Coleman v. Grand: Characterizing Conduct as “Sexual Harassment” Is Protected Opinion When Supported by Disclosed Facts Under New York Defamation Law

Introduction

In Coleman v. Grand (No. 21-800, decided Nov. 3, 2025), the U.S. Court of Appeals for the Second Circuit affirmed summary judgment against a defamation claim brought by jazz saxophonist Steven Douglas Coleman against his former pupil, saxophonist Maria Kim Grand. The suit arose from a seven-page, single-spaced letter Grand emailed in 2017 to approximately forty friends and colleagues following the rise of the #MeToo movement. The letter, anonymizing Coleman as “X,” recounted their complex personal and professional relationship from 2011 to 2016, and included Grand’s characterization that she had been “sexually harassed inside a professional relationship.”

The district court granted Grand’s motion for summary judgment, holding there was no triable issue on actual malice and, independently, that the challenged statements were non-actionable opinion under New York law. On appeal, Coleman argued (1) misapplication of New York defamation law, (2) erroneous determination that there was no genuine dispute as to actual malice, and (3) improper characterization of the statements as opinion rather than fact. The Second Circuit (Judge Sullivan, joined by Judge Chin) affirmed on the ground that Coleman challenged only non-actionable opinions, not the disclosed factual predicates of those opinions. Judge Menashi dissented, concluding that the letter contained specific, provable factual accusations of sexual harassment that a jury could evaluate.

Summary of the Opinion

The majority held that the statements Coleman identified are non-actionable opinions “supported by disclosed facts” under New York defamation law. Applying the holistic “reasonable reader” inquiry and the three-factor test from Davis v. Boeheim (precision, verifiability, and context), the court concluded that:

Grand’s characterization that she was “sexually harassed inside a professional relationship” was her personal assessment of fully disclosed interactions, not a distinct factual assertion implying undisclosed facts.
Statements such as “he convinced me to be intimate,” “he would call me in the middle of the night and never take no for an answer,” the timing assertion that harassment “started” in 2013, and a supposed insinuation that she was “forced” to a hotel room were either non-defamatory, subjective, idiomatic, or, taken in context, reflected Grand’s perception rather than verifiable, defamatory fact.
None of New York’s narrow per se categories applied; Coleman did not challenge the letter’s disclosed factual recitals but instead disputed Grand’s labels and inferences.

The court did not reach the district court’s separate actual-malice ruling or the applicability of New York’s anti-SLAPP statute as amended in 2020. It noted that it had held the appeal pending the New York Court of Appeals’ decision in Gottwald v. Sebert (2023) on anti-SLAPP retroactivity, which held the amendments are not retroactive. The panel ultimately resolved Coleman’s appeal on opinion-versus-fact grounds alone.

Judge Menashi dissented, arguing that Grand’s letter alleged specific quid pro quo and retaliatory conduct—classic sexual harassment—that is precise, verifiable, and commonly adjudicated by juries. In his view, the majority improperly reclassified specific factual allegations as mere “feelings,” and its approach conflicts with how courts treat similar accusations in other legal contexts.

