Norman Wang v. University of Pittsburgh

UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT

No. 25-1816

NORMAN WANG,

Appellant

v.

UNIVERSITY OF PITTSBURGH; UNIVERSITY OF PITTSBURGH

MEDICAL CENTER; UNIVERSITY OF PITTSBURGH PHYSICIANS;

SAMIR SABA; MARK GLADWIN, et al.

On Appeal from the U.S. District Court, W.D. Pa.

Judge Marilyn J. Horan, No. 2:20-cv-01952

Before: SHWARTZ, BIBAS, and PHIPPS, Circuit Judges

Argued: Mar. 2, 2026; Filed: July 7, 2026

OPINION OF THE COURT

BIBAS, Circuit Judge. Academic debate is a full-contact

sport. Debating ideas and questioning arguments can bruise

people’s egos. But like boxing, academic debate bars low

blows. Academics may not harm one another by recklessly disregarding the truth. Administrators may not demote professors

for calling out discrimination. And managers may not create or

tolerate hostile work environments. A culture that cancels instead

of counsels sacrifices persuasion on the altar of power. In doing

so, it forgets that the remedy for disfavored speech is more

speech, not coerced silence.

Dr. Norman Wang is an academic cardiologist. He worried

that race-based affirmative action in medical schools and hospital residencies discriminates against some minorities in favor of others, could violate the law, and harms the very doctors

whom it tries to help. Though most commentators opposed his

views, some research supported it. After he published an article expressing those views, his bosses (also mostly doctors) allegedly denounced it as false and racist, fanned the flames on social media, demoted him at the hospital system, and got his article

retracted.

Wang sued the university, hospital system, doctors, journal,

and publisher for defamation and retaliation. The District Court dismissed some claims on the pleadings and granted summary

judgment against Wang on the rest. That was partly wrong.

True, Wang has not plausibly alleged state action. But he has

plausibly alleged defamation against five defendants, and there are disputed facts about whether his employers retaliated

against him for opposing their allegedly illegal discrimination. Thus, we will let many of Wang’s claims go forward.

I. AFTER WANG CRITIQUES RACIAL PREFERENCES,

HIS BOSSES STRIKE BACK

A. Wang’s article questions race preferences in cardiology

Pittsburgh enjoys a public-private partnership to train doctors. The University of Pittsburgh (Pitt) is a public university with a medical school. It partners with a private hospital system, the University of Pittsburgh Medical Center (UPMC), and

its subsidiary, University of Pittsburgh Physicians (which

employs doctors in the UPMC hospital system). Medical

2

students study at the university and join residents and fellows in some training at the hospital system.

Wang, like many of his colleagues, is both a professor at the

medical school and a cardiologist at the hospital system. He

has worked at both for almost two decades, serving patients

and training residents and fellows. He no longer teaches regularly at the medical school. Nine years ago, he worked his way

up to become director of a cardiology fellowship program at

the hospital system. As program director, he helped select and

train fellows.

Admission to medical schools, residencies, and fellowships

is a competitive process based on grades, standardized-test

scores, and “holistic” factors. App. 2642. Around 2018, Wang

noticed that the American Heart Association, American College of Cardiology, and Accreditation Council for Graduate

Medical Education (the hospital’s accreditor for residency and

fellowship programs) were pushing another factor: race.

Wang was concerned. Though he is “very much for diversity,” he feared that these diversity programs “look[ed] overtly discriminatory and quota-based” and failed to “treat people as

individuals.” App. 2467, 2469, 2490. So in March 2020, he critiqued these practices in the peer-reviewed Journal of the

American Heart Association. His seventeen-page article included four tables, four figures, lots of data, and 108 endnotes full of scholarly sources. After explaining how these programs may

violate the law and harm members of minority groups, he concluded: “Long-term academic solutions and excellence should

not be sacrificed for short-term demographic optics.” App.

2642.

3

B. Wang’s bosses demote and ostracize him

When Wang first published his article in March, it did not

make a stir. All that changed in the summer of 2020. On July

29, a doctor emailed the president of the hospital system’s Physicians Services Division to flag and complain about Wang’s

conclusions (not his data). The president looped in the medical school’s dean, who agreed, adding that the article was “pseudoscholarly … with handpicked data,” as well as “incredibly offensive and racist.” App. 2657, 2662.

At the same time, another doctor emailed Dr. Kathryn Berlacher, who pinged Dr. Samir Saba. Like Wang, Berlacher and

Saba work at both the hospital system and the medical school.

Saba is Wang’s superior in both workplaces; Berlacher is his

colleague at the medical school but at the hospital gives guidance to him as director of the cardiology fellowship program,

an umbrella that oversees Wang’s subspecialty program. The

doctor who contacted Berlacher said she “can’t argue with

facts” but objected to the “quotations he chose to use, the biased conclusions he made from the data, etc.” App. 2655. Berlacher

agreed that, while “we can’t argue w[ith] facts,” “[i]t’s v[ery] clearly offensive and racist.” App. 2654. Saba concurred,

expressing his “overwhelming feeling … that [Wang] cannot

continue in [his leadership] role.” App. 2673. He looped in Dr. Mark Gladwin, chair of the medical school’s Department of

Medicine and head of training at the hospital system, on the

email chain. Gladwin lamented that, though Wang’s article

“was written before the George Floyd events, [the] timing of

the print publication could not be worse.” App. 2667. Later,

Saba replied that “[e]motions are very high, understandably.”

Id. So on July 31, two days after getting those complaints,

4

Berlacher and Saba met with Wang and fired him from his program directorship at the hospital system.

At that meeting, Wang raised his concerns about racial and

ethnic preferences in “our [graduate-medical-education] contracts with the fellows” at the hospital system and said he “just wanted us to follow the law.” App. 961-63. Wang did not explicitly state that the hospital was acting illegally. At some point, though, Saba told him: “you know what we’re trying to do

here, Norm,” App. 2490–91. Saba also said “that the views that

[Wang had] expressed [we]re not in line with the views and

policies that [we] were trying to create,” App. 2498.

Once Wang had been fired, the next question was whether

to rebut Wang’s article or just smear it. Gladwin and the medical-school dean urged university and hospital-system leaders

to “take[ ] the high road” by publishing an opposing paper identifying specific errors. App. 2657. They did not.

Instead, Berlacher, Saba, and other doctors launched a socialmedia campaign to discredit Wang’s article. Two days after

meeting with him, Berlacher tweeted that it “misinterprets data and misquotes people” and “is scientifically invalid and racist.” App. 145. Next, Berlacher authored a tweet on the hospital system’s Cardiology Department account (which she controlled)

accusing Wang of using “misquotes, false interpretations, and

racist thinking.” App. 145, 233. Saba reposted that tweet. That fusillade fanned “a firestorm on [T]witter, [F]acebook[,] and in all national cardiology circles.” App. 3013. Doctors called for Wang to be cut off from training to “#RetractRacists” and protect students. App. 2783. One fellow emailed Berlacher, describing an “uncomfortable” experience that he had had with an

5

“intense grilling” by Wang as evidence that he disliked minorities. App. 2701. Berlacher took this email as confirmation that it was “unsafe” for fellows to work for him and that he needed

to be removed from his directorship and all educational roles

“immediately.” App. 2700.

So they did. On August 4, two days after their tweets, Berlacher and Saba emailed Wang that “any educational environment in which you partake is inherently unsafe, increasing our

learners’ risk for undue bias and harm.” App. 2706. They

banned him from any role teaching fellows, residents, or medical students. Berlacher had already tweeted to #medtwitter,

reassuring other doctors that the students were safe. Over the

next three days, the medical school’s dean and Gladwin sent

hospital- and school-wide emails, denouncing Wang’s article

as hurtful and offensive.

As a result, Wang suffered. He was uncomfortable going to

work because long-time colleagues stopped talking to him. He

cleaned out his office. Other doctors worried that he might kill himself, and the police even called to see if he was being harassed.

C. Hospital and university doctors also get the journal

to retract Wang’s article

On August 3, Berlacher, Saba, the medical-school dean, and

four other hospital-system doctors emailed the Journal of the

American Heart Association’s editor-in-chief, urging the journal to retract Wang’s article. They decried it as “deeply offensive and disturbing,” with “several misquotes” and “proven

scientific falsehoods.” App. 2708. They attached to the email

an eight-page “point-by-point rebuttal” that they “[we]re

6

planning to submit[ ] for publication.” Id. Starting July 31, hospital-system employees also sent back-channel messages to the

American Heart Association, the journal’s proprietor.

Two days after Berlacher, Saba, and others emailed the editorin-chief, the Association and the journal’s editors retracted the article. The retraction notice said: “[T]he University of Pittsburgh Medical Center (UPMC)[ ] has notified the Editor-inChief that the article contains many misconceptions and misquotes and that together those inaccuracies, misstatements, and selective misreading of source materials strip the paper of its scientific validity.” App. 237. The Association also put out a

press release that day. The next day, an updated notice added

two alleged misquotations as a “sample of the misconceptions

and misquotes identified in correspondence from UPMC.”

App. 163. “Based on errors such as these, the editors” were

retracting the article. App. 164. Later that month, an editor’s note on behalf of the journal, lead-authored by Dr. Marc Simon

of the hospital system and medical school, accused “the retracted article … [of] misrepresent[ing] facts” and “evidence.” App.

195.

Yet the following months brewed a backlash to the backlash. Multiple institutions objected that Wang had been disciplined without due process for exercising his academic freedom. The university’s Tenure and Academic Freedom Committee asked the medical-school dean to reinstate Wang. The

American Association of University Professors thought the

same. The U.S. Department of Education started investigating

whether the university’s “campaign of denunciation” violated

Title VI and other laws. App. 3223. The medical-school dean

said that “there was a lot of consternation” over potentially

7

“limiting academic freedom” and “this was clearly a big issue

for the faculty senate.” App. 2147. So in late October, under

pressure, he told Wang that he had never lost any faculty status or privileges at the university and notified Berlacher, Saba, and Gladwin that they had no authority to bar Wang from university

activities.

D. The District Court rejects Wang’s defamation and

retaliation claims

Within a week of the first complaint about his article, Wang

was demoted, his article was retracted, and he went viral on

Twitter for being a dishonest racist. So Wang sued a range of

defendants on three theories. First, he claimed defamation by

the university, the hospital system, its physician practice, Berlacher, Saba, Simon, the American Heart Association, and the

journal’s publisher (Wiley Periodicals). But the District Court dismissed that claim. Even though most of the statements were

defamatory statements of fact or opinions based on undisclosed

facts, it reasoned, the Association’s and publisher’s statements were true and Wang (a limited-purpose public figure) had

failed to plead that the rest of the defendants acted with actual malice.

Second, Wang brought various civil-rights retaliation

claims against the university, the hospital system, and its physician practice, as well as Berlacher, Saba, and Gladwin. The

District Court dismissed his Title VI claim against the hospital for failing to plead that a primary purpose of the federal funds was to fund employment, and the Title VI and 42 U.S.C. § 1981

claims against the university because any adverse action had

come from the hospital system, not the university. Wang also

8

brought claims under Title VII and the Pennsylvania Human

Relations Act (PHRA). The District Court dismissed these

claims against the university as to Wang’s statements in his article but not as to his statements in his July 31 meeting with Berlacher and Saba. Later, the court granted summary judgment

against Wang on all remaining civil-rights claims, reasoning

that Wang had not engaged in any protected activity at the July 31 meeting.

