HEANEY, Senior Circuit Judge.

David Price, a special agent of the Federal Bureau of Investigation, was assigned to the Pine Ridge Reservation (Reservation) in South Dakota during the early 1970s. While on assignment on the Reservation, he played a role in investigating the Wounded Knee occupation and a shootout on the Reservation during which two F.B.I. agents were killed. Leonard Peltier, a member of the American Indian Movement (A.I.M.), was convicted of their killings. Peter Matthiessen, a well-known author, decided to write a book about these events, dividing the profits between himself, the publisher, a film company and the Leonard Peltier Defense Committee. The book, In the Spirit of Crazy Horse, is generally sympathetic to Indian views about these incidents and the treatment of Indians by the federal government. The author made a number of statements that Price considered defamatory. Price brought suit for defamation, intentional infliction of emotional distress, false light invasion of privacy, and prima facie tort. He seeks compensatory damages of $25,000,000, punitive damages, costs and fees. In light of the enormity of these claims, Viking withdrew the book from circulation.

Successive rulings by the district court narrowed the dispute to specific defamation claims. The remaining claims involve twenty statements made in Crazy Horse and found in the appendix to this opinion. They are divisible into five categories: Price's relation to the perjury of Louis Moves Camp and involvement in seeing that criminal charges against him were dropped; misconduct regarding the testimony and affidavits of Myrtle Poor Bear; the withholding of information or gross negligence regarding the homicide investigation of Anna Mae Aquash; harassment of Indian people; and general statements about Price's character.

Both sides have engaged in extensive and acrimonious discovery surrounding the historical facts and the editorial process. Nearly four years after Price first brought his claims, and after legal costs of over one million dollars to the defendants, the district court finally dismissed the remaining defamation claims on constitutional grounds. Price v. Viking Penguin, Inc., 676 F. Supp. 1501 (D.Minn. 1988). Price appeals the district court's judgment regarding his defamation claims and asks us to remand this case for trial by jury. We decline to do so. We begin by setting forth the applicable law.

I. FREEDOM OF SPEECH AND STATE DEFAMATION LAW [5] A. The First Amendment

Our founding fathers created a vibrant democracy, relying on virtually unregulated dissension and self-examination to provide our republic with stability and our citizens with satisfaction. In unqualified language applicable to the states, they allowed "no law * * * abridging the freedom of speech, or of the press * * *." U.S. Const. amend. I. This liberty, as much as any other, has given us our self-definition. It continues to favorably distinguish the United States from most societies in the world, and it is always the vehicle by which we appreciate the past and deliberate over our future. Though the claims raised in this appeal are particular, our review is influenced throughout by the place even this one dispute occupies in our nation's interest. Our most influential leaders have commented on the importance of the right, as well as on the proper role of judges in protecting it.

To reach sound conclusions on these matters, we must bear in mind why a state is, ordinarily, denied the power to prohibit dissemination of social, economic and political doctrine which a vast majority of its citizens believes to be false and fraught with evil consequence.

Those who won our independence believed that the final end of the state was to make men free to develop their faculties; and that in its government the deliberative forces should prevail over the arbitrary. They valued liberty both as an end and as a means. They believed that freedom to think as you will and to speak as you think are means indispensable to the discovery and spread of political truth; that without free speech and assembly discussion would be futile; that with them, discussion affords ordinarily adequate protection against the dissemination of noxious doctrine * * * the remedy to be applied is more speech, not enforced silence.

Whitney v. California, 274 U.S. 357, 374-75, 47 S.Ct. 641, 647-48, 71 L.Ed. 1095 (1927) (Brandeis, J., concurring). Thomas Jefferson once admonished us: "We have nothing to fear from the demoralizing reasonings of some, if others are left free to demonstrate their errors * * * these are safer corrections than the conscience of a judge." Id., n. 3 (quoting Beard's report in 123 The Nation 8 (1926)).

B. Political Speech and First Principles in Defamation Law

State defamation law limits free speech to protect an individual's reputational interests. As the Bill of Rights became applicable to the states, the first amendment became increasingly viewed as a limit on state defamation law. New York Times v. Sullivan, 376 U.S. 254, 84 S.Ct. 710, 11 L.Ed.2d 686 (1964), established the modern rule that, even when the defendant's remarks are both defamatory and false, where the remarks are directed at a public official and related to her official conduct, the plaintiff must also prove by clear and convincing evidence that the false remarks were made with actual malice or reckless disregard for the truth. Id. at 279-80, 84 S.Ct. at 725-26.

The motivating factor in the Court's analysis was protection for criticism of public officials and speech regarding issues of political concern. The New York Times standard was constructed in light of three truths about public speech. First, false statements would necessarily occur in the course of a vigorous public debate. Second, absent protection for even false statements, destructive self-censorship would result. Third, the legal standards for defamation must protect defendants from the self-censorship imposed by threats of litigation. Id. at 271-72, 278-79, 84 S.Ct. at 720-22, 724-26. The Court felt that debate on matters of public concern "should be uninhibited, robust, and wide-open, * * * [though] it may well include vehement, caustic, and sometimes unpleasantly sharp attacks on government and public officials." New York Times, 376 U.S. at 270, 84 S.Ct. at 720. The weaker state interest in the reputation of individuals has nevertheless been accommodated; where the plaintiff is not a public figure, a different balance is struck. Philadelphia Newspapers, Inc. v. Hepps, 475 U.S. 767, 776-77, 106 S.Ct. 1558, 1563-64, 89 L.Ed.2d 783 (1986); Gertz v. Welch, 418 U.S. 323, 343-47, 94 S.Ct. 2997, 3008-10, 41 L.Ed.2d 789 (1974).

Supreme Court decisions have created two threshold issues: first, whether the plaintiff is a public figure, and second, whether the challenged statements are allegations of fact. We begin by considering the plaintiff's status, which determines the applicable law.

C. Price's Status

While Price has argued his case under the New York Times standard, he is reluctant to concede that he is a public figure for these purposes. Under the circumstances, however, we do not hesitate in agreeing with the district court that Price was a public figure and that the challenged statements relate to his official conduct. 676 F. Supp. at 1510-11.

Price occupied a prominent role in public affairs within the Reservation. He played a substantial role in the investigation of crimes on the Reservation, including the shoot-out. While he is by no means the principal character in the book or its focus, he was the object of public notoriety over these events, having been singled out for criticism by judges and others. In addition, news accounts and editorials about this event have appeared in a variety of forums and have discussed Price's conduct. See Defendants' Exhibits nos. 35-42, 62-64, 69, 95, 96, 116, 118 (Lincoln Nebraska Journal, New York Times, The Palm Beach Post, Minneapolis Star Tribune, San Antonio Express, Sioux Falls Argus Leader, St. Paul Press Dispatch, Aberdeen American News, Rapid City Journal, The Los Angeles Daily Journal, People, Twin Cities Reader, Newsweek, The Washington Star, and The Nation). Serious accusations about Price have been directly made in other books. See Weyler, Blood of the Land (Everest House 1982); Brand, The Life and Death of Anna Mae Aquash (Lorimer Co. 1978). Serious questions about his conduct have been raised by many members of Congress. Congressional Correspondence to the President, Dec. 17, 1980 (Defendants' Exhibit no. 97).