Analysis

Precedents Cited and Their Influence

Davis v. Boeheim, 24 N.Y.3d 262 (2014): The Court of Appeals’ three-part test—(1) precise meaning, (2) verifiability, and (3) full and social context—governs the fact/opinion inquiry. The Second Circuit leaned on Davis’s “holistic approach,” emphasizing tone, purpose, and the reader’s expectation. It found Grand’s letter framed itself as a personal narrative (“my experience,” “get everything off my chest”) with ample disclosed facts, flagging opinion rather than undisclosed fact.
Gross v. New York Times Co., 82 N.Y.2d 146 (1993): A pillar of New York’s defamation doctrine distinguishing non-actionable “opinion supported by disclosed facts” from actionable “mixed opinion” implying undisclosed facts. The majority repeatedly invoked Gross to stress that where the factual basis is laid bare, a speaker’s characterization (here, “sexual harassment”) is treated as conjecture the reader may accept or reject. The dissent cited Gross to argue that even when embedded in detailed narratives, accusations like “corruption” (or here, “sexual harassment”) can be actionable statements of fact.
Steinhilber v. Alphonse, 68 N.Y.2d 283 (1986): Recognizes that an opinion accompanied by its factual basis is typically non-actionable, the principle the majority applied to Grand’s narrative.
Springer v. Almontaser, 904 N.Y.S.2d 765 (App. Div. 2d Dep’t 2010) and Jacobus v. Trump, 51 N.Y.S.3d 330 (Sup. Ct. 2017), aff’d, 64 N.Y.S.3d 889 (App. Div. 1st Dep’t 2017): Support that terms like “harassed” or idiomatic phrases (“begged”) can, depending on context, be imprecise, figurative, and not verifiable assertions of fact. The majority cited these to explain why Grand’s word choices pointed to subjective perception.
Alianza Dominicana, Inc. v. Luna, 645 N.Y.S.2d 28 (App. Div. 1st Dep’t 1996) and Chiavarelli v. Williams, 681 N.Y.S.2d 276 (App. Div. 1st Dep’t 1998): These cases show the other side of the line: accusations of harassment or specific sexual misconduct premised on undisclosed or false facts can be actionable. The Second Circuit distinguished Alianza and Chiavarelli because Grand disclosed the facts and Coleman did not challenge those facts in his suit.
Celle v. Filipino Reporter Enterprises Inc., 209 F.3d 163 (2d Cir. 2000); Dillon v. City of New York, 704 N.Y.S.2d 1 (App. Div. 1st Dep’t 1999); Chau v. Lewis, 771 F.3d 118 (2d Cir. 2014): Establish defamation elements, the need for defamatory meaning beyond mere affront, and the centrality of context. The court used these to reject statements that could not reasonably expose Coleman to public contempt or that, given context, would be read as opinion.
Liberman v. Gelstein, 80 N.Y.2d 429 (1992) and Harris v. Hirsh, 643 N.Y.S.2d 556 (App. Div. 1st Dep’t 1996): Define New York’s narrow per se categories and restrict “professional capacity” defamation per se to statements directly bearing on professional competence rather than general character. The court concluded none of the per se categories applied here.
Melius v. Glacken, 943 N.Y.S.2d 134 (App. Div. 2d Dep’t 2012): Illustrates that labels like “extortionist” can be non-actionable opinion if tethered to disclosed facts—another analogy for Grand’s “sexual harassment” label in this context.
Gottwald v. Sebert, 40 N.Y.3d 240 (2023): New York’s highest court held the 2020 anti-SLAPP amendments are not retroactive. The panel noted it had awaited this ruling but ultimately decided Coleman on opinion-versus-fact grounds and did not apply anti-SLAPP.

Legal Reasoning

The court’s analytic core is the “reasonable reader” test, taking a holistic view of the letter’s content, tone, and purpose. It closely parsed the specific statements Coleman challenged and asked whether those statements would be read as distinct factual assertions or as Grand’s personal characterization of disclosed events.

“He convinced me to be intimate with him.” The court found this neither defamatory nor factual in the relevant sense. “Convinced” is inherently subjective—reflecting Grand’s internal state of persuasion—especially within a personal narrative suffused with her feelings. In context, it reads as her perception, not a verifiable claim of coercion.
“I was sexually harassed inside a professional relationship.” This characterization, the court held, is a paradigmatic opinion-with-disclosed-facts. Grand recounted concrete episodes (e.g., anger when she refused sex, late-night calls, an incident of kissing while she slept, crude statements about sexual “owed” favors). Coleman, however, did not challenge these episodes as false; he contested the label “sexual harassment.” That label, the court emphasized, is an evaluative judgment the reader is free to accept or reject in light of the disclosed facts.
“By September 2013 … that period is when the sexual harassment started.” The court again treated this as an evaluative timeline label, not an independent factual charge. Assertions about love, desire for intimacy, and the precise start date of “harassment” were either non-defamatory feelings or the same protected characterization of disclosed interactions.
“He would call me in the middle of the night and never take no for an answer.” Read idiomatically and in context, this describes Grand’s perception of persistence—not a literal claim of forced sex (which the letter did not allege). It is, the court held, the kind of figurative, subjective phrasing New York treats as opinion.
Implied claim she was “forced to go” to his hotel room. The court rejected this reading of the letter. Grand expressly acknowledged she had agreed to share the room in advance, and nothing in the letter suggested physical compulsion. At most, the letter conveyed Grand’s belief that professional opportunities hinged on intimacy—again, an evaluative inference drawn from disclosed facts.

Two doctrinal themes underlie the opinion’s result:

Disclosed facts versus undisclosed facts. New York protects “pure” opinions and opinions that disclose their factual bases. It disfavors “mixed” opinions implying unknown facts. The court repeatedly emphasized Grand’s detailed narrative and her express purpose to share “everything” and start a “conversation.” That framing, plus specific anecdotes, signaled that she was not hinting at secret evidence; readers could evaluate her conclusions for themselves.
Focus on the plaintiff’s pleaded theory and statements actually challenged. The court underscored that Coleman did not contest the key factual incidents themselves but attacked Grand’s labels and inferences. Appellate review, the court said, centers on the “precise statements” the plaintiff claims are false and defamatory—not on re-casting the case to reach different assertions (a point pressed in response to the dissent).

