806

Argued and submitted June 24, 2021; decision of Court of Appeals affirmed in part and reversed in part, judgment of circuit court reversed, and case remanded to circuit court for further proceedings June 23, 2022

Tom LOWELL, dba Piano Studios and Showcase, Respondent on Review, v. Matthew WRIGHT and Artistic Piano, an Oregon corporation, Petitioners on Review. (CC 13CV04582) (CA A162785) (SC S068129) 512 P3d 403

Plaintiff, a piano store owner, brought a defamation action against defen- dants, a competing piano store and its employee, based on an allegedly libelous Google review, which was deleted before litigation began and never recovered. Defendants moved for summary judgment based on a First Amendment public comment defense. The trial court granted the motion, concluding that without the verbatim text of the review, it could not properly evaluate defendants’ First Amendment claim. Plaintiff appealed, and the Court of Appeals reversed and remanded, concluding that first, the absence of the text of the review was not dispositive of the summary judgment motion; second, the Google review was on a “matter of public concern,” satisfying one of the prongs of defendants’ asserted First Amendment defense; third, a speaker’s motive or purpose in speaking is relevant to whether speech is on a “matter of public concern”; and fourth, to over- come defendants’ First Amendment defense, plaintiff need not show that defen- dants acted with “actual malice” because defendants were nonmedia, following Wheeler v. Green, 286 Or 99, 59 P2d 777 (1979). Held: (1) The lack of the exact wording of the review did not entitle defendants to summary judgment; (2) the review was on a “matter of public concern” based on Neumann v. Liles, 358 Or 706, 369 P3d 1117 (2016); (3) a speaker’s motive or purpose in speaking is not relevant to whether the speech is on a matter of public concern for the purposes of the First Amendment public comment defense; and (4) a private-figure plaintiff need not show “actual malice” to overcome a First Amendment public comment defense when the defendant is nonmedia, adhering to Wheeler. The decision of the Court of Appeals is affirmed in part and reversed in part. The judgment of the circuit court is reversed, and the case is remanded to the circuit court for further proceedings.

On review from the Court of Appeals.* ______________ * On appeal from Jackson County Circuit Court, Dan Bunch, Judge. 306 Or App 325, 473 P3d 1094 (2020). Cite as 369 Or 806 (2022) 807

Tracy M. McGovern, Frohnmayer, Deatherage, Jamieson, Moore, Armosino & McGovern, P.C., Medford, argued the cause and filed the briefs for petitioners on review. Also on the briefs were Casey S. Murdock and Alicia M. Wilson. Linda K. Williams, Portland, argued the cause and filed the brief for respondent on review. Eugene Volokh, UCLA School of Law, First Amendment Clinic, Los Angeles, California, argued the cause and filed the brief for amici curiae Institute for Free Speech; Electronic Frontier Foundation; Professors William Funk, Ofer Raban, and Kyu Ho Youm; Howard Bashman; Scotusblog, Inc.; and Professors Glenn Harlan Reynolds and Eugene Volokh. Also on the brief was Owen Yeates, Institute for Free Speech, Washington, D.C. James Abernathy and Rebekah Millard, Freedom Foundation, Olympia, Washington, filed the brief on behalf of amicus curiae Freedom Foundation. Before Walters, Chief Justice, and Balmer, Flynn, Duncan, Nelson, and Garrett, Justices, and Nakamoto, Senior Judge, Justice pro tempore.** NAKAMOTO, S. J. The decision of the Court of Appeals is affirmed in part and reversed in part. The judgment of the circuit court is reversed, and the case is remanded to the circuit court for further proceedings. Balmer, J., concurred and filed an opinion in which Garrett, J., joined. Flynn, J., concurred and filed an opinion.

______________ ** DeHoog, J., did not participate in the consideration or decision of this case. 808 Lowell v. Wright

NAKAMOTO, S. J. Plaintiff’s libel per se claim is based on a Google review, written by the manager of plaintiff’s business com- petitor, that subsequently was removed from the internet without a trace. The Court of Appeals reversed a grant of summary judgment to defendants. We resolve three dis- puted legal questions: (1) whether plaintiff may reach a jury on his libel claim when the text is no longer available; (2) whether the First Amendment’s public comment defense is available in these circumstances and, relatedly, whether a defendant speaker’s identity or motive is part of the court’s inquiry on the defense’s availability; and (3) whether Oregon should require a plaintiff claiming defamation to prove that the defendant acted with a heightened culpable men- tal state, “actual malice,” in all cases when the speech is on a “matter of public concern” protected under the First Amendment, abolishing the distinction that requires such proof only when the defendant is a member of the media. The Court of Appeals concluded that the trial court had erred because plaintiff’s evidence of the allegedly defam- atory statements sufficed to create a question of fact for trial on his claim and the lack of the review’s printed text did not affect the analysis of defendants’ First Amendment defense. Lowell v. Wright, 306 Or App 325, 334-35, 473 P3d 1094 (2020). Putting aside the First Amendment defense, we, like the Court of Appeals, conclude that the lack of a copy of the review is not fatal to plaintiff’s libel claim and that two of the three allegedly defamatory statements in the review are actionable. To decide whether defendants were entitled to sum- mary judgment based on their First Amendment defense, the threshold question is whether the review about plain- tiff’s store is subject to First Amendment protection as con- taining statements on a matter of public concern. As the Court of Appeals recognized, in Neumann v. Liles, 358 Or 706, 369 P3d 1117 (2016), this court held that a review of a wedding venue contained speech on a matter of public con- cern protected by the First Amendment’s public comment defense, and the review of plaintiff’s business in this case is similar to the review in Neumann. Although plaintiff Cite as 369 Or 806 (2022) 809

argues that a speaker’s motive may affect the availability of the defense, an argument that we reject, neither party has asked this court to overrule Neumann’s holding. Accordingly, Neumann controls, and we are compelled to follow it in this case. Finally, we decline to overrule our precedent rec- ognizing the media/nonmedia distinction and to impose an across-the-board heightened proof-of-fault requirement on defamation plaintiffs in cases involving the First Amend- ment. Defendants and amici argue that we ought to abolish the distinction, in part, they assert, because it is sometimes difficult to discern whether a given speaker, such as a blog- ger, is a media or a nonmedia defendant. This case does not offer an opportunity for careful examination of that issue, considering that defendants are not “media” under any defi- nition and acknowledge that they are “nonmedia” defen- dants, and defendants have not persuaded us to abandon our precedent and to alter Oregon common law. Ultimately, we conclude that the trial court erred in granting defendants’ summary judgment motion and enter- ing a general judgment of dismissal. We affirm the decision of the Court of Appeals in part and remand the case to the trial court. I. BACKGROUND We are reviewing the trial court’s ruling granting defendants’ motion for summary judgment on plaintiff’s claim for defamation. Accordingly, we recount the facts in the light most favorable to plaintiff as the nonmoving party, including reasonable inferences that may be drawn from the facts adduced. ORCP 47 C. Plaintiff Lowell owns and operates Piano Studios and Showcase (Piano Studios), a piano store in Medford. On September 3, 2012, defendant Wright and his wife visited plaintiff’s store. At that time, Wright was employed as a general manager by defendant Artistic Piano, another piano store in Medford.1 On the day of the visit, Wright was off 1 Defendants Wright and Artistic Piano make no arguments independent of each other. This opinion will distinguish between them only when needed to clar- ify the facts. 810 Lowell v. Wright

work. Wright testified that he did not tell his boss, Werner, the owner and operator of Artistic Piano, about his plans to visit Piano Studios. Soon after the visit, Wright posted a Google review about Piano Studios. In December 2012, while browsing internet pages mentioning his business, plaintiff found Wright’s review and became upset. The review was not posted under Wright’s name, but instead under “Amazing Impressions” (Wright’s unrelated photography business). Plaintiff eventually found a phone number associated with Amazing Impressions and called it, instructing his employee, Norling, to listen to the phone call and take notes. Wright answered and eventually hung up on plaintiff. After the phone call, Wright spoke with Werner and showed him the review. According to Wright, he had told Werner previously about his visit to Piano Studios but had not mentioned his review. Once Werner looked over the review, he suggested that Wright take it down. Wright removed it without saving a copy. Plaintiff also did not save a copy of the review before it was removed. Wright composed the review on a home laptop, which he disposed of, explain- ing that it had become old and inoperable. Despite a diligent search and a request to Google, the parties were unable to recover a copy of the review during litigation. Although the actual text of the review is unavail- able, four people read the review and testified in depositions regarding its contents: Lowell, Norling, Wright, and Werner. Although they could not remember the review verbatim, they largely agreed that it contained the following para- phrased content and that the quoted language (or something extremely close to it) appeared in the review itself: Wright walked around the store for 45 minutes before a salesperson spoke to him, and the store “smelled like grandma’s attic.” When he eventually spoke with Wright, the salesman told Wright that a Yamaha C-7 piano displayed on the showroom floor was about five years old. The salesman also told Wright that plaintiff can sell new Steinway pianos.2 However, 2 The witnesses remember the exact wording of the Steinway comment dif- ferently. Plaintiff and Norling remembered the review stating that plaintiff’s salesman said that the store “can sell new Steinway pianos”; Wright recalled Cite as 369 Or 806 (2022) 811

plaintiff cannot sell new Steinway pianos, and “[there] were no new Steinways in the showroom,” which is “like a Chevy dealer not having any Chevrolets on the lot.” Finally, Wright had been warned about plaintiff’s store and now knew that it was true that “this guy can’t be trusted.” According to plaintiff, Wright not only made false statements—he made up the entire conversation at Piano Studios. Plaintiff’s theory is that defendants’ purpose in having Wright go to Piano Studios and write the review was to “cybersmear,” a practice whereby one business pseud- onymously writes about a competitor on the internet to lower the competitor’s reputation and thereby attract more business for itself. In support, plaintiff provided testimony from the salesman on duty the day Wright was in Piano Studios. Plaintiff’s sales force kept time logs in which they recorded interactions with potential customers. The sales- man’s time log from that day does not reflect a conversa- tion with Wright. And, the salesman testified, Wright never conversed with him about Steinways or the Yamaha piano on display. In further support of his theory, plaintiff empha- sizes that Wright was the manager of Artistic Piano when he wrote the review, that the review was written under the name “Amazing Impressions” rather than under Wright’s own name, and that the review included no details that would suggest that its writer had specialized knowledge or a potential ulterior motive. In 2013, plaintiff filed a defamation action against Wright and Artistic Piano.3 He alleged that Wright acted as Artistic Piano’s agent in writing the Google review. Plaintiff asserted that the review “purported to describe the personal experience of an actual customer” but that “Wright was not a bona fide potential customer.” Plaintiff alleged that three

the wording as “is a Steinway dealer.” The former comment would indicate that plaintiff is capable of selling while the latter would indicate that the plaintiff is allowed to sell because he had a dealership agreement with Steinway, authoriz- ing him to sell new Steinway pianos. The significance of this difference, plaintiff explained, is that if the salesman said that the store “can sell new Steinway pianos,” that would be true, but if the salesman said that the store “is a Steinway dealer,” that would be false. 3 The action also included an unfair trade practices claim, which is not at issue on review. 812 Lowell v. Wright

