Attributrion as Architecture in the Digital Age

The quest for authenticity—the ability to claim “this has value to others and so confers value on me”—is a longstanding human need made even more desirable in a digital age of infinite reproducibility.¹ Non-fungible tokens (NFTs)—by now ancient history—are one recent example of this phenomenon;² generative artificial intelligence (AI) is another, giving rise to fundamental questions relating to the nature of authorship and what level of contribution should lead to legal rights in output.³

But authenticity in and of itself is only part of the consideration. It is not enough for an individual to be secure in the knowledge that they created a work; others must know it too and change their appreciation based on that knowledge. Public attribution and its partner concept of reputation help in many instances to fuel creativity. The creative tree that falls in the forest might provide a personal outlet for expression, but having others know that it was your tree provides additional satisfaction independent of any desire for monetary reward.

As one high-profile example, consider the widespread reaction in 2020 when social media influencers on TikTok failed to credit the originator of the viral Renegade dance.⁴ As the (then) fourteen-year-old creator of the choreography told the New York Times: “I was happy when I saw my dance all over. But I wanted credit for it.”⁵ Taking private pleasure in one’s influence on culture is sometimes not enough; one must also be known to others as the one responsible for that influence. In short, many creators want credit conferred by others and added to their reputational accounts.

Despite this impulse, U.S. copyright law and trademark law have never quite determined how to accommodate the human desire for attribution.⁶ Copyright law seems satisfied with the ability of copyright owners (who are not always the original authors) to require attribution as part of the exercise of their reproduction and other rights, with the Visual Artists Rights Act (VARA)⁷ and the Digital Millennium Copyright Act (DMCA)⁸ providing areas of focused attention. Trademark law’s facilitation of source identification gives it a conceptual overlap with attribution,⁹ but its focus on commercial activity and the concept of the consumer makes it a difficult tool to use in practice, even without the U.S. Supreme Court’s apparent rejection of its use to achieve attribution of creative expression.¹⁰

Nevertheless, there are multiple examples of creative communities that have attempted to develop systems of attribution or systems to adjudicate perceived attributional transgressions.¹¹ For example, many open-source software licenses take a generally permissive approach to use but require downstream developers to deliver notices with any derivative works that provide attributional information, a requirement that persists even without evidence that users of the derivative work read such notices.¹² The six licenses developed by Creative Commons for use with all kinds of creative works vary in their permissiveness, but all six require attribution to be given to the creator.¹³ The antiplagiarism norms of academia are another example, where students are instructed on how and when to cite authority. In each case, the “rules” on attribution depend on the nature of the community. In fandom communities, “attribution is a norm so strongly entrenched in fandom that it is often mistaken for a legal rule,”¹⁴ while other communities may have weaker norms.¹⁵

Indeed, if attribution and the authenticity/reputation it brings along are community-mediated concepts, implementation should arguably remain the domain of such communities. When the community exists online, the implementation of attributional norms can be baked into the platform: Twitter’s (now X’s) development of retweeting functionality¹⁶ and projects such as the Content Authenticity Initiative, which aims to provide protocols for embedding provenance information into digital content, are examples of making attribution a part of platform infrastructure.¹⁷

The claim in this Article is a modest one. Attribution is often an important component of creative work, both for creators and for audiences.¹⁸ Attribution in physical spaces depends either on external legal rules or on shared understandings and practices of a community. Digital platforms and other online spaces provide the opportunity to reify practice by building attribution into the infrastructure, thus allowing both for alignment with the norms of the community and adaptation and improvement when the circumstances require. As attribution becomes, for many creators, one of the primary problems for generative AI models to address, we should consider whether, to borrow from a well-known comparison in legal scholarship, attribution should be less like a rule and more like a standard.¹⁹

I. The Perceived Importance of Attribution²⁰

Attribution can have several functions. It can allow audiences or consumers to group certain outputs and assign meaning to items in that group; it can be a way of building reputational status; and it can facilitate a sense of justice, in that giving credit (and not misattributing) is the “right” thing to do.²¹ In the online and social media world, where content is often shared and repurposed, attribution “can be seen as an important way that people distinguish remixing from ‘theft.’”²² Attribution thus has an expressive value that is separate from, but closely tied to, its instrumental value. By giving credit, we convey that credit is a norm in the relevant community.

A. Attribution for Creators

Several studies of online activity have considered norms around attribution.²³ These studies suggest some dominant themes. One theme is that attribution is an important value for online creators and community participants, even if attribution plays out inconsistently across different modes of production.²⁴ A second theme is the psychological motivation behind seeking or giving attribution—the satisfaction of being acknowledged for one’s prosocial activity—which may motivate future creativity.²⁵ (Indeed, even the small act of reposting with credit can be seen as its own form of creativity that allows online communities to flourish.)

