Chief Justice PHILLIPS
delivered the opinion of the Court,
in which Justice ENOCH, Justice BAKER, Justice ABBOTT, Justice HANKINSON, Justice O’NEILL, and Justice GONZALES joined.
We must decide whether a media defendant sued for defamation by a public official is entitled on the facts of this record to summary judgment on the issue of actual malice. Because the defendant produced [417]*417evidence negating actual malice as a matter of law, and because the plaintiff did not produce controverting evidence raising a fact issue, we affirm the summary judgment granted by the court of appeals. 995 S.W.2d 152.
I
When this claim arose, Charles Dean Huekabee was presiding judge of the 247th District Court of Harris County, which by statute gives preference to family law matters. See Tex. Gov’t Code § 24.424. Judge Huekabee claims that Respondent, HBO, defamed him by broadcasting the documentary Women on Trial on its premium cable channel. This hour-long program chronicled four southeast Texas cases in which family courts granted custody of a child to the father after the mother accused the father of child abuse. Three of these cases arose in Harris County, and Judge Huekabee presided over two of them. Judge Huekabee principally claims that the documentary defamed him in its report on his decision regarding the custody of four-year-old Wayne Hebert. See In re the Marriage of Sandra Hebert and Michael Hebert, No. 84-18892, In re John Hebert and Wayne Hebert, Minor Children, No. 88-14873 (consolidated cases)(247th Dist. Ct., Harris County, Tex. Mar. 13-15,1988)(“Hebert ”).
The Hebert case began in 1988, when Sandra Hebert discovered that Wayne had sustained an injury to his penis. The day before, Wayne had returned from visiting Michael Hebert, his father and Sandra’s ex-husband. Wayne had gone with Michael to visit his grandmother’s home in Louisiana. Believing that Michael caused Wayne’s injury, Sandra consulted with her friend Sherry Turner, a Houston police officer who specialized in sexual abuse cases. Turner, interviewing Wayne alone, videotaped Wayne’s statement that Michael had injured him while taking a bath. In two other videotaped interviews, Wayne also told social worker Cheryl Bennett and Child Protective Services caseworker Wilma Smith that Michael caused the injury. After investigating further, Smith concluded that Wayne had been abused, but that the abuser could not be identified. Because Michael was a Houston police officer, the Houston Police Department’s Internal Affairs Department also investigated the incident and likewise determined that the abuser could not be identified.
Alleging that Michael had abused Wayne, Sandra moved to modify the custody order to restrict Michael’s visitation rights. After a three-day hearing in March 1988, Judge Huekabee rendered a temporary order that not only made Michael rather than Sandra the managing conservator of Wayne, but went on to deny Sandra all access to her child, even though Michael had not sought either of these changes. Sandra unsuccessfully sought a writ of mandamus from the court of appeals to overturn the temporary order. See Hebert v. Huckabee, No. A14-88-00511-CV, 1988 WL 73789 (Tex.App.—Houston [14th Dist.], July 14, 1988, orig. proceedingXnot designated for publication). She did not seek a subsequent modification of the order, and it was still in effect when Women on Trial was broadcast in 1992.
In late 1990, Lee Grant, the director of Women on Trial, first began work on a documentary about divorce. Hoping to examine how once happily married couples later ended up in bitter divorces, Grant secured her husband’s production company, Joseph Feury Productions (JFP), to produce the film. Grant assigned JFP employee Virginia Cotts to find suitable stories for the program. In March 1991, Cotts met in Houston with Joleen Reynolds, the leader of Citizens Organized for Divorce Ethics and Solutions (CODES), a support group for men and women who felt that the Houston family courts had treated them unfairly. Reynolds discussed a number of cases with Cotts, including the Hebert case. After meeting with [418]*418Reynolds, Cotts wrote a three-page summary of Sandra Hebert’s situation.
Sandra Hebert’s story was included along with several others submitted by Cotts and Grant to HBO in April 1991. In her summary of the Hebert story, Cotts included the following bullet points: (1) “Police ex-husband abused son”; (2) “Corrupt Judge gave custody to father/abuser”; (3) and “Sandy lost all rights to see her child.” Sandra Hebert’s story particularly impressed the HBO executives. After reading it, HBO vice-president Sheila Nev-ins wrote on her copy: “Great story. Do at once.” Nevins’s assistant, Cis Wilson, wrote: “Great, sad story.” After considering the proposal, HBO agreed to purchase the film. Throughout the rest of the film’s production, Cotts and Grant regularly met with Wilson and Nevins.
Cotts and Grant both came to Houston to film interviews. In addition to Joleen Reynolds, Sandra Hebert, and her current and former attorneys, they also interviewed Ivy Raschke, another woman who had been denied access to her children by Judge Huckabee after accusing the children’s father of abuse. Cotts also continued her research into other allegations of impropriety in the Harris County family courts, including those reported by local print and broadcast media.
In September 1991, JFP delivered a “rough cut” of the film to HBO. Cotts’s contemporaneous status report revealed tension between Lee Grant and Sheila Nevins over the film’s direction. Grant apparently wanted to present a broad picture of divorce that showed both the fathers’ and the mothers’ perspectives, but Nevins wanted a narrower piece that focused on mothers who believed the family court system had treated them unfairly. Nothing in the status report, however, indicated that Grant, Cotts, or anyone at HBO believed anything in the documentary to be false or entertained serious doubts about the truth of any of the film’s allegations.