Impact

Coleman clarifies, within the Second Circuit’s application of New York law, how courts will treat #MeToo-era personal narratives circulated to limited audiences:

Safe harbor for fully disclosed, evaluative narratives. When an accuser recounts specific events and openly characterizes them (e.g., as “sexual harassment”), New York law treats the characterization as non-actionable opinion—so long as the statement does not suggest undisclosed facts and the plaintiff does not dispute the disclosed incidents themselves. This gives speakers a measure of protection when they present a candid, factual narrative coupled with personal conclusions.
Not a blanket immunity for harassment accusations. The court reaffirms that accusations can be actionable if they (a) rest on false specifics; (b) are framed as factual assertions without disclosing the basis; or (c) imply undisclosed damaging information. Context remains everything: more formal or declarative announcements, unadorned allegations, or statements to the press implying hidden proof may remain actionable (see Alianza, Chiavarelli).
Litigation strategy: challenge facts, not just labels. Plaintiffs contemplating defamation claims against #MeToo accusers must identify and contest specific factual statements, not merely dispute characterizations. Merely disagreeing with the term “sexual harassment” will not suffice if the plaintiff does not also challenge the factual underpinnings as false.
Per se defamation remains narrow. The panel reiterated that New York’s per se categories are limited. Professional harm resulting from reputational injury does not, by itself, transform allegations into per se defamation unless the statements directly impugn professional competence in the recognized way.
Anti-SLAPP posture deferred. Although amici urged robust First Amendment protection and the district court applied an actual-malice lens, the Second Circuit did not decide anti-SLAPP issues, and New York’s 2020 amendments are not retroactive (Gottwald). Defendants should still consider anti-SLAPP defenses where applicable, but Coleman shows many such suits can be resolved on opinion-versus-fact grounds alone.

The Dissent’s Counterpoint

Judge Menashi would have allowed a jury to decide whether Grand’s claims are true, viewing them as specific, verifiable allegations: sex tied to mentorship and work, anger and retaliation when sex was refused, and particular incidents (e.g., uninvited kissing while she slept). He analogized to Title VII harassment doctrine, noting that courts routinely adjudicate whether alleged conduct was “unwelcome” and tied to professional conditions. He criticized the majority for treating detailed accusations as subjective “feelings,” and pointed to concrete professional fallout (lost band members and gigs) as evidence that readers took the statements as factual.

The majority’s rejoinder is two-fold: (1) Title VII pleading labels are distinct from whether, in a defamation suit, a reader would understand “sexual harassment” as a new factual assertion or as an evaluative conclusion based on disclosed facts; and (2) Coleman did not contest the factual episodes as false, only the characterizations drawn from them. On that litigation record, the panel concluded that New York’s opinion doctrine controlled.

Complex Concepts Simplified

Defamation (New York): A false statement of fact, published to a third party without privilege, causing special damages or falling within narrow per se categories. Context matters, and statements must reasonably convey a defamatory meaning to an average reader.
Opinion versus fact: Only statements of fact are actionable. New York protects: (a) opinions that disclose the facts on which they rely; and (b) opinions that do not imply undisclosed facts. If the speaker’s evaluative label (e.g., “harassment,” “extortion”) is clearly tethered to disclosed incidents, the reader can judge for herself, and the label is typically non-actionable.
“Mixed opinion”: An opinion that implies the existence of undisclosed defamatory facts. This is actionable because it suggests the speaker knows damaging, unshared information.
Reasonable reader, holistic context: Courts read the entire publication, including tone, purpose, and social setting, to assess how an average reader would understand the statements.
Per se defamation: In New York, limited to imputing unchastity (historical category), serious crime, loathsome disease, or statements that directly injure one’s business or profession by impugning competence. Many reputational harms do not qualify as per se.
Actual malice and anti-SLAPP: “Actual malice” (knowledge of falsity or reckless disregard) is a heightened fault standard in some defamation contexts. New York’s 2020 anti-SLAPP amendments extend actual malice to certain public-participation suits, but they are not retroactive (Gottwald). Coleman was decided without reaching these issues.

Conclusion

Coleman v. Grand is a significant clarification of New York’s opinion/fact distinction as it applies to #MeToo-era personal narratives. The Second Circuit held that when a speaker discloses the factual basis of a sensitive account and uses an evaluative label such as “sexual harassment,” the law treats that label as protected opinion—unless the speaker implies undisclosed facts or the plaintiff contests the factual underpinnings as false. The decision underscores a practical lesson: in defamation litigation over #MeToo communications, outcomes often turn not on the label but on whether the plaintiff has identified and challenged concrete factual assertions.

At the same time, Coleman carefully preserves the other side of the line: specific, false allegations of misconduct—especially if premised on undisclosed facts—remain actionable. The dissent highlights the stakes, warning against collapsing detailed misconduct allegations into mere “feelings.” Going forward, Coleman will likely guide both authors and litigants: those sharing experiences should disclose the facts and speak in personal terms; those suing should pinpoint false factual statements rather than only disputing characterizations. In the broader legal landscape, Coleman reflects New York’s enduring commitment, rooted in Gross and Davis, to robust debate about disclosed facts—leaving the ultimate moral and social judgment to the “competition of ideas,” not defamation liability.