Third, and relatedly, some community members perceive a difference between attribution and credit. For example, the authors of a 2011 mixed-methods study concluded that “young users see an important, if currently under-appreciated and under-theorized, difference between credit and attribution”: Attribution represents the reference to a contribution, whereas credit “can involve an explicit acknowledgment, an expression of gratitude, and an expression of deference, in a way that simple attribution [cannot].”²⁶ In other words, “attribution” reflects a more formal—and, for some, less satisfying—set of relationships, whereas “credit” or “acknowledgment” derives from more personal, community-based engagement.²⁷

Even creators who believe in extensive sharing and repurposing of material will still use the vocabulary of property to describe their efforts, talking in terms of “my photograph” or “my writing”; in some cases, they might characterize unauthorized use by others as “stealing” or “theft.”²⁸ Such creators would happily let others use their work without compensation, so long as attribution is given; the fault lies not in copying but in taking credit for another’s efforts. The motivating factor is not, as copyright law sometimes assumes, a desire for monetary compensation but a sense of propriety or justice in the relations between individuals.²⁹

The desire for attribution from a creator’s perspective comes, at least in part, from an effort to build one’s reputation. For a particular community to hold a creator in esteem (or disregard), it has to be able to build a canon of that creator’s works, which requires that their name be attached to those works.³⁰ For such artists, the first payment they receive, as Rebecca Tushnet has noted, is in “credit”³¹ (although the “cash” may come later); recognition of their ideas or of a particularly felicitous and original turn of phrase is what brings immediate satisfaction. Although some artists talk about reputation in a property sense—echoing Othello, which characterizes one’s “good name” as something that can be “filche[d]”³²—reputation is not something that an individual can own. Rather, reputation is a social creation that derives from communications among members of a particular group: a judgment based on information known to the community.³³ Thus, to the extent that creators desire credit due to reputational concerns, they want that attributional information to be communicated to others so that their canon of works can be constructed.

As any experience with social media will illustrate, a decision to employ pseudonymity does not require foregoing the opportunity to accumulate reputational capital. A pseudonym, used consistently, can build up the same sort of reputational credit as one's given name. Examples abound: the authors George Eliot and Mark Twain, the recording artists Eminem, Jay-Z, and Lady Gaga, and actors who have adopted stage names all have used pseudonyms as a repository for reputation. Even if audiences never learn (or only discover later) the identity of the person behind a pseudonym, reputational capital still accumulates.

Indeed, pseudonyms can be used strategically as a means of accumulating reputational signals. An author who is well-known in a particular genre but who wants to try her hand at a different genre free from prejudgment or bias may create the second work under a pseudonym.³⁴ Such authors can thus attempt to compartmentalize their reputations. To the extent they are successful in keeping under wraps the fact that the two names are related, they can guard against reputation spillovers in both, much as a large company might decide to use different trademarks for different types of products. This, of course, means that attribution is not necessarily stable, nor does it necessarily correlate with what some might say are the requirements for a truthful attribution statement. In such instances, attribution functions more like trademark law's "anonymous source" doctrine, in which a trademark is considered valid, regardless of consumer awareness of the corporate entity behind it, so long as a consumer recognizes the mark as an indicator of the source of goods or services.³⁵

In addition to a single author deploying multiple names, attributional complexity can also occur when a work is attributed to someone other than the individual who authored it. The practice of attributing one work to another (typically more well-known) author stretches back at least to biblical times. Note, for example, Saint Jerome's advice to a mother on the education of her young daughter: "Let her avoid all the apocryphal books, and if she ever wishes to read them, not for the truth of their doctrines but out of respect for their wondrous tales, let her realize that they are not really written by those to whom they are ascribed."³⁶ Erasmus's 1516 introduction to the second volume of his edition of Jerome's writings explains that it was not uncommon for “zealous booksellers intentionally to misuse the names of famous writers for their own profit,” or for other writers to “cloak their designs under influential names” to gain further traction for their ideas.³⁷ Indeed, Erasmus noted, “Nothing is easier than to place any name you want on the front of a book.”³⁸ This phenomenon is an example of what some have called a “multiple use name”—an attributional device that can be used by multiple authors so as to obscure authorship. “Carolyn Keene,” a pseudonym used by the multiple writers of the Nancy Drew series and “Alan Smithee,” a title used by directors who wish to disavow their work, are but two examples.³⁹ The functional requirement for attribution is therefore not validity but stability.⁴⁰

Apart from building reputation, a second reason to seek attribution is for reasons of justice—that it is “fair” to give credit to one’s sources, even if nothing more comes of it.⁴¹ A writer who discovers that her material has been used without permission or credit on a little-read website by a complete stranger is still likely to feel some measure of unhappiness at what she perceives as a transgression. Indeed, studies have shown that perceptions of fairness often motivate our actions, even when the result is suboptimal. For example, in a classic type of study (typically referred to as the “ultimatum game”), one participant is asked to divide an amount of cash with another participant.⁴² If the second participant rejects the division, both parties get nothing.⁴³ Hence, it should be in the second participant’s interest to accept whatever division the first participant offers, since they will at least walk away with something. But in these studies, the second participant sometimes rejects the offer, presumably because they perceive that the division has not been fair.⁴⁴ This sense is likely similar to what motivates the desire for attribution even when compensation is not at stake—put otherwise, a creator might sometimes reject distribution of their work if attribution were not provided, even if such distribution would bring economic benefits.