In November 1991, Grant and Cotts returned to Houston and videotaped Judge Huckabee. While Judge Huckabee stated that he could not talk specifically about the Hebert case because it was pending in his court, he did agree to talk about it in “hypothetical” terms. He then explained that all of his decisions in this and other cases were based on the best interests of the children. HBO did not include these statements in the final version. Instead, it aired this response by Judge Huckabee to a question about a “hypothetical” version of the Hebert case:
I have to do what’s best for the child. If someone, is, uh, brainwashing the a child to the same extent that it causes psychological and emotional problems with the child, especially coupled with some physical abuse, in my opinion the child has to be removed from that situation.
The broadcast also aired Judge Hucka-bee’s explanation of his criteria for determining when a mother in that situation could see her children again:
Well, if its [sic] a person who has mental health problems, they’re going to have to seek mental health, uh, care. If its [sic] a person sexually abusing a child, they’re probably going to have to seek mental health care.
Finally, the broadcast aired Judge Hucka-bee’s statement that he took the decision to deny access to a parent very seriously, but that he was satisfied that he had made the correct decision in every case in which he had done so.
The filmmakers also interviewed Dr. Kit Harrison, a psychologist appointed by Judge Huckabee in Hebert and in many other cases. Four months after Judge Huckabee rendered the temporary order denying Sandra Hebert access to Wayne, Dr. Harrison issued a report concluding that Michael had not caused Wayne’s injury. Rather, the report concluded that Wayne’s older brother John committed the abuse while Wayne was in Sandra’s custody. Based on this belief, Dr. Harrison agreed with Judge Huckabee’s decision to [419]*419transfer Wayne to his father’s custody and deny Sandra access to the child. Although the final version mentioned Dr. Harrison’s recommendation approving of the judge’s order, it did not detail Dr. Harrison’s reasons.
Finally, Cotts and Grant interviewed Houston attorney Randy Burton, an outspoken critic of the Houston family courts. Among other things, Burton accused the Harris County family court judges of practicing cronyism and disregarding the best interests of the children before them.
After these interviews, Cotts and Grant recut the film to include some of the new footage. From April to September 1991, HBO and JFP’s lawyers reviewed the film, finally allowing the film to air in October 1991. HBO also agreed to indemnify JFP should a judgment arise from the film in excess of JFP’s errors-and-omissions insurance coverage.
Women on Triol aired on October 28, 1992. In addition to the Sandra Hebert and Ivy Raschke segments, the film included two other stories. In one, another Harris County family district court judge, Allen Daggett, had transferred custody of Mary Frances Parker’s child to her ex-husband, a convicted rapist, even though she claimed that he was abusing the child. In the other, Sherry Nance was convicted of murdering her ex-husband and his father after a Bee County jury awarded custody of her son to the ex-husband. Nance claimed that she killed her ex-husband to save her son from continuing sexual abuse. The documentary did not name the judge in the Bee County case.
Judge Huckabee sued HBO, JFP, Grant, and Burton, claiming that they had defamed him both by particular statements and by portraying him in general as a judge who knowingly disregarded children’s best interests. In addition to his claim that the entire documentary defamed him, Judge Huckabee alleged as false and defamatory these statements: (1) the Houston family courts were “filled” with cases “irrational in their decisions” and “medieval in their punishment”; (2) “[wjomen who charge their husbands with abuse are often viewed as mentally unstable and routinely lose custody of their children”; (3) all the rulings depicted in the documentary happened in one courthouse; and (4) Randy Burton’s conclusions in the film that the Houston family courts “were the last bastion of the good ole’ boy system” and that the judges in those courts were guilty of “conscious indifference to the child” and “legalized child abuse.” Judge Huckabee also alleged that the film’s description of the Hebert case omitted important facts that would have led viewers to conclude that his HebeH order was justified.
After discovery, HBO moved for summary judgment asserting that: (1) HBO published the film without actual malice; (2) Judge Huckabee’s claim actually pleaded a cause of action for false light invasion of privacy, which Texas law does not recognize; (3) all the statements concerning Judge Huckabee were literally or substantially true; (4) these statements were constitutionally protected statements of opinion; and (5) Women on Triol was privileged as a fair and reasonable comment on, or criticism of, an official act of a public official and a matter of public concern. See Tex. Civ. Peac. & Rem.Code § 73.002(b)(2). After the trial court denied HBO’s motion for summary judgment, HBO appealed as a media defendant who was denied a motion for summary judgment “arising under the free speech or free press clause of the First Amendment to the United States Constitution, or Article 1, Section 8, of the Texas Constitution, or Chapter 73” of the Texas Civil Practice and Remedies Code. Tex. Civ. PraC. & Rem.Code § 51.014(a)(6). The court of appeals, whose jurisdiction was not challenged, reversed and rendered judgment for HBO on the sole ground that HBO negated one essential element of Judge Hucka-bee’s case by conclusively proving that it broadcast Women on Trial without actual [420]*420malice. 995 S.W.2d 152. We granted Judge Huckabee’s petition for review under our jurisdiction to hear cases appealed under section 51.014(a)(6). See Tex. Gov’t Code § 22.225(d).
II
To recover for defamation, a public figure or public official, such as Judge Huckabee, must prove that the defendant published a false and defamatory statement with actual malice. WFAA-TV, Inc. v. McLemore, 978 S.W.2d 568, 571 (Tex.1998); Carr v. Brasher, 776 S.W.2d 567, 569 (Tex.1989); see also New York Times Co. v. Sullivan, 376 U.S. 254, 279-80, 84 S.Ct. 710, 11 L.Ed.2d 686 (1964). As we resolve this case solely on the issue of whether HBO negated actual malice as a matter of law, we assume without deciding that the documentary either expressly or implicitly made false statements about Judge Huckabee. We also do not reach the issue of whether any of these statements, even if false, were not defamatory because the documentary’s overall portrayal of Judge Huckabee was substantially true. See McIlvain v. Jacobs, 794 S.W.2d 14, 16 (Tex.1990).
