The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent

Free Speech

The Duty to Correct Your Own Libelous Posts, and the Single Publication Rule

I publish something about you on Jan. 1, but I don't learn that it's false until Jan. 2. You then sue me for not taking down the post—should my liability turn on my mental state as of Jan. 1, or as of the time you sue?


(For the full draft PDF of the article from which this is excerpted, with footnotes, see here.)

The few courts that have considered the issue have so far largely declined to hold people liable based on knowingly maintaining libelous material online. Rather, those courts have tended to require that the plaintiff show that the defendant had a culpable mens rea as of the time the material was initially posted.[1] And the main reason for this has been the single publication rule.

Historically, common-law libel cases used to follow the "multiple publication rule":

Each time a libelous article is brought to the attention of a third person, a new publication has occurred, and each publication is a separate tort… . [I]f a newspaper printed an article and that newspaper was purchased by ten individuals, each communication of the defamatory article was a tortious act resulting in the injured party having ten causes of action.[2]

It followed that each distribution of a libelous item would restart the statute of limitations.[3]

In the mid-1900s, courts began to shift to the "single publication rule":

(3) Any one edition of a book or newspaper, or any one radio or television broadcast, exhibition of a motion picture or similar aggregate communication is a single publication.

(4) As to any single publication,

(a) only one action for damages can be maintained;

(b) all damages suffered in all jurisdictions can be recovered in the one action; and

(c) a judgment for or against the plaintiff upon the merits of any action for damages bars any other action for damages between the same parties in all jurisdictions.[4]

On its face, the single publication rule limits how many lawsuits can be brought based on a particular publication. But many courts have also applied this logic to deciding when the lawsuits must be filed, concluding that the statute of limitations begins to run when the single publication is first distributed. In the typical Internet libel case, that happens when a Web page is first published; the clock isn't reset each time the page is displayed on a reader's computer.

It's not clear whether this was contemplated by the drafters of the Restatement (Second) of Torts, who said that the single publication rule was "[a]n exceptional rule … applied in cases where the same communication is heard at the same time by two or more persons." That is the opposite of online distribution, where the work is distributed at different times to different people, and where courts must therefore decide when the statute of limitations begins to run. Still, as a policy matter, starting the statute of limitations when an item is first posted strikes me as good policy (though with a twist I'll note below).

But some courts have read the single publication rule as dictating that the publisher's mental state must likewise be determined solely as of the date of the single publication. The cause of action accrues at the time of publication, the logic goes. If at that time, the publisher believes (or, in private-figure cases, reasonably believes) that the statement is true, then it doesn't matter what the publisher later learns. All that matters is what the publisher knew (or should have known) as of when the cause of action accrued.

This, it seems to me, is mistaken, for two related reasons.

[1.] Formally, a cause of action generally accrues only "once all of the elements of an action … are present." Under the modern rule, where the speaker's culpable mens rea is an element of a defamation claim, a libel claim thus doesn't accrue until the speaker becomes culpable.

If in March WPIX reported (based on reasonable belief) that Starlight Rainbow had mistreated a student, and in August WPIX learned that the guilty party was actually Cynthia Rainbow, then any libel cause of action would not have accrued in March, because the negligence element was absent. The action would only have accrued in August, and the single publication rule would have kicked in only then. Thus, even applying the single publication rule, the mens rea for libel liability—under the § 577(2) theory that "[o]ne who intentionally and unreasonably fails to remove defamatory matter that he knows to be exhibited on … chattels … . under his control" is liable for "continued publication"—should be determined as of August, when the negligence element was satisfied.

Some cases do say that, under "the single publication rule, … a plaintiff's cause of action accrues only once, at the time of publication," but that's an oversimplification: The cause of action accrues only once, and that is usually the time of publication. But no cause of action should be said to accrue before all the elements are satisfied. Thus, in trade libel cases, where damages are an element of the tort, the cause of action doesn't accrue until damages arise. Likewise, in the rare ordinary libel cases where the mens rea element isn't satisfied until after publication, the cause of action shouldn't accrue until the mens rea is present.

[2.] Functionally, the single publication rule was designed to prevent multiple lawsuits, and to prevent long-delayed lawsuits filed after evidence may have been lost and the key events forgotten. Without it, one article in one issue of a newspaper could lead to many lawsuits, and a statement in a book (or on a web site) could lead to a lawsuit decades after it was published. "A newspaper article published forty years ago whose veracity is called into question today could subject the publisher to a defamation suit."

For these purposes, treating the publication date as being the date of first publication makes sense. If my online article about you is published on January 1, 2025, and the statute of limitations for libel is a year, then on January 2, 2026 you can no longer sue.

Yet say I learn the article is mistaken on January 2, 2025 (because you tell me), I refuse to correct it, and you sue me on January 10, 2025, well within the statute of limitations. It's hard to see then why the single publication rule should measure my mental state solely as of January 1. My site's being available continuously might not count for statute of limitations purposes. But its being correctable continuously should indeed count for determining whether I'm continuing to publish the article with actual malice.

[1] See Rainbow v. WPIX, Inc., 179 A.D.3d 561, 563 (2020); Pippen v. NBCUniversal Media, LLC, 734 F.3d 610, 614–15 (7th Cir. 2013); see also Lakireddy v. Soto-Vigil, No. A138675, 2014 WL 1478693 (Cal. Ct. App. Apr. 16, 2014) (concluding that "failure to remove a Web site posting," even "once substantial indications of falsity existed," is not a "republication" of the original posting and thus cannot lead to liability).

[2] Restatement (First) of Torts § 578 cmt. b (1938).

[3] Restatement (First) of Torts § 899 cmt. c (1938) ("A cause of action for defamation is complete at the time of publication.").

[4] Restatement (Second) of Torts § 577A; see also Uniform Single Publication Act § 1 (adopting this rule); id. Refs. & Annos. (noting that the act has been adopted in Arizona, California, Idaho, Illinois, New Mexico, North Dakota, and Pennsylvania).