B. Attribution for Audiences

Few audiences are likely overtly motivated by literary theory’s turn away from the author and toward the audience as an additional source of meaning. Nevertheless, what was once referred to as “remix culture”—and now is just considered “culture”—has conditioned audiences to think of themselves as interpreters and retellers of creative material. A creator’s work is to be enjoyed and then used as source material for commentary, parody, or other forms of creativity. In meme culture, for example, the value that emerges from the content is not related to the authorship of the underlying image or the originator of what later becomes a meme. Indeed, few know the identities of those individuals. Rather, the value lies in the collective building of meaning; all who share the meme become authors.⁴⁵

Attribution still has value in a culture in which the seeds of creativity are being farther and farther flung. Audiences use attribution in the way in which they use brand names: to find the works they want in the marketplace (more of one artist, less of another) and as an organizational tool for reputational and aesthetic assessment. Attribution is what facilitates discussion of whether a new novel or painting represents a departure from the author’s or artist’s previous work. Likewise, a writer’s past body of work can be used as a proxy for evaluating the persuasiveness or truth value of another work. Some manuscripts are labeled “Shakespeare” and others are labeled “Marlowe,” and that labeling allows readers to do something interpretive with the texts.

In some cases, attribution information allows audiences to evaluate the content of creative production. An opinion piece or a scientific article will likely be interpreted differently and seen as more (or less) authoritative if one knows the “true” author of the piece and can evaluate it against the other information known about that person. As Seth Kreimer has written, the astronomer Carl Sagan could preface each remark with an account of his entire previous corpus, but, even in an academic seminar, it seems simpler just to sign his name. Conversely, identification makes other communications available to listeners. When evaluating an argument of Richard Posner regarding “efficiency,” it is useful to be able to refer to his other work.⁴⁶

But this is not, of course, true in all cases, given that an individual can build up reputation under a screen name on social media without any reader knowing the “real” identity of its author.⁴⁷ The inventor of Bitcoin is typically credited as “Satoshi Nakamato,” even though, as of this writing, it is uncertain whether that represents a single individual or a collective.⁴⁸ The street artist Banksy has been credited with more than one hundred works of art; as of this writing, the artist’s “true” identity remains unknown.⁴⁹

Thus, under this function, the name attached to a work requires stability but not biography. As with URLs, which substitute for the IP addresses that are technically directing our requests on the Internet, a name is simply a useful mnemonic that allows reference to authors and creators more easily than would a system that assigned them sequential numbers. So long as the attribution system allows us to organize all works by the same creator together, the system has achieved its purpose.

Of course, to the extent that any creator hopes to engage with such real-world niceties as monetary payment or the federal tax system, the attributional signal must be linked to some other information. At the most basic level, if that “something” involves a measure of accountability, whether legal, norm-based, or economic, the label allows the location of a physical individual against whom such accountability will be charged.⁵⁰ As Harold Love pointedly notes, “There was never any doubt as to where royalty cheques for Barthes, Foucault, Lacan, Kristeva and Derrida ought to be sent: they reasserted heroic authorship even in their questioning of it.”⁵¹ An attributional system that fulfills this function must therefore not only allow audiences to group works by author but also create a map that shows, in some respects, the relationship between authorial identities—that Stephen King shares a physical instantiation with Richard Bachman, as does Lady Gaga with Stefani Germanotta.

Reasserting these connections publicly, however, thwarts the attributional choices from the author’s perspective, since some authors use pseudonyms specifically to weaken the connection between two authorial personalities, much as a company might use separate brands for different products or markets.⁵² Indeed, to the extent that the second authorial personality is a repository of different reputational signals from the first, it would be a mischaracterization to equate them: Richard Bachman is not Stephen King, even though the same fingers typed the manuscript in each case.

No matter what the function, however, our choice largely depends on the authorship questions raised earlier. Imagine that we have a disputed text, heretofore unattributed, and the question on the table is whether that text was authored by Shakespeare. To answer this question, we must have a shared understanding of what it means to author a text. Does it mean that the individual put pen to paper (or, today, fingers to keyboard)? Or that the individual instructed a group of assistants on how to create the final product and authorized the result?⁵³

The choice informs, to some extent, what the attributional signal should mean to those who encounter it. If it means the individual who puts pen to paper, then the signal represents nothing more than a mark of provenance. Those familiar with an artist’s oeuvre might think it odd that the fourth novel reads nothing like the first or that the 2006 painting was done in a pop-art style while the 2003 painting drew from the Impressionists, but so long as the artist herself created each work, the attributional signal has done its job. The signal is physical, not conceptual. If the signal represents something more like a mark of consistent quality, however, then we might not be as disturbed by the contributions of assistants so long as the name, like a trademark, can reliably be used by audiences. We know what “Warhol” represents, and if we get those qualities, we have not been misled—the attributional signal represents authorization, not authorship.⁵⁴