The challenged statements from Crazy Horse all concern the discretionary performance of Price's duties. In context, criticism of the actions Price took in his official capacity reflect on the imperatives and conduct of the F.B.I. and the government generally, implicating the type of public debate at the core of the first amendment. We therefore apply the standards set forth in New York Times. See Rosenblatt v. Baer, 383 U.S. 75, 85, 86 S.Ct. 669, 675, 15 L.Ed.2d 597 (1966) ("the public official designation applies at the very least to those * * * who have, or appear to the public to have, substantial responsibility for or control over the conduct of government affairs."); Stevens v. Tillman, 855 F.2d 394 (7th Cir. 1988) (school principal), cert. denied, ___ U.S. ___, 109 S.Ct. 1339, 103 L.Ed.2d 809 (1989); Gray v. Udevitz, 656 F.2d 588 (10th Cir. 1981) (police officer).

D. Fact and Opinion

At the outset, we must also consider whether the challenged statements are opinion and therefore absolutely protected. The separation of opinion from fact acts as a limit on government power.

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.

Gertz v. Welch, 418 U.S. at 339-40, 94 S.Ct. at 3006-07; see also Masses Publishing Co. v. Patten, 244 F. 535, 539-40 (S.D.N.Y. 1917) (per Hand, J.). This is a question of law, and we have recently established how we are to proceed. Janklow v. Newsweek, Inc., 788 F.2d 1300 (8th Cir.) (en banc) ( Janklow II) (reversing in part, Janklow v. Newsweek, Inc., 759 F.2d 644 (8th Cir. 1985) ( Janklow I)), cert. denied, 479 U.S. 883, 107 S.Ct. 272, 93 L.Ed.2d 249 (1986).

In Janklow II, we established a four-part framework for distinguishing fact and opinion, drawing on Judge Starr's plurality opinion and the concurring opinion of Judge Bork in Ollman v. Evans, supra note 3. The four criteria are tools to help us determine from the totality of circumstances whether the first amendment protects a statement sufficiently that resort to the jury is unnecessary; they are: the specificity of the statement, its verifiability, the literary context and the work's public context. This enterprise is only superficially linguistic. Ultimately, we must decide — not whether a statement in isolation is by virtue of its phrasing factual — but rather whether, when taken in context, the statement functions and would be understood as an unqualified assertion of fact rather than as an element of an opinion. Id. at 994 (Bork, concurring). In this light, "fact" and "opinion" take on special meanings, and we therefore pause and consider the four factors.

Specificity

The precision and specificity with which an assertion is made may reflect the extent to which it actually recites specific factual events. Specificity also goes to the singularity of a statement's meaning. Where a statement or phrase is susceptible of more than one meaning, we will not presume either that the phrase means what the plaintiff asserts it does or that it is factual where it can be understood as an opinion. Letter Carriers, supra note 3, 418 U.S. at 283-84, 94 S.Ct. at 2780-81; Secrist v. Harkin, 874 F.2d 1244, 1247, 1250 (8th Cir. 1989). We do not recognize defamation by implication. Janklow II, 788 F.2d at 1304. Accord Fudge v. Penthouse Intern., Ltd., 840 F.2d 1012, 1016 (1st Cir.), cert. denied, ___ U.S. ___, 109 S.Ct. 65, 102 L.Ed.2d 42 (1988).

Verifiability

Assertions whose elements are unverifiable, including statements regarding motive, are "intrinsically unsuited" to serve as a basis for libel. Janklow II, 788 F.2d at 1302, 1304. Where quantification for a general assertion is impossible, allowing any fact-finder to decide its truth or falsity invites the exercising of personal dispositions regarding "the contents of the statement, its author, or its subject." Ollman, 750 F.2d at 981; see also, id. at 1006-08 (discussion by concurring judges).

Specificity and verifiability thus function as limits on majoritarian bias and government intervention. By examining each statement for "a precise core of meaning for which a consensus of understanding exists," Ollman, 750 F.2d at 979, we hope to avoid both liability conditioned on public sentiment, and putting judges and juries in the role of declaring what the "truth" is about disputed social issues.

Literary Context

The third factor to be examined is the condition and tone of the work as a whole. Various considerations include cautionary or qualifying language, language or style which signals opinion, the type of publication, the location of the statement or work within the publication, and the intended audience. For example, the challenged article in Janklow II appeared in a magazine, and "the magazine's generally freer style of personal expression and the article's transparently pro-Banks posture would signal the reader to expect a fair amount of opinion." Id. In Ollman, it was significant that an editorial was challenged. 750 F.2d at 986.

Public Context

Lastly, we consider whether the statements played a role in a public debate. Where core values of the first amendment are implicated, even some false statements of fact must be protected. "[I]t cannot be avoided if the political arena is to remain as vigorous and as robust as the first amendment and the nature of our polity require." Ollman, 750 F.2d at 1002 (Bork, concurring). Statements made in the course of a political debate are also more likely to be understood as opinion. Deupree v. Iliff, 860 F.2d 300 (8th Cir. 1988); Koch v. Goldway, 817 F.2d 507, 508 (9th Cir. 1987) (per Kennedy, J.).

* * * * * *

Judgments about context often determine a statement's proper classification. As a practical matter, every opinion involves implied or asserted facts, and choices about which facts to assert are always evaluative. Stevens v. Tillman, 855 F.2d 394, 398 (7th Cir. 1988), cert. denied, ___ U.S. ___, 109 S.Ct. 1339, 103 L.Ed.2d 809 (1989). We must therefore always ultimately focus on the context from which both the dispute and the statements arise, remaining sensitive to our republic's interest in robust debate and the protection of unpopular viewpoints.

Even when a statement is subject to verification, however, it may still be protected if it can best be understood from its language and context to represent the personal view of the author or speaker who made it. Thus we reject the suggestion, advanced by the plaintiffs in this case, that any "question of fact" which can be decided by a jury can be actionable as defamation. Such a test ignores the underlying purposes of the fact/opinion distinction, and would lead to results that could not be reconciled with the developing case law in other circuits [citing Janklow II].

Potomac Valve Fitting v. Crawford Fitting, 829 F.2d 1280, 1288 (4th Cir. 1987).

E. Reckless Disregard and Neutral Reporting

Once a plaintiff establishes that there were false and defamatory statements of fact made concerning him, the conduct of the defendants in preparing the challenged work becomes relevant. A plaintiff may prevail if she can show that the defendant, in making a particular statement, acted with actual malice, reckless disregard for the truth or had a high degree of knowledge of probable falsity. Herbert v. Lando, supra note 5, 441 U.S. at 156-57, 99 S.Ct. at 1638-39.

