Video Games and the Right of Publicity

The Rise of the Publicity Tort

The right of publicity is historically linked to the law of privacy (Warren & Brandeis, 1890). As privacy law evolved into distinct causes of action, one claim that received both statutory and common-law protection was the appropriation of a person’s name or likeness for commercial purposes (Prosser, 1960). However, the privacy-based appropriation tort did not provide a right that could be assigned to others to facilitate commercial exploitation.

In 1953, the U.S. Court of Appeals for the Second Circuit recognized a property-based right of publicity in Haelan v. Topps Chewing Gum (1953). The case involved two competing chewing gum companies using baseball cards with major league player images to promote their products. As the Haelan court put it,

We think that . . . a man has a right in the publicity value of his photograph, i.e., the right to grant the exclusive privilege of publishing his picture . . . This right might be called a “right of publicity.” (Haelan v. Topps Chewing Gum, 1953, p. 868)

As it was defined in the Restatement (Third) of Unfair Competition (1995), the right of publicity prevents “appropriat[ing] the commercial value of a person’s identity by using without consent the person’s name, likeness, or other indicia of identity for purposes of trade” (sec. 46). The right eventually gained recognition in many U.S. jurisdictions. Currently, most states recognize either a statutory or common-law right of publicity (McCarthy, 2003). Despite its ubiquity, the full contours of the tort were frequently less than clear. As one perceptive analysis put it, “Unfortunately, rather than fill Haelan’s vacuum with considered analysis, judges and lawmakers sidestepped the tough issues and increasingly adopted an attitude of ‘if value, then right’” (Dogan & Lemley, 2006, p. 1173).

The First Amendment and the Right of Publicity

From its inception, the publicity tort had an ambiguous relationship with the First Amendment. A bewildering array of First Amendment tests have been adopted by various courts in publicity cases.

Although there are certain types of speech outside the scope of the First Amendment, such as obscenity (Roth v. United States, 1957), and others that receive only moderate protection, such as advertising (Central Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 1980), most speech receives significant protection. Fully protected speech is judged under the strict scrutiny test, which makes it difficult for government to regulate it. Under the strict scrutiny test, government can only regulate speech if it can demonstrate that its regulation is narrowly tailored to advance a compelling interest (United States v. Playboy Entertainment Group, 2000). Strict scrutiny is a highly protective standard that challenges government to provide powerful justifications for any attempt to regulate protected speech.

In the realm of intellectual property, the closest legal analogue to the right of publicity, First Amendment protection is more attenuated. In copyright law, for example, the First Amendment is largely not a factor because the U.S. Supreme Court has concluded that certain built-in elements of copyright doctrine such as fair use adequately protect free expression interests. The Court has stated that copyright is not quite “categorically immune” from First Amendment scrutiny, but that “copyright’s built-in free speech safeguards are generally adequate” (Eldred v. Ashcroft, 2003, p. 220). As discussed later, some courts considering First Amendment claims in publicity cases have borrowed doctrinal elements from copyright’s fair use doctrine.

The only Supreme Court case to consider the clash of publicity rights with the First Amendment, Zacchini v. Scripps-Howard Broadcasting Co. (1977), held simply that a television station had no First Amendment right to broadcast a performer’s entire act. Hugo Zacchini, “The Human Cannonball,” sued a station that broadcast video of his 15-second act, and the Supreme Court held that the First Amendment did not protect the station:

The broadcast of [Zacchini’s] entire performance, unlike the unauthorized use of another’s name for purposes of trade or the incidental use of a name or picture by the press, goes to the heart of [Zacchini’s] ability to earn a living as an entertainer. (Zacchini v. Scripps-Howard Broadcasting Co., 1977, p. 576)

This case appears to be sui generis in that most publicity cases do not include reproduction of a plaintiff’s entire performance, but rather involve some more fragmentary borrowing of elements of a plaintiff’s persona. Moreover, as one astute commentator has pointed out, “Zacchini was an odd publicity case . . . because, unlike most of them, it involved information in a form that was closely akin to the sorts of works of authorship protected by copyright” (Zimmerman, 1998, p. 49). Thus, Zacchini appears to provide little guidance as to the First Amendment standard in more typical publicity cases.