The criteria that lead to the attributional signal are therefore related to the purpose for which we use it. If we use attribution as an indicator of style, then it matters less whether the artist painted every inch of the canvas or merely added the finishing touches to an assistant's work. If, by contrast, we use attribution as a way of helping us to interpret a work against others, it might matter less whether an author or artist uses a pseudonym, so long as he or she does so consistently. Audiences will be able to collect these works, evaluate one against the other, and draw conclusions about whether to believe the asserted claims or certify the work as art. Others—those who find biographical information to be important to interpretation—will bristle at the use of pseudonyms full stop. How, they might ask, can we possibly know what weight to give Writer X's economic theory without knowing if Writer X has an advanced degree in economics or is making things up based on an article they once read?

Here lies the importance of discursive or interpretive communities, which serve as a measure (or a proxy) of what is important to audiences.⁵⁵ For example, compare the appropriation art of Richard Prince or Sherrie Levine, which often consists of reframing or recontextualizing the photographs of others with only minor (or no) changes to the work, with the kind of photographic technique employed by photographers who reproduce famous fine art pieces for art history textbooks and the like.⁵⁶ In both cases, the second photograph is almost indistinguishable from the original. But in the first example, we attribute the second photograph to Prince or Levine, while in the second example, the second artist (the photographer) drops out, and we call the photograph a "Rembrandt" or a "Van Gogh" or whatever the case may be.⁵⁷ The difference is due primarily to the fact that a discursive or interpretive community—museums, art critics, and the like—has formed around Prince's and Levine's work, validating that these works communicate a different message or meaning from that of the original, even though the physical attributes of the work are unchanged. The attribution follows.⁵⁸

C. The Indeterminacy of Attribution

The organizational assistance the use of attribution provides should not, however, obscure the ways in which attribution affects evaluation and judgment.

First, the indeterminacy of authorship means that audiences can sometimes derive only limited information from an attribution signal. The name of a celebrity on the cover of a ghostwritten autobiography functions as an implicit statement by the celebrity that the story of her life in its pages is the one she wished to tell but does not indicate who put pen to paper. William Wordsworth, to take another example, often relied on the journals of his sister, Dorothy, as source material for his own work, despite his public persona as an author with “a unique ability to translate [his] experience for the masses of less gifted consumers.”⁵⁹ Likewise, as Catherine Fisk has written, although it is widely known that Clara Driscoll, the head of the Women’s Glass Cutting Department at the Tiffany studios, was responsible for the design or creation of most Tiffany lamps, the lamps still retain value because of the Tiffany name, which connotes a particular quality and style rather than a true statement of creative fact.⁶⁰ Judicial law clerks and political speechwriters agree, by virtue of their employment, that the words they write will be credited to their employers. Indeed, they typically aim to write in a corporate authorial style so that future assistants can later assume the task without it being obvious that a change in writer has occurred.⁶¹

Second, to the extent that audiences use attribution as a reputational signal, they should be conscious of what Robert Merton has termed the “Matthew effect.”⁶² Merton noted in 1968 that the amount of published scientific research was so great that it would be impossible to manage without relying on the professional reputation of the articles’ authors to winnow down the pile.⁶³ This leads to two results: it makes it more difficult for lesser-known authors to come to the attention of the community, and it accords better-known authors acclaim that might not be deserved on the merits of any particular piece of writing. The rich therefore get richer based largely on the fact that they can rest on their reputational laurels. Thus, audiences should keep in mind that an attributional signal might be used simply to take advantage of the Matthew effect, even though it provides little factual information about genealogy, such as when a famous name is attached to a work to facilitate its distribution and acceptance.⁶⁴

In some instances, then, the desires of authors and audiences coincide. Authors desire attribution to receive credit and build their reputations, and audiences desire attribution so that they can identify the works they wish to engage with and make critical judgments about, thus contributing to whether they will continue to engage with those works going forward. More narrowly, attribution allows discursive communities to engage in efficient conversations about a particular work. Rather than asking a friend whether she has read that new novel involving a certain list of characters and plot devices, she can simply ask whether her friend has read Author's latest book. Similarly, audiences can self-identify as followers or compatriots of a particular creator or stylistic school by saying that they are fans of Andy Warhol's work or that they try to emulate Eminem's performance style.

In other instances, however, the interests of authors and audiences can diverge. When an author writes under a pseudonym, this act hides certain information from the audience in ways that some might find to be harmful.⁶⁵ An audience might not care that Lewis Carroll's given name is Charles Dodgson, or that Stephen King wrote novels as Richard Bachman. Indeed, audiences might resist attempts to link the two identities, particularly given that the Internet does not easily forget such revelations.⁶⁶ But they very well might care when the putative Native American author Forrest Carter turns out to be the segregationist Asa Carter or when a young Chicano novelist named Danny Santiago was actually a white, much older writer named Daniel James.⁶⁷ Indeed, audiences will sometimes use terms like “hoax” or “deception” to describe their reactions to revelations in such cases. We have yet to see fully how this will play out with the cornucopia of content generated by AI but presented under the name of a human author.

