fact. There is however no rule of law with which we are familiar which states that evidence of negligence, without any indication that defendant had doubts about the truth of his publication, can somehow add up to a whole greater than the sum of its parts. See, e.g., Hardin v. Santa Fe Reporter, Inc., 745 F.2d 1323, 1326 (10th Cir.1984). If a publisher admits to some doubts, or admits that in one instance he didn’t care whether what he was printing was true or false, then evidence of acts of negligence toward the truth, as well as intent to harm, may be relevant to accumulating the clear and convincing proof a plaintiff needs. Goldwater v. Ginzburg, 414 F.2d 324, 337, 342 (2d Cir.1969), cert, denied, 396 U.S. 1049, 90 S.Ct. 701, 24 L.Ed.2d 695 (1970). But we know of no case which has found malice without any evidence of malice. See also Woods, 791 F.2d at 488-489.

Saenz has submitted evidence that he and Morris do not care for each other, that the article was not a model of fairness, that Morris did not write from the precise center of the political spectrum, and that Playboy did not greatly extend itself in checking what Morris said about Saenz. None of these things amount to evidence that either Morris or Playboy “in fact entertained serious doubts as to the truth of [their] publication.” St. Amant, 390 U.S. at 731, 88 S.Ct. at 1325. Certainly nothing on this record approaches clear and convincing proof of such doubts. Even if the article did not consist almost entirely of protected political criticism, and even if the one phrase which does not could carry a claim for libel per quod, defendants would still be entitled to judgment as a matter of law.

It has been and continues to be argued that present constitutional restrictions upon common law libel concepts place an unwarranted burden upon a citizen’s participation in the public arena and, thus, themselves disserve the cause of a robust political process. See Epstein, Was New York Times v. Sullivan Wrong?, 53 U.Chi.L. Rev. 782 (1986). Certain it is that those restrictions permit artful calumny invulnerable to redress through the forum of the courts. Certain it is that serial reliance upon past published sources can create a burgeoning and self-perpetuating mythology largely immune to judicial remedy since the reliance obviates malice. For example, Saenz strenuously denies that he ever had an office at the Montevideo police station “where torture reportedly took place”; Morris can point to a published account describing such an office on which he relied, and Playboy relied on Morris. Saenz here presents himself as a law enforcement professional who served his country properly and well, and whose career has been blighted by an unfair article. That may well be so — defendants do not seek to present evidence to support the truth of what plaintiff reads into the article — and, if so, Saenz joins a goodly company of public servants who have been pummeled by abusive charges. See Rosenblatt, 383 U.S. at 88, 86 S.Ct. at 676 (Douglas, J., concurring). The constitutional balance which has been struck does not, however, permit the use of the libel laws for the vindication he here seeks.

CONCLUSION

Defendants’ motion for summary judgment is granted.

1 . New Mexico law is not entirely clear on this point. Marchiondo II seems to set up a two-stage analysis for determination of whether a statement is fact or opinion. First the court determines whether the statement is "unambiguous;" unambiguous opinion is a question of law. Statements which could be either fact or opinion, however, go to the jury for interpretation. 649 P.2d at 472. If the Marchiondo II court meant to apply that analysis to all expressions of opinion, then it is squarely at odds with the position of most federal courts, which have held that a statement which could be either fact or opinion is probably opinion as a matter of law for that reason. See, e.g., Buckley, 539 F.2d at 895.

However, Marchiondo II did not involve a public figure plaintiff, and actually appears to distinguish between political opinion which was fully protected as a matter of law, 649 P.2d at 469, and fair comment about a private individual, 649 P.2d at 471-472, with the two-stage analysis applying only to the latter. Thus under New Mexico law political opinion, at least, apparently presents a question of law for the court. In any case, we are of course not bound by the Supreme Court of New Mexico on a question of federal constitutional law. See Martin v. Hunter's Lessee, 14 U.S. (1 Wheat.) 304, 4 L.Ed. 97 (1816); Brown & Williamson Tobacco Corp. v. Jacobson, 644 F.Supp. 1240, 1257 (N.D. 111.1986). And federal courts unanimously hold that in the context of political criticism, the difference between fact and opinion is a question of law for the court. See Janklow, 788 F.2d at 1305 n. 7; Oilman, 750 F.2d at 978, and cases cited by both. Indeed, this court has already held as a matter of law that the article’s reference to Saenz as a "drifter" and to his lack of qualifications for office are opinion and so not actionable. Saenz, No. 81 C 5723, slip op. at 9 (N.D.I11. June 14, 1984).

2 . The Oilman approach, by using the term "average reader" (and also "reasonable reader," e.g., 750 F.2d at 979) in factors used to determine a question of law, produces an analysis which in its language gives the appearance of judicial determination of questions of fact, and has been criticized for usurping the jury’s role. Janklow, 788 F.2d at 1307 (Bowman, J., dissenting) (“Beauty is in the eye of the beholder, and it would appear that the result to be obtained through application of the Oilman factors is in the eye of the judge”). However, what the Oilman court meant by "average reader” was a reader who was neither extraordinarily skeptical nor extraordinarily credulous. 750 F.2d at 979 n. 16. The standard thus seems to closely resemble one phrased as "no rational reader could find a statement of fact here.” Cf. Greenbelt Cooperative Publishing Ass'n v. Bresler, 398 U.S. 6, 14, 90 S.Ct. 1537, 1542, 26 L.Ed.2d 6 (1970) (“even the most careless reader must have perceived that the word [blackmail] was no more than rhetorical hyperbole”). The latter phrasing, though more cumbersome, would have more clearly indicated a legal determination. In any case, the Oilman court used the "average reader” and "reasonable reader” language as a shorthand means of articulating in more detail the scope of First Amendment protection: reasons why the statement at least should have been taken as opinion and thus should be protected. It did not purport to determine the actual reaction of actual average readers, nor did it make such actual reaction the standard for separating fact from opinion. 750 F.2d at 979 n. 16. We use the language in the same way. As Sullivan makes clear, the actual reaction of actual readers does not determine whether or not a statement is constitutionally protected. 376 U.S. at 289-291 n. 28, 84 S.Ct. at 730-732 n. 28; see part II supra.