Actual malice in a defamation case is a term of art. Unlike common-law malice, it does not include ill-will, spite, or evil motive. Casso v. Brand, 776 S.W.2d 551, 558 (Tex.1989). Rather, to establish actual malice, a plaintiff must prove that the defendant made the statement “with knowledge that it was false or with reckless disregard of whether it was true or not.” New York Times, 376 U.S. at 279-80, 84 S.Ct. 710. Reckless disregard is also a term of art. To establish reckless disregard, a public official or public figure must prove that the publisher “entertained serious doubts as to the truth of his publication.” St. Amant v. Thompson, 390 U.S. 727, 731, 88 S.Ct. 1323, 20 L.Ed.2d 262 (1968). Finally, to prevail at trial, a plaintiff must establish actual malice by clear and convincing evidence. See Bose Corp. v. Consumers Union of U.S., Inc., 466 U.S. 485, 511, 104 S.Ct. 1949, 80 L.Ed.2d 502 (1984); Casso, 776 S.W.2d at 558.
In Texas, under our traditional summary judgment procedure, defendants can obtain summary judgment only if they conclusively negate one of the elements of the plaintiffs claim. Tex.R. Civ. P. 166a(c); Phan Son Van v. Pena, 990 S.W.2d 751, 753 (Tex.1999); Casso, 776 S.W.2d at 556.1 A libel defendant can negate actual malice as a matter of law by presenting evidence that he or she did not publish the statement with knowledge of its falsity or reckless disregard for its truth. McLemore, 978 S.W.2d at 574; Casso, 776 S.W.2d at 559. Once the defendant has produced evidence negating actual malice as a matter of law, the burden shifts to the plaintiff to present controverting proof raising a genuine issue of material fact. See Tex.R. Civ. P. 166a(c); Phan Son Van, 990 S.W.2d at 754; City of Houston v. Clear Creek Basin Auth., 589 S.W.2d 671, 678 (Tex.1979).
Respondents and various amici aligned with their position suggest that we abandon our traditional summary judgment standard in public-figure defamation cases and adopt the federal summary judgment standard established in Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). Under this standard, “the appropriate summary judgment question will be whether the evidence in the record could support a reasonable jury finding that the plaintiff has shown actual malice by clear and convincing evidence or that the plaintiff has not.” Id. at 255-56, 106 S.Ct. 2505. Respondents and amici reason that because a plaintiff must satisfy the clear-and-convincing standard to prevail at trial, logic dictates that this standard should also apply at the summary judgment stage. Adopt[421]*421ing the clear-and-eonvincing standard at the summary judgment stage, they argue, would also align Texas practice with most other states.2 Finally, they contend that the heightened evidentiary standard is needed at the summary judgment stage to protect media defendants from the costs associated with defending groundless defamation actions. Failure to protect media organizations against these costs, they assert, will lead to self-censorship, thereby compromising the First Amendment’s guarantee of a free press.
We decline to adopt the clear-and-convincing requirement at the summary judgment stage. In Casso v. Brand, 776 S.W.2d 551 (Tex.1989), we held that neither the United States Constitution nor the Texas Constitution mandated a special summary judgment procedure in public-figure defamation cases. Id. at 555-57. We concluded that the United States Supreme Court’s requirement that a plaintiff come forward with sufficient proof to allow a jury finding of actual malice by clear- and-convincing evidence was based merely on federal procedure. See id. at 555-56. Although we recognized the importance of “encouraging free and untrammeled expression on matters of public concern or interest,” we believed that the plaintiffs heavy burden of proving actual malice at trial adequately protected these important liberty interests. Id. at 557. To some extent, we based our holding in Casso on the different role of summary judgment in the Texas and federal systems. See id. at 556. At that time, our state’s summary judgment practice served only the limited purpose of “ ‘eliminatfing] patently unmer-itorious claims and untenable defenses,’ ” id. (quoting Clear Creek Basin Auth., 589 S.W.2d at 678 n. 5), while in the federal system it played an “integral part” in “ ‘securing] the just, speedy and inexpensive determination of every action.’ ” Id. (quoting Celotex Corp. v. Catrett, 477 U.S. 317, 327, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986) (quoting Fed.R.Civ.P. 1)). Although our recent adoption of the no-evidence summary judgment as an alternate procedure in Texas obviated, to some extent, the differences in summary judgment procedure between the two systems, our holding in Casso was also consistent with practical considerations, which remain valid today.
One consideration is the difficulty in adapting review under a heightened evi-dentiary standard to Texas summary judgment practice. Requiring the trial court to determine at the summary judgment stage whether a reasonable juror could find the evidence to be clear and eonvinc-[422]*422ing suggests that the trial court must weigh the evidence. See Anderson, 477 U.S. at 266, 106 S.Ct. 2505 (Brennan, J., dissenting); Moffatt v. Brown, 751 P.2d 939, 944 (Alaska 1988) (“ ‘[T]he clear-and-convincing test inevitably implicates a weighing of the evidence, an exercise that intrudes into the province of the jury.’” (quoting Dairy Stores, Inc. v. Sentinel Publishing Co., 104 N.J. 125, 516 A.2d 220, 236 (1986))). Texas law has always emphasized that trial courts must not weigh the evidence at the summary judgment stage. See Gulbenkian v. Penn, 151 Tex. 412, 252 S.W.2d 929, 931 (1952); 3 McDonald, Texas Civil Practice § 18.26, at 499 (Allen et al., eds.1992). Instead, a trial court’s only duty at the summary judgment stage is to determine if a material question of fact exists. See Gulbenkian, 252 S.W.2d at 931. Unless constitutionally mandated, we see no reason to upset this traditional demarcation between fact-finder and judge by requiring trial courts to weigh the evidence at the summary judgment stage.