II. The Law of Attribution

Despite the importance of the various functions of attribution to audiences, U.S. law plays a limited regulatory role. Part II provides here a brief overview of the judicial treatment of attribution, focusing on U.S. copyright law and trademark law.⁶⁸

A. Copyright Law

To qualify for copyright protection in the United States, a work must be an “original work[ ] of authorship fixed in [a] tangible medium of expression.”⁶⁹ The originality requirement, as the U.S. Supreme Court has noted, presents a low, but not nonexistent, hurdle.⁷⁰ Abraham Zapruder’s film of the Kennedy assassination, for example, was held to be protected by copyright even though Zapruder had done little more than set up his tripod in a particular location, pointed his film camera toward the motorcade route, and let the camera run.⁷¹ (The court held that the result met the originality requirement because Zapruder chose the type of camera, film, and lens for his shoot as well as the time and place to set up his camera.⁷²) One federal court suggested that a work can be protected by copyright even if some aspects of it resulted from accident, such as the “shock caused by a clap of thunder.”⁷³ At its outer edges, then, authorship becomes more of an act of authorization or of claiming than it does intentional creative activity.⁷⁴

U.S. copyright law is concerned largely with economic rights, which creates a tenuous link between copyright ownership and authorial attribution. For example, it is the employer or other entity for which a work was prepared that owns the copyright, not the original author, provided that the work qualifies as a “work made for hire.”⁷⁵ Additionally, although the fact that rights under copyright law are both alienable and descendible provides authors with a significant amount of flexibility in controlling and monetizing their rights in a work, it also means that copyright ownership is sometimes not a useful indication of authorship.⁷⁶

U.S. copyright law also does not require that the author herself has to be the individual who fixed the work; the statute specifically provides that the fixation can take place “by or under the authority of the author,”⁷⁷ and distinguishes the copyrightable “work,” which is the product of an author, from the material “copy” in which that work is fixed, which need not be.⁷⁸ This means that the owner of the copyright must have been involved in the work’s creation but need not have been the one who committed the work to paper or film, so long as they authorized that act. Consider, for example, a photograph of a human subject. The subject of the photograph typically has no claim to copyright, although they may have claims arising from common law or statutory provisions such as the right of publicity. The person who took the photograph—the person who pushed the shutter button or today’s equivalent in digital photography—may also not have any claim to copyright, if what they did was merely take the last step in a creative process effectuated by another. Rather, it is the person who composed the photograph, made the decisions that led to its creation as a creative work, and then directed its completion who is deemed the photograph’s author for copyright purposes.⁷⁹

All of this is fairly uncontroversial as a matter of U.S. copyright law. But consider, for example, the common scenario of a tourist in a new city who wishes to have her photograph taken in front of a famous landmark. Relying on the kindness of a passerby, she hands the stranger her camera (or phone), poses in front of the landmark, and asks him to take a photograph of her, which he does. Who is the author of the photograph for copyright purposes? The stranger has composed much of the shot and pressed the button but has done so at the tourist’s request, and perhaps pursuant to her instructions (“Make sure you get the monument in the frame.”) Was the photograph fixed under the authority of the camera’s owner, simply through the request of “Can you please take a photograph of me?” The legal questions might be murky, but one thing is clear: The camera owner undoubtedly would consider the photo to be her property. The stranger would receive, at most, her gratitude, not compensation or a request for a license.

Despite this murkiness, U.S. copyright law does focus on attributional rights in two relatively limited respects. First, VARA provides the author of a work of visual art with rights relating to attribution and the integrity of the work.⁸⁰ Unlike the other rights provided by the Copyright Act, the rights under VARA cannot be transferred, although they can be waived by the author in writing.⁸¹ Second, the DMCA includes a provision relating to what it refers to as “copyright management information” (CMI): information, such as the title of the work or the name of the author or copyright owner, that is conveyed in connection with copies of a work.⁸² The DMCA prohibits “knowingly and with the intent to induce, enable, facilitate, or conceal infringement,” providing, distributing, or importing for distribution false CMI; intentionally removing or altering CMI; or distributing works with altered or removed CMI with the knowledge that it has been removed or altered, “knowing, or . . . having reasonable grounds to know, that it will induce, enable, facilitate, or conceal” infringement of any right under copyright law.⁸³ Some courts, relying on the DMCA’s legislative history, have held that CMI must refer to a “component of an automated copyright protection or management system” to avoid a conflict with trademark law.⁸⁴ (In other words, if CMI referred to any indicator of source, a logo would receive protection under the DMCA.) Other courts, relying on the statutory language, have concluded that the provision is not so limited, such that a claim alleging that a photograph was copied from a Twitpic page without including an attribution to the photographer was sufficient to withstand a motion to dismiss.⁸⁵ In any event, as John Tehranian has noted, the “onerous scienter requirement” of these provisions makes it “extraordinarily difficult to prevail on a section 1202 claim.”⁸⁶

Absent interests that align with these two statutory schemes, an author who is also the owner of the copyright in a work can vindicate attribution interests by requiring attribution as a condition of licensing (or, as discussed later, issue the work under a Creative Commons or other license that requires attribution).⁸⁷ Authors who are not also copyright owners may, as the next section discusses, be out of luck.

