SCHWELB, Associate Judge:

This appeal arises from an action for libel. The plaintiffs are Guilford Transportation Industries, Inc. (Guilford), Timothy Mellon, and David Fink. Guilford operates a freight railroad system in New England, consisting of the Maine Central, the Boston and Maine, and Springfield Terminal railroads. Mellon and Fink founded Guilford in 1981 and own it. Mellon is a member of Guilford's Board of Directors, and Fink manages Guilford's operations. The sole defendant is Frank Wilner, the author of an "Op-Ed" column which appeared on June 2, 1997 in the Journal of Commerce, a trade publication said to be widely read by persons in the railroad industry and by regulatory officials. In this column, a copy of which is appended to this opinion, Wilner discussed an attempt by Guilford to purchase or lease from Amtrak certain rail operations in the northeastern United States.

In their complaint, which was filed on July 8, 1997, the plaintiffs allege that Wilner's column is defamatory and contains demonstrably false statements of fact. Wilner responds that the representations of which the plaintiffs complain are all either true, or incapable of bearing a defamatory meaning, or statements of opinion protected by the First Amendment. In a comprehensive oral ruling in which she focused on the potential of this type of suit for chilling constitutionally protected speech, Judge Ellen S. Huvelle granted summary judgment in favor of Wilner. The plaintiffs now appeal.

"At the heart of the First Amendment is the recognition of the fundamental importance of the free flow of ideas and opinions on matters of public interest and concern." Hustler Magazine, Inc. v. Falwell, 485 U.S. 46, 50 (1988). If the First Amendment's guarantees of freedom of speech and of the press are to ensure that these rights are meaningful not simply on paper, but also in the practical context of their exercise, then a newspaper Op-Ed column discussing a subject of public interest must surely be accorded a high level of protection, lest the expression of critical opinions be chilled. This is so because

[t]he reasonable reader who peruses [a] column on the editorial or Op-Ed page is fully aware that the statements found there are not "hard" news like those printed on the front page or elsewhere in the news sections of the newspaper. Readers expect that columnists will make strong statements, sometimes phrased in a polemical manner that would hardly be considered balanced or fair elsewhere in the newspaper. . . . That proposition is inherent in the very notion of an "Op-Ed page." Because of obvious space limitations, it is also manifest that columnists or commentators will express themselves in condensed fashion without providing what might be considered the full picture. Columnists are, after all, writing a column, not a full-length scholarly article or a book. This broad understanding of the traditional function of a column like Evans and Novak will therefore predispose the average reader to regard what is found there to be opinion.

Ollman v. Evans, 242 U.S. App. D.C. 301, 317, 750 F.2d 970, 986 (1984) (en banc) (plurality opinion), cert. denied, 471 U.S. 1127 (1985) (citation and footnote omitted) (discussing a column in the Washington Post by Rowland Evans and Robert Novak).

Although the Supreme Court has made it clear, since Ollman was decided, that statements of "opinion" are not constitutionally protected if they assert provably false and defamatory facts, see Milkovich v. Lorain Journal Co., 497 U.S. 1 (1990), the constitutional principles that animate the passage we have quoted from Ollman remain equally compelling and equally good law today. See, e.g., Moldea v. New York Times Co., 306 U.S. App. D.C. 1, 2, 22 F.3d 310, 311, cert. denied, 513 U.S. 875 (1994) ( Moldea II) (modifying Moldea v. New York Times Co., 304 U.S. App. D.C. 406, 15 F.3d 1137 (1994) ( Moldea I)); Partington v. Bugliosi, 56 F.3d 1147, 1154 (9th Cir. 1995). "Unless persons, including newspapers, desiring to exercise their First Amendment rights are assured freedom from the harassment of lawsuits, they will tend to become self-censors." Washington Post Co. v. Keogh, 125 U.S. App. D.C. 32, 35, 365 F.2d 965, 968 (1966), cert. denied, 385 U.S. 1011 (1967). "[S]elf censorship affecting the whole public is `hardly less virulent for being privately administered.'" Id. (quoting Smith v. California, 361 U.S. 147, 154 (1959)). Moreover, in the first instance, "it is the court, not the jury, that must vigilantly stand guard against even slight encroachments on the fundamental constitutional right of all citizens to speak out on public issues without fear of reprisal." Myers v. Plan Takoma, Inc., 472 A.2d 44, 50 (D.C. 1983) (per curiam) (adopting Superior Court opinion of Weisberg, J.).

This is not an easy case, for the plaintiffs have submitted affidavits alleging that Wilner's article contains provably false statements of fact and that Wilner has admitted that he knew these statements to be false. After carefully examining the article in light of all of the plaintiffs' allegations, however, we conclude that the foregoing authorities are dispositive of this appeal, and that imposition of liability upon Wilner would tend to chill the robust debate and discussion which form an essential part of our constitutional tradition. Accordingly, we affirm the judgment in Wilner's favor.

I. THE FACTS

A. " Guilford's Tempestuous Past."

This controversy began in May of 1997, when Guilford announced its intention to acquire Amtrak's train operations in the northeastern United States. The prospect of a privatized Amtrak sparked debate throughout the transportation industry. On June 2, 1997, Wilner, an economist who had frequently written on transportation issues, entered the fray with the publication of the column, entitled Guilford's Tempestuous Past, which precipitated this litigation. At that time, Wilner was the Chief of Staff to one of the Commissioners of the Surface Transportation Board, which had replaced the Interstate Commerce Commission as the federal agency that regulates railroads and enforces the railway labor laws.

In "Guilford's Tempestuous Past," Wilner began with the rhetorical inquiry "Who is Guilford?" His response was that Guilford was a company created in 1981 when Timothy Mellon, the "38-year-old scion of the Mellon banking fortune . . . tired of a rail-tie manufacturing operation and coveted three of the nation's most ineligible railroads — bankrupt Boston Maine, comatose Delaware Hudson and marginally profitable Maine Central." According to Wilner, Mellon and Fink, "a former Pennsylvania Railroad operating officer and fourth-generation railroader," purchased the three railroads and unified them into the "4,000-mile rail system" that became Guilford. Mellon was described in the column as a recluse "with a master's degree in city planning from Yale and an interest in Indian rights."

Wilner then suggested that Guilford itself was a business venture which could only succeed by extracting concessions from labor:

Crucial to making Guilford profitable was cooperation from its 12 disparate labor unions. Quickly, a bitter labor-management conflict was ignited when Guilford bolted from traditional national wage and benefits negotiations. Local negotiations with the Brotherhood of Maintenance of Way Employees culminated in an almost three-month strike that required Congressional intervention.

Guilford tried another tack, leasing its component railroads — BM, DH, and Maine Central — to a tiny and specialized BM subsidiary, Springfield Terminal, which enjoyed a concessionary labor contract with the United Transportation Union [UTU].

Before the lease, Springfield Terminal employed just 53 workers. Suddenly it became the de facto operator of the entire Guilford system, employing 2,000 people — and Guilford asserted that all its employees, represented by a dozen different unions, had become covered by Springfield Terminal's UTU contract providing for lower wages and greater work-rules flexibility.

The Interstate Commerce Commission — with a zealous pro-management bias in those days — sanctioned the lease agreement and referred certain labor issues to binding arbitration. When a neutral arbitrator ruled in favor of employees — concluding that existing collective bargaining agreements could no more be scrapped than existing contracts for locomotive fuel — the ICC partially overturned the award.