We are reminded that the majority in Anderson insisted that its standard did not require trial courts to weigh evidence at the summary judgment stage. With all due respect, we agree with Justice Brennan’s dissenting opinion on this point:
I simply cannot square the [majority’s] direction that the judge “is not himself to weigh the evidence” with the direction that the judge also bear in mind the “quantum” of proof required and consider whether the evidence is of sufficient “caliber or quantity” to meet that “quantum.” I would have thought that a determination of the “caliber and quantity,” i.e., the importance and value, of the evidence in fight of the “quantum,” i.e., the amount “required,” could only be performed by weighing the evidence.
Anderson, 477 U.S. at 266, 106 S.Ct. 2505 (Brennan, J., dissenting)(emphasis in original). Several commentators have agreed that trial judges cannot determine the “caliber and quantity” of evidence without performing some of the functions of a finder of fact. See Issacharoff & Loewenstein, Second Thoughts About Summary Judgment, 100 Yale L.J. 73, 85 (1990); Mullenix, Summary Judgment: Taming the Beast of Burdens, 10 Am. J. Trial Anvoc. 433, 462 (1987); (“So replete is the decision with contradictory pronouncements that opposing counsel can in the future legitimately cite Anderson’s dicta for completely repugnant propositions.”); Stempel, A Distorted Mirror: The Supreme Court’s Shimmering View of Summary Judgment, Directed Verdict, and the Adjudication Process, 49 Ohio St. L.J. 95, 115-16 (1988).
Furthermore, the elear-and-convincing standard provides little guidance regarding what evidence is sufficient for a plaintiff to avoid summary judgment. See Anderson, 477 U.S. at 270, 106 S.Ct. 2505 (Rehnquist, J., dissenting). We have defined clear and convincing evidence as “ ‘that measure or degree of proof which will produce in the mind of the trier of fact a firm belief or conviction as to the truth of the allegations sought to be established.’ ” In re G.M., 596 S.W.2d 846, 847 (Tex.1980)(quoting State v. Addington, 588 S.W.2d 569, 570 (Tex.1979)); see also Tex. Civ. Prac. & Rem.Code § 41.001(2). Clearly, this standard is vague. Accordingly, we have been reluctant to require it except in those “extraordinary circumstances” when that degree of proof is mandated by constitutional or statutory requirements. Ellis County State Bank v. Keever, 888 S.W.2d 790, 792 (Tex.1994)(clear and convincing proof not required for malicious prosecution); accord Rhodes v. Cahill, 802 S.W.2d 643, 645 n. 2 (Tex.1990)(adverse possession); Sanders v. Harder, 148 Tex. 593, 227 S.W.2d 206, 209 (1950)(trespass to try title). On a cold summary judgment record, without having observed a single witness, it would take keen insight to forecast accurately whether probative evidence would or would not produce a “firm belief or conviction” in the mind of the trier of fact. The distinction, in a paper record,
[423]*423between evidence that will merely raise a fact issue and evidence that will be clear and convincing is generally subtle, if not wholly subjective.
Because of the difficulty faced by a tidal judge in applying the clear-and-convincing standard at the summary judgment stage, Justice Rehnquist predicted that Anderson would “cause the decisions of trial judges on summary judgment motions in libel cases to be more erratic and inconsistent than before.” Anderson, 477 U.S. at 272-73, 106 S.Ct. 2505 (Rehnquist, J., dissenting); see also id. at 258, 106 S.Ct. 2505 (Brennan, J., dissenting)(“I am unable to divine from the Court’s opinion how these evidentiary standards are to be considered, or what a trial judge is actually to do in ruling on a motion for summary judgment.” (emphasis in original)); Mulle-nix, supra, at 461; Stempel, supra, at 180-SI. Although we cannot empirically determine whether this prediction has in fact come to pass, we see no reason to risk such an outcome by departing from our traditional summary judgment standard, especially when the heightened standard of proof adequately safeguards the First Amendment rights of defamation defendants at the trial and appellate stages. After a record has been established at trial, courts must independently review the record to determine if the jury’s finding of actual malice was, as a matter of law, supported by clear and convincing evidence. See Doubleday & Co. v. Rogers, 674 S.W.2d 751, 755 (Tex.1984)(eiting Bose Corp., 466 U.S. at 511, 104 S.Ct. 1949). We believe it obvious that this determination may be more easily and accurately made after a trial on the merits. Cf Stempel, supra, at 177 (“A judicial decision overturning a jury verdict after trial and deliberation is based on a far more extensive data base than a grant of summary judgment_”).
We therefore believe that if a fact issue exists at the summary judgment stage, the evaluation about whether a reasonable jury would find the plaintiffs evidence to be clear and convincing is best made after the facts are fully developed at trial. That most other jurisdictions have accepted Anderson should not compel us to adopt a standard that is contrary to our traditional jurisprudence and difficult to apply in practice. As Respondents and amici have presented no authority that would constitutionally require it, we decline to adopt the clear-and-convineing standard at the summary judgment stage of a public-figure defamation case. To the extent that they hold or suggest to the contrary, we disapprove of the decisions in Rogers v. Cassidy, 946 S.W.2d 439, 446 (Tex.App.—Corpus Christi 1997, no writ); Hill v. Herald-Post Publishing Co., 877 S.W.2d 774, 781 (Tex.App.—El Paso), rev’d in part on other grounds, 891 S.W.2d 638 (Tex.1994); Schauer v. Memorial Care Sys., 856 S.W.2d 437, 446 (Tex.App.—Houston [1st Dist.] 1993, no writ).