B. Trademark Law

In many ways, an author's name works like a trademark or brand name: it signals to us the person who has declared responsibility for a particular communication, allowing us both to attribute our reactions to that communication and to find additional communications by the same author should we desire more. In the same way we might use the Pepsi trademark to ensure that we can find the cola that we like at the grocery store or the Volvo trademark to find the car that was recommended by a consumer magazine, we can use an author's name to find books that we'll like, based on our past experience with those authors or the recommendations we get from book reviewers, advertising, or like-minded friends.

When commerce consisted primarily of buyers getting goods directly from sellers—in other words, when transactions were face-to-face—trademarks were unnecessary.⁸⁸ If you walked to Mary's to purchase your bread and to Sam's to purchase your yarn, all you needed to do if you had a complaint about the product or wanted to purchase more was to return to Mary's or Sam's. Once commerce expanded, however, with customers more removed from the manufacturers of goods and intermediary vendors becoming the frontline of engagement, customers needed a memory device to identify the manufacturer responsible for the goods they purchased: a symbol etched on the bottom of a clay pot or a brand burned into the side of a leather good.⁸⁹ These symbols served as a shorthand for the manufacturer or artisan, allowing the customer to know whom to blame for shoddy workmanship or whom to look for on the next trip to the mercantile.⁹⁰ This was particularly useful in a world where the consumer might never meet the craftsman in person, and it enabled manufacturers, as commerce continued to develop, to communicate information to potential consumers in the form of advertising.⁹¹

In the economic language of today's trademark law discourse, trademarks reduce consumers' search costs. If an individual tries a new type of soda at a party, enjoys it, and wants to buy some of it on her next trip to the supermarket, she is going to have to spend a long time reading labels, convincing store owners to let her taste the product, or committing trial and error if the cans are not marked with the name of the product. By recognizing trademarks as a type of intellectual property, and restricting competitors from using those marks to deceive consumers about the product they are getting, trademark law enables consumers to find the products they want without incurring the kinds of costs just described.

The nature of trademarks today has changed in many ways from the simple artisan- or guild-based marks of the past. Whereas a trademark once identified the individual potter who created the bowl or the woodworker who crafted the rocking chair, a trademark now tends to signify a group of characteristics or qualities that remain relatively consistent over time. Given modern outsourcing practices, we can no longer say that trademarks identify manufacturers of goods, if we take "manufacturer" to mean the company that actually produces the good. Companies do, of course, supervise the quality control process—and, according to U.S. trademark law, must do so to avoid a finding that they have abandoned their trademark⁹²—but they need not physically do the manufacturing themselves. Indeed, one economist referred over twenty years ago to what he called the "Nike model," by which he meant a U.S. company whose primary U.S. activity is brand management, not manufacturing.⁹³

Franchises are another example of how trademarks refer to a level of quality rather than an actual source of production. The McDonald's trademark does not, of course, communicate to consumers that every hamburger nationwide emerged from a single kitchen. Rather, it signals to consumers that the food, preparation, service experience, décor, and other elements of the transaction will be consistent from location to location, such that the experience at a McDonald's in California will not differ all that greatly from the experience at a McDonald's in New York.⁹⁴

Both trademarks and personal names thus serve what I have previously referred to as a denotative function: they are used to refer to a product or artistic creation or a person at a particular point in time but, on their own, they do not necessarily provide verifiable information beyond that reference.⁹⁵ Indeed, courts have rejected arguments by those who have complained that name changes have made it difficult to determine the “true” identity of those with whom they are dealing.⁹⁶ A company that gives a loan to Susan Anderson has engaged in a perfectly legitimate transaction even if she was once known as Susan O’Leary; likewise, a company can be sued under its current name even if was known by a different name in the past.

Given these parallels, it is perhaps not surprising that some creators have attempted to vindicate attributional interests using federal trademark and unfair competition law in the United States—specifically, section 43 of the Lanham Act, which provides a cause of action for any person who “believes that he or she is or is likely to be damaged” by the act of another who, “in connection with any goods or services,” uses in commerce a name that is likely to cause confusion, mistake, or deception about the “affiliation, connection or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person.”⁹⁷ Some of these claims arose in circumstances where the artist was seeking a disclaimer of attribution to remove their name from work that was not theirs, under a theory of passing off. In other words, the claim was that the other party was acting no differently from a company that falsely brands its soda as Pepsi. The British comedy troupe Monty Python, the author Stephen King, and the singer Tom Waits were three such artists.