"Malice" does not refer to ill will. It is instead another name for the other formulations of the standard for culpability, all of which inquire into the defendant's knowledge prior to publication. Harte-Hanks Communications, Inc. v. Connaughton, ___ U.S. ___, ___, n. 7, 109 S.Ct. 2678, 2685 n. 7, 105 L.Ed.2d 562 (1989). The primary focus must be on the defendant's attitude toward the truth of the statements, rather than on the defendant's attitude toward the plaintiff. Id.; Herbert v. Lando, supra note 5, 441 U.S. at 199-202, 99 S.Ct. at 1660-1662 (Stewart, J. dissenting); Letter Carriers v. Austin, supra note 3, 418 U.S. at 281-82, 94 S.Ct. at 2779-80 ("a shorthand expression of the `knowledge of falsity or reckless disregard of the truth' standard."). We focus on the information within the defendant's possession; a failure to investigate cannot sustain liability unless there was already good reason to doubt the truth of the information in hand. St. Amant v. Thompson, 390 U.S. 727, 733, 88 S.Ct. 1323, 1326, 20 L.Ed.2d 262 (1968); Curtis Publishing Co. v. Butts, 388 U.S. 130, 153-54, 87 S.Ct. 1975, 1990-91, 18 L.Ed.2d 1094 (1967) (plurality). Thus, no distinction is drawn between hot and cold news. Barger v. Playboy Enterprises, Inc., 564 F. Supp. 1151, 1156 (N.D.Cal. 1983), aff'd 732 F.2d 163 (9th Cir.), cert denied, 469 U.S. 853, 105 S.Ct. 175, 83 L.Ed.2d 110 (1984).

This Court has added to this test the concept of "neutral reporting." We have held that the recitation of official actions or statements by public bodies, if substantially accurate, is protected, even if the implications are harmful. Janklow I, 759 F.2d at 649 (preserved by notation of our court in Janklow II, 788 F.2d at 1301 n. 2). The privilege protects a journalist who "believes, reasonably and in good faith, that his report accurately conveys the charges made." Edwards v. National Audubon Society, Inc., 556 F.2d 113, 120 (2d. Cir.), cert. denied, 434 U.S. 1002, 98 S.Ct. 647, 54 L.Ed.2d 498 (1977). It does not protect one who "in fact espouses or concurs in the charges made by others, or who deliberately distorts these statements to launch a personal attack of his own on a public figure." Id.; Cianci v. New Times Pub. Co., 639 F.2d 54 (2d Cir. 1980). In Cianci, a magazine published an article accusing the Mayor of Providence of committing rape. The Cianci court held that the accusations were factual and declined to apply Edwards because the New Times failed to present the other side of the story. 639 F.2d at 69. In Janklow II, we distinguished Cianci on the first ground, because there the article misrepresented the timing of the events and expressly accused Cianci of obstruction of justice, relying on the erroneous sequencing of events as proof of the charge. Janklow II, 788 F.2d at 1305-06. Janklow II also refined the neutral reporting test, for we allowed the privilege even though we found that the article was "transparently pro-Banks." Id. at 1304. Thus, we focus on whether the reports were accurate reflections of what was said or done. Evidence of the author's general disposition toward his topic does not establish whether he espoused each particular allegation.

F. Summary Judgment

In this case, the district court has granted summary judgment for the defendants after extensive discovery but before trial. To reverse this judgment, we must find that Price could show that precise factual statements were false. Philadelphia Newspapers, 475 U.S. at 775, 106 S.Ct. at 1563. In the event that some of the statements can be proven false, defamatory and factual, we must finally decide "whether the evidence in the record could support a reasonable jury finding * * * that the plaintiff has shown actual malice by clear and convincing evidence * * * ." Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255-56, 106 S.Ct. 2505, 2513-14, 91 L.Ed.2d 202 (1986). Despite the posture of the case, we make an independent judicial evaluation of the evidence. Secrist v. Harkin, 874 F.2d at 1251; Liberty Lobby, Inc. v. Dow Jones Co., Inc., 838 F.2d 1287, 1293 (D.C. Cir.), cert. denied ___ U.S. ___, 109 S.Ct. 75, 102 L.Ed.2d 51 (1988); Berry v. National Broadcasting Company, Inc., 480 F.2d 428, 432 (8th Cir. 1973) (not analyzing evidence in light favorable to the nonmover), cert. denied, 418 U.S. 911, 94 S.Ct. 3203, 41 L.Ed.2d 1157 (1974). We begin our evaluation of the challenged statements by describing their literary context and the quality of the author's investigation.

II. CRAZY HORSE [38] A. Its Contents

Crazy Horse is a long work divided into three parts. The first part surveys the conditions on the Reservation and the rise of A.I.M. The author makes clear from early on that the F.B.I. and the federal government are regarded as villains by some Indians. He writes that heretofore press coverage had been one-sided, Crazy Horse at 196-97, and sets out to tell these Indians' side of the dispute. Introductory Note at 557.

The book presents the views of members of A.I.M., that the F.B.I. targeted them for harassment because the F.B.I. associated A.I.M. with communism and other leftist ideologies. The author reviews A.I.M.'s occupation of Wounded Knee in 1973 and the resulting prosecutions of Dennis Banks and Russell Means. According to the book, Price supervised a witness, Louis Moves Camp, at their trial. Later, Moves Camp's testimony was entirely discredited. Eventually, the book relates, all charges against Means and Banks were dismissed by the trial judge for government misconduct. In dismissing the case, the court criticized the prosecution and the F.B.I. on the record for their conduct with regard to Moves Camp's perjury. Crazy Horse at 99 (quoting Judge Nichol directly).

Later chapters in the first part discuss conditions on the Reservation in between the occupation of Wounded Knee and the 1975 shootout. The book details violence on the Reservation, criticizing Bureau of Indian Affairs (B.I.A.) police and the F.B.I. Crazy Horse at 132 (quoting Al Trimble, former Superintendent of the B.I.A.).

The second part of Crazy Horse begins with two chapters devoted to the 1975 shootout. The first details the incident from the perspective of the Indians who were there. The second describes events as portrayed in the F.B.I.'s records and reports. Matthiessen acknowledges that obtaining a neutral view of events is difficult. "[T]he public had to choose between the propaganda of the authorities and the rhetoric of the A.I.M. spokesman, neither of which gave an accurate account of what had happened." Crazy Horse at 195. In sum, two F.B.I. agents went onto the Jumping Bull property on the Reservation. A shootout began during which the agents were wounded. Someone advanced and executed them at close range. The encampment was surrounded, and during subsequent shooting, an Indian was killed.

The rest of the second part of the book concerns itself with the investigation and trials of the suspected killers. The book quotes Reservation Indians, describing the course of the investigation. These quoted accounts sometimes portray Price as harassing local citizens. See e.g. Crazy Horse at 227-29, 243-44, 247. The F.B.I. focused on four suspects in the agents' killings. One was never tried. Two of the suspects, Butler and Robideau, were acquitted of the charges by a jury in Cedar Rapids. Leonard Peltier was convicted in a subsequent trial in Fargo, after extradition from Canada. His conviction was affirmed by this Court. United States v. Peltier, 585 F.2d 314 (8th Cir. 1978), cert. denied, 440 U.S. 945, 99 S.Ct. 1422, 59 L.Ed.2d 634 (1979). In the extradition proceeding, the government used the now discredited affidavits of Myrtle Poor Bear, a witness supervised by Price. Crazy Horse includes the comments of Judge Ross of this Circuit during the oral argument of Peltier's appeal, criticizing the conduct of the F.B.I. in securing Poor Bear's affidavits. Crazy Horse at 372. The book also discusses the discovery of the body of a Jane Doe, later known to be Anna Mae Aquash, an A.I.M. member whose murder remains unsolved. Price is criticized with respect to the investigation of her death.