statements in the review were false and defamatory asser- tions of fact: “a. That a Yamaha C-7 piano serial number F4910127 on the showroom floor was misrepresented to Wright as being about 5 years old, when in fact said piano was at least 15 years older and less valuable, and this misrepresenta- tion of the age of the instrument was purposely made in an effort to cheat Wright; “b. That [plaintiff] misrepresents that he sells new Steinway Pianos, when he actually doesn’t; and “c. That the above misrepresentations are proof that ‘this guy can’t be trusted.’ ” In their answer, among other defenses, defendants asserted that the First Amendment precluded liability for libel. In 2016, defendants filed a motion for summary judgment on the libel claim. They argued that (1) plain- tiff could not prevail because he could not prove that the statements were defamatory; (2) any statements made were protected by the First Amendment public comment defense (under which statements on a matter of public concern that are not susceptible to being proved true or false are not actionable); and (3) plaintiff could not prove that defendants acted with actual malice. Plaintiff opposed the motion, providing the deposi- tion testimony about the review’s content. He also submitted evidence regarding the falsity of the Yamaha and Steinway statements. Plaintiff submitted a copy of the Yamaha pia- no’s tag, which indicated that it was manufactured in 1990, to support the salesman’s testimony that he never falsely told Wright that the piano was five years old. Plaintiff also introduced evidence that he can and does sell new Steinway pianos even though plaintiff was not a Steinway dealer, including documents that he contended related to the store’s sales of Steinway pianos. The trial court granted summary judgment to defendants. In a letter opinion, the court addressed the sig- nificance of the missing verbatim text for plaintiff’s claim that he was defamed by the can’t-sell-new-Steinways remark in the review. The court indicated that, as to that statement, Cite as 369 Or 806 (2022) 813

plaintiff lacked sufficient evidence for a trial on what the review had said and whether it was false. The trial court did not fault the sufficiency of plaintiff’s evidence that the review stated that the salesman had lied to Wright about the age of the Yamaha piano in the showroom. And the “this guy can’t be trusted” statement, the court concluded, was so clearly one of subjective opinion that it was not actionable in a defamation claim. The trial court went on to conclude that, because defendants had asserted a First Amendment defense, it was required to determine whether the allegedly false state- ments about the pianos were constitutionally protected expressions of opinion. Without a copy of the review, the trial court concluded, it was not possible to determine whether the Yamaha and Steinway statements were constitution- ally protected or instead actionable statements implying an assertion of objective fact and that, under those circum- stances, judgment for defendants was appropriate. The trial court did not address other issues that the parties had raised in connection with defendants’ First Amendment defense. The court then entered a general judg- ment dismissing the action. Plaintiff appealed, and the Court of Appeals reversed and remanded. The Court of Appeals first held that the absence of the text of the review was not dispositive with respect to the defamatory remarks. Lowell, 306 Or App at 334. Next, the court considered whether defendants were entitled to sum- mary judgment based on their First Amendment defense. To be protected, the court stated, the speech must be on a matter of public concern that does not imply an assertion of objective fact about plaintiff. Id. at 335-36. The court con- cluded that defendants’ speech was on a matter of public concern, but only after concluding that, although “a speak- er’s motive or purpose in speaking is relevant to whether speech is protected by the First Amendment,” id. at 339 (emphasis in original), plaintiff had failed to meet his bur- den to establish proof of defendants’ motive to denigrate his business for private financial advantage. Id. at 342. The court then addressed whether the review implied objective facts about plaintiff and held that the Yamaha and Steinway 814 Lowell v. Wright

statements could be sufficiently factual to be actionable, but that the statement “this guy can’t be trusted” was not. Id. at 343. Finally, the Court of Appeals addressed defendants’ argument that plaintiff was not entitled to a trial because he had not made a sufficient showing of their actual malice. Following this court’s holding in Wheeler v. Green, 286 Or 99, 593 P2d 777 (1979), the court held that plaintiff need not establish that defendants acted with actual malice to overcome the First Amendment defense because that stan- dard applied only to media defendants, a status defendants lacked, and that, in any case, plaintiff supplied sufficient evidence to establish actual malice for the purposes of sum- mary judgment. Lowell, 306 Or App at 348. We allowed defendants’ petition for review. II. ANALYSIS A. Libel Actions in Oregon and the First Amendment Defense, Generally To put the issues we decide in perspective, we pro- vide some basic law governing libel actions in Oregon. As this court explained in Neumann, 358 Or at 711, defamation has long been recognized as tortious in Oregon. Indeed, the rem- edy clause, Article I, section 10, of the Oregon Constitution, has specified from its adoption that Oregonians shall have a remedy for injury to reputation, providing that “every man shall have remedy by due course of law for injury done him in his person, property, or reputation.” In Horton v. OHSU, 359 Or 168, 180, 376 P3d 998 (2016), the court explained that, textually, the “clause’s focus on providing remedies for specified types of injuries implies that it was intended to guarantee some remedy for those injuries, and not merely be a guarantee of procedural regularity for whatever injuries may, at the moment, enjoy legal protection.” A statement that “would subject the plaintiff to hatred, contempt or ridicule,” that tends to “diminish the esteem, respect, goodwill or confidence in which [the plain- tiff] is held,” or that brings about “adverse, derogatory or unpleasant feelings or opinions against” the plaintiff can be the basis for a defamation claim. Neumann, 358 Or at 711 Cite as 369 Or 806 (2022) 815

(quotation and citation omitted, brackets in original). And, as relevant to this case, a statement also is defamatory if it falsely “ascribes to another conduct, characteristics or a condition incompatible with the proper conduct of his lawful business[.]” Id. at 711-12. A court decides whether a state- ment is capable of a defamatory meaning, and, if so, the jury decides whether the statement did have a defamatory mean- ing. Brown v. Gatti, 341 Or 452, 459, 145 P3d 130 (2006). Alleged defamatory statements that the plaintiff engaged in misconduct or dishonesty in conducting the plaintiff’s business are actionable per se. Neumann, 358 Or at 712. Under Oregon law, to establish a prima facie claim of libel per se, subject to defenses, plaintiff must prove that defendants made a defamatory statement about him and published the statement to a third party. Id. at 711. Because the claim is libel per se, plaintiff is not obliged to prove a third element: that the defamatory statement caused pecu- niary loss or special harm. Brown, 341 Or at 458. But it is now well established that state law def- amation actions can be affected by the First Amendment rights of speakers. The Supreme Court first extended consti- tutional protections in a libel case over 50 years ago in New York Times Co. v. Sullivan, 376 US 254, 84 S Ct 710, 11 L Ed 2d 686 (1964). Before that decision, the Court had recog- nized libel as one of the “well-defined and narrowly limited classes of speech, the prevention and punishment of which have never been thought to raise any Constitutional prob- lem.” Chaplinsky v. New Hampshire, 315 US 568, 571-72, 62 S Ct 766, 86 L Ed 1031 (1942). Since New York Times was decided, the Supreme Court has repeatedly revisited the intersection between defamation claims and the First Amendment freedoms of speech and of the press and articulated additional constitu- tional principles that may apply in common-law defamation cases, sometimes focusing on the identity of the speaker, the identity of the plaintiff, or the nature of the speech involved. At the same time, the Court has noted that competing interests are involved. For example, the Court in Milkovich v. Lorain Journal Co., 497 US 1, 110 S Ct 2695, 111 L Ed 2d 1 (1990), as it had in earlier cases, recognized not only 816 Lowell v. Wright

the First Amendment’s protection “of free and uninhibited discussion of public issues,” but also the “ ‘important social values’ ” undergirding common-law defamation and strong societal interests “ ‘in preventing and redressing attacks upon reputation.’ ” Id. at 22 (quoting Rosenblatt v. Baer, 383 US 75, 86, 86 S Ct 669, 15 L Ed 2d 597 (1966)). In Gertz v. Robert Welch, Inc., 418 US 323, 94 S Ct 2997, 41 L Ed 2d 789 (1974), the Court also recognized the value of reputation in several ways, noting that states have a legitimate interest in compensating “individuals for the harm inflicted on them by defamatory falsehood” and that “the individual’s right to the protection of his own good name ‘reflects no more than our basic concept of the essential dignity and worth of every human being—a concept at the root of any decent system of ordered liberty.’ ” Id. at 341 (quoting Rosenblatt, 383 US at 92 (Stewart, J., concurring)).4 As relevant here, the First Amendment interacts with plaintiff’s claim in two distinct ways. First, defen- dants raise the First Amendment public comment defense, which places federal constitutional constraints on whether a state must require that a plaintiff in a defamation action show that the defamatory statements are provably false. Second, defendants argue that plaintiff must show that they acted with actual malice, a standard that comes from First Amendment doctrine and that some state courts have cho- sen to apply broadly across defamation actions. We discuss each aspect of First Amendment doctrine in detail. The Supreme Court explained the falsity require- ment in Philadelphia Newspapers, Inc. v. Hepps, 475 US 767, 106 S Ct 1558, 89 L Ed 2d 783 (1986), and Milkovich. In Hepps, the Supreme Court considered a libel claim brought by a convenience store franchise stockholder against a newspaper over an article alleging that the stockholder had

4 Recently, Justice Thomas has called into question whether, in New York Times and the cases that followed, the “constitutional libel rules” adopted by the Court by virtue of the First and Fourteenth Amendments properly displaced the common law of libel developed by the states. McKee v. Cosby, ___ US ___, ___, 139 S Ct 675, 678-82, 203 L Ed 2d 247 (2019) (Thomas, J., concurring in denial of certiorari). In his view, the states “are perfectly capable of striking an acceptable balance between encouraging robust public discourse and providing a meaning- ful remedy for reputational harm.” Id. at 682. Cite as 369 Or 806 (2022) 817

mob ties. 475 US at 769. At issue was whether the plaintiff could recover without a showing that the statements were false. Id. The Court held “that, at least where a newspaper publishes speech of public concern, a private-figure plain- tiff cannot recover damages without also showing that the statements at issue are false.” Id. at 768-69. The Hepps Court reserved the question of whether a private-figure plaintiff could recover against a nonmedia defendant with- out making a showing of falsity. Id. at 779 n 4. In Milkovich, the Court reaffirmed its holding in Hepps, explaining, “Foremost, we think Hepps stands for the proposition that a statement on matters of public concern must be provable as false before there can be liability under state defamation law, at least in situations, like the present, where a media defendant is involved.” 497 US at 19-20. Milkovich held that a media defendant’s speech was protected under the First Amendment as public comment when two criteria were met: (1) the speech was on a matter of public concern and (2) the speech could not reasonably be interpreted as stating facts or was not susceptible to being proved true or false. Id. The Court again reserved whether a nonmedia defendant could raise the public comment defense. Id. at 20 n 6. Separately, the Supreme Court has held that states cannot allow a private-figure plaintiff to prevail in a defa- mation claim against a media defendant without making a showing of some level of fault. Gertz, 418 US at 347. The states are left to define for themselves the appropriate stan- dard of fault. Id. Defendants in this case urge us to hold not only that plaintiff must prove defendants’ fault, even though they are not media defendants, but also that the standard of fault should be “actual malice,” or reckless disregard for the truth, which the Court has required only for defamation actions brought by public officials and public figures against media defendants. See New York Times, 376 US at 279-80; Curtis Publishing Co. v. Butts, 388 US 130, 87 S Ct 1975, 18 L Ed 2d 1094 (1967). In sum, without the overlay of any state law, fed- eral constitutional law currently recognizes two different areas of defamation law in which a state may make rules that distinguish between media and nonmedia defen- dants within an acceptable range. With respect to falsity, 818 Lowell v. Wright

federal constitutional law announced by the Supreme Court demands that a state require a private-figure plaintiff to show falsity when a media defendant’s speech is protected by the public comment defense. The Court’s decisions do not speak to whether that requirement applies when the defen- dant is nonmedia. With respect to fault, a private-figure plaintiff may not collect damages from a media defendant without some showing of fault. But Gertz does not announce a federal rule for fault in a defamation case brought by a private-figure plaintiff against nonmedia defendants. This court has also issued decisions involving proof of falsity and of fault in cases involving both the First Amendment and Oregon defamation law. In Neumann, the court established for the first time that a private-figure plaintiff bears the burden of proving falsity of statements in defamation cases involving the First Amendment public comment defense, even when the action is brought against nonmedia defendants. See 358 Or at 716 (announcing framework for applying Milkovich against factual back- drop with no media defendant). And, in a series of cases decided in the 1970s, this court held that the Gertz fault requirement for private-figure plaintiffs applies only to media defendants. Harley-Davidson v. Markley, 279 Or 361, 371, 568 P2d 1359 (1977); Adams v. State Farm Mutual Auto. Ins. Co., 283 Or 45, 51-52, 581 P2d 507 (1978); Wheeler, 286 Or at 110. B. Sufficiency of Evidence in Absence of Text On review, defendants rely primarily on Neumann to argue that the trial court correctly granted their motion for summary judgment. In challenging the reversal by the Court of Appeals, defendants ask this court to address three issues, the first of which is whether a plaintiff may reach a jury in a libel claim based on a writing posted on a pub- licly available site when the writing itself is no longer avail- able. Defendants urge that, without the actual text of the review, plaintiff cannot establish the content of the defama- tory statements or, at least, cannot establish enough content to permit a court to competently analyze defendants’ First Amendment public comment defense. We conclude that the lack of the text is not fatal to plaintiff’s libel claim. Cite as 369 Or 806 (2022) 819