Third, Wang brought a First Amendment retaliation claim

under 42 U.S.C. § 1983 against those same defendants. Applying Monell’s municipal-liability standard, the District Court

dismissed the claims against the university as failing to allege action under a policy or by policymakers. See generally Monell

v. Dep’t of Soc. Servs., 436 U.S. 658 (1978). It also dismissed the claims against the hospital system and its physician practice, and later granted summary judgment to the individual

defendants, reasoning that the hospital and the physician practice were not state actors, that Berlacher and Saba were not acting as employees of the university, and that Gladwin was not

involved in Wang’s removal. Wang now appeals all these adverse

rulings.

II. THE DEFAMATION CLAIM SURVIVES

AGAINST FIVE DEFENDANTS

To make out a defamation claim, Wang must plausibly

plead and prove that a defendant’s published communication

was defamatory and sufficiently linked to him. 42 Pa. Cons.

Stat. § 8343(a). We review the District Court’s dismissal de

novo, taking the complaint’s factual allegations in the light

most favorable to Wang. Fowler v. UPMC Shadyside, 578 F.3d

9

203, 206, 210–11 (3d Cir. 2009) (discussing Ashcroft v. Iqbal,

556 U.S. 662, 678 (2009)).

The District Court erred in dismissing Wang’s whole defamation claim. He has plausibly pleaded that the challenged

statements were defamatory. Defendants either did not argue

or cannot show that their statements were substantially true.

Wang appears to be a limited-purpose public figure who must

plead that the defendants spoke with actual malice, but he

pleads enough to clear that hurdle as to Berlacher, Saba, the

Association, the university, and the hospital system. So we will let much of Wang’s defamation claim proceed.

A. The challenged statements asserted facts or were

based on undisclosed facts

A statement can be defamatory if “it tends so to harm the

reputation of another as to lower him in the estimation of the

community or to deter third persons from associating or dealing with him.” Tucker v. Fischbein, 237 F.3d 275, 282 (3d Cir.

2001) (Alito, J.) (citation omitted). Statements of fact can be defamatory. So too can statements of opinion that “give[ ] rise to the inference that there are undisclosed facts that justify the forming of the opinion.” Meyers v. Certified Guar. Co., 221

A.3d 662, 670 (Pa. Super. Ct. 2019) (quoting Restatement

(Second) of Torts § 566 cmt. b (1977)). To see if a statement

implies plausibly defamatory facts, we look at it in context.

Graboff v. Colleran Firm, 744 F.3d 128, 136 (3d Cir. 2014).

The District Court rightly found that Berlacher’s and Saba’s

tweets were opinions based on undisclosed facts, so they could

amount to defamation. They launched a campaign to ostracize

him, making his article go viral and generating a backlash.

10

Berlacher tweeted from her personal account: “@PittCardiology I’m looking at you. What do we stand for? What do you

think of this OPINION piece that misinterprets data and misquotes people? @JAHA_AHA this is scientifically invalid and

racist.” App. 145, 235. Next, the hospital-system’s official

Twitter account @PittCardiology tweeted that Wang’s article

used “misquotes, false interpretations and racist thinking to

defend a single person’s conclusion.” App. 145, 231. Saba then

retweeted that on behalf of one of the hospital system’s institutes.

We cannot dismiss these statements as mere academic

jousting. Of course, science is a process of revision, so “tentative scientific conclusions” based on disclosed data and methods are not actionable. Pacira Biosciences, Inc. v. Am. Soc’y of Anesthesiologists, Inc., 63 F.4th 240, 247 (3d Cir. 2023). But

these were not unfalsifiable “disagreements about the reliability of the methodology and data underlying the statements.” Id. Rather, these tweets accused Wang of misquoting and misreading data and sources—both of which are falsifiable claims. So

they were more hit job than academic argument.

The Association made defamatory statements, too. It issued

two retraction notices (through the publisher) plus a press release. The first notice alleged that the paper contained “many misconceptions and misquotes and that together those inaccuracies, misstatements, and selective misreading of source materials strip the paper of its scientific validity.” App. 143. It gave no examples. The companion press release called Wang’s article

“a misrepresentation of the facts,” containing “deliberate misinformation [and] misrepresentation.” App. 144. Later that

month, the journal’s editorial board published an “Editor’s

11

Note,” lead-authored by Simon, accusing Wang of “misrepresent[ing] facts” and “evidence.” App. 145.

The “sting” of the tweets was that Wang’s article was

sloppy, riddled with errors, scientifically invalid, and motivated by racism. See Oweida v. Trib.-Rev. Publ’g Co., 599 A.2d

230, 239 (Pa. Super. Ct. 1991); accord Schiavone Constr. Co.

v. Time, Inc., 847 F.2d 1069, 1083 (3d Cir. 1988) (applying

analogous New Jersey law). Though “a simple accusation of

racism” is not enough in Pennsylvania, these tweets went further and “could be construed to mean that [Wang] was acting

in a racist manner” and that his racism led him to distort facts and evidence in connection with his work as an academic and

a doctor. MacElree v. Phila. Newspapers, Inc., 674 A.2d 1050,

1055 (Pa. 1996). In context, they implied that the article was

full of errors driven by motivated reasoning. The Association’s notices, press release, and article likewise accuse him of

spreading half-truths and lies. Wang plausibly alleges “the kind of harm which has grievously fractured his standing in the

community of respectable society.” Tucker v. Phila. Daily

News, 848 A.2d 113, 124 (Pa. 2004) (internal quotation marks

omitted).

It would have been fine just to say that Wang’s article was

wrong, harmful, even dangerous. Those are normal academic

opinions. See Pacira, 63 F.4th at 245–47 (reasoning that a scientific article’s claim that a drug was “inferior” or “not superior” was an opinion, a tentative scientific conclusion that

could not be defamatory). But these statements went much further, alleging misquotations, misstatements, and outright falsehoods. Wang’s academic integrity is central to his career as a

professor, researcher, and doctor. And in Pennsylvania, “any

12

language which unequivocally, maliciously and falsely imputes

to an individual or corporation [a] want of integrity in the conduct of his or its business is actionable.” Cosgrove Studio &

Camera Shop, Inc. v. Pane, 182 A.2d 751, 754 (Pa. 1962).

When a journal retracts a paper for its “many” misstatements

and inaccuracies, a reasonable reader infers that the journal

found many falsehoods and errors because someone can objectively disprove the article’s statements and claims. See Pacira, 63 F.4th at 246. So such statements can be actionable.

B. Defendants cannot win on truth yet

Defamation requires falsity. Defendants may raise truth as

an affirmative defense and prove that their statements were

true. 42 Pa. Cons. Stat. § 8343(b)(1). Courts may not dismiss

on this ground, though, unless a statement’s truth is apparent

on the complaint’s face. See Bethel v. Jendoco Constr. Corp.,

570 F.2d 1168, 1174 (3d Cir. 1978). Berlacher, Saba, and Simon

forfeited the truth defense at this stage by not raising it below. The District Court found that the Association’s and publisher’s statements were true, but in doing so mistakenly looked to the

second retraction notice, which defendants had attached as an

exhibit to their motion to dismiss. But the complaint was ambiguous; though the language it quoted appeared in both the first

and second retraction notices, the date it gave was that of the first retraction notice. Wang disputed that he was referring to the second retraction notice. Given that ambiguity, the District Court should not have relied on that exhibit, as it was “material beyond the complaint.” Dunn v. Castro, 621 F.3d 1196, 1204

n.6 (9th Cir. 2010).

13

Yet even if we read the complaint’s fuzzy language as incorporating the second retraction, making it fair game, the District Court erred by accepting truth as a defense. The defendants

identified only two of Wang’s statements as even potentially

misleading (and thus truthfully criticized as misquotations), yet neither bears the weight put on it.

The first of the “misquotes” unravels under scrutiny. App.

163. Wang’s article quoted a source stating that Ohio State University “simply made it a priority to rank [underrepresented in medicine] applicants more aggressively than in previous years,

thus achieving success in matching them regardless of recruiting efforts, with the implication being that [the school] accepted less competitive applicants in an effort to increase diversity.” Id. (first alteration in original). That quotation was accurate. Yet the Association objected that it was “taken out of context

and falsely implies that the Ohio State University (OSU) program ranked racial and ethnic groups underrepresented in medicines with lower academic credentials in order to achieve diversity.” Id. Even so, “[t]he full statement” identified by the Association still supports Wang’s reading: Underrepresented minorities, on average, scored slightly lower on the U.S. Medical

Licensing Exam. App. 163. And though the source cautioned

that the Licensing Exam scores are not perfect predictors of

success and that the observed difference was not statistically

significant, it admitted that the exam “scores are considered the most objective variable for comparison.” App. 205. Wang’s

quotation was fair, accurate, and supported his conclusion that medical-education programs were giving explicit racial preferences that arguably violated civil-rights law.

14

The second allegation of misquotation is closer to the mark.

Wang hypothesized that doctors from underrepresented minority groups do not improve medical access for underserved populations. He wrote that “a 1995 California study … demonstrated [that] primary care physicians who were not board certified were 1.6 times more likely to work in rural underserved

areas when compared with board-certified counterparts.” App.

420. That was not a quotation. But the substance of the paraphrase was off. The study actually found: “Non-board-certified

family and general physicians were 1.6 times more likely than

other non-board-certified primary care physicians to locate in

rural underserved areas.” App. 210.

No party pressed any other misquotations, misstatements,

or misrepresentations. Yet these two “misquotes” were presented as “[a] sample of the misconceptions and misquotes,”

falsely implying that there were many more. App. 163. There

were not “many” other misquotations—just one modest error.

App. 237. Defendants exaggerated greatly, and the District

Court should not have considered documents beyond those in

the complaint, so the truth defense cannot defeat Wang’s defamation claim at the dismissal stage.

C. Wang, a limited-purpose public figure, pleads actual

malice for some

Wang appears to be a limited-purpose public figure. He is

not famous. But he “voluntarily inject[ed] himself … into a

particular public controversy and thereby bec[a]me[ ] a public

15

figure for” discussing affirmative action in medicine. Gertz v. Robert Welch, Inc., 418 U.S. 323, 351 (1974).

As a public figure, Wang must have plausibly pleaded that

defendants spoke “with ‘actual malice’—that is, with

knowledge that [each statement] was false or with reckless disregard of whether it was false or not.” N.Y. Times Co. v. Sullivan, 376 U.S. 254, 280 (1964). That does not require malice in

the ordinary sense of ill will. Because defendants rarely admit lying or recklessness, “objective circumstantial evidence can

suffice to demonstrate actual malice.” Schiavone, 847 F.2d at

1090.