Despite this lack of precision, lower courts in the past have often rejected claims that use of an individual’s persona in an expressive work violated the right of publicity. For example, the Florida Supreme Court interpreted that state’s misappropriation statute not to forbid the portrayal of real persons associated with a nautical disaster in the slightly fictionalized film “The Perfect Storm” (Tyne v. Time Warner Entertainment Co., 2005). The Florida court held that the purpose of the statute was “to prevent the use of a person’s name or likeness to directly promote a product or service because of the way that the use associates the person’s name or personality with something else” (Tyne v. Time Warner Entertainment Co., 2005, p. 808). Thus, use of an individual’s persona that was not equivalent to a false endorsement could not result in liability. The court specifically noted the First Amendment concerns associated with liability for an expressive use of a plaintiff’s identity and pointed out that it was bound to interpret state law in a way that created no constitutional conflict. Commentator J. Thomas McCarthy has pointed out that this constitutional avoidance approach—narrowing publicity rights to avoid First Amendment conflicts—has been popular among many courts (at least until the rise of the transformative use approach discussed later; McCarthy, 2003). In essence, the avoidance approach does not offer any “test” to protect free speech interests, but instead narrows the elements of the right of publicity to avoid collisions with free expression values.

The Florida court quoted approvingly from a 1979 California Supreme Court decision in a publicity case involving a film about silent film star Rudolph Valentino,

While few courts have addressed the question of the parameters of the right of publicity in the context of expressive activities, their response has been consistent. Whether the publication involved was factual and biographical or fictional, the right of publicity has not been held to outweigh the value of free expression. (Guglielmi v. Spelling-Goldberg Prods, 1979)

A number of other cases have reached similar results (see, for example, Cardtoons, L.C. v. Major League Baseball Players Association, 1996 [baseball cards]; Dora v. Frontline Video, 1993 [documentary film]; Hicks v. Casablanca Records, 1978 [film]; Hoffman v. Capital Cities/ABC, Inc., 2001 [feature article]; Lerman v. Flynt Distributing Co., Inc., 1984 [photograph]; Matthews v. Wozencraft, 1994 [book]; Rogers v. Grimaldi, 1989 [movie title]; Ruffin-Steinback v. DePasse, 2001 [mini-series]).

Although the quoted statement above may have represented an emerging consensus in 1979—and even later, as the cases cited above demonstrate—courts continued to struggle with the proper analytical framework for considering the First Amendment limits of the publicity right. They have applied a variety of tests to draw the boundary. These tests range from the actual malice standard used in defamation (Hoffman v. Capital Cities/ABC, Inc., 2001), to a “predominant purpose” test created by the Missouri Supreme Court (Doe v. TCI Cablevision, 2003), to a “wholly unrelated” test for publicity rights in movie titles (Rogers v. Grimaldi, 1989), to an exhaustive step-by-step weighing of interests between the right of publicity and constitutional freedom of speech (Cardtoons, L.C. v. Major League Baseball Players Association, 1996). As well, the overriding factor in some cases was whether the use of the persona was newsworthy (Zimmerman, 1998).

Despite these varying formulations, the analysis that has become one of the primary tests recently is the transformative use inquiry imported from copyright law. Transformative use was first applied as a proxy for First Amendment interests in a publicity case in a 2001 California Supreme Court decision (Comedy III Productions, Inc. v. Saderup, 2001). The case involved the unlicensed sale of lithographs and T-shirts bearing the image of The Three Stooges. The court acknowledged the images were expressive works rather than advertising, but nonetheless concluded that this did not automatically confer First Amendment protection. The court recognized both the First Amendment value of uninhibited expression about celebrities, who often perform important symbolic functions in culture, and the value of publicity rights.

To mediate the tension between these two interests, the California high court borrowed the transformative use test, first articulated in copyright fair use doctrine in 1994 by the U.S. Supreme Court in Campbell v. Acuff-Rose Music, Inc. (1994). The Comedy III court provided scant theoretical justification for the importation of this alien doctrine, other than noting that borrowing the entire fair use analysis would be unworkable. The court also pointed out that the more a celebrity image was altered, the less likely the borrowing would necessarily harm the economic value of the persona. However, the California high court did not provide any particularly salient analysis of why altering the image would change its First Amendment value. Ultimately, the court concluded that the plaintiff prevailed as the defendant had not added any significant transformative elements to his depiction of the Stooges.