Wilner also described certain litigation that arose following the lease to Springfield Terminal. He wrote that the "employees demanded new union elections that resulted in a scrapping of [UTU's] single-representation contract with Springfield Terminal in favor of nine separate craft unions." This, in turn, led to more problems for Guilford:

As Guilford — through Springfield Terminal — began negotiating new agreements, work stoppages followed, as did a bankruptcy filing by DH (which now is owned and operated by Canadian National), appeals to the federal courts of previous ICC decisions and even an allegation by Guilford that arbitrators had been unduly influenced by the New England congressional delegation.

By the early 1990's the cats were herded, but the anti-establishment Mr. Mellon and Mr. Fink already made headlines with equally chaotic legal fisticuffs with Amtrak.

By law, freight railroads are required to host Amtrak trains and to maintain track on those routes to passenger-train specifications. Amtrak pays the cost. But in 1987 Amtrak asserted it was forced to discontinue a New York-Montreal run because of poor maintenance over [a] Boston Maine track that had slowed passenger-train speeds to 10 miles per hour.

Wilner next related how Amtrak ultimately seized the track by using "an extraordinary power granted it by Congress to condemn private property." After summarizing the rather confusing litigation which ensued, Wilner's article concluded:

Guilford has a knack for grabbing headlines. With its latest offer to buy or lease Amtrak's Northeast line, it is bound to keep the story alive.

B. The plaintiffs' allegations of defamation.

The plaintiffs claim that Wilner's column contains numerous defamatory statements, both express and implied. The principal focus of their case is on the allegation that the column falsely portrays them as hostile and antagonistic to labor. The plaintiffs do not allege that Wilner has stated this directly, but they argue instead that the idea is implied by the column's tone and choice of words:

By juxtaposing `quickly,' `bitter conflict' `ignited' and `traditional national negotiations,' the Article falsely indicates that Guilford caused [`ignited'] a `bitter conflict' and `strike' by untowardly and precipitously [`bolting'] from standard [`traditional'] negotiations to further its `crucial' need to extract `cooperation' from its `12 disparate labor unions.'

Elaborating on the theme that Wilner has described them as anti-labor, the plaintiffs claim that they have been implicitly accused of violating the Railway Labor Act (RLA), 45 U.S.C. § 151 et seq., by inciting labor unrest, by causing strikes and work stoppages, and by failing to maintain their tracks properly. They point out that management is under a continuous obligation to make every reasonable effort to avoid any action that might interrupt rail service or impede the free flow of commerce, see 45 U.S.C. § 152 (First), and they assert that the column effectively accuses them of having failed to carry out this obligation. In support of these allegations, the plaintiffs filed, inter alia, a sworn declaration by Herbert R. Northrup, Professor Emeritus of Management at the Wharton School of the University of Pennsylvania, whom the plaintiffs have presented as an expert on labor relations and the RLA. In his declaration, Northrup stated, inter alia, that

• any reader familiar with the railroad industry or the RLA would conclude from reading Wilner's Journal of Commerce, June 2, 1997, article that the plaintiffs are anti-union and anti-labor and have violated their duties and obligations under the RLA, by failing to exert every effort to make and maintain labor agreements and to avoid interruptions in service as a result of a labor dispute.

• in fact there is no evidence that plaintiffs have failed in their obligations under the RLA, or have exhibited anti-labor biases or ignited labor strikes and work stoppages.

• the article would reasonably tend to injure plaintiffs' reputation within the railroad industry and cause customers to shift their business from Guilford trains to trucks which compete with Guilford for much of its business.

Professor Northrup also synopsized what the plaintiffs view as provably false statements of fact in Wilner's column. According to Professor Northrup,

• any expert who has studied the RLA or who was familiar with the history of the railroad industry or who followed Guilford's history would know the following statements, among others, in the Article are false:

• the strike which occurred in 1986 on the Maine Central did not "quickly" occur after Guilford acquired the Maine Central. Negotiations had been ongoing for years before the strike occurred in 1986. Also, the 1986 strike did not arise from Guilford's attempts to obtain concessions from labor;

• Guilford did not "bolt" from national handling. A railroad must affirmatively opt into national handling and Guilford did not opt into national handling with respect to the labor negotiations on the Maine Central from 1982 through 1986;

• The Springfield Terminal leases had a transportation purpose and the ICC so found. Also, protective conditions ensured that employees of the leased railroads who went to work for Springfield Terminal would not suffer a reduction in compensation or a loss of seniority;

• the union representative elections at the Springfield Terminal in late 1991 and early 1992 did not result in "scrapping" existing labor agreements; the existing labor agreement with the UTU remained in effect for its full six year term;

• negotiations with the newly elected unions at Springfield Terminal after 1991 did not result in work stoppages; agreements were successfully completed without incident;

• the bankruptcy of the Delaware and Hudson did not occur after the union representation elections at Springfield Terminal.

• The Article misstates "Who is Guilford," which is expressly the question the Article promises to answer.

The plaintiffs further allege that Wilner has defamed Mellon and Fink by describing them as eccentric and strange. The plaintiffs insist that Mr. Mellon is not "reclusive," that neither he nor Mr. Fink has "coveted" any railroads, and that Wilner's assertions to the contrary are verifiably false. The plaintiffs also argue that they are depicted in the column as ineffective businessmen, and that the reference to Mr. Mellon as a "recluse" who "tired of his rail tie manufacturing business," and who engaged in "chaotic legal fisticuffs," conveys to the reader the defamatory implication that Mellon is not a man of good character, that he is unstable, and that he has an irresponsible attitude towards proper business practices and rail safety.

In opposition to Wilner's motion for summary judgment, the plaintiffs also submitted evidence designed to show knowledge on Wilner's part that his column contained false factual allegations. F. Colin Pease, Guilford's executive vice-president, related in an affidavit that on the day following the publication of the column, he and Fink telephoned Wilner to complain that Wilner had done a "hatchet job" on the plaintiffs. According to Pease, Wilner "justified the false statements by telling us that he had been given 12 hours to submit the article." Pease stated that Wilner refused to identify the person or persons who gave Wilner the deadline. Pease claimed that during a second telephone conversation, "[t]he defendant later offered to make it up to Guilford by writing a column for another publication, Railway Age, which would be favorable to Guilford." Eight months later, in a sworn deposition, Pease went even further, asserting that "I called Frank Wilner back and when he answered the phone, Frank agreed that the items in the article were false." (Emphasis added.)

Finally, according to the plaintiffs, Wilner's column did more than bruise corporate egos. Customers who ship their goods by rail are understandably sensitive about potential labor disputes that could result in an interruption of rail service. The plaintiffs claim that after the article was published, a number of Guilford clients called to voice concern over the company's labor practices. One such customer, Joseph Kittredge, stated in a sworn declaration that "[a]ny firm which has an anti-labor or anti-union reputation presents a real and unacceptable risk of interruption in service," and that after reading Wilner's column, he contacted Fink and warned him that, in light of Wilner's allegations, shippers were likely to reallocate their business away from Guilford to other forms of transportation. Elaborating on Mr. Kittredge's account, Fink stated in a sworn declaration that in 1997, following the publication of Wilner's column, Guilford suffered a substantial decline in its business. According to Fink, the column was "the only material adverse event during this time and . . . the only reasonable explanation for this drastic drop-off of performance." On the basis of this type of evidence, the plaintiffs assert that Wilner's column has ascribed to them conduct and characteristics that would tend to injure them "in [their] trade, profession or community standing," Howard Univ. v. Best, 484 A.2d 958, 988 (D.C. 1984), or "adversely affect [their] fitness for the proper conduct of [their] lawful business." Wallace v. Skadden, Arps, Slate, Meagher and Flom, 715 A.2d 873, 877 (D.C. 1998).