1. Wang pleads Berlacher’s and Saba’s actual malice. As

a threshold matter, Saba never raised his actual-malice argument before the District Court. But since the court ruled on that issue anyway, we address it and conclude that Wang properly

pleaded enough objective circumstances against both Berlacher and Saba. Recall the complaint’s allegations, which we

take as true, along with the documents integral to or explicitly relied on by the complaint. See In re Burlington Coat Factory

Sec. Litig., 114 F.3d 1410, 1426 (3d Cir. 1997). Wang published his peer-reviewed article in March 2020. Four months later,

doctors alerted hospital-system and university staff to Wang’s

article and views. Panicked, the doctors and staff “criticized

[Wang’s] conclusions.” App. 140. Berlacher and Saba quickly

reacted to their criticisms by tweeting and retweeting socialmedia posts accusing him of misinterpreting data, misquoting

sources, and racist thinking. App. 140, 141, 145. Yet Wang’s

seventeen-page article was supported by 108 endnotes. And

Wang alleges that the “citations and quotations in [his] article were clearly attributed, easily verified, and correct.” App. 150.

16

Mental states, including malice and intent, may be alleged

generally, as long as the plaintiff pleads enough supporting

facts to make inferring the mental state plausible. See Fed. R. Civ. P. 9(b); Biro v. Conde Nast, 807 F.3d 541, 544–45 (2d Cir. 2015). From Wang’s allegations, we could draw several plausible inferences. One is that Berlacher and Saba did not adequately investigate their claims that Wang had misquoted his

sources. That inference supports dismissal, for a mere “failure to investigate” is not enough to show actual malice. McCafferty v. Newsweek Media Grp., Ltd., 955 F.3d 352, 359 (3d Cir.

2020).

But it is just as plausible to infer that Berlacher and Saba

knew, or at least recklessly disregarded the possibility, that

their tweets were flat wrong. Several facts support that inference. First, the things that they said Wang got wrong were, as

Wang alleges, easy to verify. Berlacher and Saba could easily

have checked Wang’s sources, all readily accessible and

properly cited, and found that his citations and quotations were almost uniformly accurate. Second, the tweets themselves imply

that Berlacher and Saba did in fact check Wang’s sources. As

we explained above, some statements of opinion “give rise to

the inference that there are undisclosed facts that justify the … opinion.” Meyers, 221 A.3d at 670 (cleaned up). Alleging that

someone repeatedly “misquote[d]” his sources implies that the

speaker checked them. App. 145.

These facts distinguish Wang’s claims from most actualmalice cases, which usually turn on allegations that journalists did a shoddy job. See, e.g., McCafferty, 955 F.3d at 359; Biro, 807 F.3d at 546. Wang does not just allege that Berlacher and

Saba relied on bad sources (as in Biro) or failed to interview

17

people familiar with the subject (as in McCafferty). Nor does

he allege that Berlacher and Saba failed to “read every” relevant source on the question. Church of Scientology Int’l v. Daniels, 992 F.2d 1329, 1334 (4th Cir. 1993). Rather, Wang’s defamation claim fits squarely into the space left open by McCafferty. McCafferty had not alleged that the defendant “made up”

the allegedly defamatory material or “published facts contrary

to information it otherwise knew.” 955 F.3d at 360. Wang, by

contrast, alleges facts—especially Berlacher’s and Saba’s own

tweets—supporting an inference that they had checked Wang’s

citations and quotations and found them (with one minor exception) substantially accurate. Nonetheless, they trashed his article on Twitter, telling the public that Wang had repeatedly misquoted his sources, knowing that was “contrary” to the information they had. Id. A complaint alleging that a speaker directly contradicted information that he had can survive a motion to

dismiss. See id.; Tholen v. Assist Am., Inc., 970 F.3d 979, 984– 85 (8th Cir. 2020).

We could also infer that Berlacher and Saba intentionally

avoided learning the truth. “[P]urposeful avoidance of the

truth” shows actual malice. Harte-Hanks Commc’ns, Inc. v.

Connaughton, 491 U.S. 657, 692 (1989). When a speaker

chooses “not to listen” to an objective source that could verify a story, that choice supports finding actual malice. Id. at 684. Actual malice is even more apparent where facts show that a

speaker or publisher decided—before investigating—to run

with a particular narrative. See id.

Wang’s article was peer-reviewed. That means it “was vetted pursuant to the usual procedures for review of all articles in JAHA.” App. 143. That peer-review vetting process is designed

18

to weed out the egregious errors that Berlacher and Saba accused Wang of making. And the allegations plausibly suggest that

they “doubted the veracity of [their] story.” Tucker, 237 F.3d at 286–87. Given that, a cite check was the only thing that could

have confirmed what they tweeted. If they failed to check the

citations in the face of those doubts, then they “purposefully

avoided further investigation with the intent to avoid the truth.” Michel v. NYP Holdings, Inc., 816 F.3d 686, 703 (11th Cir.

2016).

Particularly given those concerns, the speed of Berlacher

and Saba’s stampede strongly suggests a reckless rush. Within

a couple of days of the first complaints, they demoted Wang.

Within two days of that, they lambasted him on social media.

No one seems to have paused to do a proper cite check; neither

of the alleged errors was a misquotation, and one was not even

wrong. Berlacher and Saba’s coordinated campaign of denunciation was thus a stark departure from normal academic practice. The allegations in the complaint suggest that the doctors were reacting to Twitter outrage, not reasoned scholarly inquiry backed by facts. This circumstantial evidence “permit[s] the

conclusion that the defendant[s] in fact entertained serious

doubts as to the truth of [their] publication[s].” St. Amant v. Thompson, 390 U.S. 727, 731 (1968).

To be sure, each inference is plausible. And two of them

suffice to let the defamation claims against Berlacher and Saba go forward. Because we must draw “all reasonable inferences”

in Wang’s favor at this stage, even if we thought he would

struggle to “ultimately prevail on the merits,” we would let his claim proceed. Connelly v. Lane Const. Corp., 809 F.3d 780,

790–91 (3d Cir. 2016).

19

2. Wang fails to plead Simon’s actual malice. The case

against Simon is inadequate. True, Simon, as lead author for

the journal’s editorial note, denounced Wang’s article as “a

misrepresentation of evidence” with “misrepresented facts.”

App. 195. He “deeply regret[ted]” publishing it and was “grateful to the medical community for assisting in correcting this

error.” Id. But he largely relied on the allegations of error that they had gotten from the hospital system, Wang’s employer.

Plus, Simon (unlike Berlacher and Saba) never accused Wang

of misquoting his sources—an accusation that would have

been verifiable with a simple cite check. The most he said was

that Wang had misrepresented the facts and evidence. Though

that statement (an opinion premised on undisclosed facts) can

still be defamatory, it is harder to disprove, making actual malice harder to allege. What is more, the complaint contains little suggesting that Simon rushed to retaliate against Wang without

verifying his suspicions, at least as compared to Berlacher and Saba. So Wang has failed to plausibly allege that Simon wrote

with actual malice.

3. The allegations against the publisher are likewise inadequate. Alleging that a publisher of another’s statement acted

with actual malice is hard. Wang must plausibly allege more

than that it radically departed from its usual standards. McCafferty, 955 F.3d at 359–60. He does not. The publisher just

passed on information from the Association, the hospital system, and its doctors, which it reasonably thought reliable.

Again, mere failure to investigate, on its own, is not enough.

Tucker, 237 F.3d at 286.

4. Wang pleads the Association’s actual malice. The closest call is about the Association. On the one hand, in places it

20

simply relayed the hospital’s and doctors’ statements, which it reasonably thought reliable. But the Association went further,

and it did so hastily. Within two days of being asked to retract the article, it issued not only a retraction, but also a press release in its own name and voice. It “denounce[d] the views expressed

in the article” as “contrary to our organization’s core values and historic commitment to promoting diversity and inclusion.”

App. 166. It called the criticism for “misrepresentation”

“right.” Id. And it went much further, accusing him of “deliberate misinformation or misrepresentation”—intent that it

could hardly know. App. 167. In its updated retraction statement, the Association did not merely pass along information as

a publisher; it expressly accused Wang of lying.

Nor did the Association notify Wang of any errors or give

him a chance to rebut those charges before publishing the accusations. That is questionable in the academy. Cf. Bd. of Regents v. Roth, 408 U.S. 564, 573 (1972) (explaining that an academic

might be entitled to “notice and an opportunity to be heard”

and “to refute the charge” when officials challenge his “reputation, honor, or integrity” (internal quotation marks omitted)). Although a departure from professional standards, standing

alone, does not prove actual malice, it may contribute to that

inference when coupled with inflammatory accusations and a

deficient investigation. Thus, the Association’s allegedly false charges of intentional deception—made hastily, without a thorough review, and without giving Wang a chance to rebut

them—are enough to plead actual malice. Cf. McCafferty, 955

F.3d at 359–60.

21

D. The university and hospital system may be vicariously

liable for defamation

Once the District Court found that Wang had failed to state

a defamation claim against the individual defendants, it reasoned that their employers could not be vicariously liable either. But with the defamation claim against two individuals back in

the case, vicarious liability is back on the table.

Employers can be liable for defamatory statements that

their employees make within the scope of their employment.

To be within the scope of their employment, conduct must be

the kind of action that the worker is hired to do, must happen

substantially within the authorized work time and space, and

must be at least partly motivated to serve the employer. Restatement (Second) of Agency § 228 (1958); Fitzgerald v. McCutcheon, 410 A.2d 1270, 1272 (Pa. Super. Ct. 1979).

Wang claims that Berlacher and Saba made their statements

as part of their employment by the university or hospital system. He plausibly alleges that, as professors and leaders of the cardiology department, it was within the scope of their employment to “accuse[ ] [Wang] of professional incompetence, and

academic dishonesty.” App. 149. But none of Wang’s factual

allegations names the physician practice or statements made on

its behalf, let alone plausibly alleges its actual malice or vicarious liability. So the District Court properly dismissed the defamation claim against the physician practice.

* * *

In sum, Wang plausibly alleged that Berlacher and Saba,

speaking for the university and hospital system, defamed him,

22

as did the Association. It is plausible that they either knew they were overclaiming the whole time, or at least buried their heads in the sand to avoid checking his citations and investigating

their doubts. That is enough for now. The New York Times v.

Sullivan actual-malice standard is supposed to be a shield for

free expression, not a sword to silence disfavored speech.

III. WANG’S CIVIL-RIGHTS RETALIATION CLAIMS SURVIVE

Three different civil-rights laws converge to protect Wang’s

calling out what he reasonably believed was illegal activity. At multiple stages, for multiple reasons, the District Court rejected this claim. But text and precedent support it.

No one may retaliate against an employee “because he has

opposed any practice made an unlawful employment practice

by” Title VII. 42 U.S.C. § 2000e-3(a). Section 1981 creates a

cause of action for “retaliation against a person who has complained about a violation of another person’s contract-related

‘right.’ ” CBOCS W., Inc. v. Humphries, 553 U.S. 442, 445, 452

(2008); 42 U.S.C. § 1981(a). Pennsylvania also bars “discriminat[ion] … against any individual because [he] has opposed

any practice forbidden by” the PHRA. 43 Pa. Cons. Stat.

§ 955(d). Despite some differences, Title VII, the PHRA, and

§ 1981 all require the same three elements for a prima facie case of retaliation: protected activity by Wang, adverse employment

action against him, and a causal connection between the two.

Est. of Oliva ex rel. McHugh v. New Jersey, 604 F.3d 788, 798

& n.14 (3d Cir. 2010); Wilkerson v. New Media Tech. Charter

Sch. Inc., 522 F.3d 315, 318, 320 (3d Cir. 2008).