Since Wright removed his Google review, it is no longer available. In lieu of the unavailable writing, plaintiff relies on the testimony of four people who read the review to supply the content of the statements that he contends were defamatory. We reject defendants’ contention that plaintiff cannot proceed to trial on his libel claim without a copy of the review. Defendants argue that the loss of the text means that a court cannot competently apply the constitution- ally mandated test to discern whether the public comment defense is available. In Neumann, this court held that, to determine whether speech was capable of defamatory mean- ing (and thus whether it satisfied one of the prongs of the public comment defense), a court had to look to three factors: (1) the “general tenor of the entire work”; (2) the specific con- text and content of the statements including figurative and hyperbolic language; and (3) whether the statement itself is sufficiently factual to be proved true or false. 358 Or at 718. In defendants’ view, because the review is lost, a court lacks sufficient information to apply the test and properly perform its gatekeeping role. For instance, defendants doubt whether a court can discern the “general tenor” of the work without the text. Having a copy of the review would certainly sim- plify matters, but factfinders have long been asked to weigh competing or incomplete evidence and to make credibility determinations. The task is no different here. It is an arti- fact of how libel cases are typically litigated that we might be tempted to think that a copy of the at-issue writing would be required. Because, definitionally, libel has been put into writing, the parties are typically able to supply the court with the writing. But we can easily find similar defamation cases in which no such exact record was available, and the case was nonetheless able to be tried to conclusion. The Court of Appeals aptly compared this libel case to slander cases. Lowell, 306 Or App at 334. Slander cases can be litigated without the benefit of verbatim records of what words were spoken. See, e.g., Worley v. OPS, 69 Or App 241, 243, 686 P2d 404, rev den, 298 Or 334 (1984) (relat- ing the factual background, including that the allegedly 820 Lowell v. Wright

defamatory statements were spoken in a staff meeting, with- out verbatim quotes of the statements). In slander cases, a plaintiff need not prove that the defendant spoke exactly the words alleged in the complaint, but only that the words are in substance the same, i.e., “so many of the words alleged in the declaration as constitute the sting of the charge.” Swift & Co. v. Gray, 101 F2d 976, 981-82 (9th Cir 1939) (internal quotation marks omitted). There is no reason that such a standard should be categorically inapplicable to libel cases because the medium of the speech is different.

Here, four people—plaintiff, Norling, Wright, and Werner—read the review and testified as to its contents. Their accounts largely agree, and they all seem to agree that they have not collectively forgotten a substantial component of the review (e.g., an additional customer service complaint or remark about the store’s instruments). The point of dis- agreement concerns the exact wording of the statement in the review about Steinway piano sales. Defendants’ wit- nesses recall its wording in one way, and plaintiff’s wit- nesses recall it in another.

But, at summary judgment, the trial court must view the evidence in the light most favorable to plaintiff, leaving it to the eventual factfinder to make credibility determinations. ORCP 47 C. And the testimony of the four witnesses is sufficient for a factfinder to find the facts, includ- ing whether the review conveyed that plaintiff’s salesman had misrepresented to Wright that plaintiff was a Steinway dealer or that plaintiff could sell Steinway pianos.

We turn to defendants’ additional argument that the text of the review is necessary to apply First Amendment protections properly. Although it certainly is easier to dis- cern the “general tenor” of a piece or the use of hyperbolic or figurative language with the text in hand, and those are important factors to determine whether a statement implies an assertion of fact, nothing in Neumann or the case law from which it derives suggests that the fact that exact word- ing is disputed means that a trial court must throw up its hands, declare defeat, and grant summary judgment to the defendants. 358 Or at 718-19. Competent evidence going to Cite as 369 Or 806 (2022) 821

the general tenor of the review, its use of language, and the nature of the allegedly actionable statements was available. Thus, a constitutional inquiry was certainly possible, albeit not as straightforward as it would be if the trial court could assure itself that it had all the information that ever existed about the allegedly defamatory writing. Accordingly, we reject defendants’ argument that they were entitled to summary judgment because plaintiff could not produce the exact wording of the review. C. The First Amendment Defense We turn to the second issue on review: whether defendants are entitled to assert their public comment First Amendment defense. As explained above, under Milkovich, for a statement to be protected under the First Amendment, it must (1) be on a matter of public concern and (2) not be susceptible to being proved true or false. 497 US at 19. The parties’ dispute here centers on the public concern prong and, in particular, whether the identity of the speaker or the speaker’s motive can affect whether speech is on a matter of public concern. We begin by describing the parties’ arguments in some detail, as they circumscribe our analysis here. First, plaintiff does not ask us to overrule Neumann. Plaintiff briefly discusses Neumann but argues that it is distinguish- able from the present case. He argues that, because of the procedural posture in which the question about the nature of the statements in Neumann arose, the decision had little to do with whether the speech was on a matter of public con- cern in a constitutional sense. The defendant in Neumann had asserted protection under the Oregon Anti-Strategic Lawsuits Against Public Participation (anti-SLAPP) stat- ute, ORS 31.150. In plaintiff’s view, the court’s comments about the statements at issue being on a matter of public interest concerned the statutory standard under the anti- SLAPP statute, which shifts a modest burden of production to the plaintiff when the allegedly actionable communi- cations are made “in a place open to the public or a pub- lic forum in connection with an issue of public interest[.]” ORS 31.150(2)(c). 822 Lowell v. Wright

Plaintiff also offers reasons to doubt the result of Neumann’s application to the present matter. Plaintiff would have us conclude that speaker identity and motive can affect the public concern analysis and that, under the circum- stances of this case, Wright’s identity as an employee for a competitor and his alleged motive of sinking plaintiff’s busi- ness renders his review not on a matter of public concern. Plaintiff submits that the “attacks by a competing business are not issues of public interest” and argues that the Court of Appeals correctly concluded that the speaker’s motivation is relevant to whether the speech is on a matter of public con- cern. Plaintiff also argues that the Court of Appeals incor- rectly concluded that the record was insufficient to establish a factual issue for trial on motivation; he highlights the fact that Wright concealed who he was when he composed the review as a basis to find that defendants were not speak- ing on a matter of public concern. In response, defendants advance a simple argument: The identity or motive of the speaker is irrelevant to the determination of whether the speech is on a matter of public concern, and Neumann con- trols the outcome here. To address whether the public concern prong of the public comment defense applies under the circumstances of this case, we address both interrelated aspects of the par- ties’ arguments: (1) the degree to which Neumann controls and (2) whether a speaker’s identity and motive affect the public concern analysis. We conclude that Neumann con- trols in the absence of plaintiff’s request that we overrule it, although we have doubts about its approach, and that a speaker’s identity and motive do not affect whether a matter is of public concern. 1. Neumann v. Liles Neumann was a libel action involving a wedding guest’s negative online consumer review of a wedding venue posted on Google Reviews. 358 Or at 708. As discussed, the defendant filed a special motion to strike the plaintiffs’ claim under Oregon’s anti-SLAPP statute, which applies to cases involving written statements presented “in a place open to the public or a public forum in connection with an issue of public interest,” among others. Id. at 709, 725; see Cite as 369 Or 806 (2022) 823

also ORS 31.150(2)(c). The trial court granted the motion and dismissed the claim, and the Court of Appeals reversed. Neumann, 358 Or at 709. Rejecting the defendant’s conten- tion that his review was hyperbolic and mere opinion, the Court of Appeals concluded that some statements in the review were capable of a defamatory meaning and that the plaintiffs had adduced sufficient evidence, if credited, to permit a factfinder to determine that the defendant’s state- ments were defamatory. Id. at 710; see also ORS 31.150(3) (providing that, if a defendant establishes grounds for the motion under subsection (2), the burden shifts to the plain- tiff to establish the probability that “the plaintiff will prevail on the claim by presenting substantial evidence to support a prima facie case”). This court allowed review “to determine how an actionable statement of fact is distinguished from a constitutionally protected expression of opinion in a defama- tion claim and whether the context in which a statement is made affects that analysis.” Neumann, 358 Or at 710. This court began by reviewing the Supreme Court’s decision in Milkovich. 358 Or at 713-16. Neumann described Milkovich as having two prongs for First Amendment pro- tection to apply: first, whether the statements were on a matter of public concern and, second, whether the state- ments were susceptible to true-false analysis. Id. at 714 (cit- ing Milkovich, 497 US at 19-20). Notably, Neumann did not mention that the Court had reserved judgment in Milkovich as to whether its First Amendment limitations applied to nonmedia defendants in defamation cases. See Milkovich, 497 US at 20 n 6 (“In Hepps the Court reserved judgment on cases involving nonmedia defendants, and accordingly we do the same.” (Internal citation omitted.)). Additionally, Neumann’s discussion of whether the wedding review was on a matter of public con- cern was limited to the following: “[Plaintiff] has not disputed that [the defendant’s] state- ments involve matters of public concern, and we readily conclude that they do. [The defendant’s] review was posted on a publicly accessible website, and the content of his review related to matters of general interest to the public, particularly those members of the public who are in the market for a wedding venue.” 824 Lowell v. Wright

358 Or at 720 (citing Unelko Corp. v. Rooney, 912 F2d 1049, 1056 (9th Cir 1990), cert den, 499 US 961 (1991)). The court then turned to the legal question before it: How should a court determine whether an allegedly defam- atory statement is susceptible to a true-false analysis, the second prong of the test in Milkovich? The court in Neumann adopted the approach the Ninth Circuit had crafted in Unelko Corp., 912 F2d at 1053, soon after the Supreme Court decided Milkovich. Neumann, 358 Or at 716-19. That approach involves a three-part inquiry, which we described as “(1) whether the general tenor of the entire publication negates the impression that the defendant was asserting an objective fact; (2) whether the defendant used figurative or hyperbolic language that negates that impression; and (3) whether the statement in question is susceptible of being proved true or false.” Id. at 719. Through that framework, a court considers the work as a whole, the context of the state- ments, and the statements themselves to determine whether a factfinder could conclude that the statements imply a false assertion of objective fact. Id. The Neumann court explained in detail the choice to adopt the Unelko approach, but it did not address a fac- tual difference between the two cases: The Neumann defen- dant was nonmedia (a wedding guest who wrote an online review), 358 Or at 708-09, and the Unelko defendants were media (CBS and Andy Rooney, a broadcaster best known for his appearances on 60 Minutes), 912 F2d at 1050. In Unelko, the Ninth Circuit mentioned but did not dwell on the Supreme Court’s language in Milkovich reserving the question of whether the First Amendment defense applied to nonmedia defendants. See id. at 1056 (quoting Milkovich’s announcement of the rule as applying “at least in situations, like the present, where a media defendant is involved,” Milkovich, 497 US at 19-20, and applying the test without acknowledging CBS and Rooney as media defendants). In short, the Neumann court did not note that whether the defendant is media or not could affect the analysis, even to reject the idea. The court went on to apply Unelko to the wedding guest’s online review, agreeing with the Court of Appeals Cite as 369 Or 806 (2022) 825

that some of the statements were capable of a defamatory meaning, but stating that the question remained “whether they are nevertheless protected under the First Amendment.” Neumann, 358 Or at 719-20. The court addressed each prong of the Milkovich test, and, concluding that the statements were protected expressions of opinion, it reversed the Court of Appeals. Id. at 722.