In support of his motion for summary judgment, Wilner submitted extensive materials which contested most, if not all, of the plaintiffs' allegations. Relying heavily on the "wire service defense," see Winn v. United Press Int'l, 938 F. Supp. 39, 44 n. 6 (D.D.C. 1996), aff'd mem., 1997 U.S. App. Lexis 19025 (D.C. Cir. July 14, 1997), Wilner produced numerous prior articles and documents in which Guilford had been described as hostile to labor, and which, according to Wilner, provided substantiation for all of the claims in the column.

C. Judge Canan's ruling.

After the plaintiffs instituted their lawsuit, Wilner filed a motion to dismiss the complaint for failure to state a claim upon which relief may be granted or, in the alternative, for summary judgment. On December 19, 1997, Honorable Russell F. Canan, the judge to whom the case was initially assigned, denied outright the motion to dismiss the complaint, and he denied the motion for summary judgment without prejudice to refiling upon the completion of discovery. The judge recognized that the First Amendment concerns in the case were "substantial," but he concluded that, at least "at this early stage of the litigation," the defendant had not shown that the column was incapable of bearing a defamatory meaning. The judge stated:

[R]ead in context, including the defendant's apparent prominence in the field and his official position at the time, this [c]ourt believes . . . that the . . . Article can be read regarding the hostility to labor and that plaintiffs have acted irresponsibly and antagonistically regarding labor contrary to the public interest in public safety.

And, that these statements in the Article would have the tendency, which is part of the standard at this stage, to injure plaintiffs in their trade, profession, or community standing. And, to lower them in the estimation of the community in which they sit, the railway transportation community.

D. Judge Huvelle's ruling.

Judge Canan's ruling having kept the case alive, the parties conducted discovery, and Wilner ultimately filed a second motion for summary judgment. On January 15, 1999, following extensive briefing and argument, Judge Huvelle delivered a comprehensive oral ruling in which she granted Wilner's renewed motion. Because we agree substantially with significant parts of Judge Huvelle's decision and place some reliance on her analysis in our disposition of the appeal, we set forth her oral opinion in some detail.

(1) First Amendment implications of the suit.

After briefly addressing a number of other matters, Judge Huvelle focused upon the First Amendment aspects of the case. She emphasized, citing the Keogh decision, supra, 125 U.S. App. D.C. at 35, 365 F.2d at 968, that "[t]he threat of being put to the defense of a lawsuit . . . may be as chilling to the exercise of First Amendment freedoms as fear of the outcome of the lawsuit itself." Agreeing with Judge Oberdorfer's comments in Novecon, Ltd. v. Bulgarian-American Enter. Fund, 977 F. Supp. 45, 47 (D.D.C. 1997), Judge Huvelle stated that "summary procedures are particularly important to guarantee robust discussion of [matters of] public controversy and to protect media defendants from the threat of chilling freedom of the press and speech." Citing this court's decision in Myers, supra, 472 A.2d at 50, the judge emphasized that in the first instance it was the court, and not the jury, that must be vigilant against encroachments upon the free and unfettered exercise of First Amendment rights.

After outlining the elements of an action for defamation, Judge Huvelle explained that in order to constitute a libel, a statement must be capable of bearing a defamatory meaning and must be provably false. The judge characterized the question whether the requisite showing on these issues had been made as a "pure question of law," and she "[did] not think that [she] could decide these questions by reference to the affidavit[s] of Mr. Northrup or Mr. Kittredge." She noted, however, that the court must examine a challenged writing in the "entire context in which it was published." The judge stated:

The context here I think is somewhat clear and undisputed. [ Moldea II) stresses, even after Milkovich, the importance of context, because it is in part the setting of the speech in question that makes [its] hyperbolic nature apparent and which helps determine the way in which the intended audience will receive [it].

Here we know that this article was prepared in response to the announcement of a private company, Guilford, of [its] plan to buy Amtrak's northeast corridor. The article appeared in a trade journal on an opinion page. It relates undisputably to a matter of public concern.

Acknowledging the teaching of Milkovich, supra, 497 U.S. at 18-19, the judge stated that there is "no wholesale definiti[ve] exemption for anything that might be labeled opinion, because to have such an exemption for opinion would ignore the fact that expressions of opinion may imply an assertion [of verifiably false] objective fact." Relying on Washington v. Smith, 317 U.S. App. D.C. 79, 80, 80 F.3d 555, 556 (1996), however, the judge stated that "a statement of opinion is actionable only if it has an explicit or implicit factual foundation and is therefore objectively verifiable." Paraphrasing Milkovich, supra, 497 U.S. at 20, the judge opined that "the First Amendment protects statements of imaginative expression and rhetorical hyperbole in order to assure that the public debate will not suffer for lack of these statements, which have traditionally added much to the discourse of our nation." The judge relied on the Supreme Court's holding in National Ass'n of Letter Carriers v. Austin, 418 U.S. 264, 282-87 (1974), that epithets such as "scab" and "traitor," when used in a union publication, were not actionable. She also cited Greenbelt Coop. Publ'g Ass'n, Inc. v. Bresler, 398 U.S. 6, 11-15 (1970), in which the Court held that the use of the word "blackmail" was not libelous when the context made it clear that the accusation was hyperbolic and that the plaintiff was not actually being accused of a crime. Concluding her summary of the applicable legal principles, the judge stated:

We also know from Moldea II . . . that where challenged material addresses a public controversy such as [the] one that we have here, the First Amendment provides breathing space for journalists to criticize and interpret the actions and decisions of those involved. Where a journalist is interpreting facts and information, he cannot be held liable for defamation unless no reasonable person could find that the interpretation was supported by the underlying facts.

(2) The claim that Wilner accused the plaintiffs of general hostility to labor.

Judge Huvelle then turned to what she called "the gist of the action," namely, the allegation that Wilner had characterized Guilford as "unfair, hostile, [and] antagonistic to labor." Although Wilner did not explicitly describe the plaintiffs in this manner, the judge assumed, for purposes of her decision, that the language of the column "could be reasonably capable of this defamatory interpretation." The judge concluded, however, that even if the allegation of anti-labor bias was fairly implied in the column, that charge "is not provably false. It is much more of a subjective assessment of a company's attitude and conduct towards labor." Relying on Emde v. San Joaquin County Cent. Labor Council, 143 P.2d 20, 23 (Cal. 1943), Judge Huvelle stated that "the preponderance of judicial decision[s] supports the rule that the use of the word `unfair' in connection with labor controversies does not impute want of moral integrity or business capacity but merely is a characterization of an employer who refuses to conduct his business in the manner desired by the union." The judge cited a number of additional authorities to the same general effect, including Montgomery Ward Co. v. McGraw-Hill Pub. Co., 146 F.2d 171 (7th Cir. 1944), and Gregory v. McDonnell Douglas Corp., 552 P.2d 425, 428 (Cal. 1976).

The plaintiffs argued that, with the exception of the Montgomery Ward case, the cited authorities were distinguishable upon the common ground that they all involved litigation between parties to a labor dispute. Relying on Myers and Gregory, Judge Huvelle emphatically rejected this contention:

. . . [T]his is no reason not to apply these cases, even though this is not a per se labor dispute, it is an issue of public dispute or controversy, and the holdings of these cases have just as much applicability to a commentator in the labor field as to a union involved in a labor dispute.

[It] seems to this [c]ourt that the author must be given the same freedom to express an opinion as a labor union, and that this is the heart of providing First Amendment protection to a person, to promote — to enable him to promote public debate. Mere criticism does not constitute defamation, obviously. The First Amendment was designed to protect articles such as this one, that serve to promote debate in a free society.

(3) The claim that Wilner accused the plaintiffs of violating the RLA.