Once a plaintiff makes out a prima facie case, defendants

may give legitimate, non-retaliatory reasons why they took

23

their adverse actions. McDonnell Douglas Corp. v. Green, 411

U.S. 792, 802 (1973). The plaintiff may then show that those

reasons are pretextual. Id. at 804. Reviewing de novo the District Court’s dismissal of Wang’s claims relating to his article and its grant of summary judgment on his claims relating to

comments made during the July 31 meeting, we hold that these

claims should go forward. Burtch v. Milberg Factors, Inc., 662

F.3d 212, 220 (3d Cir. 2011); Moore v. City of Phila., 461 F.3d 331, 340 (3d Cir. 2006), as amended (Sept. 13, 2006).

A. Wang suffered for criticizing his employers both in

his article and at the July 31 meeting

1. Wang opposed policies that he reasonably thought were

unlawful. Wang criticized what he believed was the hospital’s

and university’s illegal practice of giving racial preferences. He made arguments like the ones that the Supreme Court later

accepted in Students for Fair Admissions v. President & Fellows of Harvard College, 600 U.S. 181 (2023). That criticism

was protected. Moore, 461 F.3d at 341. Yet the District Court

disagreed because he had never called out the university or hospital system by name, and “advocacy, unconnected to employment practices,” is not enough. Curay-Cramer v. Ursuline

Acad., 450 F.3d 130, 136 (3d Cir. 2006). That was mistaken.

To start, it is not clear that an employee must oppose his

own employer’s illegal actions. All the statute seems to require is “oppos[ing] any practice made … unlawful.” 42 U.S.C.

§ 2000e-3(a) (emphasis added). We have not passed on that issue, nodding toward it only in dictum. Curay-Cramer, 450 F.3d at

135. And we need not decide it here because, in context,

Wang’s article plausibly opposed the university’s and hospital

24

system’s admissions and hiring practices. He worked for those

institutions. His article opposed “racial and ethnic preferences within graduate medical education training programs” industry-wide and nationwide. App. 415. He wrote that the accrediting bodies’ diversity requirements, which the university and

hospital system had to follow, may be illegal.

Discovery bore out Wang’s complaint. Gladwin read his article as criticizing affirmative-action practices, “[i]ncluding in graduate medical education” and the “selection of fellows in

cardiology” at the hospital system and physician practice. App. 1339–40. In a school-wide email, Gladwin decried it as “antithetical to our values.” App. 2764. Because these comments

“identif[ied] [Wang’s] employer and the practice—if not specifically, at least by context,” the District Court erred in finding that Wang’s claims in the article were not protected activity.

Curay-Cramer, 450 F.3d at 135.

At the July 31 meeting, Wang testified that he raised his

concerns about racial and ethnic preferences in “our [graduate

medical education] contracts with the fellows” at the hospital

system, and that he “just wanted us to follow the law.” App.

961–63. Saba told him, “you know what we’re trying to do

here, Norm.” App. 2490–91. Saba also told him “that the views

that [he had] expressed [we]re not in line with the views and

policies that they were trying to create.” App. 2498. They understood his criticism perfectly and took it personally. Because

there is a genuine dispute of material fact over whether Wang’s comments were protected conduct, the District Court erred in

granting summary judgment on Wang’s Title VII/PHRA/

§ 1981 retaliation claim.

25

2. The university and hospital system took non-trivial adverse actions against Wang. An adverse action is one that “well might have dissuaded a reasonable worker from making or supporting a charge of discrimination.” Burlington N. & Santa Fe

Ry. Co. v. White, 548 U.S. 53, 68 (2006) (cleaned up). As the

District Court recognized, the demotion and ban on contacting

fellows, residents, and students are enough for the hospital system. See id. at 68; Gilbert v. Des Moines Area Cmty. Coll., 495 F.3d 906, 917 (8th Cir. 2007).

Wang also alleged that his bosses and colleagues at the university and hospital had retaliated against his speech by harassing him to “isolate[ ] [him] in his work environment in hopes

that he would resign.” App. 104. Viral tweets called him racist. School-wide emails from leaders acting as both hospital and

university employees denounced him and his views as “antithetical to our values and deeply hurtful to many of our URM

faculty.” App. 2764.

His superiors at both the university and hospital were aware

of the environment. Saba “predict[ed] … that the social environment here will become untenable for Norm [Wang] and this

will lead to him leaving UPMC or to his reallocation to a peripheral site.” App. 597. The blowback was so intense that a colleague feared that Wang might kill himself, telling Berlacher:

“One thing that crossed my mind … [i]s the issue of his overall suicide risk .… it must not be easy to get all this push back and come back to work.” App. 3051–52 (second elision in original). With all that Wang endured, “a reasonable jury might well conclude that this pattern of harassment might dissuade a reasonable worker from bringing or supporting a charge of discrimination.” Moore, 461 F.3d at 348.

26

The multi-front reputation-bashing that impugned Wang’s

integrity as an academic is the sort of conduct that “might

[have] impair[ed]” Wang in current or future “employment situations,” and “well might have dissuaded a reasonable worker

from making or supporting a charge of” retaliation. Nelson v.

Upsala Coll., 51 F.3d 383, 387 (3d Cir. 1995) (first two quotations); Burlington, 548 U.S. at 68 (third one) (internal quotation marks omitted). Advocating for a paper’s retraction for academic misconduct can harm Wang’s “reputation in the academic community” by “diminish[ing] his academic reputation, thus denying him opportunities” for future employment and publications.

Wu v. Thomas, 863 F.2d 1543, 1545 (11th Cir. 1989) (upholding

professors’ claims of Title VII retaliatory harassment). An academic’s reputation is his livelihood, so the harm to Wang’s reputation is non-trivial.

The retraction itself could likewise dissuade a reasonable

academic from reporting discrimination. In academia, publications and citations are the coin of the realm. They affect salary, promotions, and job opportunities. It is “publish or perish,” so losing a publication is no small matter—not to mention the

stigma of a retraction. Theidon v. Harvard Univ., 948 F.3d 477, 483 n.4 (1st Cir. 2020).

In response, the university (though not the hospital system)

says Wang cannot impute its employees’ actions to it. But that

confuses § 1983’s state-action inquiry with Title VII’s. Under

Title VII, we routinely impute employees’ harassment to their

employers when senior employees know about it. See Huston

v. Procter & Gamble Paper Prods. Corp., 568 F.3d 100, 107

(3d Cir. 2009). And even with dual employers, under Title VII

we do not limit liability to the employer who had more control.

27

Faush v. Tuesday Morning, Inc., 808 F.3d 208, 215 (3d Cir.

2015). Wang raised a genuine dispute about whether Pitt “management knew or should have known about the harassment,

but” instead of stopping it, fanned the flames. Jensen v. Potter, 435 F.3d 444, 453 (3d Cir. 2006) (Alito, J.), overruled in part and on other grounds by Burlington, 548 U.S. at 60, 64. At the

very least, his superiors at the university (Saba and Gladwin)

knew of the retaliation and “failed to undertake prompt corrective measures within [their] control.” EEOC v. Glob. Horizons,

Inc., 915 F.3d 631, 641 (9th Cir. 2019) (quoting EEOC guidance). Because employees “sufficiently senior in [the university’s] governing hierarchy, or otherwise in a position of administrative responsibility” over Wang, knew of his harassment

and publicly accused him of academic misconduct, there is a

genuine dispute of material fact over whether the university is responsible for the adverse actions. Huston, 568 F.3d at 107.

3. There is enough for causation. As the District Court

rightly found, there are genuine disputes about causation and

possible legitimate, nondiscriminatory reasons for the adverse

actions. Title VII (and thus the PHRA and § 1981) requires only but-for causation. Univ. of Tex. Sw. Med. Ctr. v. Nassar, 570

U.S. 338, 360 (2013). That is easily met here, as Wang’s opposition to race-based discrimination led to everything that happened to him in the summer and fall of 2020. Thus, Wang’s

Title VII/PHRA claims against the hospital and University proceed. Because Wang’s § 1981 claim against Pitt was dismissed

before discovery even though it was sufficiently pleaded, we

will vacate the District Court’s dismissal of that claim and remand for further proceedings.

28

B. The Title VI claim survives against the hospital, but

not the university

Title VI outlaws discrimination (and by extension, retaliation) by recipients of federal funds. 42 U.S.C. § 2000d; 28

C.F.R. § 42.107(e) (protecting complainants from retaliation);

34 C.F.R. § 100.7(e) (same). Title VI retaliation claims are governed by the same framework as Title VII, PHRA, and § 1981

claims. See Woodson v. Scott Paper Co., 109 F.3d 913, 920 (3d

Cir. 1997); Hankins v. Temple Univ. (Health Scis. Ctr.), 829

F.2d 437, 441 (3d Cir. 1987). The university and hospital system both take federal money. To state a claim under Title VI,

though, Wang must also plead that “a primary objective of the

Federal financial assistance is to provide employment.” 42

U.S.C. § 2000d-3.

The District Court dismissed the Title VI claim against the

university for lack of retaliation; the reasoning above applies equally to revive this part of the claim. And it dismissed the

Title VI claim against the hospital system for failing to allege the required primary purpose. Yet Wang did allege that the hospital system “receives federal funds to employ residents and fellows in its residency, G[raduate] M[edical] E[ducation], and

fellowship programs.” App. 137. Medical residency straddles

the line between education and employment. Cf. Doe v. Mercy

Cath. Med. Ctr., 850 F.3d 545, 556, 559 (3d Cir. 2017) (finding it plausible that residency programs are education and residents are employees). At this stage, all Wang need do is allege that

the hospital system gets federal money to fund residents. He

barely clears that bar.

29

But his allegations against the university come up short. He

alleges only that it “receives federal funds that, among other

things, assist it in providing a medical education for students at” the school. App. 137 ¶ 6. That statement does not allege a

primary purpose of funding employment. 42 U.S.C. § 2000d-3.

The District Court dismissed Wang’s Title VI claim on other

grounds, so Wang will have an opportunity to amend.

IV. WANG’S FIRST AMENDMENT RETALIATION CLAIM FAILS

Wang also sued the university, hospital system, physicians

practice, Berlacher, Saba, and Gladwin under 42 U.S.C. § 1983

for violating his First Amendment rights. The District Court

dismissed the claims against the institutional defendants for

lack of state action and later granted summary judgment to Berlacher and Saba for the same reason and to Gladwin because

he was not involved in demoting Wang. Even if Wang could

make out a First Amendment retaliation claim, those rulings

were right.

A. Most defendants did not act on Pennsylvania’s behalf

Because the First Amendment does not restrict private parties, only “state actors” can be liable. Leshko v. Servis, 423 F.3d 337, 339 (3d Cir. 2005). People and business count as state actors when the state encourages or joins in the activity, or is “integrally related” to the actor, so that the “source fairly can be said to be the state.” Id. at 340. What is more, § 1983 has a tighter standard for vicarious liability than Title VII. Title VII follows agency law, often holding employers vicariously liable and

imputing supervisors’ actions to them. Moody v. Atl. City Bd.

of Educ., 870 F.3d 206, 215–16 (3d Cir. 2017). Yet for his

§ 1983 claim, Wang must show that “the state has exercised

30

control over the particular conduct that gave rise to the plaintiff’s alleged constitutional deprivation.” Kach v. Hose, 589

F.3d 626, 649 (3d Cir. 2009).