The transformative test is rooted in a Harvard Law Review article by prominent federal judge Pierre Leval (Leval, 1990). Transformative use was firmly established in copyright doctrine through the influential U.S. Supreme Court decision in Campbell. There the Court held that by adding the transformative ingredients of “new expression, meaning, or message” (Campbell v. Acuff-Rose Music, Inc., 1994, p. 579), a copyright borrower was more likely to be a fair user. Since that decision, the transformative doctrine, which is not technically one of the four statutory fair use factors, has come to dominate the analysis in most courts (see, for example, Netanel, 2011, noting that “fair use doctrine today is overwhelmingly dominated by the Leval-Campbell transformative use doctrine,” p. 736). However, the doctrine as applied in the copyright context has been criticized by some legal scholars as murky and manipulable. As one noted scholar put it, “At the end of the day, characterizing a use as transformative may be nothing more than a conclusion based on some unconscious, inarticulable balancing of social costs and benefits” (Cotter, 2010, p. 725).

Transformative Use in Hart and Keller

Both factually and analytically, the similarities between the July 2013 decision by the U.S. Court of Appeals for the Ninth Circuit in Keller v. Electronic Arts, Inc. (2013) and the opinion rendered two months earlier by the Third Circuit in Hart v. Electronic Arts, Inc. (2013) are remarkable. Each case involved right of publicity claims filed by college football players who objected to the use of their alleged digital avatars and biographical data in a series of successful video games. Each dispute produced a two-to-one decision in favor of the respective plaintiff-athletes, with all six judges agreeing that the transformative use standard provided the correct legal test, but then fracturing badly on its actual application and the results it produced.

Indeed, the contrasting analyses of the majority and dissenting opinions in these cases—each supposedly adopting the same transformative standard—demonstrate just some of the dangers the transformative use approach creates for resolving the conflict between publicity rights and First Amendment interests. As a threshold matter, the judges in both Keller and Hart framed the interests at stake in similar fashion. “We must balance the right of publicity of a former college football player against the asserted First Amendment right of a video game developer to use his likeness in its expressive works,” wrote Judge Jay Bybee for the Keller majority (Keller v. Electronic Arts, Inc., 2013, p. 1271). Likewise, Judge Joseph Greenaway reasoned for the Hart majority that “we must balance the interests underlying the right to free expression against the interests in protecting the right of publicity” (Hart v. Electronic Arts, Inc., 2013, p. 149). The dissenting judges in both cases also agreed these constituted the respective interests that needed to be balanced (see Keller v. Electronic Arts, Inc., 2013 [Thomas, J., dissenting], asserting that “any right of publicity must be balanced against the constitutional protection afforded by the First Amendment,” p. 1284; Hart v. Electronic Arts, Inc., 2013 [Ambro, J., dissenting], writing that “[t]his claim requires us to balance the competing interests implicated by the right of publicity and the First Amendment,” p. 171).

In adopting the transformative use test to resolve this tension, the Third Circuit in Hart spent more ink explaining its choice than did the Ninth Circuit in Keller. Specifically, the Hart majority opined that the transformative use standard “provides courts with a flexible—yet uniformly applicable—analytical framework” (Hart v. Electronic Arts, Inc., 2013, p. 163) that “maintains a singular focus on whether the work sufficiently transforms the celebrity’s identity or likeness, thereby allowing courts to account for the fact that misappropriation can occur in any market segment, including those related to the celebrity” (Hart v. Electronic Arts, Inc., 2013, p. 163).

The Hart majority added that the transformative use test “recognizes that if First Amendment protections are to mean anything in right of publicity claims, courts must begin by considering the extent to which a work is the creator’s own expression” (Hart v. Electronic Arts, Inc., 2013, p. 163). It drew support for the proposition that the focus must be on the defendant-creator’s expression from Justice Lewis Powell’s Zacchini dissent, in which Powell wrote that “we should direct initial attention to the actions of the news media: what use did the station make of the film footage?” (Zacchini v. Scripps-Howard Broadcasting Co., 1977, p. 581). The Hart majority repackaged this query, under the transformative use test, as “what use did the [defendant] make of the [celebrity identity]?” (Hart v. Electronic Arts, Inc., 2013, p. 164).