The third part concerns the post-conviction incarceration of Leonard Peltier and returns to consider at greater length earlier themes regarding the economic condition of the Indian people. It includes interviews with the three men tried for the murders, Peltier, Butler and Robideau. Finally, the book argues for a new trial for Peltier based on government misconduct and using documents obtained under the Freedom of Information Act. Peltier's subsequent appeal based on these documents was denied by a district court, and we affirmed. United States v. Peltier, 800 F.2d 772 (8th Cir. 1986) (per Heaney, J., denying a new trial), cert. denied, 484 U.S. 822, 108 S.Ct. 84, 98 L.Ed.2d 46 (1987).

B. Preparation

Peter Matthiessen has written at least eighteen other books, one of which received the National Book Award in 1978. Matthiessen's preparation was extensive, as the number of documents in the exhibits demonstrate. Primary sources included court transcripts, court decisions, F.B.I. records, congressional investigations, legislative resolutions, books, articles and personal interviews. The text and footnotes in the book explain the background and sympathies of quoted sources. See, e.g., Crazy Horse at 459-60. He also consulted with journalists of varying dispositions, and examined the fruits of their investigations. Matthiessen Affidavit at 7-8 (April 21, 1987). In addition, Matthiessen lists other sometimes conflicting sources for information about the same events. Crazy Horse at 577-79.

Though there was disagreement over the author's economic explanation of government policies, no one Matthiessen circulated the manuscript to indicated that any statements regarding Price were inaccurate. Moreover, Matthiessen indicates that he excluded from his manuscript unfavorable information about Price that he discredited. Matthiessen Affidavit at 9, 172. The author explained his one-sided selection of material to the reader:

Although this book argues and supports the cause of the traditional Indian people and their allies and their long struggle with U.S. Government, I originally wished it to include as many views as possible on both sides of the question. I discovered, however, that federal judges, prosecutors, and law enforcement personnel are rarely willing to comment on a case that is still being contested; also, their essential views are everywhere throughout the text, in field reports, memos, trial transcripts, court rulings, and many other documents, and any attempt to balance the argument with "official" comment would sink an already long book under the weight of rhetoric and repetition.

Crazy Horse at 577; see e.g., id. at 497-98 (unsuccessful attempts to interview the United States Attorney).

C. Tone

Books can be written in a variety of styles, often telling stories or asserting viewpoints. Crazy Horse does both, mixing legends, news accounts and social criticism with the retelling of personal interviews. It is dedicated "for all who honor and defend those people who still seek to live in the wisdom of Indian way." Many of the events are retold in the words of the participants. The author adopts some of their general preconceptions about others, and Matthiessen reports Indian versions of events. See e.g., id. at 250 (pistol whipping), 67-83 (Wounded Knee occupation).

While the author is frank about his sympathies, he also, with regard to nearly every important incident, presents contrary views. Often, even the statements Price challenges are made with cautionary language. See e.g., appendix to this opinion, paragraphs 10( o) ("alleged," "disputed"), 10(i) ("suspicions"), 10(m) ("suggestions"), and 10(s) ("speculations"). Nor does Matthiessen always espouse specific claims.

The author also managed to locate and interview David Price. As Matthiessen explained, "Fairly or not, Price had come to personify the most cynical abuses of the F.B.I.'s regressive attitudes toward Indians, and it seemed important to hear his side of the story." Crazy Horse at 460. Once Price began talking, the author "warned him that my sympathies in the Pine Ridge confrontations lay with the traditional Indians. `The Indian situation deserves sympathy,' Price assured me, reverting to a soft and thoughtful tone, `but it's hopeless. All the F.B.I. was trying to do was to stop the Indians out there from killing each other.'" Id. at 462. Price repeatedly indicated that he knew that he was blamed for a lot of bad acts, and that his reputation had already been spoiled. Id. at 461 ("I got worked over extremely heavily to say the least * * * I've just had too many ringers thrown at me."), 465 ("Don't take it from me, because I'm supposed to be this heinous villain. . . ."), 470 ("Without the allegations, I would be nobody.").

Price also took the opportunity to level allegations of his own. He accused A.I.M. of planning the ambush to get press attention. He said he thought the planner was not an Indian but would not name names. Id. at 467. He suggested that a particular Indian was involved in or had knowledge of Anna Mae Aquash's murder. Id. at 469-70. He also criticized the acquittal of Butler and Robideau. Id. at 468.

At the end of the conversation, the author concludes the report of the interview with mixed comments about Price:

To judge from his speech, David Price is not a stupid man, but one whose intelligence is severely limited by preconceptions — either that, or he is a liar who not only believes but is angered and moved by his own lies. Despite his awful sentimentality — about Myrtle Poor Bear, about the idealism of his fellow agents — his outrage about Ron Williams's death is genuine; and he does not whine about his villain's role, but seems to accept it in the line of duty.

* * * * * *

[T]his man on the other end of the wire was only another casualty of the new Indian wars, and he would be soiled for the rest of his days by deeds done in the belief that the end justified the means * * *.

* * * * * *

[He is also] the man regarded wrongly as the villain of the tale * * *.

Crazy Horse at 471-72. Against this backdrop, we analyze the challenged statements from Crazy Horse in topical categories.

III. THE CHALLENGED STATEMENTS [57] A. Louis Moves Camp

Moves Camp testified for the prosecution against A.I.M. members charged with crimes connected to the occupation of Wounded Knee. Crazy Horse at 93-99. The government called Moves Camp as a rebuttal witness who allegedly, quoting a defense attorney, "filled in every gap in the prosecution's case, directly connecting the defendants with the alleged offenses — it was amazing!" Crazy Horse at 93. According to the book, Moves Camp also confirmed the government's political theory by testifying that A.I.M. was connected with an international communist conspiracy, saying that agents from Russia, China and Czechoslovakia had attended the first meeting of the A.I.M.-sponsored International Indian Treaty Council. Id.

The book alleges that Moves Camp was located by Price. Id. at 93-94. The author relates that at trial there was testimony that Price and his partner, Williams, had met daily with Moves Camp from August 5 through August 10. During this time period, the witness completed affidavits, foreshadowing his testimony, that were turned over to the federal prosecutor. The prosecutor requested a lie detector test, and the F.B.I. refused. Crazy Horse at 94. The defense eventually impeached Moves Camp by showing that he had not been present at Wounded Knee at the time of some of the events he purported to describe. He had been in California appearing on television and at college campuses. United States v. Banks, 383 F. Supp. 389, 394 (D.S.D. 1974), appeal dismissed, 513 F.2d 1329 (8th Cir. 1975).