Wang has not made that showing. Though affiliated with

the public university, the hospital system and physician practice are private corporations. They, not the university, are responsible for their employees’ actions. See id. at 646. Their academic affiliation alone is not enough to make them liable; “[a]ction taken by private entities with the mere approval or acquiescence of

the State is not state action.” Am. Mfrs. Mut. Ins. Co. v. Sullivan, 526 U.S. 40, 52 (1999).

Wang’s allegations trying to link the hospital system’s and

physician practice’s actions to the university are conclusory.

True, the people who disciplined him and retaliated against him were employed by both the hospital system and university. Yet

dual employment is not enough for state action. See Borrell v.

Bloomsburg Univ., 870 F.3d 154, 160 (3d Cir. 2017). The key

question is whether each employee “was wearing his [hospitalsystem] hat or his [university] hat when he decided to” take

adverse action. Id. Berlacher, Saba, and Gladwin were all wearing their hospital-system hats. In dismissing Wang from his

hospital-system directorship and restricting his access to hospital-system residents and fellows, Berlacher and Saba acted as hospital-system employees. They testified that the decision to

remove Wang as director and bar him from contacting residents

and fellows was made without consultation with anyone from

the university. Berlacher’s and Saba’s tweets, as well as the

tweet from the hospital system’s account, neither stated the

university’s views nor were made on its behalf. See Lindke v.

Freed, 601 U.S. 187, 191 (2024) (treating social-media posts

31

as state action only when the official who posted had actual

authority to speak for the state and purported to do so on social media). And Berlacher did not even have supervisory authority

over Wang at the university.

Saba’s purported bar on Wang’s interacting with medical

students unaffiliated with the hospital system is a closer case. True, university officials knew about the supposed bar yet did

not lift it for two months. Even so, that is not enough for state action. To start, it was effectively a restriction in name only, as Wang never taught classes at the medical school, mostly teaching residents and fellows at the hospital. And Saba never purported to bar him on behalf of the university. Three days later, Berlacher and Saba sent another email from Saba’s hospitalemail address, telling Wang: “[W]e can no longer have you

serve in any medical education role in the institution, specifically pertaining to the general cardiology and electrophysiology fellows, while also including medical students and residents.” App. 861. In context, they were wearing their hospitalsystem hats. Saba had no authority to take away his university

privileges. He did not consult with the medical-school dean.

The dean later reversed Saba’s decision, telling Berlacher,

Saba, and Gladwin that “personnel decisions regarding Dr.

Wang at [the hospital] are not applicable to the University.” App. 912.

Nor is Gladwin responsible for that supposed bar. Though

he emailed the medical school that “[w]e have taken immediate

action and removed [Wang] from [his] leadership position,” he

said nothing about interacting with medical students and did

not in fact remove Wang from any positions. App. 2764. He,

Berlacher, and Saba all testified without contradiction that the

32

decision was made without Gladwin. Gladwin’s after-the-fact

defense of Berlacher’s and Saba’s actions is not enough to create state action, as he “played no affirmative part in depriving” Wang of his rights. Rizzo v. Goode, 423 U.S. 362, 377 (1976).

B. Wang’s claims against the university fail

That leaves the public university itself. We have called Pitt

an “instrumentalit[y] of the state, both in name and in fact.”

Krynicky v. Univ. of Pittsburgh, 742 F.2d 94, 103 (3d Cir.

1984). Thus, we have previously held, “the University of Pittsburgh [is a] state actor[ ] when [it] make[s] employment decisions.” Molthan v. Temple Univ. of Commw. Sys. of Higher

Educ., 778 F.2d 955, 961 (3d Cir. 1985). Yet the District Court and the parties treat it as a “municipal subdivision” subject to Monell liability. Frazer v. Temple Univ., 25 F. Supp. 3d 598,

607 (E.D. Pa. 2014); Monell, 436 U.S. at 691.

We need not decide whether that was right. Even if it was,

on the face of Wang’s complaint, the university is not plausibly liable under § 1983. That provision does not allow vicarious liability. Parkell v. Danberg, 833 F.3d 313, 330 (3d Cir. 2016). To hold the university liable, Wang must show either that its policy or custom directly caused his constitutional harm or that it knowingly acquiesced in constitutional violations. Id.; Monell, 436 U.S. at 694. He alleges neither. On policy or custom, Wang

does not argue that the university’s policies harmed him, but

implies just the opposite—that the employees violated the university’s policies by infringing on his academic freedom. Plus, the crucial decisions were made by Berlacher and Saba;

Gladwin and the medical-school dean just relayed and commended others’ actions. And though Wang argues that the either

33

the university “implemented” or the dean “ratified” Berlacher

and Saba’s bar on interacting with medical-school students, he

never says how. App. 324–25; City of St. Louis v. Praprotnik,

485 U.S. 112, 124 (1988). Although the medical-school dean

knew of the bar, it apparently applied only at the hospital-system and was decided by Wang’s superiors there. Neither

Gladwin nor the dean controlled Wang’s hospital-system employment, and Wang did not interact with medical students at the

university. In short, Wang’s conclusory allegations do not plausibly allege the university’s “participation in or actual

knowledge of and acquiescence in the wrongful conduct.”

Chavarriaga v. N.J. Dep’t of Corr., 806 F.3d 210, 222 (3d Cir.

2015). So the District Court properly rejected Wang’s First

Amendment retaliation claim.

* * * * *

The First Amendment is part “of a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open.” N.Y. Times v. Sullivan,

376 U.S. at 270. And civil-rights protections “serve as a spur

or catalyst to cause employers to self-examine and to selfevaluate their employment practices and to endeavor to eliminate, so far as possible, the last vestiges of discrimination.” McKennon v. Nashville Banner Publ’g Co., 513 U.S. 352, 358

(1995) (cleaned up).

Taking Wang’s allegations as true, Berlacher, Saba, the Association, the university, and the hospital system did not just exercise their own academic freedom; they defamed and retaliated

against Wang for calling out what he reasonably believed was

their own discrimination. His defamation claim may proceed

34

against those five, as may his civil-rights retaliation claims

against all appellees. And the District Court should grant him

leave to amend his Title VI claim against the university. But his First Amendment claim fails for lack of state action. We will

thus AFFIRM IN PART, VACATE IN PART, and REMAND to the District Court for further proceedings consistent with this opinion.

35

Counsel for Appellant

Shawn M. Rodgers

GOLDSTEIN LAW PARTNERS

John R. Renner [Argued]

Michael E. Rosman

Michelle A. Scott

CENTER FOR INDIVIDUAL RIGHTS

Counsel for Appellee Wiley Periodicals Inc.

Elizabeth Seidlin-Bernstein

Michael L. Berry [Argued]

BALLARD SPAHR

Counsel for Appellee American Heart Ass’n, Inc.

Howard A. Chajson

Steven L. Ettinger [Argued]

John C. Conti

Jeffrey J. Wetzel

DICKIE MCCAMEY & CHILCOTE

Counsel for Appellee Marc Simon

Bryon M. Chowka

Robert J. Ridge [Argued]

CLARK HILL

Counsel for Appellee University of Pittsburgh

Jeremy D. Engle

MARCUS & SHAPIRA

Jeremy W. Brinster

Brian Boynton [Argued]

Bruce M. Berman

Molly M. Jennings

WILMER CUTLER PICKERING HALE & DORR

Counsel for Appellees University of Pittsburgh Medical Center, University of Pittsburgh Physicians, Samir Saba, Mark Gladwin, and Kathryn Berlacher

Emilie R. Hammerstein

Taylor Brailey

Terrence H. Murphy

Theodore A. Schroeder [Argued]

LITTLER MENDELSON

37

SHWARTZ, Circuit Judge concurring in part, dissenting in

part.

I agree with the disposition my colleagues reached on

all claims except Wang’s defamation claims. I conclude the

defamation claims fail for two reasons. First, the challenged

statements are non-actionable opinions of academic research

and scientific methodology based on disclosed facts. Second,

even if those statements were actionable, Wang is a limited

purpose public figure, but his complaint does not allege facts

from which to infer that the challenged statements were made

with actual malice. Accordingly, I would affirm the District

Court’s order dismissing Wang’s defamation claims.

I1

Wang is a cardiologist affiliated with the University of

Pittsburgh Medical Center (“UPMC”) and the University of

1

The Majority’s fact section merges allegations from

the operative complaint (the “FAC”) with facts revealed during

discovery and characterizes some facts differently from my

view. Because the defamation claims were dismissed under

Federal Rule of Civil Procedure 12(b)(6), this fact section

relies only on the factual allegations and other materials that may be considered when evaluating such a ruling. “In

reviewing a motion to dismiss under Rule 12(b)(6), we treat as

true all well-pleaded facts in the complaint, which we construe in the ‘light most favorable to the plaintiff.’” Santomenno ex

rel. John Hancock Tr. v. John Hancock Life Ins. Co. (U.S.A),

768 F.3d 284, 290 (3d Cir. 2014) (citations omitted).

1

Pittsburgh School of Medicine (“Pitt”). 2 At UPMC, Wang

serves as a doctor and previously directed the fellowship

program in clinical cardiac electrophysiology. At Pitt, Wang

is a faculty member.

Wang wrote an article entitled “Diversity, Inclusion,

and Equity: Evolution of Race and Ethnicity Considerations

for the Cardiology Workforce in the United States of America

from 1969 to 2019” (the “Article”). App. 139 (FAC ¶ 20), 178-94. He submitted the Article to the Journal of the American

Heart Association (“JAHA”). JAHA is published by the

American Heart Association (“AHA”) and Wiley Periodicals,

Inc. 3 (“Wiley”). In March 2020, AHA and Wiley published

the Article in JAHA after the “normal review and vetting

process.” App. 139-40 (FAC ¶ 21).

The Article “trac[ed] the history of the use of race and

ethnicity as factors in determining admission into medical

schools, residency programs, and fellowships.” App. 139

(FAC ¶ 20). It “asserted that the medical profession had not

been successful in reaching its goals of increasing the

percentages of underrepresented races and ethnicities in the

medical profession generally, and cardiology in particular.”

App. 139 (FAC ¶ 20). The Article argued that, “to achieve

those goals,” “programs . . . applied different standards to

applications by members of underrepresented races and

ethnicities and raised questions about the legality,

2

UPMC is the sole owner of Defendant University of

Pittsburgh Physicians (“UPP”), which employs physicians on

behalf of UPMC, including Wang. I refer to both entities

collectively as “UPMC.”

3

AHA and Wiley are private corporations.

2

effectiveness, and wisdom of doing so.” App. 139 (FAC ¶ 20).

The Article contained 108 endnotes with embedded

hyperlinks.

In late July 2020, “various individuals on social media”

criticized the Article, App. 140 (FAC ¶ 22), including Dr.