Importantly, the Third Circuit also considered and rejected an alternative test created by the Second Circuit in Rogers v. Grimaldi (1989). In that case, the Second Circuit held that the use of Ginger Rogers’ name in a movie title was protected by the First Amendment unless the “title was ‘wholly unrelated’ to the movie or was ‘simply a disguised commercial advertisement for the sale of goods or services’” (Rogers v. Grimaldi, 1989, p. 1004). The Hart majority spurned the Rogers test, labeling it a “blunt instrument, unfit for widespread application in cases that require a carefully calibrated balancing of two fundamental protections: the right of free expression and the right to control, manage, and profit from one’s own identity” (Hart v. Electronic Arts, Inc., 2013, p. 157). If Rogers were applied to the facts in Hart, the majority wrote, it would lead to a victory for Electronic Arts and “threaten[] to turn the right of publicity on its head” (Hart v. Electronic Arts, Inc., 2013, p. 157) by denying Ryan Hart any control over his likeness and profits from it in the precise area—college football—where he earned “hard-won celebrity” (Hart v. Electronic Arts, Inc., 2013, p. 158).

The Ninth Circuit majority in Keller spent no time debating between tests. Applying the substantive law of California, it simply chose the transformative use standard because the California Supreme Court already had determined a dozen years before in Comedy III Productions v. Saderup (2001) that transformative use was the proper test (Keller v. Electronic Arts, Inc., 2013, pp. 1273-1274). The Keller majority buttressed this logic by noting that two California appellate court decisions (Kirby v. Sega Am., Inc., 2006; No Doubt v. Activision Publishing, Inc., 2011) involving right of publicity claims that, like Keller, centered on video games also had adopted the transformative use test. The dissenting judges in both Keller and Hart offered no pushback to the respective majorities—both agreed transformativeness was the correct rule.

Turning to how adopting the transformative use standard defeated the First Amendment interests, the Third Circuit majority in Hart began by framing the question as whether plaintiff Ryan Hart’s identity—specifically, the combination of his likeness and biographical information—was “sufficiently transformed” (Hart v. Electronic Arts, Inc., 2013, p. 165). It quickly found that “based on the combination of both the digital avatar’s appearance and the biographical and identifying information—the digital avatar does closely resemble the genuine article” (Hart v. Electronic Arts, Inc., 2013, p. 166). Given such abject lack of transformativeness, the majority then turned its attention to “the context within which the digital avatar exists” (Hart v. Electronic Arts, Inc., 2013, p. 166). Here, it also found a lack of transformative context, pointing out that Hart’s avatar did exactly what the real-life plaintiff did—namely, play college football in a stadium.

With two strikes against Electronic Arts, the Third Circuit turned to one final factor present in the video game—the ability of players to alter the appearance of Hart’s avatar. Was this factor sufficient to make the use of Hart’s avatar transformative? No, the Third Circuit concluded. It reasoned that “interactivity cannot be an end unto itself” (Hart v. Electronic Arts, Inc., 2013, p. 167) and it emphasized both that the game’s default depiction-image before any interactivity on the part of a user was of Hart’s likeness and that Hart’s “unaltered likeness is central to the core of the game experience” (Hart v. Electronic Arts, Inc., 2013, p. 168). Finally, the Third Circuit majority refused to expansively consider the other creative elements in the video game—the game as a whole—when applying the transformative use test. In rejecting this approach, it reasoned that emphasis must narrowly be on the celebrity’s identity and that “wholly unrelated elements do not bear on this inquiry” (Hart v. Electronic Arts, Inc., 2013, p. 169).

Two months later, the Ninth Circuit majority similarly applied the transformative use test in Keller, finding that “use of Keller’s likeness does not contain significant transformative elements” (Keller v. Electronic Arts, Inc., 2013, p. 1276). As with the Third Circuit majority in Hart, the Keller majority rejected Electronic Arts’ arguments that the transformative qualities of the game as a whole and its interactive nature made a difference. Instead, what mattered most was that the video game “realistically portrays college football players in the context of college football games” (Keller v. Electronic Arts, Inc., 2013, p. 1279).

Thus, the majorities in both Hart and Keller focused narrowly on the application of the transformative use test on (a) the physical resemblance of the avatars and their biographical information to that of the plaintiffs and (b) the contextual similarity of video game surroundings—specifically, football games played in football stadiums—with the context in which the plaintiffs achieved their celebrity and fame. The other elements of the game as a whole, as well as the interactivity variable, were given short shrift.