In addition, the book discusses an incident that occurred while Moves Camp was in the agents' custody. One night when the two F.B.I. agents and Moves Camp were staying in Wisconsin, they went out drinking. The F.B.I. agents went to bed, leaving Moves Camp in a bar with a young girl. Moves Camp was detained the next day by River Falls, Wisconsin, police investigating a rape. In the morning, Price received a phone call and "to quote Ron Williams, `he indicated it was from some type of law enforcement-related person, but did not specify the individual.'" Crazy Horse at 96. According to the book, Price returned to River Falls without Williams, assuring Moves Camp, "Don't worry." No rape charges were filed. The author notes that Price and Williams failed to file standard field reports describing these events. In addition, he argues that the defense established that Price conferred for several hours with the state prosecutor, as well as River Falls police, and Matthiessen alleges that Price made it clear "to his fellow lawmen that Moves Camp was a crucial witness for the prosecution in the Means-Banks trial. . . ." Id. at 97. He also quotes the federal prosecutor, however, to the effect that "he saw no reason to believe that what had happened `wasn't a consensual act between two adults. I have seen no evidence yet that there was any rape. . . .'" Id. The author goes on to exonerate Williams with the phrase, "[He] cannot be implicated in the alleged cover-up of the disputed rape." Id. at 97. Matthiessen makes no similar exoneration for Price.

Price alleges that the challenged statements defame him by suggesting that he suborned perjury and that he covered up an alleged crime committed by an important witness. Appellant's Brief at 22-23. At the outset, we must decide whether the challenged statements are specific and verifiable. Price does not dispute that Moves Camp committed perjury, nor that Price and Williams chaperoned, interrogated and helped prepare Moves Camp for trial. See Affidavit of David Price at 6-7 (Mar. 30, 1987) (Price Aff. no. 1); Affidavit of David Price at 3 (May 13, 1987) (Price Aff. no. 2); see also, Report of Agent Williams (Sept. 30, 1974) (Defendants' Exhibit no. 34); Affidavit of Moves Camp (Aug. 11, 1974) (Defendants' Exhibit no. 16). Price only claims that Moves Camp came to the F.B.I. and that "we did the best we could to attempt to determine whether Moves Camp was telling the truth." Price Aff. no. 1 at 7; Price Aff. no. 2 at 3. In addition, Price says that he had only minimal contact with the Wisconsin prosecutor about the rape charge and did not discuss it with the police. Price Aff. no. 1 at 6-7.

We begin with Price's challenge to the author's conclusionary statement about Moves Camp's testimony:

More serious than Louis Moves Camp's lies was the all but inescapable conclusion that Agent Price and perhaps Agent Williams had knowingly prepared this man to give false testimony; at the very least, they found his story so convenient that they had not bothered to find out if it was true.

Crazy Horse at 98; complaint, paragraph 10(c). Initially, this statement seems specific and verifiable. On its face, the statement seems to allege that as a historical fact, Price did one of two things. Either he suborned perjury or he failed to investigate.

The second part of the statement, alleging that there was not an adequate investigation, would ordinarily be considered factual, especially in light of its seemingly absolute language — "not bothered to find out if it was true." In context, however, we feel the second part is an assertion of an opinion. It is a qualitative judgment about the agency's motivation, effort and effectiveness. For this reason, it is unverifiable. Evaluating the agency's effort requires subtle judgments about what steps were required for an adequate investigation. Our conclusion that this statement is opinion is buttressed by the surrounding text omitted from Price's complaint. The next two sentences of Crazy Horse state:

More serious still was the likelihood that Assistant U.S. Attorney Hurd had also been aware that Moves Camp's testimony might be false even before he put him on the stand. Judge Nichol later concluded that Hurd did not have foreknowledge, but the whole lurid episode, which turned out to be the climax of the trial, was a fatal discredit to the prosecution.

Crazy Horse at 98. The entire discussion in context is a consideration of possibilities and likelihoods with respect to Moves Camp's testimony. The author concludes with the opinion that the only certainty is that the episode discredited the prosecution. Our conclusion that the second part of the challenged statement is opinion contributes to our view that the first part of the challenged statement is opinion as well. It is phrased in even more cautionary terms. The author describes the failure to investigate as "the very least" that can be said. He describes it as short of an inescapable conclusion that the F.B.I. gave Moves Camp the false information that Moves Camp used to testify, implying that it is unverifiable.

In our view, the entire statement is opinion. So long as the underlying facts on which the opinion is based are substantially accurate, it is protected. Janklow II, 788 F.2d at 1305-06. Moreover, Judge Nichol's condemnation on the record arguably went further. He surmised that the F.B.I.'s "investigation" consisted of giving Moves Camp liquor, putting him up at plush resorts and overcompensating him for his short service as a witness in the amount of $2,074.50. United States v. Banks, trial transcript at 21,749-50 (Defendants' Exhibit no. 26).

Even if we were to view the failure to investigate as factual, Price's claims would still fail. Price has not related one thing he or anyone else did to confirm Moves Camp's testimony. Price only denies that nothing was done, failing to refer to any evidence. Price Aff. no. 1 at 7. Nor was the prosecution able to put any evidence before Judge Nichol to show that they investigated Moves Camp's story. Banks, trial transcript at 21,749. Furthermore, there is no evidence to show that the author entertained a subjective knowledge of probable falsity with respect to his comments on the investigation. He has relied on the public record. Summary judgment is appropriate for factually unsubstantiated claims, especially in this context.

Next, we consider Price's claim that he is accused of obstructing justice with regard to the rape. At the outset, we are confronted with a problem in fixing a precise meaning to the statements. While Price attaches an implication to the sequence of events as the book presents them, it is obvious that more than one interpretation is possible. The rape charges might have been dropped for insufficient evidence. The alternative, while no doubt fueled by the author's failure to exonerate Price as he exonerated Williams, is inexplicit. The author describes it as an "alleged" coverup of a "disputed" rape. In this respect, Crazy Horse closely mirrors the sentiments of the presiding judge. Judge Nichol found the conduct of Price in returning to River Falls outrageous, and Price's account incredible. Judge Nichol concluded that Price had likely influenced local authorities and also commented on the failure to investigate Moves Camp's story.

The mere fact that the F.B.I. were — those words that I guess Mr. Lane put up on the blackboard: don't worry we'll go down, we'll go down to the police station, to Moves Camp. They do go down to the police station. There is no direct testimony that they said to the district attorney: now we don't think you should prosecute this case, but they did indicate that he was a witness. Certainly people in Wisconsin, twenty miles from here, or Hudson is twenty miles, and the little town they were in is another twenty miles; I think they probably read the Twin Cities' papers. They probably knew there was a Wounded Knee trial going on. And they indicated he was a witness in the Wounded Knee trial. I don't think they had to tell them that they would prefer they didn't prosecute. I think they got the message. But it's hard for me to believe that the F.B.I., which I have revered for many years, has fallen to that low an estate. . . . [With respect to confirming Moves Camp's testimony] * * * Since when does the head of the F.B.I. start telling the prosecution that they can't have a lie detector test if they want to? * * * [W]ell, I don't know just what their checking was. A lot of it, I know, must have taken place in bars over there in Wisconsin; maybe some other places that we haven't heard about. But to me it is almost incredible.