Samir Saba and Dr. Katheryn Berlacher, who are both

affiliated with UPMC and Pitt. On August 2, 2020, Saba

retweeted a Twitter post about the Article and wrote the Article “uses misquotes, false interpretations, and racist thinking to

defend a single person’s conclusion[s]. We are outraged that

@JAHA_AHA published this shameful and infuriating piece.”

App. 145 (FAC ¶ 43) (quoting App. 233). Later that day,

Berlacher retweeted a Twitter post about the Article and added

that it “misinterprets data and misquotes people” and is

“scientifically invalid and racist.” App. 145 (FAC ¶ 44)

(quoting App. 231, 235). 4

On August 5, 2020, AHA and Wiley retracted the

Article “[d]espite the fact that [it] was vetted pursuant to the usual procedures for review of all articles in JAHA.” App. 143

(FAC ¶ 39). The Article remains available on JAHA’s website

but contains a label noting its retracted status. In connection with the retraction, Wang contends that AHA and Wiley made

three defamatory statements against him. First, AHA and

4

Later on August 6, Saba emailed UPMC employees

and “asserted that the Dean of UPSOM, Dean Anantha

Shekhar, sought the retraction of [the Article].” App. 145

(FAC ¶ 43). Wang initially claimed this statement was

defamatory, see App. 150 (FAC ¶ 74), but omitted it from his

list of defamatory claims in his briefing, thereby abandoning it, see Appellant’s Br. at 35-37.

3

Wiley issued a statement on August 5 that accompanied the

Article’s retraction (the “Retraction Statement”). It stated, in relevant part:

The author’s institution, the University of

Pittsburgh Medical Center (UPMC), has notified

the Editor-in-Chief that the article contains many

misconceptions and misquotes and that together

those inaccuracies, misstatements, and selective

misreading of source materials strip the paper of

its scientific validity. . . . The Editor-in-Chief

deeply regrets publishing the article and offers

his apologies.

App. 143 (FAC ¶ 39), 237. The Retraction Statement did not

provide examples of misconceptions or misquotes from the

Article but noted JAHA would “be publishing a detailed

rebuttal.” App. 237. The next day, AHA and Wiley posted an

updated Retraction Statement (the “Updated Retraction

4

Statement”) 5 with virtually the same paragraph 6 as well as two specific “misconceptions and misquotes identified in

correspondence from UPMC” to serve as a “sample” of the

Article’s errors. 7 App. 163-64.

5

Even though the Updated Retraction Statement was

not explicitly mentioned in the complaint, we may consider it

because it: (1) was explicitly referenced in a document

referenced in the complaint, (2) was attached to the motion to

dismiss, and (3) is undisputedly authentic and integral to the

complaint. See Buck v. Hampton Twp. Sch. Dist., 452 F.3d

256, 260 (3d Cir. 2006) (recognizing that courts can consult

“documents that are attached to or submitted with the

complaint, and any matters incorporated by reference or

integral to the claim, items subject to judicial notice, matters of public record, orders, [and] items appearing in the record of the case” (alteration in original) (citations and quotation marks

omitted)); Schmidt v. Skolas, 770 F.3d 241, 249 (3d Cir. 2014)

(“[A] document integral to or explicitly relied upon in the

complaint may be considered without converting the motion to

dismiss into one for summary judgment.” (citations and

internal quotation marks omitted)); see also In re Burlington

Coat Factory Sec. Litig., 114 F.3d 1410, 1426 (3d Cir. 1997)

(“Plaintiffs cannot prevent a court from looking at the texts of the documents on which its claim is based by failing to attach

or explicitly cite them.”).

6

The Updated Retraction Statement uses “void” rather

than “strip.” App. 163, 237.

7

One example claimed that Wang “falsely implie[d]

that the Ohio State University (OSU) program ranked racial

and ethnic groups underrepresented [classified as

underrepresented minorities or “URM”] in medicines with

5

Second, AHA issued a statement (the “AHA

Statement”), providing, in relevant part:

The [Article] has rightfully drawn criticism for

its misrepresentations and conclusions. As an

organization focused on the relentless pursuit of

longer, healthier lives for everyone everywhere,

the [AHA] denounces the views expressed in the

article and regrets its role in enabling those views

to be promoted. Those views are a

misrepresentation of the facts and are contrary to

our organization’s core values and historic

commitment to promoting diversity and

inclusion in medicine and science. . . . While

[JAHA] and the other AHA scientific journals

are editorially independent of the Association,

lower academic credentials in order to achieve diversity,”

App. 163, but JAHA noted that a complete recitation of the

source Wang relied on for the statement said that “USMLE

score comparison shows no significant difference between

URM and non-URM fellows matched to the OSU program,”

App. 163.

The second example claimed Wang “falsely implie[d]

that working in an underserved area may be attributable to [an] inability to secure a job in other areas because of low

professional qualifications” based on a study comparing boardcertified doctors to non-board-certified doctors. App. 163-64.

The Updated Retraction Statement explained that the source

showed that the study compared two different groups of nonboard-certified doctors.

The Updated Retraction Statement included the sources

it relied upon in endnotes with embedded hyperlinks.

6

we take our responsibility to ensure factual

accuracy seriously. The independent editors of

JAHA and the [AHA] are reviewing the journal’s

peer-review and publication processes to ensure

future submissions containing deliberate

misinformation or misrepresentation are never

published. The journal can and will do better.

App. 166-67; see also App. 144 (FAC ¶ 39).

Third, JAHA published another article entitled “A Path

Forward,” which “stated that [Wang’s] institution had

requested the article be retracted ‘[based on] specific scientific errors as well as misleading and incomplete quotations.’” App.

144 (FAC ¶ 39) (quoting App. 170-71).

Separately, Wang also identified as defamatory an

article published by JAHA and authored by Dr. Marc Simon, a

UPMC and Pitt colleague. This piece, entitled “Equity,

Diversity, and Inclusiveness in Cardiovascular Healthcare and

Medicine,” (the “Simon Article”), App. 145 (FAC ¶¶ 45), 195-96, claimed that the Article “misrepresented facts” and

“evidence” “[in its efforts] to argue against affirmative action in the field of cardiology,” App. 145 (FAC ¶ 45) (quoting App.

195). The Simon Article’s sources were set forth in endnotes

with hyperlinks.

Wang filed defamation claims against Defendants,

alleging that: (1) Saba and Berlacher defamed him in their

August 2 tweets, (2) Simon defamed him in the Simon Article,

(3) UPMC and Pitt defamed him via their employees’

statements, (4) AHA defamed him in the Retraction Statement,

the AHA Statement, and A Path Forward, and (5) Wiley

7

defamed him in the Retraction Statement and A Path Forward.

II 8

In assessing a defamation claim, a “court must

determine: (1) whether the defendants have harmed the

plaintiff’s reputation within the meaning of state law; and (2) if so, whether the First Amendment nevertheless precludes

recovery.” Marcone v. Penthouse Int’l Mag. for Men, 754 F.2d

1072, 1077 (3d Cir. 1985) (citations and quotations omitted).

A

To succeed on a defamation claim under Pennsylvania

law, a plaintiff must show, among other things, “[t]he

defamatory character of the communication” at issue. 42 Pa.

8

We exercise plenary review of a district court’s order

granting a motion to dismiss for failure to state a claim. Burtch v. Milberg Factors, Inc., 662 F.3d 212, 220 (3d Cir. 2011).

To survive a motion to dismiss, a complaint must

contain sufficient factual matter, accepted as

true, to state a claim to relief that is plausible on

its face. A claim has facial plausibility when the

plaintiff pleads factual content that allows the

court to draw the reasonable inference that the

defendant is liable for the misconduct alleged.

Sheridan v. NGK Metals Corp., 609 F.3d 239, 262 n.27 (3d

Cir. 2010) (citations and quotations omitted); see also James v. City of Wilkes-Barre, 700 F.3d 675, 679 (3d Cir. 2012) (“[W]e

disregard rote recitals of the elements of a cause of action, legal conclusions, and mere conclusory statements.”).

8

Cons. Stat. § 8343(a)(1). 9 “A communication is defamatory if

it tends so to harm the reputation of another as to lower him in the estimation of the community or to deter third persons from

associating or dealing with him.” Birl v. Philadelphia Elec.

Co., 167 A.2d 472, 475 (Pa. 1960) (citations omitted); Graboff

v. Colleran Firm, 744 F.3d 128, 136 (3d Cir. 2014). Statements

of opinion, as opposed to statements of fact, may be viewed as

having such an effect if they are based on undisclosed facts.

Remick v. Manfredy, 238 F.3d 248, 261 (3d Cir. 2001); Parano

v. O’Connor, 641 A.2d 607, 609 (Pa. Super. Ct. 1994) (holding

that “expressions of . . . opinion” were “not actionable unless they imply undisclosed, false and defamatory facts”).

In the course of academic debate, criticisms and

characterizations of scientific methodology and scholarly

analysis are better understood as statements of opinion rather

than fact. See ONY, Inc. v. Cornerstone Therapeutics, Inc.,

720 F.3d 490, 497 (2d Cir. 2013) (“[W]hile statements about

contested and contestable scientific hypotheses constitute

assertions about the world that are in principle matters of

verifiable ‘fact,’ for purposes of the First Amendment and the

laws relating to fair competition and defamation, they are more closely akin to matters of opinion, and are so understood by the relevant scientific communities.”); Pacira BioSciences, Inc. v.

9

A Pennsylvania defamation plaintiff must also show:

(1) the statement’s publication by the defendant, (2) the

statement’s application to the plaintiff, (3) the understanding by the recipient of its defamatory meaning, (4) the

understanding by the recipient of it as intended to be applied to the plaintiff, (5) special harm resulting to the plaintiff from its publication, and (6) the lack of a conditionally privileged

occasion. 42 Pa. Cons. Stat. § 8343(a).

9

Am. Soc’y of Anesthesiologists, Inc., 63 F.4th 240, 246 (3d

Cir. 2023) (determining that “tentative scientific conclusions” were opinions in libel claim); Saad v. Am. Diabetes Ass’n, 123

F. Supp. 3d 175, 178 (D. Mass. 2015) (holding that medical

journal’s expression of concern about “the reliability of the

data” in professor’s articles was protected opinion).

The Majority holds that accusations of “misquoting and

misreading data and sources” are “falsifiable,” thereby

rendering them “more hit job than academic argument.”

Majority at 11. Setting aside the Majority’s characterization of these statements as a “hit job,” one academic accusing another

of “misreading” data is not actionable because it reflects an

opinion about the other’s work based on disclosed facts. See

Pacira, 63 F.4th at 247 (“[M]ere disputes about the reliability of a scientific study’s disclosed methodology cannot create an

actionable falsehood . . . as such disputes do not address

whether the statements themselves are verifiable.”); ONY, 720

F.3d at 497 (declining to recognize a cause of action where

“plaintiff allege[d] that the inferences drawn from . . . data were the wrong ones”).

That is what happened here. First, AHA and Wiley

criticized the Article for “contain[ing] many misconceptions

and misquotes and that together those inaccuracies,

misstatements, and selective misreading of source materials

strip the paper of its scientific validity.” App. 143, 163-65,

237. 10 Likewise, the Simon Article stated that the Article

“misrepresented facts in its effort[] to argue against affirmative

10

The AHA Statement and A Path Forward made

similar claims.