Monty Python was able to successfully invoke section 43 in the 1970s when the ABC television network, contrary to the group’s expectations, aired what was originally ninety minutes of Monty Python’s Flying Circus episodes by cutting twenty-four minutes of material to remove objectionable content and make room for commercials.⁹⁸ Monty Python claimed that this “mutilated” the group’s work and rendered many of the jokes meaningless.⁹⁹ As a result, the court concluded, ABC had conveyed that Monty Python was the source of a work that was “a mere caricature of [the group’s] talents” in its bowdlerized form and thus should be preliminarily enjoined from broadcasting the edited episodes again.¹⁰⁰

Stephen King’s involvement with the work attributed to him was more attenuated than Monty Python’s.¹⁰¹ Written in 1970 and published in 1975, King’s short story “The Lawnmower Man” told the story of a man chopped up by a lawnmower controlled psychokinetically by an evil “lawnmower man” hired to take care of the lawn.¹⁰² King assigned the film and television rights to the story to Great Fantastic Picture Corporation in 1978; the agreement allowed Great Fantastic to adapt the story in any way it desired.¹⁰³ Great Fantastic then transferred its rights in 1990 to another film company named Allied, which sought to produce the story as a feature-length film.¹⁰⁴ Once preproduction work was under way, Allied began advertising the film as “Stephen King’s The Lawnmower Man” and as “based upon” a short story by Stephen King.¹⁰⁵ After learning about the film in October 1991, King, through his lawyer, conveyed to the film’s distributor his “outrage[]” over the possessory credit, which he believed was a “complete misrepresentation.”¹⁰⁶

King filed suit in 1992, claiming that the use of his name violated section 43(a) of the Lanham Act, and secured a preliminary injunction against any use of his name in connection with the film.¹⁰⁷ On appeal, the Second Circuit held that the district court did not err in granting the injunction with regard to the possessory credit.¹⁰⁸ The testimony in the district court had suggested that such credits are typically “given to the producer, director or writer of the film; and that the credit at a minimum refers to an individual who had some involvement in, and/or gave approval to, the screenplay or movie itself,” which was not the case here.¹⁰⁹ By contrast, the appellate court held, the district court should not have enjoined use of the “based upon” credit.¹¹⁰ The film was indeed based on the short story to some degree, given that the core of the story appeared in the final scene of the movie¹¹¹—indeed, said the court, “King would have cause to complain if he were not afforded the ‘based upon’ credit.”¹¹² (King had not made much noise about this credit in any event.) The case eventually settled, with the defendants prohibited from using Stephen King’s name in connection with the film and a requirement that new sleeves or stickers be used to hide King's name from existing inventory.¹¹³

Tom Waits's relationship to the work he challenged was even more attenuated; unlike Monty Python and Stephen King, he had no connection whatsoever to the work.¹¹⁴ Waits is a well-known recording artist with a particularly distinctive voice—as the court described it, citing a fan, it sounds “like how you’d sound if you drank a quart of bourbon, smoked a pack of cigarettes and swallowed a pack of razor blades. . . . Late at night. After not sleeping for three days.”¹¹⁵ Notably, Waits had a long-standing policy against doing commercial endorsements, believing that they compromised artistic integrity.¹¹⁶ When Frito-Lay, the snack chip manufacturer, planned to introduce a new product, SalsaRio Doritos, it hired an agency to develop the ad campaign.¹¹⁷ The agency liked Waits’s song “Step Right Up” (ironically, an indictment of corporate shilling), found a singer who could closely imitate Waits’s voice, and recorded a radio commercial in which the lead singer touted the flavor of the new chip in the style of the song.¹¹⁸ Waits heard the commercial over the air in California and was “shocked,” fearing that his fans would think that he had sold out.¹¹⁹ (In court, Waits explained, “[I]t embarrassed me. I had to call all my friends, that if they hear this thing, please be informed this is not me. I was on the phone for days.”¹²⁰) Waits brought suit under, among other things, the Lanham Act, alleging that Frito-Lay and its ad agency had falsely implied that he had endorsed SalsaRio Doritos,¹²¹ and the Ninth Circuit upheld the jury verdict on the claim (although it vacated the damages award as duplicative).¹²²

One example of the use of the Lanham Act to compel attribution involved a little-known actor named Paul Smith, who agreed to star in a film produced by an Italian film company.¹²³ His contract promised that he would get star billing in the credits and advertising.¹²⁴ The film, “Convoy Buddies,” was distributed in the United States by Edward Montoro and his company.¹²⁵ Montoro, however, removed Smith’s name from the credits and advertising and substituted the name “Bob Spencer.”¹²⁶ Smith’s litigation theory was that Montoro had engaged in reverse passing off under the Lanham Act by promoting Smith’s work under another name, depriving him of the reputational benefits his performance might bring.¹²⁷ The district court dismissed Smith's complaint, holding that Smith's allegations did not state a claim under the Lanham Act.¹²⁸ The appellate court disagreed, noting that

[s]ince actors' fees for pictures, and indeed, their ability to get any work at all, is often based on the drawing power their name may be expected to have at the box office, being accurately credited for films in which they have played would seem to be of critical importance in enabling actors to sell their "services," i.e., their performances.¹²⁹