In short, while the main issue in, and the clear holding of, Neumann was that courts in Oregon would use the Unelko three-part inquiry to determine whether speech was capable of defamatory meaning under Milkovich, the Neumann court decided two other issues with compara- tively little to no discussion: (1) that the First Amendment falsity requirement applied when a private-figure plaintiff made claims against nonmedia defendants and (2) that a scathing internet review was on a matter of “public con- cern.” Neumann allowed the defendant to raise the First Amendment public comment defense because his speech was on a matter of public concern and his review could not reasonably be interpreted as asserting fact. 358 Or at 722. It appears that the parties did not dispute that the review of the wedding venue was of public concern, a term of art in First Amendment defamation doctrine.

In Dun & Bradstreet, Inc. v. Greenmoss Builders, 472 US 749, 105 S Ct 2939, 86 L Ed 2d 593 (1985), the Court had the first occasion to consider how the First Amendment applies to a defamation suit between private-figure plain- tiffs and nonmedia defendants and used “public concern” as a term of art in the defamation context. A building com- pany sued a credit reporting service for defamation for mis- reporting to five subscribers that the company had filed for bankruptcy. Id. at 751. The Court considered how best to balance the state’s interests in protecting its law of defa- mation and First Amendment values given the context. It noted that the First Amendment interest at issue was “less important than the one weighed in Gertz. We have long rec- ognized that not all speech is of equal First Amendment importance. It is speech on matters of public concern that is at the heart of the First Amendment’s protection.” Id. at 758 (internal quotations omitted). The Court explained that the 826 Lowell v. Wright

First Amendment was created to protect the interchange of ideas to effectuate the political and social will of the people and that speech concerning public affairs was the “essence of self-government.” Id. at 759. Accordingly, that speech was worthy of the First Amendment’s greatest protection. Id. On the other hand, speech of only private concern deserved less protection. Id. The Court quoted this court’s decision in Harley-Davidson, 279 Or at 366, extensively and favorably to support that point. See Dun & Bradstreet, 472 US at 760 (“ ‘[There] is no threat to the free and robust debate of public issues; there is no potential interference with a meaningful dialogue of ideas concerning self-government; and there is no threat of liability causing a reaction of self- censorship by the press. The facts of the present case are wholly without the First Amendment concerns with which the Supreme Court of the United States has been strug- gling.’ ”). When balanced against a much weaker First Amendment interest, the state’s interest in protecting its citizens from defamation was much stronger. Id. The Court in Dun & Bradstreet readily concluded that the speech at issue was of private, not public, concern. Id. at 762, 764 (Burger, J., concurring), 774 (White, J., con- curring).5 It applied the “public concern” test (examining the content, form, and context of the speech) originally devel- oped in public employment cases. Id. at 761. The Court noted that the speech was wholly false, clearly injurious, and made available to only five subscribers. Id. at 762. Additionally, there was “simply no credible argument that this type of credit reporting require[d] special protection to ensure that ‘debate on public issues [remained] uninhibited, robust and wide-open.’ ” Id. (quoting and citing New York Times, 376 US at 270). The Court also noted that the speech at issue was unlikely to be chilled based on the rule announced because

5 Dun & Bradstreet was a plurality decision, but five members of the Court (three signing on to the lead opinion and Justices Burger and White concurring separately) agreed that the speech at issue was on a matter of private rather than public concern and saw that as dispositive in distinguishing the case from Gertz. Justices Burger and White wrote separately to express disagreement with the trend in the case law of “constitutionalizing the entire law of libel and slan- der” but agreed that the lead opinion was correct given that case law. Id. at 764 (Burger, J., concurring), 766, 774 (White, J., concurring). Cite as 369 Or 806 (2022) 827

its motivation was profit, which would be best served with accurate information. Dun & Bradstreet, 472 US at 762-63. We draw two lessons from the Supreme Court’s dis- cussion of public concern. First and foremost, the legal rule is that a reviewing court discerns whether speech is on a matter of public concern by looking to its content, form, and context. Id. at 761. Second, Dun & Bradstreet discusses mat- ters of public concern as the “heart of the First Amendment’s protection” because they protect the interchange of ideas required for effective self-government. Id. at 759. Defendants essentially contend that a negative review of a business posted on the internet is categorically speech on a matter of public concern and that we should follow Neumann. Defendants seem to assume that the internet can breathe constitutional importance into speech posted in a way that makes it publicly available. We doubt that sweep- ing proposition. The internet is revolutionary. Scholars, legal practitioners, and laypeople alike have urged that its revo- lutionary nature requires an entirely different set of rules, urging that speech on the internet enjoys unusual protec- tion from influence or restraint. But many innovations were once revolutionary. The telegram allowed messages to travel in a matter of minutes or hours, instead of the days it took for letters to reach their destinations. But a statement com- municated by telegram is no different from the same one communicated by letter. So too with the internet. In prac- tice, there is no difference between a statement being posted on social media, Google reviews, on a sign carried around outside the plaintiff’s home, or written in the sky: The state- ment is the same no matter how it reaches the public. The touchstone principle in evaluating whether speech is on a matter of public concern is whether the speech must be protected to ensure the continuance of vigorous debate on public issues and, by extension, self-governance. Dun & Bradstreet, 472 US at 761-62. The idea that negative remarks about a business’s practices are always (or almost always) necessary to ensure vigorous debate on public issues or are part of self-governance is doubtful. Although such remarks may be made in public or be on a subject that a member of the public finds interesting, those circumstances 828 Lowell v. Wright

do not automatically render them of constitutional impor- tance. None of this is to say that a customer’s review of a business cannot be on a matter of public concern while hew- ing closely to Dun & Bradstreet. Rather, instead of assum- ing that a customer review of a business is on a matter of public concern, a careful and more nuanced approach would likely be to examine the “content, form, and context * * * as revealed by the whole record.” Id. at 761. 2. Application of Neumann We now turn to whether Neumann applies to the present case. Although the court’s analysis on public concern consists of little more than a citation to Unelko, Neumann, 358 Or at 720 (citing Unelko, 912 F2d at 1056), we cannot agree with plaintiff’s characterization of Neumann as hold- ing that, for purposes of the anti-SLAPP statute as opposed to the First Amendment, the wedding review was on a matter of public concern. The court’s focus on the First Amendment belies that characterization. The court explained at the outset of its analysis that the “determination of the legal sufficiency of Neumann’s defamation claim hinges on whether Liles’s statements are protected under the First Amendment[.]” 358 Or at 711. After reviewing state common law and some of the Supreme Court’s First Amendment deci- sions concerning defamation actions, especially Milkovich, the court stated that, “to determine whether a defamatory statement is protected under the First Amendment, the first question is whether the statement involves a matter of public concern.” Id. at 718. If so, “then the dispositive ques- tion is whether a reasonable factfinder could conclude that the statement implies an assertion of objective fact.” Id. at 718-19. Accordingly, considering that plaintiff has not asked us to overrule Neumann, the court’s determination that the online wedding review in Neumann was on a matter of pub- lic concern that triggered the First Amendment is relevant. That negative online review of a wedding venue is in many respects like the negative online review of plaintiff’s piano store in this case, and both cases involved private-figure plaintiffs and nonmedia defendants. In view of plaintiff’s arguments, we apply Neumann and conclude that, under Cite as 369 Or 806 (2022) 829

that case, the online review in this case was on a matter of public concern. 3. The impact of speaker motive and identity on public concern We now turn to the impact of the speaker’s motive or identity on the public concern analysis. Although the Court of Appeals did not treat it as dispositive in this case, it declared that “a speaker’s motive or purpose in speak- ing is relevant to whether speech is protected by the First Amendment.” Lowell, 306 Or App at 339 (emphasis in orig- inal). To support its conclusion, it cited public employment cases from which Dun & Bradstreet derived its public con- cern test, including the following: “[W]hen a public employee speaks not as a citizen upon matters of public concern, but instead as an employee upon matters only of personal interest, absent the most unusual circumstances, a federal court is not the appropriate forum in which to review the wisdom of a personnel decision taken by a public agency allegedly in reaction to the employee’s behavior.”

Connick v. Myers, 461 US 138, 147, 103 S Ct 1684, 75 L Ed 2d 708 (1983). The Court of Appeals further reasoned that the motive of the speaker was properly part of the content, form, and context analysis, as part of the context of a statement, in public employment cases in Oregon. Lowell, 306 Or App at 340 (citing one of its earlier cases). The Court of Appeals also cited Harley-Davidson, 279 Or at 363, 366, as support. In that case, this court held that a fake customer complaint written by a competitor and sent directly to a mutual dis- tributor was not speech on a matter of public concern. The Court of Appeals acknowledged that this court “gave little explanation of its specific reasoning” but concluded that the court derived its result from the touchstone principle that to be speech on a matter of public concern, the speech must be in the interest of democratic dialogue and this court determined that “the interest in democratic dialogue [was] non-existent.” Lowell, 306 Or App at 340. The parties disagree on whether the Court of Appeals was correct. Plaintiff argues that the Court of Appeals was 830 Lowell v. Wright

correct, particularly in treating public employment cases as instructive, and defendants argue the opposite. In the public employment cases, the relationship between whether the topic is one of public concern and the motive of the speaker is as follows: When the topic truly is one of public concern, it is likely that the employee is speaking in her capacity as a citizen, which is protected by the First Amendment irrespective of the fact that the government is also the speaker’s employer. But, when the topic is narrow or one of private concern, it is likely that the employee is speaking in her capacity as an employee, which any employer, regardless of whether that employer happens to be the government, may have a legitimate inter- est in controlling. Plaintiff interprets Dun & Bradstreet as the Supreme Court’s endorsement of a similar approach in the defamation context because it applied the content, form, and context approach crafted for public employment cases. Plaintiff argues that, when a writer makes nega- tive comments about a business (the content), the writer’s status—business competitor or disgruntled customer (the context)—and whether she masks her status in her delivery of those comments (the form) are instructive as to whether the comments truly are on matters of public concern. When the writer is a competitor, the logic goes, it is likely, no mat- ter how the comments are presented, that the subject is one of private concern. When the writer is a disgruntled cus- tomer, it is likely that the matter is one of public concern (albeit not one of overwhelming public concern). That reasoning is mistaken because the logic is not parallel in public employment cases and defamation cases. The public employment cases reason from the topic of the speech to determine which relationship between the par- ties (the citizen-government relationship or the employee- employer relationship) ought to control the outcome. In other words, whether speech is on a matter of public concern is a tool of a larger inquiry in public employment cases. In the context of defamation, whether speech is on a matter of pub- lic concern is the object of the inquiry. Plaintiff asks us to reason from the status of the speaker to determine whether the topic is of public concern, going so far as to suggest that, whenever a competitor or its employee speaks about Cite as 369 Or 806 (2022) 831