The dissenting judges in both Hart and Keller applied the transformative use test much more expansively to reach pro-defendant results. Judge Thomas Ambro in Hart, for example, wrote that “my colleagues limit effectively their transformative inquiry to Hart’s identity alone, disregarding other features of the work” (Hart v. Electronic Arts, Inc., 2013, p. 170). Rejecting such a narrow approach, Judge Ambro instead considered both the game-as-a-whole view and interactivity. Specifically, he reasoned that the video game

transforms Hart’s mere likeness into an avatar that, along with the rest of a digitally created college football team, users can direct and manipulate in fictional football games. With the many other creative features incorporated throughout the games, sufficient expressive transformation takes place to merit First Amendment protection. (Hart v. Electronic Arts, Inc., 2013, p. 170)

In similar fashion, dissenting Judge Sidney Thomas in Keller wrote the Ninth Circuit

majority confines its inquiry to how a single athlete’s likeness is represented in the video game, rather than examining the transformative and creative elements in the video game as a whole. In my view, this approach contradicts the holistic analysis required by the transformative use test. (Keller v. Electronic Arts, Inc., 2013, p. 1285)

In summary, one standard—transformative use—was applied in two nearly identical cases but with radically different analyses and results by the respective majorities and dissents. A cabined and confined approach to transformative use employed by the majorities in both Hart and Keller led to pro-plaintiff results, whereas the much more expansive and holistic tacks taken by the dissents resulted in opposite conclusions.

Analysis

Hart and Keller are significant cases at the intersection of publicity rights and the First Amendment in a number of ways. First, the decisions add the voices of two important federal appeals courts to the chorus of those who hold that transformative use is the proper standard to evaluate publicity claims that may violate a defendant’s free speech rights. It is true that the Ninth Circuit was using the test in a diversity context, and thus, did not make the choice of its own accord, but one can safely assume that the case will ultimately be cited by future courts in support of applying the transformative approach. Second, both majorities interpret the transformative test in a narrow manner that has the potential to result in significantly diminished First Amendment protection for future expressive users of celebrity identity. Third, the very fact that the cases each reached the same result may mean the U.S. Supreme Court will be less likely to take up the issue, because so-called “circuit splits” are among the dominant reasons the Court chooses to intervene in lower court doctrine.

Significantly, the Ninth Circuit in 2015 cited Keller favorably in once again adopting the transformative use test in a sports video game case, this one involving the likenesses of former professional football players in the Madden NFL game series (Davis v. Electronic Arts, Inc., 2015). The Ninth Circuit found the Davis case “factually indistinguishable” from Keller and thus rejected a transformative use defense by Electronic Arts (p. 1178). The court wrote,

Like NCAA Football, Madden NFL replicates players’ physical characteristics and allows users to manipulate them in the performance of the same activity for which they are known in real life—playing football for an NFL team. Neither the individual players’ likenesses nor the graphics and other background content are transformed more in Madden NFL than they were in NCAA Football. (p. 1178)

It is important to note that video games are far from the only medium to which the transformative use test has been applied in the publicity context. The test has been applied to comic books (Winter v. DC Comics, 2003), greeting cards (Hilton v. Hallmark Cards, 2009), paintings (ETW Cor. V. Jireh Pub., Inc., 2003), the appropriation of the name of a famed drug dealer by a rap artist (Ross v. Roberts, 2013), and other expressive contexts. The Ross case, by a California appellate court, suggested that the transformative test properly applies to all media. Thus, although predictions about the evolution of legal doctrine are always somewhat risky, the authors of this study see no reason the test could not ultimately be applied in additional media contexts. Ross also illustrates another important point: Although the case was decided in favor of the defendant, that fact alone certainly does not refute the central thesis of this article. That is, the claim here is that the transformative test is a dangerous addition to publicity law even though it will unquestionably result in some victories for defendants. The danger of the test’s highly subjective nature is that, over time, it will result in fewer victories for expressive defendants than would our proposed solutions, as well as chilling other speakers who do not wish to test the uncertainties of transformativeness.