10

action in the field of cardiology” and included the

“misrepresentation of evidence.” App. 145, 195-96.

In making these claims, each author explicitly referred

to the Article—the subject of their criticism—and discussed

the bases for their opinions, namely the sources listed in the

Article’s endnotes. Indeed, the Updated Retraction Statement

even identified specific errors as a “sample of [Wang’s]

misconceptions and misquotes.” App. 163.

Thus, JAHA’s readers—medical professionals and

academics equipped to evaluate the challenged statements and

the sources on which they were based—had the opportunity

and information to assess for themselves the disputed

conclusions. See Pacira, 63 F.4th at 249 (determining

statements were based on disclosed facts where “[academic]

journal’s readers were provided the basis for the statements,

have the expertise to assess their merits based on the disclosed data and methodology, and thus are equipped to evaluate the

opinions the authors reached”); Cheng v. Neumann, 51 F.4th

438, 447 (1st Cir. 2022) (concluding an opinion was based on

disclosed facts and thus “not actionable because [it] provide[d] a link to the source material, which, in the context of the

[controversy], enable[d] readers to draw their own conclusions

‘based on facts accessible to everyone’” (citations omitted)).

The Majority insists “[i]t would have been just fine to

say that Wang’s article was wrong, harmful, or even

dangerous” because “[t]hose are normal academic opinions.”

Majority at 12. That is exactly what the speakers did here. In

any event, whether “normal” or not, courts should be cautious

when weighing in on academic debates because doing so chills

scientific progress and the rigorous discourse required to bring

11

it about. See Joseph v. Springer Nature, No. 20 civ. 4672

(JPC), 2021 WL 1372952, at *7 (S.D.N.Y. Apr. 12, 2021)

(declining to consider a retraction statement as actionable

where the underlying dispute concerned “a critique of a

scientific work” in a contested area of research “that the Court has no business solving”), aff’d sub nom. Joseph v. Springer

Nature Am. Inc., No. 21-959-cv, 2021 WL 6105369 (2d Cir.

Dec. 21, 2021) (summary order). As the Court of Appeals for

the Seventh Circuit held, “scientific controversies must be

settled by the methods of science rather than by the methods of litigation. . . . More papers, more discussion, better data, and more satisfactory models—not larger awards of damages—

mark the path toward superior understanding of the world

around us.” Underwager v. Salter, 22 F.3d 730, 736 (7th Cir.

1994); see also Pacira, 63 F.4th at 247-48 (holding that a

defendant’s characterization of academic work as employing a

“flawed method” was not actionable because “data and

methodology may be the basis of future scholarly debate, but

they do not form the basis for trade libel under New Jersey law” as concluding “otherwise would risk ‘chilling’ the natural

development of scientific research and discourse”). 11

11

Of course, academic disputes are not always beyond

the scope of defamation claims. Indeed, there is a marked

distinction between criticizing an academic’s work and the

academic themselves. As the Court of Appeals for the Seventh

Circuit observed, “[i]f a professor is falsely accused of

plagiarism or sexual harassment or selling high grades or other serious misconduct, rather than of having unsound ideas, he

has the same right to damages as any other victim of

defamation.” Dilworth v. Dudley, 75 F.3d 307, 310 (7th Cir.

1996).

12

Second, Saba and Berlacher’s tweets are also nonactionable opinions about an academic work based on

disclosed facts because they explicitly refer to the Article they critiqued and the reasons for their views. See Veno v.

Meredith, 515 A.2d 571, 573-75 (Pa. Super. Ct. 1986)

(concluding that an article constituted non-actionable opinion

because it “merely stated an opinion” on another article and the defendant author did not “imply that he was privy to some

additional facts, beyond [the article] which supported his

opinion”). Furthermore, although their characterization of the

Article as “racist” may be inflammatory, it is not actionable

under Pennsylvania law. See McCafferty v. Newsweek Media

Grp., Ltd., 955 F.3d 352, 358 (3d Cir. 2020) (“In Pennsylvania, ‘a simple accusation of racism’ is not enough . . . Rather, the accusation must imply more, by for instance suggesting that

the accused has personally broken the law to ‘act[ ] in a racist manner.’” (second alteration in original) (quoting MacElree v.

Philadelphia Newspapers, Inc., 674 A.2d 1050, 1055 (Pa.

1996)); 12 see also McAndrew v. Scranton Republican Publ’g

12

The Majority relies on MacElree v. Philadelphia

Newspapers, Inc., 674 A.2d 1050, 1055 (Pa. 1996). There, the

court determined that calling a local district attorney “the

David Duke of Chester County” was actionable because it

“could be construed to mean that [he] was acting in a racist

manner in his official capacity as district attorney” by “abusing his power . . . to further racism and his own political

aspirations.” Id. at 1054-55. In reaching this holding,

however, the court made clear “[t]his is not to say that every

accusation of racism is defamatory.” Id. at 1055.

Interpreting MacElree, this Court recognized that an

accusation of racism absent “specific, unlawful wrongdoing”

13

Co., 72 A.2d 780, 783 (Pa. 1950) (“It would doubtless be very

annoying for a man to be charged with bigotry of any kind, yet

it has been held that for a man to be charged with such bigotry is not defamation.”); Rybas v. Wapner, 457 A.2d 108, 110 (Pa.

Super. Ct. 1983) (“A publication which charges that an

individual is actuated by unpleasant or undesirable prejudice

may offend his sensitivities, but is not thereby libelous.”).

Because Defendant’s statements are nonactionable

opinions, I would affirm the order dismissing Wang’s

defamation claims. 13

was not actionable. McCafferty v. Newsweek Media Grp.,

Ltd., 955 F.3d 352, 358 (3d Cir. 2020). I apply this same logic to conclude that the accusations that Wang’s article were

“racist” are insufficient to support a cause of action given the lack of any assertion that he engaged in unlawful conduct.

13

Although the allegedly defamatory statements are

best understood as opinions, even if interpreted as factual, they are truthful statements about the accuracy of certain matters

Wang asserted, and are thus not actionable for three reasons.

First, Wang cited a paper discussing admission standards for

URM applicants at OSU. He quoted the paper as stating:

[W]e simply made it a priority to rank [URM]

applicants more aggressively than in previous

years, thus achieving success in matching them

regardless of recruiting efforts, with the

implication being that we accepted less

competitive applicants in an effort to increase

diversity.

App. 414 (second alteration in original). After this quote,

Wang added: “[e]ncouraging the explicit use of race and

14

ethnicity for employment reveals a lack of knowledge

regarding legal permissibility and fellow status.” App. 414.

The Retraction Statement highlighted that Wang omitted the

sentence immediately following the quoted excerpt, which

provided:

Although the measurement of an applicant’s

achievement reflected in his or her application

and, by extrapolation, the applicant’s potential

for being a successful fellow, cannot be fairly

quantified, USMLE score comparison shows no

significant difference between URM and nonURM fellows matched to the OSU program.

App. 163. In other words, the Article suggested that the

original work stood for the proposition that less qualified

candidates were accepted into the OSU program in the name

of diversity, but the original work stated that there was no

significant difference in test scores (i.e., qualification) between candidates from different racial groups.

Second, in a section of the Article relating to the

qualification of URM, Wang suggested another researcher’s

hypothesis that “working in an underserved area may be

attributable to inability to secure a job in other areas because of low professional qualifications” was bolstered by:

a 1995 California study that demonstrated

primary care physicians who were not board

certified were 1.6 times more likely to work in

15

rural underserved areas when compared with

board-certified counterparts.

App. 163-64, 420. However, as the Updated Retraction

Statement noted, this 1995 study in fact stated:

Among physicians completing postgraduate

medical education after 1974, board-certified

family physicians were [three] times more likely

to locate in medically underserved rural

communities than were other primary care

physicians. Non-board-certified family and

general physicians were 1.6 times more likely

than other non-board-certified primary care

physicians to locate in rural underserved areas.

App. 164. In other words, the original source compared rural

employment rates for non-board-certified family and general

physicians for non-board-certified primary care physicians, but Wang described it as comparing rural employment rates

between board-certified and non-board-certified primary care

physicians. While Wang did not include quotation marks, he

(incorrectly) described the results of a study and, at the very least, misrepresented what it demonstrated.

Third, the Retraction Statement’s assertion of “many”

errors does not change the truth-as-a-defense analysis. The use of the word “many” could change the meaning of a statement

such that it is defamatory—for example, one alleged act of

misconduct may be less damaging than fifty allegations,

depending on the severity of that single act. However,

“defamation case law affords leeway in the reporting of

numbers if the discrepancy does not alter the gist of what is

being reported and does not produce a different effect in the

mind of the reader.” Croce v. N.Y. Times Co., 345 F. Supp.

16

B

1

Even if any of Defendants’ statements were actionable

under state law, Wang’s claims are still prohibited under the

First Amendment’s public figure doctrine. Like my

colleagues, I conclude that Wang is a limited purpose public

figure and failed to allege actual malice as to Defendants Wiley

3d 961, 987 n.5 (S.D. Ohio 2018), aff’d, 930 F.3d 787 (6th Cir. 2019). Indeed, Pennsylvania courts recognize that “[t]he law

does not require perfect truth, so long as any inaccuracies do

not render the substance and ‘gist’ of the statements untrue. . . . The ‘gist’ of a statement is true if the effect upon a reader is the same regardless of the inaccuracy.” ToDay’s Hous. v.

Times Shamrock Commc’n, Inc., 21 A.3d 1209, 1215 (Pa.

Super. Ct. 2011); Rubin v. CBS Broad. Inc., 170 A.3d 560, 565

(Pa. Super. Ct. 2017) (“Substantial truth ‘absolve[s] a

defendant even if she cannot justify every word of the alleged

defamatory matter; it is sufficient if the substance of the charge be proved true, irrespective of slight inaccuracy in the details.’” (alteration in original) (quoting Masson v. New Yorker Mag.,

Inc., 501 U.S. 496, 516-17 (1991))). The gist of the statement

remains accurate because it conveyed that more than one

statement mischaracterized its source.

Thus, if the alleged defamatory statements are viewed

as factual rather than opinions, then truth as a defense would

apply to all of Wang’s defamation claims. See Budhun v.

Reading Hosp. & Med. Ctr., 765 F.3d 245, 259 (3d Cir. 2014)

(“While ordinarily a party may not raise affirmative defenses

at the motion to dismiss stage, it may do so if the defense is

apparent on the face of the complaint.” (citation omitted)).

17

and Simon, 14 but I part company with their conclusion that the pleadings are sufficient to establish AHA, Saba, and

Berlacher’s allegedly defamatory statements were made with

actual malice.