Judge Huvelle also addressed the plaintiffs' related claim that "the article is defamatory insofar as it charges them with failure to comply with the [RLA and] also with acting inconsistently with standard and sound labor management relations, not being a reliable freight railroad." The judge observed that "this is not what the article says," that "there is no mention of the [RLA]," and that "there's no discussion of what would constitute standard and sound labor management relations." According to the judge, "[t]he kinds of implications that [the plaintiffs argue] here . . . would amount to enlarging the sense of the published words." Perceiving "no basis upon which to find [that] defendant intended or endorsed the defamatory inference suggested by the plaintiff[s]," the judge stated, citing Saenz v. Playboy Enters., Inc., 841 F.2d 1309, 1318 (7th Cir. 1988), that "one cannot require a publisher to guarantee the truth of all the inferences a reader might reasonably draw from a publication, for this would . . . undermine the uninhibited, open discussion of matters of public concern."

The judge also expressed the view that even if there was some suggestion in the column that Guilford was an unreliable freight railroad and did not follow sound labor policies and practices, Wilner still would not be liable for defamation. As the judge saw it, the reader would understand that these perceived suggestions "represent the writer's interpretation of the facts presented, because the reader is free to draw his or her own conclusions based on these facts." The judge, citing Moldea II, supra, 306 U.S. App. D.C. at 8, 22 F.3d at 313, stated that, under these circumstances, the statements were not actionable in defamation. The judge described the uncontested facts:

Here, it's undisputed and the article reveals that this was a strike by the Brotherhood of Maintenance of Way Employees and a work stoppage by the employees at the Springfield Terminal. We also know that Guilford leased the operating rights of the BM and MEC to the Springfield Terminal which caused all employees to be covered by a union agreement which the UTU had negotiated with Springfield Terminal.

We know that arbitrations with labor occurred and Amtrak did condemn a section of the track owned by the BM. These events are not disputed. So given that these facts are true and they were presented in the context of the column, a reader would understand that such supported opinion represented Mr. Wilner's interpretation of the facts presented. Since the reader is free to draw his or her own conclusions whether plaintiffs acted inconsistently with sound and standard labor relations, this implication is not actionable in defamation.

As required by [ Washington v. Smith, supra, 317 U.S. App. D.C. at 81, 80 F.3d at 557], in cases involving opinion on matters of public concern, the writer is immune from defamation action unless plaintiff can show that the statement is so obviously false that no reasonable person could find that the review's characterizations were supportable interpretations of the underlying facts.

Even assuming, as plaintiff argues that there [are] facts that are inaccurately stated here, the gist of the facts mean that they cannot sustain their burden under [ Moldea II].

(4) Claims that Mellon and Fink were personally defamed.

The judge also addressed the individual plaintiffs' claim that "it is defamatory to say [that] Mr. Mellon and Fink are eccentric and strange." The judge pointed out that

the article does not say this. It calls them reclusive and antiestablishment. In this [c]ourt's view they are not synonymous. . . .

The [c]ourt also finds that it is not defamatory to call someone . . . reclusive, strange or eccentric or antiestablishment. Allegedly defamatory remarks must be more than unpleasant or offensive. The language must make the plaintiff appear odious, infamous or ridiculous.

The judge added that "what we have here is interpretive expression or subjective impression; [i]t's a characterization of someone's personality." Further, quoting from Liberty Lobby, Inc. v. Dow Jones Co., 267 U.S. App. D.C. 337, 350, 838 F.2d 1287, 1300, cert. denied, 488 U.S. 825 (1988), she stated that "[e]valuative statements of taste and belief can never provide the basis for a defamation suit because such statements are incapable of being proved false."

II. LEGAL DISCUSSION

A. Standard of Review.

"[W]hether summary judgment was properly granted is [a question] of law, and we review de novo a decision granting such relief." Abdullah v. Roach, 668 A.2d 801, 804 (D.C. 1995). Our en banc court has stated:

In order to be entitled to summary judgment [the moving party] must demonstrate that there is no genuine issue of material fact and that [it is] entitled to judgment as a matter of law. Super. Ct. Civ.R. 56(c); Clyburn v. 1411 K Street Ltd. Partnership, 628 A.2d 1015, 1017 (D.C. 1993). The record is viewed in the light most favorable to the party opposing the motion. Graff v. Malawer, 592 A.2d 1038, 1040 (D.C. 1991). On appeal, we must assess the record independently, but the substantive standard applied is the same as that utilized by the trial court. Northbrook Ins. Co. v. United Servs. Auto Ass'n, 626 A.2d 915, 917 (D.C. 1993).

Colbert v. Georgetown Univ., 641 A.2d 469, 472 (D.C. 1994) (en banc). On summary judgment, "[t]he evidence of the non-movant is to be believed, and all justifiable inferences are to be drawn in his favor." Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255 (1986). Indeed, while the moving party's papers must be "closely scrutinized," the opponent's materials are to be treated "indulgently." Fry v. Diamond Constr., Inc., 659 A.2d 241, 246 (D.C. 1995).

But this case implicates serious First Amendment concerns, and to prolong this litigation may inhibit protected speech. We must therefore remain mindful of our obligation as an appellate court, in a case raising First Amendment issues, "to make an independent examination of the whole record in order to make sure that the judgment does not constitute a forbidden intrusion on the field of free expression." Milkovich, supra, 497 U.S. at 17 (quoting Bose Corp. v. Consumers Union of United States, Inc., 466 U.S. 485, 499 (1984)) (internal quotation marks omitted). "The threat of prolonged and expensive litigation has a real potential for chilling journalistic criticism and comment on public figures and public affairs." Myers, supra, 472 A.2d at 50. "Because of the compelling First Amendment interest at stake, we regard summary judgment as a useful method of disposing of constitutional libel actions where appropriate." Nader v. Toledano, 408 A.2d 31, 44 (D.C. 1979), cert. denied, 444 U.S. 1078 (1980).

In Keogh, supra, 125 U.S. App. D.C. at 35, 365 F.2d at 968, the court, after recognizing the importance of summary judgment as a means of avoiding long and expensive litigation and of preventing the harassment and coercion of non-affluent parties, stated that

[i]n the First Amendment area, summary procedures are even more essential. For the stake here, if harassment succeeds, is free debate. One of the purposes of the [ New York Times v. Sullivan, 376 U.S. 254 (1964)] principle, in addition to protecting persons from being cast in damages in libel suits filed by public officials, is to prevent persons from being discouraged in the full and free exercise of their First Amendment rights with respect to the conduct of their government. The threat of being put to the defense of a lawsuit brought by a popular public official may be as chilling to the exercise of First Amendment freedoms as fear of the outcome of the lawsuit itself, especially to advocates of unpopular causes.

Our obligation to protect the right of the citizen to free expression requires us to include these considerations in the summary judgment calculus.

B. Law of the case.

The plaintiffs claim that Judge Huvelle violated the principle of "the law of the case" by declining to follow Judge Canan's earlier ruling denying Wilner's motion for summary judgment. Inviting us to examine a tabular comparison, set forth in the appendix to their brief, of Wilner's first and second motions, they argue that the two submissions are substantially identical, that Judge Huvelle did not consider any materials not previously presented to Judge Canan, that the decisions of the two judges are irreconcilable, and that Judge Huvelle erred by declining to follow Judge Canan's ruling. Wilner responds that Judge Canan explicitly made the denial of Wilner's first request for summary judgment without prejudice, that discovery was conducted between the two motions, and that the law of the case doctrine therefore is not implicated.