Ill
HBO supported its motion for summary judgment with affidavits from Lee Grant, Sheila Nevins, Cis Wilson, and Virginia Cotts. Grant’s affidavit stated that she neither believed the film to have contained a false statement nor entertained any doubts about the truth of any statement regarding Judge Huckabee. Her sources for the Sandra Hebert story included the transcript of the March 1988 hearing, information from Sandra’s current and former lawyers, and research by Virginia Cotts.
Sheila Nevins’s affidavit stated that as vice-president for documentaries and family programming for HBO, she relied on the favorable reputations for accuracy and truthfulness of both Grant and JFP and her own favorable personal experience with their earlier work. She was aware of Grant and Cotts’s efforts to ensure the film’s accuracy, and she neither believed any statement in the documentary to be untrue nor harbored any doubts about the film’s truthfulness. Cis Wilson’s affidavit contained similar statements.
[424]*424HBO presented two extensive affidavits from Virginia Cotts. In her first affidavit, Cotts explained the steps she took in researching the stories presented in Women on Trial. To ensure that the film’s account of the Hebert case was accurate, she (1) reviewed the transcript from the March 1988 hearing, (2) interviewed Sandra and her attorneys, (3) viewed all three videotapes of Wayne Hebert, (4) reviewed articles in the Houston press describing problems in the family courts, and (5) read Dr. Harrison’s deposition in the Hebert case. In all, Cotts reviewed over two thousand pages of documents in connection with the Texas cases. From this extensive review, Cotts stated that she believed that the film’s depiction of the Hebert case was accurate and that she had no doubts regarding this account.
Cotts’s second affidavit detailed her reasons for doubting Dr. Harrison’s conclusion that Wayne’s brother was the abuser, such as (1) her own viewing of the videotapes in which Wayne identified his father as the abuser; (2) the improbability of Harrison’s theory that Wayne’s older brother John had injured him using a favorite toy; (3) the fact that Wayne’s initial description of events was similar to stories that John had told Cotts about abuse from his father; (4) Dr. Harrison’s own statement in a scholarly paper that children often recant after disclosing sexual abuse; and (5) the fact that Wayne had sustained a similar injury once before. Cotts buttressed her conclusion by attaching her own notes from the Harrison interview, indicating that she did not believe his explanation even as the interview was in progress.
Because these affidavits are from interested witnesses, they will negate actual malice as a matter of law only if they are “clear, positive, and direct, otherwise credible and free from contradictions and inconsistencies, and [able to be] readily controverted.” Tex.R. Crv. P. 166a(c); see also Casso, 776 S.W.2d at 558. In actual malice cases, such affidavits must establish the defendant’s belief in the challenged statements’ truth and provide a plausible basis for this belief. See McLemore, 978 S.W.2d at 574; Carr, 776 S.W.2d at 571. As all four of HBO’s affidavits satisfied the Rule 166a(c) requirements, HBO negated actual malice as a matter of law.
Thus, the burden shifted to Judge Huckabee to present evidence to raise a fact issue. He offered six categories of allegedly controverting evidence: (1) HBO and JFP’s alleged desire to portray him in an unflattering light; (2) editorial choices by HBO and JFP that left a false impression of events; (3) the filmmakers’ disregard for Judge Huckabee’s and Dr. Harrison’s explanations for Judge Huckabee’s order; (4) JFP’s and HBO’s alleged purposeful avoidance of the truth; (5) HBO’s extensive legal review of the film, the film’s many rewrites, and the indemnification agreement between HBO and JFP; and (6) HBO’s and JFP’s decision to air the film despite the knowledge that it contained inaccurate statements. In determining whether the evidence presents a fact issue, we assume that all facts favorable to the nonmovant are true and indulge all reasonable inferences in that party’s favor. See Phan Son Van, 990 S.W.2d at 753. Even under this lenient standard, we are persuaded that Judge Huckabee has not raised a genuine issue of material fact on any of his categories.
1. HBO’s desire to portray Judge Huckabee in an unflattering light.
In claiming that JFP and HBO intended to portray him unfairly, Judge Huckabee first points to Virginia Cotts’s three-page summary of the Hebert case describing him to HBO executives Sheila Nevins and Cis Wilson as a “corrupt judge.” He also points to Cotts’s September 1991 status report regarding the disagreement between Grant and Nevins over the film’s artistic direction. Neither of these documents, however, indicates actual malice. While Cotts’s original memo might suggest personal ill-will toward [425]*425Judge Huckabee, nothing in either of these documents suggests that Cotts or Grant had any doubts about the truth of the broadcast. See Carr 776 S.W.2d at 571; Casso, 776 S.W.2d at 558.
Likewise, Nevins’s insistence that the filmmakers focus on divorce from the women’s perspective is no evidence of actual malice. Without more, mere evidence of pressure to produce stories from a particular point of view, even when they are hard-hitting or sensationalistic, is no evidence of actual malice. See Tavoulareas v. Piro, 817 F.2d 762, 796 (D.C.Cir.1987)(en banc); Perez v. Scripps-Howard Broadcasting Co., 35 Ohio St.3d 215, 520 N.E.2d 198, 204 (1988)(both holding that editorial pressure to produce sensationalistic stories is not evidence of actual malice); see also Harte-Hanks Communications, Inc. v. Connaughton, 491 U.S. 657, 667, 109 S.Ct. 2678, 105 L.Ed.2d 562 (1989)(“Nor can the fact that the defendant published the defamatory material in order to increase its profits suffice to prove actual malice.”). Although evidence that HBO directed Grant to produce a sensational story without regard for its truth would raise a fact question, Judge Huckabee has not produced any such evidence. See Tavoulareas, 817 F.2d at 796.