another business, the speech is of purely private concern. That is mistaken. Defendants correctly note that the rules arising from the public employment context are intended to balance interests where the two players at issue each have two statuses, citizen-employee and government-employer. In deciding whether the topic of the speech was of public or private concern, the court decides which status controls (the citizen-government or employee-employer) and thereby which interests ought to control (a citizen’s interest in her free speech rights or an employer’s interest in appropriately regulating the workplace). Here, the object of the inquiry is to determine whether the speech is on a topic of public or private concern. The Supreme Court’s decision in Garcetti v. Ceballos, 547 US 410, 126 S Ct 1951, 164 L Ed 2d 689 (2006), supports that analysis. In that case, a public employee wrote a memo on what he believed to be serious internal misconduct pur- suant to his job duties. Id. at 414. He was reassigned and denied a promotion soon after. Id. at 415. He sued, citing First Amendment protections for speech relating to mat- ters of public concern because whether the department was corrupt was a matter of public concern. Id. The Supreme Court held that the memo was not protected by the First Amendment because it was written pursuant to his official duties, i.e., the topic of the speech was part of his duty as an employee, so he wrote it as an employee, not as a citizen. Id. at 421. Thus, the government’s interests as an employer regulating a workplace controlled. Whether the employee, in addition to having a job duty, was motivated to act because he believed that he was addressing an issue of public con- cern was not relevant in determining whether the speech was First Amendment-protected. Id. Focusing on the motive of the speaker also distracts from the question central to the inquiry: Does the speech bear on public discourse, self-governance, or the ordering of society? There is no reason to suppose that the exact same words delivered in the exact same way have different ram- ifications for those areas of constitutional concern because the speaker’s motive is different in the two cases. A useful example of that comes from a Connecticut case, Gleason v. Smolinski, 319 Conn 394, 125 A3d 920 (2015). In that case, 832 Lowell v. Wright

the defendants, family members to a missing person, posted missing person signs around the plaintiff’s neighborhood. Id. at 396-97, 125 A3d at 927. The defendants believed that the plaintiff, the missing person’s girlfriend, was involved in his being missing or dead. Id. at 396, 125 A3d at 927. Although there was substantial evidence that some of the defendants’ motive in posting the signs was to harass the plaintiff so that she would divulge information about the missing person, the Connecticut Supreme Court concluded that the matter was of public concern because it related to a police investigation and finding a missing person. Id. at 433 & n 33, 125 A3d at 949 & n 33. The fact that the defendants were partially motivated by a desire to harm the plaintiff did not detract from the speech being on a matter of public concern, a police investigation of a missing person. Additionally, adoption of the Court of Appeals’ analy- sis would be inconsistent with First Amendment values, which include “secur[ing] the widest possible dissemina- tion of information from diverse and antagonistic sources.” New York Times, 376 US at 266 (internal quotation omitted). Motives like competition, hatred, and ridicule may under- lie speech vital to the public discourse, which the First Amendment is intended to protect. Allowing the govern- ment to scrutinize the motive of a speaker to determine the scope of that speaker’s protection from government enforce- ment against her under the First Amendment invites the government to define the contours of the public discourse according to the motives it finds the worthiest. Rather than giving expression the “breathing space” it needs to survive, id. at 271-72, we would suffocate it. 4. Whether the speech is susceptible to true-false analysis Having concluded that Neumann applies to this case such that the speech is on a matter of public concern and that motive and speaker identity do not alter that outcome, we turn to the last step in applying the First Amendment public comment defense, in view of defendants’ position that we are unable to make that determination on this record. That last step is whether the speech is sus- ceptible to being proved true or false under Milkovich. The First Amendment precludes liability for statements that a Cite as 369 Or 806 (2022) 833

reasonable factfinder could not find to imply an assertion of objective fact. The Court of Appeals applied the Unelko test adopted in Neumann to resolve that question, and we agree with its conclusions.

As discussed above, the three-part inquiry for dis- cerning whether speech is susceptible to a true-false analysis is as follows: “(1) whether the general tenor of the entire publication negates the impression that the defen- dant was asserting an objective fact; (2) whether the defen- dant used figurative or hyperbolic language that negates that impression; and (3) whether the statement in question is susceptible of being proved true or false.” Neumann, 358 Or at 719. In the present matter, we examine how the three- part inquiry applies to three different statements: (1) the remark that plaintiff misrepresented whether his business could sell new Steinway pianos, (2) the statement that the salesman misrepresented the age of the Yamaha C-7 piano, and (3) “this guy can’t be trusted.”

We agree with the Court of Appeals that the Steinway and Yamaha statements are sufficiently factual to be actionable but that “this guy can’t be trusted” is not. Lowell, 306 Or App at 343. Starting with the last prong of the Unelko test, whether plaintiff’s business misrepre- sented its ability to sell new Steinway pianos and whether the salesman misrepresented the age of the Yamaha piano are factual matters with truth values. Straightforward sets of facts would make Wright’s statements in the review either true or false. Applying the two other prongs of the Neumann inquiry does not alter that result with respect to those two statements. Based on the record of the contents of the review, nothing in the review’s “general tenor” would negate the impression that the writer really was asserting that the business and its employees were misrepresenting facts about pianos to customers. Although there is some evi- dence that the review used evocative language, such as the store “smelled like grandma’s attic,” that language is not so figurative or hyperbolic as to undermine a reader’s impres- sion that the review is alleging that plaintiff’s business lied to the writer about selling new Steinway pianos and the age of the Yamaha piano on display. See Milkovich, 497 US at 21 834 Lowell v. Wright

(considering whether the writer’s use of “loose, figurative, or hyperbolic language” or the “general tenor of the article” negated the impression that the writer “was seriously main- taining that petitioner committed the crime of perjury”). The analysis differs with respect to the “this guy can’t be trusted” remark. In isolation, the statement is sub- jective and not susceptible to being proved true or false. And, viewing the remark in the context of the whole review does not alter that result. As the Court of Appeals explained in its opinion, Lowell, 306 Or App at 345, the remark can be best understood as a conclusion that the writer drew from the “facts” presented in the review: The business misrepre- sented whether it could sell new Steinway pianos and the age of a piano on display, and therefore its owner “can’t be trusted.” The writer’s conclusion implies no facts beyond those already offered in the review and is not actionable. 5. Whether to alter defamation law in Oregon by follow- ing Obsidian Finance Because the public comment First Amendment defense is available to defendants and applies to two state- ments, we address what plaintiff’s burden is to show defen- dants’ fault and the media/nonmedia distinction that applies when a private figure plaintiff sues a nonmedia defendant. Defendants ask that, if Neumann did not already abolish the distinction between media and nonmedia defendants in defamation claims brought by private figures, we now should do so and follow the Ninth Circuit’s approach in Obsidian Finance Group, LLC v. Cox, 740 F3d 1284, 1291 (9th Cir 2014). We decline to abolish the media/nonmedia distinction. The principles of stare decisis counsel that result for a few reasons. First, the facts of the case before us are far from an ideal vehicle for considering the nuances of the question with which we are confronted, and amici’s hypo- theticals are no substitute. Second, we are not persuaded by the strength of Obsidian Finance’s logic and support to overrule our existing precedent. At the outset, we note that “the party seeking to change a precedent must assume responsibility for affirma- tively persuading us that we should abandon that precedent,” and we assume, grounded in the principle of stare decisis, Cite as 369 Or 806 (2022) 835

that “fully considered prior cases are correctly decided.” State v. Ciancanelli, 339 Or 282, 290, 121 P3d 613 (2005). Stare decisis is “a prudential doctrine that is defined by the competing needs for stability and flexibility in Oregon law.” Farmers Ins. Co. v. Mowry, 350 Or 686, 697-98, 261 P3d 1 (2011). Importantly, “[s]tare decisis does not permit this court to revisit a prior decision merely because the court’s current members may hold a different view than its pre- decessors about a particular issue. At the same time, stare decisis is not absolute.” Couey v. Atkins, 357 Or 460, 485, 355 P3d 866 (2015). This court held in Wheeler that plaintiffs must prove that a defendant acted with actual malice to obtain presumed damages only if the defendant is a media defen- dant. 286 Or at 110. That has been an established rule of the law of defamation in Oregon for decades. Wheeler and its related cases do not suffer from some of the faults that have compelled us to overrule precedent in the past. The rule was not adopted in what amounted to dicta or without explanation. See Couey, 357 Or at 485 (identifying the above as a reason to overcome the application of stare decisis). Wheeler discussed its decision in detail, relying on Gertz’s own emphasis on the fact that it dealt with media defen- dants. See 286 Or at 108-10 (quoting Gertz extensively). Nor was Wheeler’s analysis clearly incorrect, see Couey, 357 Or at 485 (identifying that as an additional reason to overcome the application of stare decisis): Gertz did not itself resolve whether it applied to nonmedia defendants. Defendants and amici urge that the time has come to overturn Wheeler and to abolish the media/nonmedia dis- tinction because it has become incompatible with modern times and technology. First, they argue that the distinction creates a “double standard” at odds with the Supreme Court’s current approach. They note that the Supreme Court has said that the press has no special speech privileges distinct from those of other speakers, see Citizens United v. Federal Election Comm’n, 558 US 310, 352, 130 S Ct 876, 175 L Ed 2d 753 (2010), and that all speakers, whether or not members of traditional media, should have the same standards of lia- bility. Second, defendants and amici argue that Wheeler is at odds with the majority approach of most federal appellate 836 Lowell v. Wright

courts, including the Ninth Circuit. Given the split between Oregon state law and the law binding the Ninth Circuit, defendants and amici argue that Oregon defamation defen- dants will be less protected than their out-of-state coun- terparts facing similar suits in diversity cases in federal court. Amici further argue that Wheeler conflicts with the approach of the majority of states that have addressed the question. Defendants urge, in summary, that “[t]here should be no different constitutional analysis for a news reporter[ ] (media defendant), a food-critic blogger (arguably media- defendant or non-media defendant), a private citizen review on Google or Yelp (non-media defendant), and any other per- son who posts a consumer review that is accessible to the public.” We begin with defendants’ concerns about the changing times and later address their arguments based on case law. Notably, defendants’ arguments have little to say about the facts of this case, with good reason: Wright is not a blogger or a food critic. He is an individual who wrote a review of a retailer and posted it to the internet, just as millions of other Americans do. Artistic Piano is just a store that sells musical instruments and accessories. At bottom, defendants argue that the media/nonmedia distinction should be abandoned as clearly wrong because it presents a difficult line-drawing problem, but they make that argu- ment in a case in which the line is entirely clear: Defendants are not media under any workable definition. We also note that the resolution of the issue may make little practical difference under the circumstances of this case: Plaintiff is correct that he could make the required showing of actual malice to overcome summary judgment. Plaintiff’s theory of the case is not that the review’s statements are just false; it is that the conversation described in the review was entirely made up, and plaintiff adduced evidence in support of that theory. Because at summary judgment all reasonable infer- ences are drawn in the light most favorable to the non- moving party, a factfinder may very well conclude that the conversation reported in the review never took place. And, if defendants described an entirely made-up conversation, anything they claim was “said” in it would be made knowing that it was false or with reckless disregard for the truth, i.e., Cite as 369 Or 806 (2022) 837

with actual malice, the very fault standard for which defen- dants advocate. In short, we have before us a case in which the defendants are admittedly not media and may very well have acted with actual malice. Defendants do not engage with those facts in their efforts to persuade us to overrule the media/nonmedia distinction for purposes of determin- ing what showing of defendants’ fault plaintiff must make. Stare decisis is not mechanistic, Mowry, 350 Or at 697, but it is demanding. And it does not permit us to destabilize over 40 years of precedent on the strength of a few hypotheticals and some abstract concerns about modernity. We acknowl- edge the competing need for flexibility in a modern world and recognize that old rules can become outmoded with the passage of time, see id. at 697-98, but this case simply does not present an example of how the rule has become outmoded. It instead presents facts to which the rule can be straightforwardly applied.6 In the absence of an argument that the facts illus- trate why the existing media/nonmedia distinction ought to be overruled, defendants attempt to meet their burden to show why precedent should be abandoned by appealing to case law from other jurisdictions, particularly Obsidian Finance. We are not persuaded that defendants’ argument meets that burden, nor do we find Obsidian Finance’s logic persuasive enough to convince us that our longstanding approach ought to be abandoned. In Obsidian Finance, the Ninth Circuit acknowl- edged that the Supreme Court has not directly addressed whether its First Amendment defamation rules apply equally to the institutional press and to private, non- media defendants. 740 F3d at 1291. However, the Ninth Circuit held that a defendant blogger’s allegations that a