The narrowness of the conception of transformativeness adopted by both majorities in Hart and Keller, and now in Davis, is disturbing from a free speech perspective. Essentially, any creative effort in the production of an expressive work by a defendant is ignored unless the persona itself is significantly altered. This crabbed approach does not seem warranted in light of the seminal transformative use cases. In the locus classicus of the genre, Comedy III, the California Supreme Court defined transformative use in the publicity context as “whether the celebrity likeness is one of the ‘raw materials’ from which an original work is synthesized, or whether the depiction or imitation of the celebrity is the very sum and substance of the work” (p. 406). The quoted language clearly contemplates a larger “work” containing the celebrity persona, and it is that “work” that is being tested for transformativeness. Dissenting Judge Ambro made this very point in Hart (2013, p. 173). In other words, Comedy III suggests, unlike Hart and Keller, that the site of the transformation need not be the celebrity identity itself, but instead the new expression or meaning may consist of surrounding expressive elements that create a transformative context for the presentation of that identity. The same understanding is at work in the Sixth Circuit’s 2003 decision involving a sports painting of golfer Tiger Woods, ETW Corp. v. Jireh Publishing, Inc. (2003). There, the Sixth Circuit held that a realistic portrayal of Woods was transformative because the entire painting contained expressive elements that communicated the historic nature of Woods’ victory in the 1997 Masters. Woods’ identity itself did not need to be transformed if it was set in a transformative context. This holistic approach is also consistent with that used in copyright determinations of transformative use—the question in Campbell, the original transformative use case, was not whether the borrowed elements from the song “Pretty Woman” were sufficiently altered, but whether the new work into which those elements had been inserted was itself transformative.

Even if courts recalibrated the transformative inquiry to be more holistic, that would not necessarily solve an essential problem with transformative use doctrine—its relative indeterminacy. The transformative test is notorious for its uncertain application in copyright law. One influential scholar called transformative use doctrine, as frequently applied, “a triumph of mindless sound bite over principled analysis” (Goldstein, 2008, p. 442). Thus, its importation into publicity law does not exactly inspire confidence. What precisely counts as transformative in the copyright context has been debated endlessly by judges and scholars, and the consensus seems to be that courts have not articulated clear guidelines for identifying a transformative use. The split decisions in Hart and Keller themselves suggest that determinations of transformativeness are hardly clear-cut.

Moreover, the seeming prejudice against realistic portrayals of human identity in publicity law poses great dangers to other expressive uses traditionally understood to be protected speech, including uses of celebrity persona in songs, films, books, and other media. As discussed earlier, past cases have simply roped off (sometimes through the constitutional avoidance approach) certain areas of expressive work as protected from publicity claims. These include realistic depictions of individuals in literary works of fiction and nonfiction, in films, in songs, and in a variety of other expressive works. The growing emphasis on transformativeness may unsettle existing First Amendment understandings and chill a variety of speech.

As noted earlier, although Hart and Keller are both video game cases, the transformative test itself is being used well beyond the confines of video game litigation. The authors of this article predict that trend may well continue. Consider, for example, a documentarian who creates a film about a college athlete that uses that individual’s likeness, statistical information, and biographical data, without his or her permission. Because both documentaries and video games are fully protected speech—and thus essentially indistinguishable for First Amendment purposes—it is difficult to imagine a principled basis on which a lower court in the Third or Ninth Circuits could decline to apply the transformative test. Furthermore, the fact that the documentary about the athlete (as compared with the video game) may not be intended for entertainment purposes makes no difference in this stage of the First Amendment analysis. As Justice Scalia wrote in Brown when protecting violent video games, “we have long recognized that it is difficult to distinguish politics from entertainment, and dangerous to try” (p. 2733). Of course, it is entirely possible a court could find a doctrinal route around transformativeness, but it is nonetheless clear that Hart and Keller create a jurisprudential problem that extends beyond video game litigation.

Beyond issues of application, an even more fundamental question is why the transformative use test should be the standard at all in publicity cases. The choice itself seems odd—transformative use arose simply as a Supreme Court gloss on an already Byzantine statutory fair use analysis in copyright. As mentioned earlier, the justification for adopting the test in publicity cases seems undertheorized and almost arbitrary. Publicity law and copyright are quite different animals, as will be discussed shortly. The seminal Comedy III opinion provided no intelligible theory for plucking a single element out of the complex copyright fair use doctrine and installing it as the exclusive means for a publicity defendant to vindicate her First Amendment rights. As legal scholars Stacey Dogan and Mark Lemley have astutely pointed out, this approach “ignores the fact that the fair use calculus reflects a subtle and complex balancing of multiple interests that makes sense only in combination” (Dogan & Lemley, 2006, p. 1189).