In our view, Judge Canan's use of the phrase "without prejudice," while obviously significant, is not necessarily dispositive of the plaintiffs' "law of the case" claim. If the two judges considered the same facts and legal principles and reached diametrically opposite conclusions, then it would surely exalt form over substance to attach decisive significance to the two words in question. A reasonable construction of Judge Canan's intent in making his ruling "without prejudice" is that the judge did not mean to foreclose a different result if relevant new facts were established during discovery, or if the law changed. His order should not be read as meaning "without prejudice to another trial judge's overruling me if he or she believes that I was wrong."

In this case, the record relied on by Judge Huvelle differed at least slightly from the record available to Judge Canan, for Judge Huvelle made an explicit reference to Wilner's deposition in one part of her oral decision. The judge's citation to Wilner's testimony did not, however, appear to be crucial to her disposition. Indeed, in granting summary judgment, Judge Huvelle focused almost exclusively on the text of Wilner's column, to the exclusion of other materials. But even assuming, without deciding, that the two motions are sufficiently similar to one another to implicate law-of-the-case considerations, reversal is not warranted where, as here, the appellate court agrees on the merits with the second judge's analysis.

As Justice Holmes wrote for the Court in Messinger v. Anderson, 225 U.S. 436, 444 (1912), the law of the case doctrine "merely expresses the practice of courts generally to refuse to reopen what has been decided, not a limit to their power." The courts have thus held that "denial of a motion for summary judgment by one judge does not foreclose grant of summary judgment by another judge." 18 C. Wright, A. Miller E. Cooper, Federal Practice and Procedure § 4478, at 794 (1981 Supp. 2000) (footnote omitted). Although mutual respect between members of the same court, and consistency in their rulings, are no doubt desirable, "[j]udicial sensibilities should play no part in the disposition of suitors' rights." Dictograph Prods. Co. v. Sanotone Corp., 230 F.2d 131, 135 (2d Cir. 1956) (Learned Hand, J.). "Obviously we cannot be expected to reverse a correct decision by one [trial] judge simply because we find that it is contrary to a prior ruling by another [trial] judge in the same case, i.e., contrary to the law of the case." Parmelee Transp. Co. v. Keeshin, 292 F.2d 794, 797 (7th Cir.), cert. denied, 368 U.S. 944 (1961). We therefore agree with the Supreme Court of Connecticut that

[i]n an appeal to this court where views of the law expressed by a judge at one stage of the proceedings differ from those of another at a different stage, the important question is not whether there was a difference but which view was right.

Barnes v. Schlein, 473 A.2d 1221, 1222 (Conn. 1984) (citations and internal quotation marks omitted). If Judge Huvelle's disposition of Wilner's motion was correct on the merits — and we are satisfied that it was — we are required to affirm her ruling, see Parmelee Transp. Co., supra, 292 F.2d at 797, regardless of whether or not the judge adhered to the law of the case.

C. The merits.

(1) Reputational interests and the First Amendment.

In the District of Columbia, "a statement is defamatory "if it tends to injure [the] plaintiff in his trade, profession or community standing, or lower him in the estimation of the community." Best, supra, 484 A.2d at 989. But not every uncomplimentary publication is libelous. "[A]n allegedly defamatory remark must be more than unpleasant or offensive; the language must make the plaintiff appear `odious, infamous or ridiculous.'" Id. (quoting Johnson v. Johnson Pub. Co., 271 A.2d 696, 697 (D.C. 1970)).

Defamation is . . . that which tends to injure "reputation" in the popular sense; to diminish the esteem, respect, goodwill or confidence in which the plaintiff is held, or to excite adverse, derogatory or unpleasant feelings or opinions against him. It necessarily, however, involves the idea of disgrace; and while a statement that a person is a Republican may very possibly arouse adverse feelings against him in the minds of many Democrats, and even diminish him in their esteem, it cannot be found in itself to be defamatory, since no reasonable person could consider that it reflects upon his character.

W. Page Keeton, Prosser and Keeton on Torts § 111, at 773-74 (5th ed. 1984) (emphasis added) (footnotes omitted). Words which do not disparage a plaintiff's character are not actionable, even if special damages flow from their publication. Knight v. Blackford, 3 Mackey (14 D.C.) 177, 182-83 (Supreme Ct. D.C. 1877).

"[A] publication may convey a defamatory meaning if it `tends to lower the plaintiff in the estimation of a substantial, respectable group, though they are a minority of the total community or [of the] plaintiff's associates.'" White v. Fraternal Order of Police, 285 U.S. 273, 279, 909 F.2d 512, 518 (1990) (quoting Afro-American Publ'g Co. v. Jaffe; 125 U.S. App. D.C. 70, 75 n. 10, 366 F.2d 649, 654 n. 10 (1966)). If the publication "obviously would hurt the plaintiff in the estimation of an important and respectable part of the community, liability is not a question of majority vote." Peck v. Tribune Co., 214 U.S. 185, 190 (1909) (opinion by Holmes, J.). If Wilner's column defamed the plaintiffs in the eyes of persons connected with the railroad industry, then it is unnecessary for the plaintiffs to show that their reputations were unfavorably affected among the members of the public at large.

The trial court's threshold task in an action for defamation is to determine whether the challenged statement is "capable of bearing a particular meaning," and "whether that meaning is defamatory." Restatement (second) of Torts § 614 (1) (1977).

It is only when the court can say that the publication is not reasonably capable of any defamatory meaning and cannot be reasonably understood in any defamatory sense that it can rule, as a matter of law, that it was not libelous.

Levy v. American Mut. Ins. Co., 196 A.2d 475, 476 (D.C. 1964) (footnote omitted). The court should not, however, indulge far-fetched interpretations of the challenged publication. The statements at issue should not be "interpreted by extremes, but should be construed as the average or common mind would naturally understand [them]." 8 S. Speiser, C. Krause A. Gans, The American Law of Torts § 29:37 (1991) (hereinafter Speiser, Krause Gans). If the court determines that a statement is indeed capable of bearing a defamatory meaning, then whether that statement is in fact "defamatory and false [is a question] of fact to be resolved by the jury." Id.; Restatement (second) of Torts, supra, § 614 (1).

As both Judge Canan and Judge Huvelle recognized, modern defamation law also implicates constitutional principles. Thirty-six years ago, the Supreme Court held, for the first time, that the First Amendment to the United States Constitution may constrain the application of state defamation law, for "libel can claim no talismanic immunity from constitutional limitations." New York Times v. Sullivan, supra, 376 U.S. at 269. This is so because "[w]hatever is added to the field of libel is taken from the field of free debate." Id. at 272 (citing John Stuart Mill, on Liberty 15 (Oxford Blackwell ed. 1947)). "Even a false statement may be deemed to make a valuable contribution to public debate, since it brings about the clearer perception and livelier impression of truth, produced by its collision with error." Id. at 279 n. 19 (quoting On Liberty, supra, at 15) (internal quotation marks omitted).

Since New York Times v. Sullivan, the Court has made several attempts to balance the reputational interests protected by the common law of defamation with the guarantee of free speech that forms the constitutional and philosophic bedrock of our Republic. See, e.g., Gertz v. Robert Welch, Inc., 418 U.S. 323, 347 (1974) (holding that "so long as they do not impose liability without fault, the States may define for themselves the appropriate standard of liability for a publisher . . . of defamatory falsehood injurious to a private individual"); Philadelphia Newspapers, Inc. v. Hepps, 475 U.S. 767, 777-78 (holding that "the common law presumption that defamatory speech is false cannot stand when a plaintiff seeks damages against a media defendant for speech of public concern," and stating that "the First Amendment requires that we protect some falsehood in order to protect speech that matters") (quoting Gertz, supra, 418 U.S. at 341)). But striking an appropriate balance is not easy, for the competing interests are substantial. Children suspect, and adults know, that there is little truth to the adage that "sticks and stones may break my bones but words will never hurt me." "A defamatory statement may destroy an individual's livelihood, wreck his standing in his community, and seriously impair his sense of dignity and self esteem." Ollman, supra, 242 U.S. App. D.C. at 305, 750 F.2d at 974. Even Shakespeare's quintessential villain, Iago, spoke the literal truth when, with malice in his heart and treachery on his mind, he addressed his Commanding General:

Good name in man and woman, dear my lord,

Is the immediate jewel of their souls.