Banks, transcript at 21,749-54.

While we think there is a defamatory implication in the author's presentation, we cannot distinguish this case from Janklow II. In Janklow II, the article accurately recited historical events and perhaps intentionally left a defamatory implication hanging relating to Janklow's prosecutorial motive. We nevertheless held that the article was protected. 788 F.2d at 1306. Crucially, Price does not challenge the essentials of the author's recitation of events. While he alleges that his contact with local authorities was minimal, the record establishes that, without Moves Camp, Price returned to River Falls and visited and talked with local authorities while they were deciding whether to prosecute. Banks, 383 F. Supp. at 395.

B. Myrtle Poor Bear

Myrtle Poor Bear is acknowledged to be a mentally unstable and unreliable witness. She testified against an A.I.M. leader, Dick Marshall, and signed contradictory affidavits to support the extradition of Peltier from Canada.

The prosecution elected not to call Poor Bear as a witness at Peltier's eventual trial, despite the inclusion in her original affidavits of a supposedly eye-witness account of Peltier murdering the agents. The defense attempted to call her as a witness to exemplify their claim that the government engaged in misconduct. The comments of Judge Ross questioning the F.B.I.'s conduct are set forth, together with views of the prosecutor, Director Webster and Agent Price. Crazy Horse at 371-73, 469, 473. Price objects to six statements that concern his conduct with respect to Poor Bear, alleging that he is accused of suborning perjury. Appellant's Brief at 22.

The first statement merely recites that the F.B.I. "supplied" Poor Bear as a witness against Dick Marshall. Complaint paragraph 10(d). Matthiessen does not say that Price suborned her perjury. He notes only that Price alerted the State to her usefulness. Moreover, the term "supplied" is the description used in F.B.I. reports. F.B.I. Teletype, April 9, 1976 (Defendants' Exhibit no. 52). The remaining five statements raise two issues: was it evident that Poor Bear was unqualified, and was there misconduct with respect to her testimony and affidavits? Complaint paragraphs 10(n), 10(f), 10( o), 10(q) and 10(w).

The United States Attorney, Evan Hultman, who prosecuted Peltier, began his oral argument of Peltier's direct appeal before this Court by conceding that Poor Bear was incompetent to serve as a witness and that her affidavits were false. United States v. Peltier, No. 77-1487, transcript (April 12, 1978); Defendants' Exhibit no. 78. The affidavits were flatly contradictory. The first affidavit said that she left the site of the shootings before the agents arrived. The second was exactly the same, except the sentence saying she left was replaced with one that said that she was present and saw Peltier kill the agents. The next day she gave an agent an explicit description of events not put in any affidavit, alleging that she tried to stop Peltier, that she had been raped by nine people before the incident, and that she escaped from the scene of the killings by shooting at one of Peltier's companions. She also said that A.I.M. had a hit list, that they drugged her by putting things in her coffee, that she was Marshall's girlfriend, that Marshall tried to kill her, that A.I.M. was going to bomb a school, and that she was Peltier's girlfriend. F.B.I. Reports (Jan. 20-26 1976) (Defendants' Exhibit no. 58). She then signed a third affidavit claiming that she tried to escape from the site but was stopped and held by someone so that she was able to watch the shootings. She then alleged that she ran up to Peltier after the agents were shot, hit him, and then ran away. She has since recanted everything. See, Marshall v. State, 305 N.W.2d at 844 (Wollman, C.J., now a member of this Court, dissenting). No verification of her claims was made. Only the last two affidavits were submitted to the Canadian authorities. The Canadian prosecutor has indicated, "I did not have anything to do with the preparation of the affidavits and left this entirely up to the two agents who were controlling Myrtle Poor Bear." Defendants' Exhibit no. 91.

Whether Poor Bear was obviously unreliable or whether Price should have known that she was, are matters of opinion. The author merely comments on what he thinks must have been apparent to anyone who met her, who read the affidavits or, for that matter, who took her statements. Impressions such as this are unverifiable.

Additionally, Matthiessen asserts that Price should have been aware of her mental history, because the man who brought her to Price, Riggs, was aware of her condition. Complaint paragraph 10(q). In reply, Price argues that Riggs, her high school principal, vouched for her credibility, Price Aff. no. 2 at 5, and that Matthiessen failed to interview Riggs. The factual statements are first, that Price and Riggs knew each other, which Price concedes, and second, that Riggs knew of Myrtle's pyschological history, which is borne out by the record.

Reckless disregard is not measured by what a prudent person would have investigated or published. Secrist v. Harkin, 874 F.2d at 1252; Speer v. Ottaway Newspapers, Inc., 828 F.2d 475, 478 (8th Cir. 1987), cert. denied, ___ U.S. ___, 108 S.Ct. 1247, 99 L.Ed.2d 445 (1988). We do not believe that Matthiessen's failure to interview Riggs is fatal to the statement that Riggs knew of Poor Bear's condition and the author's qualified assertion that Riggs could have told Price. Riggs was a peripheral figure, and the Supreme Court has rejected higher showings of reckless disregard.

The second issue concerns misconduct with regard to the affidavits. Complaint paragraphs 10(f), 10( o) and 10(w). It is obvious that, in context, the statements rely on public reports. The author concludes from all the events that the conduct of the prosecutors and agents constituted misconduct and repeats that same accusation as it is made by others. The allegation that she was "exploited" is an opinion. There are only two specific and verifiable statements of fact contained in these statements: First, that Poor Bear's affidavits were false, and second, that Poor Bear alleges that Price and another agent used improper interrogation methods. Both of these statements rely on public proceedings.

In oral argument on Peltier's appeal, Judge Ross and Mr. Hultman, the prosecutor, engaged in the following exchange:

JUDGE ROSS: But anybody who read those affidavits would know that they contradict each other. And why the F.B.I. and Prosecutor's office continued to extract more to put into the affidavits in hope to get Mr. Peltier back to the United States is beyond my understanding.

MR. HULTMAN: Yes.

JUDGE ROSS: Because you should have known, and the F.B.I. should have known that you were pressuring the woman to add to her statement.

MR. HULTMAN: Your Honor, I personally was not present at that stage. I read the affidavits after they had been submitted, so I want this court to know that.

JUDGE ROSS: The Government —

MR. HULTMAN: And I don't excuse, by my remark just now to Your Honor, I don't in any way excuse what the court has just indicated. Your Honor, I have trouble with that myself, and Your Honor that is the exact reason which I did read these affidavits and put together the fact that — And that gets to the second point, Judge Gibson and Judge Ross. It was clear to me her story didn't later check out with anything in the record * * *. * * * And it was then that I personally made the decision that this witness was no witness. First of all, because she was incompetent in the utter, utter, utter ultimate sense of incompetency * * * [and because] there was not one scintilla that showed Myrtle Poor Bear was there, knew anything, did anything, et cetera.

JUDGE ROSS: But can't you see, Mr. Hultman, what happened happened in such a way that it gives some credence to the claim of the —

MR. HULTMAN: I understand, yes, Your Honor.