6 Our colleagues criticize us for not announcing a test to distinguish media from nonmedia. 369 Or at 842 (Balmer, J., concurring). The rationale for not offering a test follows from the discussion above: The facts of this case do not lend themselves to the nuances that we would need to consider to craft one, and amici’s hypotheticals are no substitute. We have no difficulty applying the dis- tinction to the facts of this case: Wright and Artistic Piano are clearly not media defendants, and therefore the plaintiff need not show actual malice to obtain relief. 838 Lowell v. Wright

bankruptcy trustee was corrupt were protected by the First Amendment, though the defendant was not a trained jour- nalist and “apparently ha[d] a history of making similar allegations and seeking payoff in exchange for retraction.” Id. at 1287, 1291. The court’s rationale, like defendants’ and amici’s argument before this court, was largely based on decisions by the Supreme Court—in contexts other than defamation claims—rejecting constitutional privileges for the institutional press greater than those available to indi- viduals engaged in the same activities. See id. at 1290 (cit- ing cases). Citing decisions from other circuits, the Ninth Circuit concluded that the First Amendment defamation rules “apply equally to the institutional press and individ- ual speakers.” Id. at 1291. We stand by our previous reasoning, that the legal context here, a defamation claim, matters for the purposes of whether and how the First Amendment must alter state common law. Defamation law in Oregon has developed over time mainly in the common-law tradition (the anti-SLAPP statute is a notable exception). The common-law libel cause of action exists to provide remedies in individual cases of harm caused by false speech, and liability is by no means a foregone conclusion, even in cases of libel per se, as a result of common-law privileges and defenses recognized in Oregon. We have already discussed in detail how the Supreme Court approaches defamation cases. In our view, the Supreme Court’s approach in state common-law defama- tion cases is different from cases that involve a jurisdiction that enacts and seeks to enforce laws against speakers that are not content-neutral or suffer from similar constitutional defects. See Hepps, 475 US at 777 (acknowledging that “a suit by a private party is obviously quite different from the government’s direct enforcement of its own laws”). And, the Court has recognized that the common-law speech-based torts are varied and call for an appreciation of the specific context at issue. See, e.g., Zacchini v. Scripps- Howard Broadcasting Co., 433 US 562, 97 S Ct 2849, 53 L Ed 2d 965 (1977). In Zacchini, the Ohio Supreme Court had held that a television news broadcast of the plaintiff’s “human cannonball” act in its entirety without the perform- er’s permission was protected as reporting on a matter of Cite as 369 Or 806 (2022) 839

public interest, relying on the Court’s decision in a false light case, Time, Inc. v. Hill, 385 US 374, 87 S Ct 534, 17 L Ed 2d 456 (1967). The Supreme Court reversed, noting that the tort in Time “involved an entirely different tort” from the plaintiff’s cause of action for appropriation of his right of publicity under Ohio law, Zacchini, 433 US at 571, and explained that the “differences between these two torts are important,” id. at 573. The Court also recognized that its line of cases involving the First Amendment and defama- tion, including Gertz, were inapplicable. Id. at 574. As a result, we are not persuaded that the Supreme Court’s First Amendment cases involving other legal con- texts are or should be determinative as to state law defa- mation cases. For example, in Bartnicki v. Vopper, 532 US 514, 121 S Ct 1753, 149 L Ed 2d 787 (2001), a case cited in Obsidian Finance, 740 F3d at 1290, the plaintiffs’ mobile phone conversation was unlawfully intercepted and recorded by an unknown third party during a labor dispute, and one defendant gave the recording to the media defendants. After the recording was broadcast over radio and published in newspapers, the plaintiffs brought claims for damages based on federal and Pennsylvania wiretap acts that pro- hibited disclosure of content of communications that a party has reason to know were obtained unlawfully. The defen- dants asserted that, if they violated the wiretap laws, their disclosures of the conversation between the plaintiffs were protected by the First Amendment. Id. at 518-21. The Third Circuit Court of Appeals evaluated the statutes and concluded that they deterred more speech than necessary to protect the privacy interests involved and there- fore reversed the district court’s denial of the defendants’ motion for summary judgment. Id. at 521-22. Applying its framework for reviewing statutes challenged under the First Amendment, the Supreme Court agreed that the stat- utes were content-neutral. Id. at 526-27. However, the Court noted that it repeatedly had held that, if the press has law- fully obtained truthful information, state action punishing publication of information of public concern will generally be unconstitutional. Id. at 527-28. The Court concluded that the call between the union president and the union’s chief negotiator, in which the president threatened use of physical 840 Lowell v. Wright

harm during ongoing negotiations over the terms of com- pensation for teachers at the public high school, was on a matter of public concern. Id. at 535. Thus, the question in the case was whether the First Amendment protected the defendants when they had reason to know that the intercep- tion of the phone call was unlawful. The Court in Bartnicki acknowledged the communication privacy interest that the statutes protected but concluded that, based on the facts of the case—particularly the kind of speech that was publi- cized and the fact that none of the defendants had performed the interception—all the defendants were protected from liability for the publication. Id.7 Similarly, Cohen v. Cowles Media Co., 501 US 663, 111 S Ct 2513, 115 L Ed 2d 586 (1991) (contract claim for breach of confidentiality), and First National Bank of Boston v. Bellotti, 435 US 765, 767, 98 S Ct 1407, 55 L Ed 2d 707 (1978) (freedom of speech challenge to a Massachusetts criminal statute forbidding banks and cor- porations from making campaign contributions or expendi- tures to influence public votes on referenda other than those affecting their property, business, or assets), arise in con- texts that are dissimilar to common-law defamation claims. Finally, we have reviewed the other circuit court decisions that the Ninth Circuit also treated as persua- sive in concluding that the First Amendment defamation rules apply both to the institutional press and nonmedia defendants. Three of the cases are not on point, because they do not involve private figure plaintiffs and nonmedia defendants, thus obviating the need to decide the question of how the First Amendment applies in a defamation claim between two wholly private, nonmedia parties. See Garcia v. Bd. of Ed. of Socorro Consol. Sch. Dist., 777 F2d 1403, 1408 (10th Cir 1985) (the plaintiffs, school board members 7 It is unclear why the Court relieved all defendants of liability. See id. at 525 n 8 (stating only that “we draw no distinction between the media respondents and Yocum[, the individual who gave the recording to the media]. See, e.g., New York Times Co. v. Sullivan, 376 U.S. 254, 265-266, 84 S. Ct. 710, 11 L. Ed. 2d 686 (1964); First Nat. Bank of Boston v. Bellotti, 435 U.S. 765, 777, 98 S. Ct. 1407, 55 L. Ed. 2d 707 (1978)”). But, as in New York Times and the Court’s other cases involving publication by the press of information it obtains on matters of public concern, it is apparent that the Court was concerned with freedom of the press. That concern, rather than an interest in treating all defendants alike in defama- tion and other cases with a First Amendment dimension, would be a likely basis for the Court’s treatment of Yocum. Cite as 369 Or 806 (2022) 841

in a defamation cross-claim against a fired school superin- tendent, were public officials); Avins v. White, 627 F2d 637, 648 (3rd Cir 1980) (the plaintiff was a public figure); Davis v. Schuchat, 510 F2d 731, 734 n 3 (DC Cir 1975) (the defen- dant was a reporter). Although another case stated that, whether the defendant is media or nonmedia is “irrelevant to the question of what level of constitutional protection that right is to receive,” In re IBP Confidential Business Documents Litigation, 797 F2d 632, 642 (8th Cir 1986), that case concerned the degree to which the First Amendment right to petition was implicated and should be protected. It is unsurprising that the Eighth Circuit concluded that the media/nonmedia distinction should be irrelevant when the context is the defendant’s exercise of the right to petition the government. And we are not persuaded by the two remaining cases, which emphasize the difficulties in defining media. See Snyder v. Phelps, 580 F3d 206, 219 n 13 (4th Cir 2009), rev’d on other grounds, 562 US 443, 131 S Ct 1207, 179 L Ed 2d 172 (2011) (stating—as to an intentional infliction of emotional distress claim—that “[a]ny effort to justify a media/nonmedia distinction rests on unstable ground, given the difficulty of defining with precision who belongs to the ‘media’ ”); Flamm v. American Ass’n of University Women, 201 F3d 144, 149 (2d Cir 2000) (“We agree that a distinc- tion drawn according to whether the defendant is a member of the media or not is untenable.”). Thus far, the Supreme Court itself has recognized that distinction in defamation cases. The values underlying defamation claims have been recognized by this court for over 150 years, see Neumann, 358 Or at 711 (citing Hurd v. Moore, 2 Or 85 (1863)), and in the common law for much longer, see Milkovich, 497 US at 11 (noting that the common-law cause of action has existed since “the latter half of the 16th century”). As discussed ear- lier, the Supreme Court has repeatedly recognized the strong interests of the states in providing remedies for harms to reputation and the individual and societal benefit in pre- venting and remedying invidious false speech. Considering the Court’s precedents in defamation cases and its overall approach to assessing the interests of speakers in light of 842 Lowell v. Wright

the legal context (and concomitant competing interests), we are not persuaded that the Supreme Court requires states to alter their common law by applying First Amendment protections in defamation cases brought by private figures against nonmedia defendants. We therefore decline to over- rule existing precedent based on Obsidian Finance. In sum, we affirm the decision of the Court of Appeals in part on different grounds. On remand to the trial court, plaintiff’s libel claim is subject to the First Amendment defense, but plaintiff will not be subject to a heightened proof-of-fault requirement based on the Supreme Court’s First Amendment cases that apply to media defen- dants in defamation cases. The decision of the Court of Appeals is affirmed in part and reversed in part. The judgment of the circuit court is reversed, and the case is remanded to the circuit court for further proceedings. BALMER, J., concurring. The majority holds that Oregonians who publish a review of a local business may be liable for defamation without any showing of fault on their part. In so doing, the opinion upholds an untenable distinction between “media” and “nonmedia” defendants without clearly articulating the differences between the two. In my view, any approach that retains that distinction must be supported by a workable test to delineate media from nonmedia, and the majority makes no attempt to provide that test. I would follow the federal and state courts that have rejected that distinction for purposes of defamation claims by a private figure and would hold that the First Amendment equally protects the media and private individuals. The majority also casts doubt on this court’s deci- sion in Neumann v. Liles, 358 Or 706, 369 P3d 1117 (2016), based on arguments that are not made by the parties and are unrelated to the case before us. I would not reach beyond the facts of this case to call into question a recent decision of this court that no party here has challenged and that was correctly decided. Cite as 369 Or 806 (2022) 843

Thus, although I agree with much of what the major- ity opinion holds—including its discussion of the missing text of the review, its conclusion regarding the role of motive in identifying speech of public concern, and its ultimate disposition—I disagree with key parts of the majority’s First Amendment analysis. I therefore respectfully concur in the judgment, but not in all of the analysis.