William Shakespeare, Othello, Act III, scene 3 (quoted in Milkovich, supra, 479 U.S. at 12). The plaintiffs in this case assert that Wilner has damaged their business and sullied their collective good name.

Imposition of liability upon an individual on the basis of his words, on the other hand, threatens to inhibit speech and to undermine the core values of a free society. "[T]he free flow of ideas and opinions is integral to our democratic form of government." Ollman, supra, 242 U.S. App. D.C. at 305, 750 F.2d at 974. The "central commitment of the First Amendment . . . is that debate on public issues should be uninhibited, robust and wide open." Bond v. Floyd, 385 U.S. 116, 136 (1966). More than two centuries ago, the French philosopher, Voltaire, anticipatorily articulated the spirit of our First Amendment when he wrote to Helvetius that "I disapprove of what you say, but I will defend to the death your right to say it." That right is surely at its zenith when it is exercised in an Op-Ed column on a subject of general public interest. Cf. Moldea II, supra, 306 U.S. App. D.C. at 16-17, 22 F.3d at 315.

In this case, balancing of the plaintiffs' reputational interests against Wilner's right to free expression is complicated by the fact that much of what is alleged to be defamatory in Wilner's column is not expressly stated. Defamation by implication, an area of law "fraught with subtle complexities," White, supra, 285 U.S. App. D.C. at 279, 909 F.2d at 518, requires careful exegesis to ensure that imagined slights do not become the basis for costly litigation.

[B]ecause the Constitution provides a sanctuary for truth, a libel-by-implication plaintiff must make an especially rigorous showing where the expressed facts are literally true. The language must not only be reasonably read to impart the false innuendo, but it must also affirmatively suggest that the author intends or endorses the inference.

Chapin v. Knight-Ridder, 993 F.2d 1087, 1092-93 (4th Cir. 1993); White, supra, 285 U.S. App. D.C. at 279, 909 F.2d at 520; Dodds v. American Broad. Co., 145 F.3d 1053, 1063-64 (9th Cir. 1998) (holding that plaintiff in defamation-by-implication case must show that the defendant intended to convey the defamatory impression); Howard v. Antilla, 191 F.R.D. 39, 44 (D.N.H. 1999) (same); Royal Palace Homes, Inc. v. Channel 7 of Detroit, Inc., 495 N.W.2d 392, 395-96 (Mich.Ct.App. 1992) (same); see also, Moore v. Sun Publ'g Corp., 881 P.2d 735, 741-42 (N.M.Ct.App. 1994) (applying heightened standard for defamation by implication although the plaintiff was not a public figure and the subject matter of the suit was not of general public interest).

Moreover, the substance of Wilner's column was fairly characterized by Judge Huvelle as "opinion." Prior to the Supreme Court's decision in Milkovich, statements classified as "opinion" were generally viewed as non-defamatory. In Gertz, supra, 418 U.S. at 339-40, the Court stated that

[u]nder 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.

"[A] majority of the federal . . . courts of appeal . . . interpreted the Gertz dictum to mean that statements of fact can be actionable defamation [but that] statements of opinion cannot." Sigal Const. Co. v. Stanbury, 586 A.2d 1204, 1209 (D.C. 1991); but cf. Myers, supra, 472 A.2d at 47 (holding that "[e]xpressions of opinion are entitled to constitutional protection unless they imply the existence of undisclosed defamatory facts as the basis of the opinion").

In 1989, in Milkovich, the Supreme Court rejected any per se rule providing blanket First Amendment protection for all statements of opinion:

[W]e do not think this passage from Gertz was intended to create a wholesale defamation exemption for anything that might be labeled "opinion.". . . . Not only would such an interpretation be contrary to the tenor and context of the passage, but it would also ignore the fact that expressions of "opinion" may often imply an assertion of objective fact.

497 U.S. at 18. The Court reasoned that no such absolute exemption for opinion was necessary, for the "vital guarantee of free and uninhibited discussion of public issues" was adequately protected by the Court's existing First Amendment jurisprudence. Id. at 22. The Court noted that liability under state defamation law for a statement on a matter of public concern may be imposed only if the statement is provably false, id. at 19-20, citing Hepps, supra, 475 U.S. at 776-77; that "imaginative expression" and "rhetorical hyperbole" are not actionable in defamation because they "cannot reasonably be interpreted as stating actual facts about an individual," id. at 20, quoting Falwell, supra, 485 U.S. at 50, 53-55; and that a public figure may recover in defamation only if the challenged defamatory statements were "made with knowledge of their false implications or with reckless disregard of their truth." Id.

Thus, although Milkovich has foreclosed any contention that there exists a blanket immunity for opinion, the Supreme Court took pains to ensure that constitutional restraints on the reach of state defamation law would remain in place so that protected speech would not be inhibited. Under Milkovich, "statements of opinion can be actionable if they imply a provably false fact, or rely upon stated facts that are provably false." Moldea II, supra, 306 U.S. App. D.C. at 10, 22 F.3d at 313. "[B]ut if it is plain that a speaker is expressing a subjective view, an interpretation, a theory, conjecture, or surmise, rather than claiming to be in possession of objectively verifiable facts, the statement is not actionable." Haynes v. Alfred A. Knopf, Inc., 8 F.3d 1222, 1227 (7th Cir. 1993).

"[A] statement of opinion is actionable only if it has an explicit or implicit factual foundation and is therefore objectively verifiable." Washington v. Smith, supra, 317 U.S. App. D.C. at 80, 80 F.3d at 556 (citation and internal quotation marks omitted). "Assertions of opinion on a matter of public concern receive full constitutional protection if they do not contain a provably false factual connotation." Id. (citation and ellipsis omitted). Finally, " Milkovich did not disavow the importance of context," Moldea II, supra, 306 U.S. App. D.C. at 5, 22 F.3d at 314; see also Masson v. New Yorker Magazine, Inc., 501 U.S. 496, 511-13 (1991), and it therefore remains critical to our inquiry that the allegedly defamatory utterances in this case appeared in an Op-Ed column in which Wilner was commenting on matters of substantial public concern.

We summarize. In order to sustain an action for defamation against Wilner, the plaintiffs must show both that his column was capable of bearing a defamatory meaning and that it contained or implied provably false statements of fact. On appeal from an order granting summary judgment, the question is whether an impartial jury could reasonably so find.

(2) The perceived imputation of hostility to labor.

Substantially for the reasons stated by Judge Huvelle, we are unpersuaded by the plaintiffs' principal contention, namely, that Wilner defamed them by characterizing them as antagonistic to labor. As Judge Huvelle noted, a majority of courts have held that the characterization of an employer as "unfair" during the course of a labor dispute is not libelous. We agree with these authorities. In Gregory, supra, 552 P.2d at 425, the Supreme Court of California stated that

statements made in the context of a labor dispute which on their face resemble statements of fact, may, depending on the circumstances, be treated as statements of opinion not subject to an action for libel. In Emde v. San Joaquin County etc. Council, supra, . . . 143 P.2d [at] 20, we expressed the principle that ". . . since such [labor] disputes, realistically considered, normally involve considerable differences of opinion and vehement adherence to one side or the other, a necessarily broad area of discussion without civil responsibility in damages is an indispensable concomitant of the controversy." . . . (143 P.2d [at] 26). Thus, we characterized as opinion, statements by a union that an employer had "hired" nonunion workers and put them on a "straight commission plan," that certain guarantees gained by the union had been "wiped out," that the status of employees "remained unchanged," and that the employer's labor policy was "destructive."

