Tweet Containing Question Mark Isn’t Defamatory–Boulger v. Woods

02-PM-300x111In 2016, the Chicago Tribune published a photo of a woman giving a Nazi salute at a Trump rally. Twitter user @voxday wrongly identified the plaintiff as that woman. Shortly afterwards, actor James Woods, who then had 350,000 twitter users, tweeted the Nazi salute photograph along with a photo of plaintiff (Portia Boulger), saying “So called #Trump “Nazi” is a #BernieSanders agitator/operative?”

Woods later deleted the tweet but did not, as requested by Boulger’s counsel, issue a formal retraction and apology. In the 11 days that the tweet was online, Boulger alleged she received hundreds of “obscene and threatening messages, including death threats.” She sued for defamation and invasion of privacy. The district court dismissed because at least some of readers of Woods’s Twitter account would view the tweet as a legitimate question. (Blog post on the district court ruling here: “Twitter Defamation Claim Defeated by a Question Mark–Boulger v. Woods“). The Sixth Circuit affirms.

Ohio courts use a four-factor test to determine whether a statement is an opinion or fact. Applying those factors, the court finds the tweet not actionable.

Specific language used: the court find that this factor is not determinative. Some readers likely viewed the tweet as a statement that the woman in question was Boulger. Other readers likely viewed Woods as asking a legitimate question. Given that the tweet is not unequivocally a statement, this factor favors a finding of non-actionability.

Verifiability: The court rejects a blanket rule that questions are by nature not verifiable statements and thus not actionable.

The context of the statement: Factors three and four look to the “general context and broader context” of the statement. The district court struggled with whether to look at Woods’s surrounding tweets, but the panel takes a different view. Reviewing Woods’s twitter feed from March 2016, the court finds that Woods often offered “his own colorful commentary” to news articles. Indeed, the court says that Woods’s other tweets tip off a reasonable reader that Woods often used “sarcasm, exaggeration, and hyperbole”—i.e., opinion.

The court also takes stock of Twitter as a medium and finds that it is used for both opinion and news.

Innocent construction: The court finds that the factors do not definitively tip the scales in favor of either opinion or fact. The court invokes Ohio’s innocent construction rule to find that the tweet gets the benefit of the doubt and is not actionable:

the tweet at issue is reasonably susceptible to both a defamatory meaning—that Woods was asserting Boulger was the woman giving the Nazi salut—and an innocent meaning—that Woods was merely asking his followers a question. Because Woods’s tweet could reasonably be read to have an innocent meaning, under the innocent construction rule the tweet, as a matter of law, is not actionable.

Concurrence: Judge Nalbandian concurred, saying that the four-factor test to distinguish fact from opinion is ill-suited to determine whether “the tweet was a statement at all”:

Asking whether a question is verifiable is like asking how tall an elephant weighs. A question is an inquiry, not an assertion, so it’s neither verifiable nor unverifiable. To be sure it might be answered, or it might imply some kind of assertion—both of which could be verifiable. But a genuine question lacks a truth value, so asking about verifiability only highlights how the Ohio Supreme Court designed the totality test to solve a different problem than the one here.

Judge Nalbandian would rather focus on the “simpler question” of whether a reasonable reader would interpret the language as a genuine question. For him, the existence of a question mark and the juxtaposition of photos is sufficient to conclude that at least some readers would view the tweet as posing a genuine question.


I blogged about the post-Charlottesville doxxing and misidentification lawsuit last week (“Post-Charlottesville Doxxing and Misidentification Creates Legal Risks–Vangheluwe v. GotNews“). There social media users wrongly identified plaintiff as the driver of the vehicle who ploughed into the crowd. The court found that the plaintiff could maintain at least some claims for misidentification (defamation) and doxxing. This is another lawsuit in the same genre.

Woods escapes liability despite his actions having a similar effect to the defendants in the Vangheluwe case. In both instances, the plaintiffs complained of a torrent of online abuse, up to and including death threats. Woods did not engage in doxxing, but you wonder whether any of his followers did.

Both the district court and the appellate panel grappled with how much context to take into account in determining whether an online statement is actionable or opinion. As the judges all note, a blanket rule immunizing questions doesn’t make sense. But they still struggle with where to draw the line and what hypothetical person is used to determine whether the tweet is a statement or a true question. A hypothetical twitter user? A hypothetical follower of this type of opinion content online?

In deciding the case, the Sixth Circuit interpreted Ohio law. This likely lessens the chances of review by the Supreme Court.


Eric’s Comments: This result is generally good news. The presence of the question mark reduces defamation exposure, which is good news because questions are a key part of exploring the truth. Obviously a question mark isn’t a “get out of defamation free” card, but it’s still a powerful signal to readers that they can’t assume the statement is factual truth.

Having said that, I’m deeply troubled by the proliferation of garbage content on social media, especially Twitter, and especially from the “#MAGA” crowd. I see this as an existential challenge to Twitter to continue to be a credible venue for news, not a hotbed of conspiracy theories, personal attacks and smears, and outright lies. This particular post may not be the poster child for Twitter’s challenges, but it’s close. Did Woods’ tweet improve the discourse in any way? If not, how can Twitter channel users like him towards better behavior?

Case citation: Boulger v. Woods, Nos. 18-3170/3220 (6th Cir. Feb. 27, 2019)

Related posts:

Post-Charlottesville Doxxing and Misidentification Creates Legal Risks–Vangheluwe v. GotNews

Twitter Defamation Claim Defeated by a Question Mark–Boulger v. Woods

Twibel Ruling: Tweeting That Someone is “Fucking Crazy” is Not Defamatory

Hyperlinking to Sources Can Help Defeat Defamation Claims–Adelson v. Harris

Using Links as Citations Helps Gizmodo Defeat a Defamation Claim–Redmond v. Gawker Media

Protip: Don’t Send Emails Threatening to “Inflict the Maximum Amount of Financial Pain” Allowed By Law

Want To Avoid Defaming Someone Online? Link To Your Sources (Forbes Cross-Post)

Social Media Rant Against Airline Employee Wasn’t Defamatory But May Be False Light–Patterson v. Grant-Herms

Calling Out Scraper for “Stealing” Data Is Not Defamatory – Tamburo v. Dworkin

A Twitter Exception for Defamation?

9th Circuit Issues a Blogger-Friendly First Amendment Opinion–Obsidian Finance v. Cox