2. Editorial choices.
Next, Judge Huckabee complains of HBO’s choice of material for the documentary. His principal complaint is that Women on Trial did not discuss much of the evidence presented at the 1988 Hebert hearing, including (1) Wayne’s initial treating physician’s testimony that Wayne had denied that his father caused the injury; (2) Child Protective Services case worker Wilma Smith’s testimony that in his videotaped interview Wayne said that his mother told him to say that his father had abused him (although Wayne still maintained that such abuse occurred); (3) Smith’s further testimony that in a subsequent interview with Wayne, he told her that his father had not abused him during the Christmas holidays, but that his father had touched his private area in July 1987; (4) Smith and social worker Cheryl Bennett’s testimony that Wayne and his brother John often fought after Wayne returned from Michael; (5) Bennett’s testimony that Sandra told her that she preferred that Michael not be allowed visitation rights and had inquired about what was necessary to terminate them; and (6) Wayne’s grandmother’s testimony that Michael had not bathed Wayne during their visit to her home. By failing to include this evidence, Judge Huckabee claims that HBO intentionally made it look like he was presented with an open-and-shut case against Michael Hebert, when in fact much of the evidence justified his order.
Further, Judge Huckabee complains about the film’s failure to clarify two facts: first, that his statements in the interview with Grant came in response to questions about a “hypothetical” case; and second, that Sandra did not move to modify the temporary order in the three years after the court of appeals denied her petition for mandamus. Because of all these omissions, Judge Huckabee claims that the viewers saw him falsely as a judge who flouted his legal duty to render decisions in the best interests of children; See Tex. Fam.Code § 153.002.
A broadcaster’s omission of facts may be actionable if it so distorts the viewers’ perception that they receive a substantially false impression of the event. See Golden Bear Distrib. Sys. v. Chase Revel, Inc., 708 F.2d 944, 949 (5th Cir.1983)(applying Texas law); Express Publishing Co. v. Gonzalez, 350 S.W.2d 589, 592 (Tex.Civ.App.—Eastland 1961, writ ref'd n.r.e.); see also Toney v. WCCO Television Midwest Cable & Satellite, Inc., 85 F.3d 383, 395 (8th Cir.1996)(Byron White, J.)(applying Minnesota law); Memphis Publishing Co. v. Nichols, 569 S.W.2d 412, 419-20 (Tenn.1978); Keeton et al., Prosser & Keeton: Law of Torts, § 116, at 117 (5th ed. Supp.1988). But see American Broad. Cos. v. Gill, 6 S.W.3d 19, 43 [426]*426(Tex.App.—San Antonio 1999, pet. denied); Evans v. Dolcefino, 986 S.W.2d 69, 77 (Tex.App.—Houston [1st Dist.] 1999, no pet.); Hardwick v. Houston Lighting & Power, 943 S.W.2d 183, 185 (Tex.App.—Houston [1st Dist.] 1997, no writ)(all holding that there is no claim for implied libel if all the facts in a story are literally true). As a public official, however, Judge Hucka-bee may recover for such an omission only by making the familiar showing that the publisher selected the material with actual malice, i.e., the awareness that the omission could create a substantially false impression. See Perez v. Scripps-Hoivard Broadcasting Co., 35 Ohio St.3d 215, 520 N.E.2d 198, 204 (1988); Dixon v. Ogden Newspapers, Inc., 187 W.Va. 120, 416 S.E.2d 237, 244 (1992); see also Brasslett v. Cota, 761 F.2d 827, 843 (1st Cir.1985); Pierce v. Capital Cities Communications, Inc., 427 F.Supp. 180, 186 (E.D.Pa.1977), aff'd, 576 F.2d 495 (3d Cir.1978); Diesen v. Hessburg, 455 N.W.2d 446, 453-54 (Minn.l990)(dieta). This standard does not, therefore, prevent liability if a media organization selectively omits facts from the record to portray falsely a judge’s opinion as arbitrary and unreasonable. Even if a defamation defendant is not persuaded by the evidence which supports a judge’s decision, he or she may not deliberately omit all reference to this evidence in order to portray the decision as arbitrary, when in fact it was not. But in the absence of evidence that the defendant selected the material to portray the judge’s record falsely, the First Amendment protects the organization’s choice of which material to include in its broadcast. See Miami Herald Publishing Co. v. Tornillo, 418 U.S. 241, 258, 94 S.Ct. 2831, 41 L.Ed.2d 730 (1974)(striking down a law requiring newspapers to give political candidates a right to reply to negative editorials); Levan v. Capital Cities/ABC, Inc., 190 F.3d 1230, 1243 (11th Cir.1999).
In this case, there is no evidence that HBO chose its material for the broadcast with actual malice. We recognize that an omission may be so glaring and may result in such a gross distortion that by itself it constitutes some evidence of actual malice. For example, when an article reported that an FBI memorandum mentioned plaintiff several times in connection with Jimmy Hoffa’s disappearance, the newspaper’s decision not to report that the memorandum also cleared plaintiff of wrongdoing was held to be evidence of actual malice. See Schiavone Constr. Co. v. Time, Inc., 847 F.2d 1069, 1092 (3d Cir.1988). In such a case, the omission so changes the character of the story that one could infer that the defendant knew, or at least suspected, that the omission would convey a false impression. Here, HBO’s omissions did not change the character of the story to such an extent. Although the facts omitted might or might not have led a reasonable viewer to suspend judgment or even to reach an opposite conclusion regarding Judge Huckabee’s order, their omission did not grossly distort the story. At most, HBO’s failure to capture accurately all the story’s details suggests an error in judgment, which is no evidence of actual malice. See Time, Inc. v. Pape, 401 U.S. 279, 290, 292, 91 S.Ct. 633, 28 L.Ed.2d 45 (1971)(magazine’s failure to convey all the subtleties of a long, complicated government document was no evidence of actual malice); El Paso Times, Inc. v. Trexler, 447 S.W.2d 403, 406 (Tex.1969). Moreover, the broadcasters did acknowledge Judge Huckabee’s explanation for his decision when they aired the portion of the interview responding to questions about the “hypothetical” Hebert case. Although the documentary did not convey Judge Huckabee’s position as strongly as it could have, the law did not require it to do so. See Levan, 190 F.3d at 1243; Brown v. Herald Co., 698 F.2d 949, 951 (8th Cir. 1983)(per curiam).