We also agree with Judge Huvelle's view that, insofar as allegations of hostility to labor are concerned, the author of an Op-Ed column is entitled to protection at least as great as that which is provided to the employee with the denunciatory picket sign. In the Montgomery Ward case, supra, which involved a column in Business Week, the court concluded that a charge of being unfair to labor was not libelous. The court stated:

If it were libelous simply to say of another that he is unfair to labor, every picketer who carries a banner with such a legend would be guilty of libel. To say that one is unfair to labor is not a statement of fact, but of an opinion. Likewise to say of one: you are reactionary, you are undemocratic, you are a nationalist, you are an isolationist, you are a New Dealer, you are a Union Leaguer, you are opposed to labor, you are a coddler of labor, is similarly to express an opinion. Calling one unfair to labor is no different. All of these expressions spring from controversies on questions of public interest. "But written abuse relating solely to political views or arguments on questions of public interest, and which do not attack the character of a person, nor impute immorality, nor violation of law, are not actionable per se." Newell on Libel and Slander, 4th Ed. p. 57.

146 F.2d at 176.

Moreover, we do not believe that those statements in the column which the plaintiffs have characterized as "provably false" are of the genre which would support a defamation case against the author of a column on the opinion page of a newspaper. The plaintiffs' focus has been on Wilner's allegation that Guilford "bolted" from national wage and benefit negotiations. According to Professor Northrup, Guilford did not bolt; rather, it declined to "opt into" national "handling." Either way, Guilford negotiated locally and not nationally. Even assuming that Wilner's use of the verb "bolted" reflects lack of precision, and treats the plaintiffs with undeserved asperity, the challenged language surely pales in comparison to "blackmail," Greenbelt Publ'g Ass'n, supra, 398 U.S. at 11-14, or "traitor" or "scab," Austin, supra, 418 U.S. at 282-87. If we were to adopt a rule of law which sustains the plaintiffs' position on this issue, then

authors of every sort would be forced to provide only dry, colorless descriptions of facts, bereft of analysis or insight. There would be little difference between the editorial page and the front page, between commentary and reporting, and the robust debate among people with different viewpoints that is a vital part of our democracy would surely be hampered.

Partington, supra, 56 F.3d at 1154.

Speaking through Professor Northrup, the plaintiffs also insist that the Springfield Terminal leases "had a transportation purpose and the ICC so found," and that Guilford's employees did not suffer a reduction in compensation or a loss of seniority. But Wilner acknowledged in his column that the ICC "sanctioned the lease agreement and referred certain labor issues to arbitration." In other words, he informed the reader that Guilford's actions had been upheld by a federal agency. To be sure, Wilner described the ICC as having a "zealous pro-management bias in those days," but he was surely within his rights in disagreeing with and criticizing the Commission. Any reasonable reader of the column would understand that Guilford took certain actions, that Wilner was apparently unenthusiastic about those actions, and that the ICC basically sustained them. This is not the stuff of which successful libel suits are made.

Finally, whether or not any of the remaining statements challenged by Professor Northrup are in fact inaccurate, Wilner's basic theme is not provably false. Whether the plaintiffs are hostile to labor is not an issue of fact susceptible of objective proof. "[I]t is the defamatory implication — not the underlying assertions giving rise to the implication — which must be examined to discern whether the statements are entitled to full constitutional protection." White, supra, 285 U.S. App. D.C. at 284, 909 F.2d at 523 (emphasis in original). "[I]n determining the gist or sting of a newspaper article to assess whether it is actionable, a court must look at the highlight of the article, the pertinent angle of it, and not to items of secondary importance." Partington, supra, 56 F.3d at 1161 (emphasis in original; citation and internal brackets and quotation marks omitted). Any inaccuracies in Wilner's column regarding Guilford's attitude vis-á-vis labor do not give rise to liability for defamation.

(3) The alleged accusation that the plaintiffs violated the Railway Labor Act.

The plaintiffs' claim that Wilner accused them of violating the RLA, while closely related to the broader issue of hostility to labor, stands on a somewhat different legal footing. "To charge one with being unfair to labor is a much broader charge than to charge one with unfair labor practices." Montgomery Ward, supra, 146 F.2d at 176. A narrower and more specific charge, focused on allegedly unlawful conduct, is more readily susceptible of a defamatory meaning than is the mere attribution of a hostile state of mind. In Montgomery Ward, for example, the court recognized that to accuse an employer falsely of violations of a state safety statute is libelous per se. Id. at 177 (citations omitted). "[T]o constitute a libel it is enough that the defamatory utterance imputes any misconduct whatever in the conduct of the [plaintiff's] calling." Restatement (Second) of Torts § 569, cmt. (e) (1977). Violation of the RLA might reasonably be viewed as "misconduct," and we shall assume for purposes of this opinion that a provably false accusation of such conduct may be defamatory.

The RLA provides, in pertinent part, as follows:

First. Duty of carriers and employees to settle disputes

It shall be the duty of all carriers, their officers, agents, and employees to exert every reasonable effort to make and maintain agreements concerning rates of pay, rules, and working conditions, and to settle all disputes, whether arising out of the application of such agreements or otherwise, in order to avoid any interruption to commerce or to the operation of any carrier growing out of any dispute between the carrier and the employees thereof.

45 U.S.C. § 152 (First). The gist of the plaintiffs' position on this aspect of the case appears to be that Wilner has falsely accused them of not carrying out their responsibilities as set forth in the quoted language. We do not agree.

In Montgomery Ward, supra, it was alleged in the Business Week article, inter alia, that the employer's labor policy was "anathema" to the United States Conciliation Service (USCS) and to the CIO; that the employer had refused to meet with the USCS and the union, had "brushed off" the union, and had given the USCS the "run-around"; and that the employer had said "no" to every request made by the union. The employer claimed that these and other allegations in the article were intended to mean, and were understood by readers to mean, that the employer had refused to bargain collectively with the duly chosen representatives of its employees, in violation of the National Labor Relations Act, 29 U.S.C. § 151 et seq. The court held, to the contrary, that Business Week had not accused the employer of any unlawful conduct, and that "the alleged libelous words cannot reasonably be said to charge the plaintiff either with violation of the National Labor Relations Act or with being unfair to labor." Montgomery Ward, supra, 146 F.2d at 175.

In the present case, as the trial judge pointed out, the RLA is not mentioned in Wilner's column at all. Guilford's obligations under that statute likewise are not addressed. We know of no provision of the RLA requiring Guilford to participate in national rather than local negotiations, and Wilner's column certainly does not provide the reader with any reason to believe that "not opting" into national negotiations is unlawful. In addition, for the reasons previously stated, we do not believe that use of the word "bolted" alters the result. Similarly, the RLA does not require employers to accede to union proposals or to refrain from litigating legal issues before agencies and courts. To the extent that there is any discussion in the column of the legality or lack thereof of Guilford's various actions, Wilner discloses that the ICC ruled in Guilford's favor on some issues and that Guilford had engaged in contested litigation with the union and with Amtrak. Judge Huvelle correctly pointed out that, notwithstanding any claimed inaccuracies as to dates and the like, the critical historical facts are undisputed, and the reader is therefore free to draw his or her own conclusions regarding whether the plaintiffs acted wrongfully.