From a First Amendment perspective, the transformative use doctrine does not connect in any significant way with the underlying theories of protecting free speech, such as individual autonomy, the marketplace of ideas, or self-government (Greenawalt, 1989). In particular, why should realistic portrayals of human identity be frowned upon while altered portrayals are highly protected? Realistic portrayals of celebrities, athletes, and others may contribute significantly to the marketplace of ideas, yet for some unarticulated reason they are regarded as unworthy of First Amendment protection. There is seemingly no convincing conceptual connection between First Amendment values and the transformative use doctrine. Even the idea that speech must prove itself worthy of protection is an odd one in First Amendment doctrine—under Supreme Court precedent, speech is generally automatically protected, without requiring an individualized demonstration of its value, unless it falls into one of the unprotected categories. As one perceptive scholar has noted,

If speakers can only claim First Amendment protection if they are first able to convince a judge or jury that their speech is sufficiently important and weighty to deserve it, it would tantamount to concluding that only speech that satisfies the personal predilections of the finder of fact is protected. (Zimmerman, 2000, p. 301)

If indeed courts are to require demonstrations of expressive value in publicity cases, the transformative test as applied in Hart and Keller seems particularly unsuited to detect that value. Trivial variation of a persona may not have much expressive value, whereas realism in the right artistic context may constitute valuable speech. It is, in fact, a false dichotomy to conclude that transformation of the persona creates valuable expression, whereas realism does not.

The only sensible way to measure artistic value is to examine the entire work in question, rather than the celebrity persona in isolation (cf. Miller v. California, 1973). Comedy III makes it clear that it is the value of the totality of the work that is at issue—it just so happened that in that case that the representation of the Stooges was the entire work.

The copyright analogy also seems misplaced because copyright law is premised on the theory that without significant protection for expressive works, authors simply will not go to the trouble to create them, resulting in the underproduction of valuable literary and artistic works. As the Cardtoons court pointed out, it seems unlikely that withholding some degree of publicity protection will lead to diminished efforts by celebrities to achieve success and establish public personae, which in any event are almost certainly less valuable contributions to the marketplace of ideas than are copyrighted works as a whole. Cardtoons further noted that “the justifications for the right of publicity are not nearly as compelling as those offered for other forms of intellectual property” (Cardtoons, L.C. v. Major League Baseball Players Association, 1996, p. 976). Thus, it makes little sense to accord publicity plaintiffs a comparable level of protection with that offered to copyright plaintiffs by applying a similar First Amendment (proxy) test.

Moreover, in copyright doctrine, the transformative use issue is just one part of the overall fair use analysis. In publicity cases in which it has been employed, transformative use is the exclusive criterion for whether the work is protected speech. Thus, transformative use, as difficult as that concept is to define and operationalize, becomes a make-or-break issue for a publicity defendant—all without any serious justification of the link between transformative use and core First Amendment values. In fact, the use of the transformative test as the exclusive route to vindication produces the perverse result that borrowers in the publicity context are accorded less protection than similarly situated borrowers in copyright law, because nontransformative borrowings in copyright are still eligible for fair use protection (Campbell v. Acuff-Rose Music, Inc., 1994). Moreover, courts in copyright cases are now routinely accepting as “transformative” uses that reproduce the borrowed work unaltered (Bunker & Calvert, 2014).

A Better Approach?

Given the problematic state of the doctrine, what can be done to infuse rigorous First Amendment values into right of publicity law? The best solution may be a return to the “categorical” or constitutional avoidance approach described earlier. As in Tyne v. Time Warner Entertainment Co. (2005), the “Perfect Storm” case, courts could simply declare that expressive works are off limits to publicity actions. This approach is also consistent with the Restatement, which states that uses of celebrity identity in “entertainment and other creative works, including both fiction and nonfiction” (Restatement [Third] of Unfair Competition, sec. 47, comment c, 1995) are generally not infringing.

The benefits of the constitutional avoidance or categorical approach include a clear and determinate rule that provides protection when speakers borrow celebrity personae for genuinely expressive purposes. Speakers are not left guessing as to whether a particular use is “transformative” or some similar after-the-fact determination; instead, protection is provided as long as the borrowed persona is used in an expressive context, as opposed to a pure merchandise context, a false endorsement setting, or as part of purely commercial speech. Protection for speech interests is provided at the “front end”—the determination of the scope of the tort—rather than through an add-on defense. This approach has the further advantage of extricating judges from complex aesthetic judgments, such as those necessitated by the transformative test. One is reminded of famed jurist Justice Oliver Wendell Holmes cautioning against judges making contested aesthetic judgments in copyright law: “It would be a dangerous undertaking indeed for persons trained only in the law to constitute themselves final judges of the worth of pictorial illustrations, outside of the narrowest and most obvious limits” (Bleistein v. Donaldson Lithographing Co., 1903, p. 251). Under the categorical approach, if the work in question falls within the category of expressive works, such as books, plays, songs, films, video games (Brown v. Entertainment Merchants Association, 2011) and other such works, the publicity tort is simply not implicated.