As a limited purpose public figure, Wang had to plead

that the allegedly defamatory statements were made with

“actual malice.” Marcone, 754 F.2d at 1081-82. “‘Actual

malice’ is a term of art that does not connote ill will or

improper motivation.” McCafferty, 955 F.3d at 359; Masson

v. New Yorker Mag., Inc., 501 U.S. 496, 510 (1991) (“Actual

malice under the New York Times standard should not be

confused with the concept of malice as an evil intent or a

motive arising from spite or ill will.”). Instead, “[t]he actual malice standard requires that the defendant make the

defamatory remarks with ‘knowledge that [the statement] was

false or with reckless disregard of whether it was false or not.’” McDowell v. Paiewonsky, 769 F.2d 942, 951 (3d Cir. 1985)

(alteration in original) (quoting N.Y. Times Co. v. Sullivan,

14

A limited public figure is typically one who “actively

participates in the public issue in a manner intended to obtain attention.” Marcone v. Penthouse Int’l Mag. for Men, 754 F.2d

1072, 1083 (3d Cir. 1985); Schiavone Constr. Co. v. Time,

Inc., 847 F.2d 1069, 1077 (3d Cir. 1988) (“[A]lthough they are

not well known throughout the country or on every issue,

[they] are nonetheless sufficiently involved in one particular

arena to qualify as public figures for that purpose.”). “For

example, a meat producer that aggressively advertised its

product in the media becomes a limited purpose public figure

for comment on the quality of its product.” Marcone, 754 F.2d

at 1084 (citing Steaks Unlimited, Inc. v. Deaner, 623 F.2d 264, 273-74 (3d Cir. 1980)).

18

376 U.S. 254, 280 (1964)). “[R]eckless disregard for the truth

means that the defendant ‘in fact entertained serious doubts as to the truth’ of the statement,” id. (quoting St. Amant v.

Thompson, 390 U.S. 727, 731 (1968)), “or that the defendant

had a ‘subjective awareness of probable falsity,’” id.

(quoting Gertz v. Robert Welch, Inc., 418 U.S. 323, 335 n.6

(1974)). This “standard is a subjective one, based on the

defendant’s actual state of mind,” though a court may rely on

“objective circumstantial evidence” to determine actual

malice. Schiavone Constr. Co. v. Time, Inc., 847 F.2d 1069,

1089-90 (3d Cir. 1988); see also Kendall v. Daily News Publ’g

Co., 716 F.3d 82, 93 (3d Cir. 2013) (“The subjective nature of

this inquiry requires that there be some evidence showing,

directly or circumstantially, that the defendants themselves

understood the potential defamatory meaning of their

statement.”).

Actual malice exists where a “publisher’s allegations

are so inherently improbable that only a reckless man would

have put them in circulation,” or “where there are obvious

reasons to doubt the veracity of the informant or the accuracy

of his reports.” St. Amant, 390 U.S. at 732; see also Schiavone, 847 F.2d at 1090 (“Where the defendant finds internal

inconsistencies or apparently reliable information that

contradicts its libelous assertions, but nevertheless publishes those statements anyway, the New York Times actual malice

test can be met.”). For example, in McCafferty we reasoned

that actual malice could be established where a publisher

“made up [a commentator’s] statements, used them to accuse

him of breaking the law when it knew he was innocent, or

published facts contrary to information it otherwise knew.”

955 F.3d at 360. In Tucker v. Fischbein, we recognized actual

malice where an attorney publicly mischaracterized a filed

19

complaint and only later attempted to explain his false

statements by claiming that he had failed to read the complaint before speaking on it. 237 F.3d 275, 285 (3d Cir. 2001). In

Schiavone, we concluded the defendant acted with actual

malice where reliable, independent sources contradicted the

veracity of a defamatory claim, and the publisher deliberately

omitted “exculpatory” information and “add[ed] editorial

comment to draw attention” from readers. 847 F.2d at 1091-92.

Absent circumstances that indicate a publisher was or

should have been on notice of the falsity of their statements,

we have generally declined to conclude that defamatory

statements were made with actual malice. See, e.g., Marcone,

754 F.2d at 1091 (concluding no actual malice was shown

where false statements were not based on facts so “inherently

improbable that [publisher] should have been put on notice of

the probability of falseness”).

2

Applying this framework to Wang’s factual allegations,

and assuming that Defendants’ statements are false and

otherwise defamatory, Wang’s allegations do not show that

any of the statements were made with knowledge or reckless

disregard of their falsity. See Marcone, 754 F.2d at 1091

(“Although the lack of truthfulness of a statement is a

prerequisite to liability, mere falsity, without more, is generally not sufficient to establish actual malice.”). Because I agree

20

with the Majority that Wang failed to establish actual malice

as to Wiley and Simon, I will not discuss those claims. 15

As to AHA, the Majority holds that Wang alleged actual

malice by relying on: (1) assertions in the AHA Statement, and

(2) AHA’s failure to give Wang an opportunity to respond.

Majority at 20-21. Neither establishes actual malice.

First, the AHA Statement—“denounc[ing]” Wang’s

views and claiming he engaged in “deliberate misinformation

or misrepresentation,” App. 166-67—was strongly worded,

but not sufficient to show actual malice. Passionate opposition to a speaker’s views does not constitute actual malice, even if motivated by “spite or ill will.” Masson, 501 U.S. at 510.

Moreover, AHA’s claim that Wang relied on “deliberate”

falsehoods does not show that it acted with knowledge of

falsity or reckless disregard for truth. Rather, it is reasonable for AHA to have surmised that if Wang misrepresented sources

15

I would not grant Wang leave to amend his

defamation claim against AHA because (1) he already had an

opportunity to amend the complaint as to this allegation, and

(2) AHA was passing along the hospital’s views, thus

undermining an actual malice determination, see, e.g., Hatfill

v. N.Y. Times Co., 532 F.3d 312, 325 (4th Cir. 2008) (no actual malice where defendant relied on source “considered an expert

in [her] field”), and hence amendment would be futile.

21

in making his argument, he did so “deliberate[ly]” to bolster

his conclusions. 16

Second, that AHA did not to provide Wang with notice

of the retraction or an opportunity to rebut the claims against him does not constitute actual malice. Indeed, we have

previously determined no actual malice existed where a

publisher failed to offer the subject of a news report the

opportunity to respond. McCafferty, 955 F.3d at 359. Even if

Wang had alleged that giving an author notice and an

opportunity to respond was standard practice in the industry

(which he did not), “an extreme departure from professional

16

Wang characterized the AHA Statement as arguing

the Article “was factually inaccurate, affirming it ‘take[s] [its] responsibility to ensure factual accuracy seriously [in the

publication of the JAHA]. AHA assured the public and the

medical community that ‘[t]he Journal can and will do

better[,’] in the future, to prevent the publication of other

articles like Plaintiff’s article, which ‘contain[ed] deliberate misinformation’ and ‘misrepresentation,’” App. 144 (FAC

¶ 39) (quoting App. 166-67 (alterations of quoted material in

original)), but this distorts the AHA Statement. For example,

the AHA Statement provides that it was “reviewing the

journal’s peer-review and publication processes to ensure

future submissions containing deliberate misinformation or

misrepresentation are never published.” App. 167. It does not

necessarily characterize the Article as an example of

“deliberate misinformation and misrepresentation,” as Wang

pleaded, App. 144 (FAC ¶ 39). Even with Wang’s framing of

this statement, however, his claim still fails because AHA’s

statement does not show that it acted with knowledge of falsity or reckless disregard for truth.

22

standards, without more, will not support a finding of actual

malice.” Tucker, 237 F.3d at 286 (citing Harte-Hanks

Commc’ns, Inc. v. Connaughton, 491 U.S. 657, 665 (1989)). 17

Wang also failed to plead actual malice as to Saba and

Berlacher. Wang contends that actual malice was shown by:

(1) the Article’s easily verifiable assertions, (2) Saba and

Berlacher’s strong opposition to Wang’s position on racebased admissions, and (3) their orchestrated “attack campaign”

against Wang, App. 143 (FAC ¶ 38). Even if true, none

constitute actual malice.

First, Saba and Berlacher’s alleged failure to review all

of the sources in Wang’s article does not demonstrate actual

malice. See Marcone, 754 F.2d at 1089 (“Failure to

investigate, without more, does not demonstrate actual

malice.”); see also Harte-Hanks, 491 U.S. at 688 (“[F]ailure to investigate before publishing, even when a reasonably prudent

person would have done so, is not sufficient to establish

reckless disregard.”). “[T]he purposeful avoidance of the truth is” distinguishable from a mere failure to investigate, HarteHanks, 491 U.S. at 692, but this case is not one in which the

publisher of the statement failed to interview a key witness or necessarily did not consult other objective evidence. Id. at 692-93. Furthermore, the complaint does not plead facts to suggest

17

By the Majority’s logic, Wiley—as a fellow publisher

of JAHA, which distributed the Retraction Statement and A

Path Forward—would also have failed to provide Wang with

notice and an opportunity to respond. The Majority properly

recognized the claim against Wiley should fail, and that the

claim against AHA should fail for the same reasons.

23

Saba and Berlacher purposefully avoided learning whether

their claims were false.

The Majority insists—without citation—that Berlacher

and Saba must have checked all of Wang’s sources and

determined they were “substantially accurate,” but still decided to “trash[]” the Article. Majority at 18. This inference is not supported by the complaint. There is no allegation that Saba

and Berlacher checked each source. Rather, the allegations

show that Saba and Berlacher claimed some sources Wang

quoted or relied on were not accurately conveyed and their

mere failure to fully investigate does not constitute actual

malice. Marcone, 754 F.2d at 1089.

Second, Wang argues that he has pleaded actual malice

by pointing at Saba and Berlacher’s opposition to his views and their characterization of the Article as “scientifically invalid and racist.” App. 145 (FAC ¶ 44). That does not show they

made their statements with knowledge or reckless disregard of

their falsity. See Masson, 501 U.S. at 510 (“Actual malice

under the New York Times standard should not be confused

with the concept of malice as an evil intent or a motive arising from spite or ill will.”). Although it “cannot be said that

evidence concerning motive or care never bears any relation to

the actual malice inquiry,” Harte-Hanks, 491 U.S. at 668, the

vitriol of a challenged statement alone does not show that it

was made with actual malice for defamation purposes. In a

similar vein, the Majority adds its view that Saba and

Berlacher’s criticism was “a stark departure from normal

academic practice.” Majority at 19. Even if this were true, as

explained above, “an extreme departure from professional

standards, without more, will not support a finding of actual

24

malice.” Tucker, 237 F.3d at 286 (citing Harte-Hanks, 491

U.S. at 665).

Third, even if Saba and Berlacher attempted to induce

others to criticize Wang’s work, that conduct would not

establish actual malice because it does not suggest they knew

or reasonably should have known their claims were false. See

McCafferty, 955 F.3d at 359 (“‘Actual malice’ is a term of art

that does not connote ill will or improper motivation.”).

Because Wang’s complaint does not establish that Saba

and Berlacher acted with actual malice, the public figure

doctrine precludes recovery even if Wang’s defamation claims

against them were otherwise actionable. 18

III

I agree with the Majority that our First Amendment

standard “is supposed to be a shield for free expression, not a sword to silence disfavored speech,” Majority at 23, and that is the very reason why I would dismiss these defamation claims.

Thus, I respectfully dissent from Part II of the Majority’s

opinion.

18

Because Wang’s defamation claims against Saba,

Berlacher, and Simon fail, so too do his defamation claims

against Pitt and UPMC predicated on vicarious liability. See

Mamalis v. Atlas Van Lines, Inc., 560 A.2d 1380, 1383 (Pa.

1989) (“A claim of vicarious liability is inseparable from the

claim against the agent since any cause of action is based on

the acts of [an agent].”).

25