The column may perhaps be read as indicating that Wilner is sympathetic to the union's point of view rather than to Guilford's, but that surely does not render the column defamatory. The reader might fairly infer from the column that Guilford bargains "tough," that it selects the forum that it deems most favorable ( e.g., local rather than national negotiations), and that it litigates vigorously in support of its position, sometimes successfully, sometimes not. Such a posture roughly parallels the conduct attributed by Business Week to the employer in Montgomery Ward, and we conclude that here, as in that case, no violation of the statute has been alleged or implied.

Under these circumstances, the plaintiffs have not satisfied the rigorous standards which the courts have applied to claims of libel by implication. See Washington, supra, 317 U.S. App. D.C. at 81, 80 F.3d at 557; White, supra, 285 U.S. App. D.C. at 279, 909 F.2d at 518; Chapin, supra, 993 F.2d at 1093. "Nothing in law or common sense supports saddling a libel defendant with civil liability for a defamatory implication nowhere to be found in the published article itself." Tavoulareas v. Piro, 260 U.S. App. D.C. 39, 58, 817 F.2d 762, 781 (footnote omitted), cert. denied, 484 U.S. 870 (1987). It is not impossible that some readers would read the column as suggesting what the plaintiffs say that it suggests, but to hold Wilner responsible for every inference a reader might reasonably draw from his article would undermine the uninhibited discussion of matters of public concern. Saenz, supra, 841 F.2d at 1318.

We are not unmindful of the plaintiffs' contention that materials presented by them in opposition to Wilner's motion for summary judgment present genuine issues of material fact. According to Professor Northrup, "any reader familiar with the railroad industry would conclude from reading [Wilner's column] that the plaintiffs . . . have violated their duties under the RLA. . . ." On the surface, this assertion, if accepted at face value, see Anderson, supra, 477 U.S. at 255, could present significant difficulties for Wilner's motion for summary judgment. Considered in context, however, Northrup's declaration cannot change the result.

The quoted sentence of Northrup's affidavit concerns the legal conclusion that, according to him, a presumably sophisticated railroader would draw from the column. The question whether the words that Wilner has written are fairly susceptible of such a meaning is one of law. See Restatement (Second) of Torts, supra, § 614 (1). The court, and not a lay expert witness (or, for that matter, a reader familiar with the railroad industry) decides such questions as whether "bolting" from national negotiations (by opting for local ones) violates the RLA. Professor Northrup's comments concern a column which says nothing directly about the RLA, and which suggests, at most, a belief on Wilner's part that the plaintiffs did not negotiate as generously as they ought to have done, and that they litigated more than they should have done. In our view, Northrup's claimed expertise in railroad matters cannot qualify him to decide whether the column is libelous.

If Professor Northrup's affidavit is supposed to suggest that the column in fact charges or implies a violation of the RLA, and if the court discerns no such allegation or implication in the column, then the court's resolution of an issue of law necessarily trumps the views of a witness, even one with Professor Northrup's qualifications. If, on the other hand, Northrup means that, even if the column does not on its face allege or imply RLA violations on Guilford's part, persons in the railroad business would believe that it does charge Guilford with contravening the statute, then Wilner cannot be held responsible for the railroaders' erroneous beliefs. To the extent that Northrup may be suggesting that Wilner is writing in a code — in "railroadese," perhaps — that railroad people will understand, even if judges do not, then we believe that such a suggestion is too remote to support a finding of defamation by implication. We agree with Judge Huvelle that the kinds of implications that the plaintiffs attribute to the column go well beyond the sense of the published words.

An affidavit in opposition to a motion for summary judgment must be based on the affiant's personal knowledge. Super. Ct. Civ.R. 56 (e). Professor Northrup has no personal knowledge, and claims none, regarding whether an unidentified railroader would interpret Wilner's column as attributing RLA violations to the plaintiffs. Arguably, Northrup could be qualified as an expert witness and permitted to express opinions on the subject of his expertise. In our view, however, the conclusion that unidentified readers might draw from the column as to the lawfulness of the plaintiffs' conduct is not a proper subject of expert testimony. An expert witness generally may not express an opinion on a question of law, see Weston v. Washington Metro. Area Transit Auth., 316 U.S. App. D.C. 321, 323, 78 F.3d 682, 684, amended on reh'g, 318 U.S. App. D.C. 142, 86 F.3d 216 (1996), and an opinion as to what others may, rightly or wrongly, believe about the law is not significantly different.

The plaintiffs' opposition to Wilner's motion for summary judgment has been constructed with considerable skill and sophistication, and the issue before us is not an easy one. But if the judgment is reversed, and if Wilner is compelled to continue to litigate the case through trial and the inevitable appeals, the potential chilling effect upon robust debate and freedom of expression may be quite substantial. For all of the foregoing reasons, we are satisfied that the entry of summary judgment against Guilford was appropriate. See Keogh, supra, 125 U.S. App. D.C. at 35, 365 F.2d at 968; Myers, supra, 472 A.2d at 50; Nader, supra, 408 A.2d at 42-44.

(4) The claim that Fink and Mellon were personally defamed.

We agree with the trial judge that the comments in the column about Mellon and Fink do not make either man appear "odious, infamous, or ridiculous." Best, supra, 484 A.2d at 989; see also Prosser and Keeton, supra, § 111, at 773-74 (defamation "involves the idea of disgrace"). The implication in Wilner's column that Mellon is a kind of dilettante, and that he, and perhaps Fink too, lack seriousness as businessmen, is the quintessential kind of commentary — provocative, perhaps, but hardly libelous — which is entitled to protection under the First Amendment. The column was, after all, written in response to Guilford's attempt to acquire and to privatize portions of Amtrak's operations. In a society which regards freedom of the press as a core value, a newspaper columnist must surely have the right to question the qualifications of the would-be privatizers, as well as their motivations and business acumen, without fear of retaliatory litigation. A jurisprudential giant has said it well:

But when men have realized that time has upset many fighting faiths, they may come to believe even more than they believe the very foundations of their own conduct that the ultimate good desired is better reached by free trade in ideas — that the best test of truth is the power of the thought to get itself accepted in the competition of the market. . . .

Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting). The plaintiffs have the means to respond to critical Op-Ed columns, and under our Constitution they may not be denied the opportunity to do so. At the same time, the right of the Wilners of the world to have their say must also be preserved.

III. CONCLUSION

In the Moldea case, the New York Times' reviewer of the plaintiff's book accused the plaintiff, inter alia, of "too much sloppy journalism" and "attack[ed] Moldea's competence as a practitioner of his chosen profession, a matter archetypically addressed by the law of defamation." Moldea I, supra, 304 U.S. App. D.C. at 409, 15 F.3d at 1140. The court initially held that "the accuracy of [some] statements in the review is sufficiently open to dispute that we cannot hold as a matter of law that no reasonable juror could find them to be false." Id. But after setting aside the trial court's order granting summary judgment in favor of the newspaper, Moldea I, the court reconsidered the issue, modified its prior opinion, and affirmed the District Court's ruling. Moldea II. The court stated:

Unfortunately, [the Moldea I] opinion failed to take sufficient account of the fact that the statements at issue appeared in the context of a book review, a genre in which readers expect to find spirited critiques of literary works that they understand to be the reviewer's description and assessment of texts that are capable of a number of possible rational interpretations. While there is no per se exemption from defamation for book reviews, our initial resolution of this case applied an inappropriate standard to judge whether the Times review was actionable.

306 U.S. App. D.C. at 3-4, 22 F.3d at 311-12. An Op-Ed column in a trade newspaper is indistinguishable in principle from a book review, and application to this record of Moldea II' s reasoning dooms the plaintiffs' action.

Affirmed.