Judge Huckabee also complains about HBO’s editing choices in the Raseh-ke segment. There, Grant interviewed Steve Raschke, Ivy Raschke’s husband, about Dr. Charles Martin, the court-appointed psychologist in the Roberts/Rasch-[427]*427ke case. During the discussion, Steve claimed that Dr. Martin had himself been accused of child abuse.3 After this discussion, the documentary cut immediately to Judge Huckabee’s photo, which led into a segment of Grant’s interview with the judge. Judge Huckabee claims that this quick juxtaposition reflected HBO’s attempt to paint him falsely as a child abuser. Despite the sudden cut to Judge Huckabee’s photo, the documentary made it clear that the alleged abuser was Dr. Martin, not the judge. On these facts, there is no evidence of actual malice.
3. Harrison and Huckabee interviews.
Next, Judge Huckabee argues that after Grant and Cotts interviewed him and Dr. Harrison, they should have been put on notice that the documentary was false. That HBO nevertheless persisted in broadcasting it, they contend, is evidence of actual malice.
That Judge Huckabee offered an explanation for his decision, however, is not evidence that the filmmakers or HBO either believed it or had reason to doubt the truth of their broadcast. Denials by public figures to media charges are part and parcel of free discussion about public affairs. The mere fact that a defamation defendant knows that the public figure has denied harmful allegations or offered an alternative explanation of events is not evidence that the defendant doubted the allegations. As the United States Supreme Court has noted, “ ‘such denials are so commonplace in the world of polemical charge and countercharge that, in themselves, they hardly alert the conscientious reporter to the likelihood of error.’ ” Harte-Hanks, 491 U.S. at 692 n. 37, 109 S.Ct. 2678 (quoting Edwards v. National Audubon Soc’y, 656 F.2d 113, 121 (2d Cir. 1977)). Moreover, as we noted earlier, the filmmakers did not reject Judge Hucka-bee’s position entirely, but broadcast a portion of the interview in which he explained the reasons for denying parents access to their children.
Dr. Harrison’s opinion that Judge Huckabee’s order was justified also did not raise a fact issue as to actual' malice. Because Dr. Harrison was an expert in the field of child psychology, Judge Huckabee argues that his report should have given the filmmakers doubts about the film’s suggestion that Judge Huckabee’s decision to take Wayne away from Sandra was unjustified. But the mere fact that an expert has a view on a dispute is not evidence that a defamation defendant who offers a different view does so with actual malice, unless the record shows that the expert’s reasoning caused the defendant to experience substantial doubts regarding the story’s truthfulness. See Peter Scalamandre & Sons, Inc. v. Kaufman, 113 F.3d 556, 562 (5th Cir.1997)(fact that an expert holds a belief does not foreclose debate on that belief). Here, Cotts’s second affidavit and her notes after interviewing Dr. Harrison make it clear that she had credible reasons for rejecting Dr. Harrison’s view of the case. Nor did the filmmakers’ failure to discuss Dr. Harrison’s theory that Wayne’s brother was the real abuser amount to a false characterization of the evidence before Judge Hucka-bee. See supra Part 111(2). To the contrary, Dr. Harrison did not file the report containing this theory with the court until four months after the initial hearing.
4. Purposeful avoidance.
Next, Judge Huckabee contends that the filmmakers purposefully avoided discovering the truth about the Hebert case. Under Harte-Hanks, evidence [428]*428showing that HBO purposefully avoided the truth would be some evidence of actual malice. See 491 U.S. at 692, 109 S.Ct. 2678. In Harte-Hanks, a newspaper published a story claiming that Daniel Con-naughton, a candidate for municipal judge, had promised two sisters, Alice Thompson and Patsy Stephens, jobs and vacations in return for making allegations of corruption against the incumbent judge’s court administrator. Id. at 660, 109 S.Ct. 2678. The newspaper’s only source for this story was Thompson. Before the newspaper published the story, Connaughton produced five witnesses who were present when Thompson claimed that Connaugh-ton offered her and Stephens the gifts. All the witnesses denied Thompson’s story. Id. at 691, 109 S.Ct. 2678. Connaughton also produced a tape recording of the conversation in which Thompson accused the administrator of corruption. Id. at 683, 109 S.Ct. 2678. The newspaper failed to listen to this recording even though it would have confirmed or denied many of Thompson’s claims, such as her claims that Connaughton had selectively turned the reeoi’der on and off during various parts of the interview and that her allegations of corruption against the court administrator had come in response to leading questions from him. Id. More importantly, the newspaper failed to interview Stephens, the one person not associated with Con-naughton who could have confirmed or denied Thompson’s allegations against Connaughton. Id. at 691-92, 109 S.Ct. 2678. According to the Court, the newspaper’s failure to consult the two sources that could have objectively verified the story was evidence that the newspaper purposefully avoided learning facts that would have shown the story to be false. Id. at 692, 109 S.Ct. 2678. Upholding a jury verdict against the newspaper, the Court held that this purposeful avoidance of the truth was enough to suggest that the newspaper doubted the story’s accuracy, and hence was evidence of actual malice. Id.