At issue in this case are two First Amendment pro- tections that apply in defamation cases, which I address in turn. The first is that a plaintiff must show that a defendant acted with some level of fault, with the specific level of fault depending on the identities of the parties. The second is that a statement must be provably false to be actionable if the statement involves matters of public concern.

As to the first protection, the Supreme Court has indicated that the level of fault that a plaintiff must show varies with the circumstances. If the plaintiff is a public fig- ure, for example, the plaintiff must show that the allegedly defamatory statement was made with “actual malice.” See New York Times Co. v. Sullivan, 376 US 254, 283, 84 S Ct 710, 11 L Ed 2d 686 (1964); Curtis Publishing Co. v. Butts, 388 US 130, 155, 87 S Ct 1975, 18 L Ed 2d 1094 (1967). A plaintiff who is not a public figure must show that a defen- dant acted with negligence (or some higher level of fault) to recover, at least in cases involving media defendants. See Gertz v. Robert Welch, Inc., 418 US 323, 347, 94 S Ct 2997, 41 L Ed 2d 789 (1974). Gertz further held that, to recover presumed or punitive damages, a plaintiff who is not a pub- lic figure must show that the defendant acted with actual malice. Id. at 349.

Following Gertz, state and federal courts have split on whether Gertz’s fault requirements apply in cases involv- ing nonmedia defendants, as described below. For its part, this court held several decades ago that Gertz did not apply to nonmedia defendants, as the majority explains. 369 Or at 818 (citing Harley-Davidson v. Markley, 279 Or 361, 371, 568 P2d 1359 (1977); Adams v. State Farm Mutual Auto. Ins. Co., 283 Or 45, 51-52, 581 P2d 507 (1978); Wheeler v. Green, 286 Or 99, 110, 593 P2d 777 (1979)). 844 Lowell v. Wright

That approach to Gertz, however, fails to engage with the increasingly difficult question of how to distin- guish clearly between media and nonmedia defendants and, for that reason, places this court at odds with recent decisions from other state and federal courts. Those cases, and the rapidly changing nature of media, communications, and public discourse, counsel that reevaluating the media/ nonmedia distinction may lead to a sounder approach to pro- tecting our “profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open.” New York Times, 376 US at 270. In short, I would reject the media/nonmedia distinction to which the majority adheres and would reconsider this court’s lim- itation of Gertz to media defendants in Harley-Davidson, Wheeler, and Adams, in light of more recent case law and the changing media landscape. As identified by the majority, one helpful recent fed- eral case is Obsidian Finance Group, LLC v. Cox, 740 F3d 1284, 1291 (9th Cir), cert den, 572 US 1142 (2014), where the Ninth Circuit held that the Gertz fault rules should apply equally to media and nonmedia defendants. 369 Or at 834. Obsidian Finance noted that, of the six federal circuit courts that had then reached the issue, all had extended the First Amendment protections of “[New York Times v.] Sullivan and its progeny,” including Gertz, to media and nonmedia defendants. 740 F3d at 1291. The Ninth Circuit agreed with that approach, observing: “The protections of the First Amendment do not turn on whether the defendant was a trained journalist, for- mally affiliated with traditional news entities, engaged in conflict-of-interest disclosure, went beyond just assembling others’ writings, or tried to get both sides of a story. As the Supreme Court has accurately warned, a First Amendment distinction between the institutional press and other speak- ers is unworkable * * *.” Id. The Court therefore applied the Gertz negligence require- ment for private defamation actions in that case involving a nonmedia defendant. The majority rejects Obsidian Finance and the cases it cites as variously inapposite or unpersuasive. 369 Or at Cite as 369 Or 806 (2022) 845

840-41. The majority is correct that some of those cases are not precisely on point here, because they considered only whether the New York Times fault rule for defamation actions by public figures extended to nonmedia defendants (as it already does in Oregon, see Wheeler, 286 Or at 110-11 (“We conclude that all defendants, not only those associated with the media, continue to be protected by the New York Times rule in cases involving comment upon public officials and public figures.”)) rather than the Gertz fault rule for plaintiffs that are not public figures. E.g., Garcia v. Bd. of Ed. of Socorro Consol. Sch. Dist., 777 F2d 1403, 1410 (10th Cir 1985), cert den, 479 US 814 (1986); Avins v. White, 627 F2d 637, 649 (3rd Cir), cert den, 449 US 982 (1980). At the same time, however, the majority disregards the significance of the cases cited by Obsidian Finance as a set as well as the persuasive reasoning of the Ninth Circuit. Together with Obsidian Finance, those cases show that recent federal court decisions have uniformly rejected the media/nonmedia distinction in the First Amendment context and have done so for two primary reasons. First, in the context of the evolving communications and media landscape, with the boundary between traditional media and new or social media disappearing, that distinction is “unworkable,” Obsidian Finance, 740 F3d at 1291, “rests on unstable ground,” Snyder v. Phelps, 580 F3d 206, 219 n 13 (4th Cir 2009), aff’d, 562 US 443, 131 S Ct 1207, 179 L Ed 2d 172 (2011), or is “untenable,” Flamm v. American Ass’n of University Women, 201 F3d 144, 149 (2d Cir 2000). Second, that distinction is at odds with the fundamental First Amendment principle that the value of speech “ ‘does not depend upon the identity of its source, whether corporation, association, union, or individual.’ ” In re IBP Confidential Bus. Documents Litigation, 797 F2d 632, 642 (8th Cir 1986), cert den, 479 US 1088 (1987) (quoting First National Bank of Boston v. Bellotti, 435 US 765, 777, 98 S Ct 1407, 55 L Ed 2d 707 (1978)). The federal courts are not alone. Several state courts have taken a similar tack, either by explicitly apply- ing Gertz to nonmedia defendants or applying a function- ally equivalent fault-based rule under state law. See, e.g., Antwerp Diamond Exch. v. Better Bus. Bur., 130 Ariz 523, 846 Lowell v. Wright

528, 637 P2d 733, 738 (1981) (applying Gertz to nonmedia defendant); Bierman v. Weier, 826 NW2d 436, 470-71 (Iowa 2013) (Hecht, J., concurring in part and dissenting in part) (listing cases and noting that 22 state and federal jurisdic- tions apply Gertz to nonmedia defendants, while only eight states, including Oregon, have held that Gertz does not apply to nonmedia defendants); Lester v. Powers, 596 A2d 65, 69 (Me 1991) (holding that Maine common law requires a showing of “fault amounting at least to negligence” in def- amation suits against nonmedia defendants); Jacron Sales Co., Inc. v. Sindorf, 276 Md 580, 592, 350 A2d 688, 695 (1976) (“[W]e conclude as a matter of state law that the Gertz hold- ing should apply to media and non-media defendants alike * * *.”); Maethner v. Someplace Safe, Inc., 929 NW2d 868, 878-79 (Minn 2019) (concluding that the limitation on pre- sumed damages in Gertz applies equally to media and non- media defendants in suits by private plaintiffs); Bender v. City of Seattle, 99 Wash 2d 582, 599, 664 P2d 492, 503 (1983) (citing Gertz and requiring a showing of fault in actions by private individuals against nonmedia defendants). At least one state has gone even farther and extended the New York Times actual malice requirement to cases involving private plaintiffs and nonmedia defendants based on statements involving matters of public concern. See Durando v. Nutley Sun, 209 NJ 235, 250, 37 A3d 449, 458 (2012) (“Today, in New Jersey the actual-malice standard protects both media and non-media defendants who make statements involving matters of public concern, regardless of whether the targets of the statements are public figures or private persons.”). The Restatement (Second) of Torts takes the same approach as many of those cases, stating that, although the precise holding of Gertz was limited to media defendants, “the principle of the Gertz decision would appear to be broad enough to cover” situations involving only private, nonmedia individuals. Restatement (Second) of Torts § 580B comment e (1977). As the Restatement explains, “It would seem strange to hold that the press, composed of professionals and causing much greater damage because of the wider distribution of the communication, can con- stitutionally be held liable only for negligence, but that a private person, engaged in a casual private conversation Cite as 369 Or 806 (2022) 847

with a single person, can be held liable at his peril if the statement turns out to be false, without any regard to his lack of fault.” Id. The Restatement goes on to explain that, even if Gertz itself is limited to media defendants, “the common law of the states is almost certain to apply the same standard” of requiring a fault showing for private defamation actions. Id. The majority responds by emphasizing that the def- amation context is unique among First Amendment appli- cations, and states, “Thus far, the Supreme Court itself has recognized that [media/nonmedia] distinction in defamation cases.” 369 Or at 841. But the Court itself has never taken up the question of whether Gertz applies to nonmedia defen- dants. To the contrary, when the Court has considered the media/nonmedia distinction, even in the context of defama- tion, the Court has consistently declined to embrace that distinction and just as often has strongly cautioned against employing it. For example, in Dun & Bradstreet, Inc. v. Greenmoss Builders, 472 US 749, 105 S Ct 2939, 86 L Ed 2d 593 (1985), the Court considered a defamation action by a private figure (a construction contractor) against a nonmedia defendant (a credit reporting agency). In resolving that case, the Court did not rely on a media/nonmedia distinction, but instead concluded that the speech at issue was not of public concern. See id. at 772-73 (White, J., concurring in the judgment). The majority discusses Dun & Bradstreet, 369 Or at 825-27, but it does not mention that, although the plu- rality opinion did not decide the issue, a majority of the Court in that case nevertheless explicitly rejected the media/nonmedia distinction in the defamation context. At least five justices agreed that, “in the context of defa- mation law, the rights of the institutional media are no greater and no less than those enjoyed by other individuals or organizations engaged in the same activities.” Id. at 784 (Brennan, J., dissenting, joined by Marshall, Blackmun, and Stevens, JJ.); see also id. at 773 (White, J., concurring in the judgment) (“I agree with Justice BRENNAN that the First Amendment gives no more protection to the press in defa- mation suits than it does to others exercising their freedom 848 Lowell v. Wright

of speech. None of our cases affords such a distinction; to the contrary, the Court has rejected it at every turn.” (Uppercase in original; emphasis added.)). The Court reaffirmed that principle more recently in Citizens United v. Federal Election Comm’n, 558 US 310, 340, 130 S Ct 876, 175 L Ed 2d 753 (2010), where the Court noted that treating media differently from nonmedia in the First Amendment context is particularly fraught. “With the advent of the Internet and the decline of print and broadcast media,” the Court observed, “the line between the media and others who wish to comment on political and social issues becomes far more blurred.” Id. at 352. As a result, the Court has “consistently rejected the proposition that the institu- tional press has any constitutional privilege beyond that of other speakers.” Id. (emphasis added). Thus, although the Court has never specifically held that Gertz applies to a non- media defendant, the majority’s brief assessment that “the Supreme Court has recognized that distinction in defama- tion cases,” 369 Or at 841, is belied by the Court’s own state- ments emphasizing that it has drawn no such line. Few would disagree that, as the Court observed in Citizens United in 2010, the line between the media and non- media is indeed “blurred” and that it has only become more so in the last decade. Even if the Court had endorsed such a distinction in the abstract, we would be left with the diffi- cult task of crafting a workable test to distinguish between types of defendants. It is no secret that, in today’s world of internet-based communication and social media, private individuals and groups who are not part of any kind of “tra- ditional media” can share similarly powerful platforms for the dissemination of speech on matters of public concern. The difficulty in drawing the media/nonmedia distinction has also been highlighted by other courts and scholars. See, e.g., Snyder, 580 F3d at 219 n 13; Flamm, 201 F3d at 149; Ryan M. Walters, When Can You Shoot the Messenger? Understanding the Legal Protections for Entities Providing Information on Business Products and Services in the Digital Age, 96 Or L Rev 185, 191 (2017) (“The bar- rier between a news organization and an individual has never been lower.”); Clay Calvert, Emma Morehart & Sarah Cite as 369 Or 806 (2022) 849