JUDGE ROSS: — the Indian people that the United States is willing to resort to any tactic in order to bring somebody back to the United States from Canada.

MR. HULTMAN: Judge —

JUDGE ROSS: And if they are willing to do that, they must be willing to fabricate other evidence. And it's no wonder that they are unhappy and disbelieve the things that happened in our courts when things like this happen.

MR. HULTMAN: Judge Ross, I in no way do anything but agree with you totally.

* * * * * *

JUDGE ROSS: We have an obligation to them, not only to treat them fairly, but not give the appearance of manufacturing evidence by interrogating incompetent witnesses.

Defendants' Exhibit no. 78 at 3-5 (emphasis added).

The author cites this exchange at length. Crazy Horse at 371-73. The words used in the book, "exploited" and "misconduct", mirror Judge Ross' commentary and are similar to statements made in two opinions of this Circuit. United States v. Peltier, 800 F.2d at 778 ("misconduct"); United States v. Peltier, 585 F.2d at 335 n. 18; see also, Marshall v. State, supra note 11, at 4-6. The conclusion that the affidavits were false comes from Judge Ross and Mr. Hultman. Poor Bear's allegations of mistreatment by Price were also made on the record. Marshall v. State, 305 N.W.2d at 845 (Wollman, C.J., dissenting). We think Matthiessen's use and repetition of these judicial proceedings is privileged under Greenbelt Cooperative Publishing Assoc. v. Bresler, 398 U.S. 6, 90 S.Ct. 1537, 26 L.Ed.2d 6 (1970) (newspaper article called the plaintiff's tenacious negotiation posture as "blackmail," as well as quoting a statement made at a public meeting to that effect). Moreover, Price's version of the events was presented. Crazy Horse at 468-69.

C. Anna Mae Aquash

Anna Mae Aquash was an A.I.M. activist wanted by the F.B.I. She was the victim of an unsolved homicide. Her body was discovered several miles from the nearest village during the winter following the shootings. She was initially unidentified. At least one F.B.I. agent arrived on the scene to assist the local investigation. Agent Price photographed her body when it was brought to town. The first autopsy attributed her death to exposure, finding only a small contusion on her head. Her hands were removed and sent to Washington, D.C., for identification. Meanwhile, she was buried as Jane Doe. After her identity was discovered, her body was exhumed for a new autopsy. The second autopsy quickly concluded that she had been executed by a bullet fired point-blank into the back of her head, which made an entry wound and left a bulge in her head where it stopped.

Price attacks six statements found in the book which allege that Aquash's death was suspicious, that Price should have recognized her because he had met Aquash before, that there were rumors that the F.B.I. killed her, and that some people thought that Price had killed her. Complaint paragraphs 10(f), 10(g), 10(i), 10(k), 10(m) and 10(x). Price reads these statements together to mean that he is being accused of murder. Appellant's Brief at 19.

Whether or not the circumstances of Aquash's death were "suspicious" is a matter of opinion, as is whether or not Price should have recognized her when the body was found. The only factual statement with regard to the latter is that Price had met Aquash before. Price acknowledges that he had met her twice before, but indicates that he could not recognize her on sight and that her body was partially decomposed when it was found. Price Aff. no. 2 at 6-9. Whether he should have recognized her after she was found is obviously speculative and unverifiable. The author offers statements from doctors and others saying that it might or might not have been possible, given the body's decomposition. Crazy Horse at 260-67. Crazy Horse offers the opinion of the doctor from the second autopsy and the man who found her that recognition would have been difficult. The author asked Price why he failed to recognize her and reports Price's answer. Id. at 469.

Next, we consider the reporting of rumors about the murder, together with the author's summation, "No one I have talked to makes the serious claim that David Price killed Anna Mae Aquash * * *." Crazy Horse at 445. Rumors that talk of the F.B.I. or white people in general but fail to mention Price are irrelevant in the absence of a direct accusation. Complaint paragraphs 10(i) and 10(m). We therefore focus on the statement: "Kunstler took this matter up with Agent Price, whom some of the Indians, at least, had suspected of involvement in the killing * * *." Complaint paragraph 10(k). This phrase appears as a transition in the book between attorney Kunstler's questioning of an A.I.M. member about Aquash's death and his questioning of Agent Price about her death. Crazy Horse at 304-08. During an offer of proof, the A.I.M. witness testified that Aquash had said Price had threatened Aquash with death if she didn't cooperate. It is immediately after this that the statement reporting suspicions about Price appears. This statement is in turn immediately followed by Price's testimony that he made no such threat. Id. at 306-07.

Whether or not Price killed Aquash is a question of fact. But obviously the author makes no such accusation. He exonerates Price of killing her. Complaint paragraph 10(m). We do not believe that Price can choose what meanings to attach to these statements where several are available. Fudge, 840 F.2d at 1016.

Whether or not people suspected Price of involvement is another matter. First, it is not clear what exactly Price is suspected of — "involvement" is a vague term. It might refer to aiding or abetting, or to hiding things in the investigation, or to making her a target by making her an informer, or merely to making her a target by questioning her, thus giving her the appearance of being an informer. These possibilities are also raised by reading the statements together. The suspicions about "involvement" are thus too vague for us to conclude that Price is accused of criminal conduct. Secrist v. Harkin, 874 F.2d at 1250-51. Surely, Price would not want to be sued for his claim that an Indian named Black Crow was involved or had information about Aquash's murder. Crazy Horse at 469-70.

We reach the same conclusion with respect to the other statements regarding Aquash. One statement quotes someone saying that he thought a law enforcement official killed her, then the excerpt in Price's complaint quotes the author saying that "Inevitably, local suspicion focused on David Price * * * ." Id. Complaint paragraph 10(i). In between these two quoted parts of the statement, the text of the book omitted from the complaint indicates that the F.B.I. in general is accused of blackmail, as well as a lack of interest and faulty work in the investigation. It is thus not clear whether local people suspected Price of murder, blackmail, or merely of poor investigating. This statement is also unspecific.

Furthermore, were we to focus on Price's interpretation, we would nevertheless affirm because Price cannot show that this statement or the others are false or made with reckless disregard. The fact at issue would be whether people suspect Price of being "involved." Price's denials of involvement are not to the point. Moreover, the author quotes sources voicing their suspicions. Crazy Horse at 267.

We do not know how authors can ever write about controversies without reporting accusations and counter-accusations. Price was given a chance to respond, which he used to accuse an Indian named Black Bear of involvement. Crazy Horse at 469. The book also includes official F.B.I. statements exonerating the investigations. Crazy Horse at 267. We believe that the author has merely recited the accusations and counter-accusations, and has strongly indicated that he believes Price is innocent of the death of Aquash. The author does not espouse either side's accusations. We believe that these statements are protected.

D. Harassment

Two of the statements listed in the complaint allege investigatory harassment. Complaint paragraphs 10(a) and 11(g). The first describes Price's conduct in evaluative terms, "open and even aggressive surveillance * * * `he even photographed our cats and dogs.'" This statement is full of rhetorical hyperbole and would be understood as an opinion.