Judge Huckabee has not presented a purposeful avoidance ease. Unlike Harte-Hanks, in which the newspaper based its story on the testimony of a single unreliable source, here the summary judgment evidence reveals that the filmmakers interviewed several people on both sides of the story, including Judge Huckabee and Dr. Harrison. They also read, among other documents, the transcript of the Hebert hearing. Such extensive research precludes a finding of purposeful avoidance. See Levan, 190 F.3d at 1243; Perk v. Reader’s Digest Ass’n, 931 F.2d 408, 411— 12 (6th Cir.1991)(both distinguishing Harte-Hanks on the ground that the stories at issue were supported by many sources). Although the filmmakers did not interview Michael Hebert, Robert Roberts, or their lawyers, they were not required to continue their research until they could find one more person who agreed with Judge Huckabee’s Hebert order.4 See Levan, 190 F.3d at 1243 (failure to track down every possible source is not purposeful avoidance). Further, unlike Harte-Hanks, no source could have easily proved or disproved the documentary’s allegations. Thus, the purposeful avoidance theory does not apply. See Levan, 190 F.3d at 1243; Perk, 931 F.2d at 412.
5. Legal Review, Rewrites, and Indemnification
Next, we turn to the evidence that Judge Huckabee believes established “institutional doubt” on the part of HBO regarding the truth of Women on Trial. According to Judge Huckabee, HBO’s extensive legal review of the film, the editorial rewrites that accompanied this review, and the indemnification agreement between HBO and JFP all suggest that HBO entertained serious doubts about the film’s content. We disagree.
[429]*429That the film underwent a lengthy legal review does not by itself provide evidence of actual malice. HBO could have wished merely to confirm the film’s controversial and potentially damaging allegations before its release. See McFarlane v. Sheridan Square Press, Inc., 91 F.3d 1501, 1512 (D.C.Cir.1996)(lawyers’ review was not evidence of actual malice). This same conclusion also applies to the indemnification agreement. Judge Hucka-bee can point to no evidence that JFP and HBO entered into the agreement because they entertained serious doubts about the film’s truthfulness. Individuals and business organizations enter into indemnification agreements for various reasons; doing so, without more, simply presents no evidence of actual malice.
6. Knowing inaccuracies.
Finally, Judge Huckabee claims that inaccuracies in Women on Trial present evidence of actual malice. First, Judge Huckabee points to this language in the film’s conclusion: “If these rulings can happen in one family courthouse in one county of one great state, what is happening in the rest of this country?” In fact, one of these cases occurred in Bee County, not Harris County. As proof that HBO knew this statement was false, Judge Huckabee points to Cis Wilson’s notes on a memo discussing HBO’s promotional strategy for the film, which stated that the rulings in the film had occurred in a “single family court.” On her copy, Wilson circled the “single family court” statement and wrote the words “Daggatt” [sic] and “Huckaby” [sic]. Wilson stated in her deposition that these notations indicated her knowledge that the memo’s “single family court” statement could be false.
Judge Huckabee claims that Wilson’s knowledge that the memo’s “single family court” statement could be false implies that she knew that the documentary’s “one family courthouse” statement could be false.5 Regardless of the falsity of the statement, however, it was not defamatory to Judge Huckabee. The documentary neither stated nor implied that Judge Huckabee presided over all of these cases. In fact, the film as a whole made it clear that different judges were responsible for the rulings portrayed therein. See Goodrich v. Reporter Publishing Co., 199 5.W.2d 228, 229 (Tex.Civ.App.—El Paso 1946, writ ref'd)(to determine whether a statement is defamatory, one must consider the publication as a whole). The statement was thus a criticism of the family courts in general and not of Judge Hucka-bee in particular and, as a result, was not defamatory. See Newspapers, Inc. v. Matthews, 161 Tex. 284, 339 S.W.2d 890, 893 (1960)(to be defamatory, a statement must be directed at the plaintiff). Because the statement was not defamatory, Wilson’s alleged knowledge of its falsity was irrelevant. See Tavoulareas, 817 F.2d at 794 (actual malice must be in conjunction with a defamatory statement).
A second knowing inaccuracy, Judge Huckabee claims, is the film’s statement that, in the Houston family courts, “[w]omen who charge their husbands with abuse are often viewed as mentally unstable and routinely lose custody of their children.” Judge Huckabee charges that the filmmakers knew that this statement was false because he had told them that in only four cases had he entered an order denying all access to one parent, only two of which were against the mothers. Despite Judge Huckabee’s protests to the contrary, media accounts on which the filmmakers had relied reported that such denials of custody were routine.6 The filmmakers’ interviews [430]*430with Burton, Reynolds, and other advocates of family-court reform confirmed these accounts. The filmmakers reasonably could have concluded, therefore, that such denials of custody occurred routinely. As we have already noted, Judge Hucka-bee’s denial of this allegation is no evidence that the filmmakers experienced substantial doubts about the film’s truth. Because Judge Huckabee did not offer any other evidence that the filmmakers seriously doubted this allegation, he failed to raise a fact issue on actual malice.7
[[Image here]]
Because HBO’s affidavits negated actual malice as a matter of law, and because none of Judge Huckabee’s proffered evidence raised a fact issue regarding actual malice, we affirm the judgment of the court of appeals.
Justice HECHT filed a dissenting opinion.
Justice OWEN did not participate in the decision.