Assuming a simple, after-the-fact test were to be necessary for publicity doctrine, the test derived from Rogers v. Grimaldi (1989) seems much more suited to the task than does transformative use. The test originally applied only to titles of expressive works, but a number of courts have extended its application to the expressive works themselves. Rogers, which requires that there be some minimal artistic relevance of the borrowed name or likeness to the borrowing work, and that the use not be explicitly misleading, is significantly more rigorous and speech-protective, and provides much improved clarity and predictability over the transformative test. Moreover, the Rogers test had its genesis in trademark law, which is arguably a much closer cousin of publicity law than is copyright (Dogan & Lemley, 2006), thus creating a better doctrinal fit.

Outcomes under the Rogers test would probably be quite similar to the constitutional avoidance approach, although Rogers could weed out cases in which use of a persona is blatantly intended to exploit the celebrity identity with absolutely no relevance to the expressive work. Unlike the transformative test, Rogers would not discriminate against realistic portrayals—the style of the portrayal is irrelevant under Rogers as long a minimal quantum of relevance to the work is present.

If transformative use is to remain the standard, it seems clear that courts should apply a holistic approach to consider the transformativeness of the entire work produced by the defendant, rather than viewing the persona in isolation and requiring transformation to take place there, as Hart and Keller did. As discussed earlier, the holistic approach seems consistent both with language from Comedy III and with standard practice in copyright fair use cases, the genesis of the transformative use test. As noted, the dissents in Hart and Keller also urged this approach. The holistic approach has the advantage of identifying added expressive value not simply through alterations to the borrowed persona, but through other creative material to which the persona may be joined to create a new and valuable expressive work.

Finally, an additional possible route to improve the transformative use test would be to import the “purpose” strand of transformative use that has gained momentum in copyright law in recent years (Perfect 10, Inc. v. Amazon.com, Inc., 2007). The purpose strand does not require any aesthetic reconfiguration of the borrowed work (thus making it compatible with realism in publicity law). Instead, it simply asks whether, regardless of any actual transformation in form, the borrower has used the borrowed materials for a different purpose than that for which they were created. In copyright law, for example, courts have been open to a fair user taking a work created for entertainment purposes and reproducing it verbatim for such purposes as news reporting or Internet search (Kelly v. Arriba Soft Corp., 2003; Nunez v. Caribbean Int’l News Corp., 2000). Courts have held such uses to be “transformative” despite the absence of any change in the form of the borrowed material.

Importing the purpose strand into publicity law would require some adjustment of the approach—clearly, celebrities who create public personae probably do not have one single purpose in mind. However, one suggestion might be to consider the putative purpose of a celebrity or public figure in creating a persona to be largely a profit-driven effort to craft a marketable image that can be monetized. That is indeed the “purpose” presupposed by the creation of the publicity tort itself. To the extent an expressive borrower uses the persona for a purpose different than purely commercial exploitation, a court could find the borrowing to be transformative and thus protected. Clearly, the purpose-driven approach has some built-in difficulties because such purposes are rarely perfectly transparent to an observer. However, if courts are insistent upon pursuing the transformative paradigm in publicity law, the purpose approach may have some promise in helping to provide additional protection for expressive users who do not necessarily qualify for transformative use through aesthetic reconfiguration of the persona.

Conclusion

Right of publicity law is clearly in a state of flux. After some early cases in which protected speech was given a wide berth, the current environment in which the transformative use test seems to be becoming the standard methodology is troubling to those concerned with robust protection for free expression. This work has critiqued the transformative test, particularly as applied in Hart and Keller. It has argued that transformativeness makes little sense as a First Amendment standard in publicity cases, and that its use dilutes important expressive values, particularly in cases involving realistic portrayals of human identity. As well, this work has contended that the conceptual connections between the publicity tort and copyright also suggest that the transformative test is a poor doctrinal fit that, perversely, leaves publicity defendants worse off than copyright defendants. The article has suggested that courts should look to other approaches to provide clear and rigorous protection for speech in the right of publicity context.