Papadelias, Plausible Pleading & Media Defendant Status: Fulfilled Promises, Unfinished Business in Libel Law on the Golden Anniversary of Sullivan, 49 Wake Forest L Rev 47, 73-83 (2014) (describing three reasons why the media/ nonmedia distinction “should be obliterated”: difficulties in defining “media” in a digital world, the democratization of media empowering private individuals to respond to defa- mation, and notions of speaker equality espoused in Citizens United, as discussed above). Even if a workable distinction between media and nonmedia might be possible, the majority makes no attempt to explain how that distinction would be made. The major- ity asserts that defendants are “not ‘media’ under any defi- nition.” 369 Or at 809. But in making that assertion, the majority does not provide any definition of media. Although dictionary definitions are of limited use in discerning the meaning of Supreme Court opinions, a few examples never- theless illustrate that defendants are “media” under at least some definitions of the term. For example, Black’s Law Dictionary defines media as certain means of commu- nication: “Collectively, the means of mass communication; specif., television, radio, newspapers, magazines, and the Internet regarded together.” Black’s Law Dictionary 1175 (11th ed 2019). Here, the allegedly defamatory statements were made through the internet, which is one of those listed “media.” More importantly, the statements were posted as a review on Google. A central purpose of posting an online review is to make the writer’s views regarding a product or service available to the broader public—including poten- tially any person in the world with an internet connec- tion and a browser. The extent of that potential reach is exactly the “means of mass communication” we think of as “media.” Under the above definition, defendants here could be deemed to have been acting as “media” insofar as they created content about a local business for broad online pub- lication. Defendants, of course, might not fit the mold of tra- ditional “media,” as the majority seems to understand that term, such as major newspapers or television broadcasters. But the very fact that defendants fit the terms of a definition of “media,” but do not seem to fit the unexpressed terms of 850 Lowell v. Wright

the majority’s definition of “media,” highlights the difficulty and ambiguity in clearly distinguishing between media and nonmedia, particularly where user-generated internet con- tent is involved.

Without an explanation of how to distinguish between media and nonmedia, the remainder of the opinion relying on that distinction is unpersuasive. Various federal and state courts have rejected that distinction, and I find Obsidian Finance’s explanation of why to do so persuasive. As a result, I conclude that that distinction should not hold in defamation cases.

If the Gertz rule were applied here, plaintiff would have to allege and prove at least negligence on defendants’ part to be able to recover for defamation at all, and he would have to show actual malice to recover for his defamation per se claim, which may proceed “without proof of specific harm.” Brown v. Gatti, 341 Or 452, 458, 145 P3d 130 (2006); see Gertz, 418 US at 350 (“In short, the private defamation plaintiff who establishes liability under a less demanding standard than that stated by New York Times[, i.e., actual malice,] may recover only such damages as are sufficient to compensate him for actual injury.”). The majority asserts that applying that actual malice standard “may make little practical difference under the circumstances of this case,” because “Plaintiff is correct that he could make the required showing of actual malice to overcome summary judgment.” 369 Or at 836. I disagree that requiring that showing would be of “little practical difference” to defendants. A heightened burden of proof for plaintiff could change whether defen- dants are liable for their alleged defamation—which is why defendants raised the actual malice argument. I agree, how- ever, that plaintiff has raised a triable issue of fact as to whether he may be able to make what would be the required showing of actual malice in this case. Thus, although I dis- agree with the majority’s analysis on this point, I concur in the judgment.

The second type of constitutional protection for speech that applies in defamation cases is the requirement that an allegedly defamatory statement be provably false Cite as 369 Or 806 (2022) 851

to be actionable. That requirement stems from Supreme Court decisions that were recently relied on by this court in Neumann. 358 Or at 713-16 (citing Philadelphia Newspapers, Inc. v. Hepps, 475 US 767, 106 S Ct 1558, 89 L Ed 2d 783 (1986); Milkovich v. Lorain Journal Co., 497 US 1, 110 S Ct 2695, 111 L Ed 2d 1 (1990)). In Neumann, we applied the requirement that statements be provably false to statements involving matters of public concern made by a nonmedia defendant. Id. at 722 (holding that the defendant’s online review was “an expression of opinion on matters of public concern that is protected under the First Amendment”). Neumann also adopted a test for whether a statement is provably false that was articulated by the Ninth Circuit in Unelko Corp. v. Rooney, 912 F2d 1049, 1053 (9th Cir 1990), cert den, 499 US 961 (1991). 358 Or at 718. I agree with the majority’s ultimate conclusion that Neumann controls this case in part and that some, but not all, of defendant’s statements are actionable under the Unelko test. I disagree, however, with other aspects of the majority’s discussion of Neumann.

The majority upholds Neumann begrudgingly, and seemingly only because plaintiff failed to ask us to over- rule it. See 369 Or at 822 (“We conclude that Neumann con- trols in the absence of plaintiff’s request that we overrule it, although we have doubts about its approach * * *.”). As part of that discussion, the majority explains at length some effects of the internet on modern communication and the extent to which the advent and development of the inter- net has changed, or not changed, defamation law in Oregon. 369 Or at 827-28. The majority also questions whether Neumann properly considered whether a media/nonmedia distinction might affect the analysis in that case, and it casts doubt on Neumann’s conclusion that the online review in that case was on a matter of public concern. 369 Or at 823-24, 824, 827-28.

But because the majority upholds Neumann and applies it to this case, the extended critique of that case’s rea- soning does not bear on the result here and is dicta. Whether online reviews should be automatically considered matters of public concern, or whether the internet can “breathe 852 Lowell v. Wright

constitutional importance” into a potentially defamatory statement, are questions that are neither presented by this case nor relevant to its outcome. 369 Or at 827. I would not express an opinion on matters so far removed from the situ- ation at hand.

Nevertheless, in response to the majority’s discus- sion, I would observe that Neumann is far more defensible than the majority suggests. In Neumann, this court “readily conclude[d]” that the statements in that case involved mat- ters of public concern. 358 Or at 720. That ready conclu- sion was not due to a lack of consideration by this court, but rather due to the clarity with which the court under- stood that the character and reputation of a local business can be of great importance to members of the surround- ing community, regardless of whether they are consumers of that business’s services or products. Indeed, it appears that it did not even occur to the litigants in that case to argue otherwise. In Unelko, the Ninth Circuit similarly held that a statement made on 60 Minutes that the prod- uct Rain-X “didn’t work” involved a matter of public concern, because the statement was “of general interest and was made available to the general public,” and because “protec- tion of statements about product effectiveness will ‘ensure that debate on public issues will be uninhibited, robust and wide-open.’ ” 912 F2d at 1056 (quoting Dun & Bradstreet, 472 US at 762 (internal quotation marks and brackets omit- ted)). There should be no doubt that the “content, form, and context” of the statements here indicate that they too involve matters of public concern. Dun & Bradstreet, 472 US at 761 (indicating that “ ‘whether speech addresses a matter of public concern must be determined by the expres- sion’s content, form, and context as revealed by the whole record’ ” (quoting Connick v. Myers, 461 US 138, 148, 103 S Ct 1684, 75 L Ed 2d 708 (1983) (alterations and brackets omitted))).

The majority also casts doubt on Neumann’s hold- ing that the First Amendment’s provably-false rule should apply to nonmedia defendants. See 369 Or at 824 (“In short, the Neumann court did not note that whether the defen- dant is media or not could affect the analysis, even to reject Cite as 369 Or 806 (2022) 853

the idea.”). As the majority recognizes, that question has not been decided by the Supreme Court, but, instead, has been expressly reserved. 369 Or at 817; id. (citing Hepps, 475 US at 779 n 4; Milkovich, 497 US at 20 n 6). But even if Neumann had not resolved the issue for purposes of Oregon law, which it did, this court should have come to the same conclusion in this case and applied the provably-false rule to defendants without regard to their media or nonmedia status. In the absence of such a rule, Oregonians who post online reviews of businesses would have to be prepared to defend the truth of their statements in court or face liability for defamation—even where their statements are not sus- ceptible to being proven either truth or false. The same arguments that weigh in favor of reject- ing the media/nonmedia distinction in the Gertz context, articulated above, also support Neumann’s rejection of that distinction. The majority offers no example of another state that allows its citizens to be sued for defamation on matters of public concern where the statement at issue does not sat- isfy Milkovich’s provably-false standard, and several state courts have taken the opposite approach and agree with Neumann. See, e.g., Dodson v. Dicker, 306 Ark 108, 111, 812 SW2d 97 (1991) (applying Milkovich in a case with a non- media defendant); Kahn v. Bower, 232 Cal App 3d 1599, 1606-07, 284 Cal Rptr 244, 248-49 (1991) (same). Because Neumann is settled law, no party chal- lenges its underlying reasoning (whereas they do chal- lenge the media/nonmedia distinction in Harley-Davidson, Wheeler, and Adams), and its First Amendment conclusions remain sound, I would not cast doubt on Neumann’s reason- ing and its application in this case. In sum, the majority adheres to a distinction between media and nonmedia defendants, for purposes of defamation claims by private figures, that has not been embraced by the Supreme Court and that has been rejected by multiple federal and state courts, numerous scholars, and the Restatement. Although the majority is correct that the Court has not yet applied Gertz to nonmedia defendants, the Court has also consistently rejected the media/nonmedia distinction, and, in its own words, refused to offer the media 854 Lowell v. Wright

“any constitutional privilege beyond that of other speakers.” Citizens United, 558 US at 352. I would not continue to hold that Gertz applies only to media defendants in the absence of a workable method for distinguishing between media and nonmedia, and the majority makes no attempt to provide one. As to Neumann, I would not cast doubt on a case that goes unchallenged by the parties, and I would not express an opinion on matters unrelated to the facts and legal arguments presented by this case. For the foregoing reasons, I disagree with parts of the majority’s First Amendment analysis and concur in the judgment. Garrett, J., joins in this concurring opinion. FLYNN, J., concurring. I agree with the concurrence that Neumann is not only controlling but correct, and that the critique of that case’s reasoning in the majority opinion is dicta. And I am persuaded by the argument in the concurrence that a media/ nonmedia distinction is in tension with the Supreme Court’s more recent First Amendment decisions. But we are not writing on a clean slate. Rather, we are being asked to over- rule our own controlling precedent on the strength of predic- tions about how the Supreme Court ultimately will rule on the question of whether the First Amendment requires proof of fault in cases like this. In my opinion, it is both prema- ture and potentially unnecessary to declare our precedent overruled.