The second statement quotes the account of a local Indian girl that Price entered her home without a warrant. This is factual, and thus we are concerned with its truth or falsity and whether the author would have reason to believe it was false. In his first affidavit, Price did not indicate that he had a warrant. Attached to his second affidavit, however, Price included a copy of the relevant F.B.I. report. The report indicates that he had a "witness" warrant. Price Aff. no. 2, Exhibit H. We are not sure if the report refers to a subpoena or a warrant, nor is it clear whether Price had a copy of either with him. Neither side has presented us with any direct evidence on this point. We believe that this may be sufficient to create a contest as to the existence of a warrant. It is, in any event, insufficient to show by clear and convincing evidence that Matthiessen reported the girl's account with reckless disregard. The report states:

EVELYN BORDEAUX then went to a hallway leading to the bedrooms and bathroom, and stood in the hallway screaming at the Agents that they needed a search warrant and that they had to have a copy of the arrest warrant.

Price has never indicated that he showed the family any documents. There is no evidence that the author knew or should have known that the girl's account was false, if it is.

E. Character Statements

The remaining statements are all character attacks on Price. Complaint paragraphs 10(s), 10(t), 10(u) and 10(v). The first quotes an attorney calling Price "corrupt and vicious." The other three statements express the view of the author that Price alters facts to fit his pre-conceptions, that he would be soiled by his acts no matter whether he paid any penalties, and that he exemplified dishonest government. All these statements are opinion. They are unspecific, unverifiable and are protected criticisms of a public official's performance of his duties. The surrounding text constitutes the author's reflections on Price and clearly signals opinion.

IV. RECKLESS DISREGARD

In St. Amant v. Thompson, the Court indicated what types of evidence might suffice to show reckless disregard in a story: if there was evidence that the story was fabricated, if the story was inherently improbable, if the sources were entirely anonymous and their reports unverified, or if there were particular reasons to doubt sources. 390 U.S. at 732, 88 S.Ct. at 1327; accord Tavoulareas v. Piro, 817 F.2d 762, 790 (D.C. Cir.), cert. denied, 484 U.S. 870, 108 S.Ct. 200, 98 L.Ed.2d 151 (1987).

Price's arguments lack the specificity called for in St. Amant. He has to show by clear and convincing evidence that Matthiessen published particular false material facts with knowledge of their probable falsity. He never contests the accuracy of Matthiessen's quotations of primary sources. In lieu of evidence that shows Matthiessen doubted particular statements, Price makes general allegations relating to Matthiessen's slant. He cites the profit-sharing agreement as evidence that Matthiessen had an incentive to defame him. He faults Matthiessen for relying on "hoodlums, criminals and movement lawyers." Appellant's Brief at 40.

On this record, Price has failed to show by clear and convincing evidence that Matthiessen published particular facts with knowledge of their probable falsity. Instead, Price has turned the legal argument into the type of political rhetoric best left to open and free speech. Far more substantial and specific claims than this have failed. See supra note 13.

Viking Penguin

The existence of reckless disregard must be proved with respect to each defendant. Cantrell v. Forest City Publishing Co., 419 U.S. 245, 95 S.Ct. 465, 42 L.Ed.2d 419 (1974); St. Amant v. Thompson, 390 U.S. at 731-32, 88 S.Ct. at 1325-26; Speer, 828 F.2d at 477; Hunt v. Liberty Lobby, 720 F.2d 631, 648 (11th Cir. 1983). Cantrell permits reckless disregard to be imputed only on a theory of respondeat superior. If Matthiessen is not an employee of Viking, independent evidence of Viking's culpability is required. Hunt v. Liberty Lobby, 720 F.2d at 649; Reader's Digest Ass'n v. Superior Court, 37 Cal.3d 244, 208 Cal.Rptr. 137, 145-46, 690 P.2d 610, 618-19 (1984).

The deposition of Viking's editor for this project, Elizabth Sifton, makes clear that Viking played only a limited role in the story's development. Sifton had worked with him on his award-winning book, The Snow Leopard, and considered him to be an "immensely meticulous, scrupulous and obsessively attentive writer who reworks and reworks and reworks his manuscripts until he has them the way he wants them." Sifton Deposition at 48-49 (Oct. 8, 1986). Because of his reputation and success, Matthiessen is in demand by publishing houses. Id. at 65-69. Sifton's role was to be a "sounding board." Id. at 80. He consulted her on structural matters and for expository decisions on how to allocate space and attention. Id. at 81, 111. Her principal concerns were length, id. at 83, 90-91, and coherence, id. at 111. Viking undertook no factual review, but the editors were aware of Matthiessen's fact checking process of looking at original documents and sending out drafts to people acquainted with the events. Id. at 114. Matthiessen clearly was not an employee of Viking such that liability could be imputed, were we to find any. Rinaldi v. Holt, Rinehart Winston, Inc., 42 N.Y.2d 369, 397 N.Y.S.2d 943, 951, 366 N.E.2d 1299, 1307 cert. denied, 434 U.S. 969, 98 S.Ct. 514, 54 L.Ed.2d 456 (1977).

V. CONCLUSION

Sometimes it is difficult to write about controversial events without getting into some controversy along the way. In this setting, we have decided that the Constitution requires more speech rather than less. Our decision is an anomaly in a time when tort analysis increasingly focuses on whether there was an injury, for in deciding this case we have searched diligently for fault and ignored certain injury. But there is a larger injury to be considered, the damage done to every American when a book is pulled from a shelf, as in this case, or when an idea is not circulated.

As a result, we have asked the plaintiff to do a great deal more than establish his disagreement with the book. We have asked him to select what he finds objectionable, to explain why it has a precise core of meaning, to persuade us that it is a factual matter capable of the jury's resolution, to show that he could demonstrate its falsity, and to convince us that he could show with clear and convincing evidence that the author had a high degree of awareness of a particular statement's probable falsity. That Agent Price cannot surmount these obstacles implies no condemnation of him. He worked diligently for the government while he was assigned to South Dakota. Local prosecutors commended him for his efforts.

Agent Price is particularly noted by our office for his enthusiam and the zeal with which he works and prepares his cases. Without referring to our records, I believe it can safely be said that a conviction has been obtained in every case prepared by David Price and prosecuted in the Western Division of this District. This alone speaks highly for the caliber of work which he has presented to our office. It is the view of our Rapid City office that Agent Price has been and continues to be highly valuable in our efforts to maintain a high rate of successful prosecutions.

Correspondence from U.S. Attorney for South Dakota to the FBI. Price. Off. No. 2. Exhibit A. Unfortunately, Price also became involved in public controversies during his service and, in the posture of this case, we must focus on the bad things that have been said.

At the same time, we recognize that the plaintiff's complaint has taken statements from their setting in a way that distorts the flavor of the book. In its entirety, Crazy Horse focuses more on public institutions and social forces than it does on any public official. The sentiments it expresses are debatable. We favor letting the debate continue.

APPENDIX

* * * * * *

It is very obvious throughout the course of examination that Myrtle Poor Bear is one who can be very easily led and her testimony is fairly fraught with contradiction. Her value as a witness is obviously very limited.