Before 2003, then, it seemed to be uncontroversial that section 43 of the Lanham Act could be used by authors and artists to vindicate their attributional interests—both in cases in which work was falsely attributed to them and in cases in which another individual was credited with their work. That view was upended, however, by the U.S. Supreme Court's 2003 opinion in Dastar Corp. v. Twentieth Century Fox Film Corporation.¹³⁰

The facts of Dastar, which I have discussed at length in previous work,¹³¹ are not particularly sympathetic. Twentieth-Century Fox was the owner of the television rights to Dwight D. Eisenhower's book Crusade in Europe, his 1948 memoir of the Allied campaign in Europe during World War II.¹³² It arranged for Time Inc. to produce a television series, also called Crusade in Europe, based on the book, for which Fox would hold the copyright.¹³³ The series, which first aired in 1949, comprised a narration of the book with film footage from various U.S. and foreign government sources as well as news footage.¹³⁴ Doubleday, which published the book, renewed its copyright in the book in 1975.¹³⁵ Fox, however, did not renew its copyright in the television series, as was required under U.S. copyright law at the time, and so the copyright expired in 1977.¹³⁶ In 1995, Dastar acquired the tapes of the television series, which was in the public domain, edited them, and added some new opening and ending material.¹³⁷ It then marketed the series on video as World War II Campaigns in Europe.¹³⁸ Dastar did not include any reference to Fox or to the original series in the video; it sold the video as "Produced and Distributed by: Entertainment Distributing" (which was owned by Dastar) or with the credit "DASTAR CORP presents."¹³⁹ The video set allegedly competed in the marketplace with a reissue of the Crusade in Europe television series on videotape, for which Fox had reacquired the rights in 1988; it subsequently licensed those rights to SFM Entertainment and New Line Home Video.¹⁴⁰ Fox, SFM, and New Line brought suit in 1998, alleging that Dastar's sale of its videos without crediting the Crusade in Europe series constituted reverse passing off in violation of section 43 of the Lanham Act.¹⁴¹ Although both the district court and the Ninth Circuit found in favor of Fox on this claim, the Supreme Court reversed.¹⁴²

The Court began by noting that section 43(a), while not a catch-all claim for every unfair trade practice, was capacious enough to include issues relating to false designations of origin of source or manufacture, geographic origin, and claims for reverse passing off in addition to passing off.¹⁴³ But, it held, Dastar had made no false claim of origin in this case because "origin," in the Court's view, referred to the manufacturer or producer of physical goods, and Dastar was indeed the producer of the videotapes labeled with its name.¹⁴⁴ Moreover, if "origin" were to be interpreted as the equivalent of authorship, claims under the Lanham Act seeking attribution would potentially raise a conflict with copyright law, which, according to the Court, permits unrestricted use of works in the public domain.¹⁴⁵ Finally, the Court noted the inherent indeterminacy of attribution discussed above, which was aimed, apparently, at providing a pragmatic justification for the Court's interpretation of congressional intent:

A video of the MGM film Carmen Jones, after its copyright has expired, would presumably require attribution not just to MGM, but to Oscar Hammerstein II (who wrote the musical on which the film was based), to Georges Bizet (who wrote the opera on which the musical was based), and to Prosper Merimee (who wrote the novel on which the opera was based). In many cases, figuring out who is in the line of "origin" would be no simple task. . . . We do not think the Lanham Act requires this search for the source of the Nile and all its tributaries.¹⁴⁶

The Court’s interpretation of “origin” thus leads to some hairsplitting. If a defendant takes a sheaf of papers constituting the plaintiff’s manuscript, removes the title page, and substitutes a new title page with her own name, she has (as one court suggested) simply repackaged the physical good and has thus arguably engaged in reverse passing off.¹⁴⁷ But if she retypes the entire manuscript and then appends her own name, she has apparently engaged in no Lanham Act violation because she is indeed the source of the physical manuscript (although she may be a copyright infringer).¹⁴⁸ To use copyright’s vocabulary, the Lanham Act’s conception of source applies to the copy, not to the work, regardless of how misaligned that might be with how communicative products are experienced.

Fox’s Lanham Act claim was almost certainly an attempt to rescue the copyright claim it lost when it failed to renew its copyright in the television series. Its strategy gives support to the Court’s intuition that Fox was trying to make an end run around the Copyright Act. But more sympathetic plaintiffs have subsequently been ensnared by Dastar’s net. Lower courts have readily dismissed claims by authors, artists, and others claiming that another individual or entity failed to provide correct attribution for a work of authorship,¹⁴⁹ even when the work was still subject to copyright.¹⁵⁰ In some instances, courts have been able to parse the plaintiff's claims to allow both copyright and trademark claims to go forward. For example, a court held that the owner of the copyright and trademarks relating to the 1979 film George A. Romero's Dawn of the Dead could pursue both sets of claims against the developer of a zombie video game because the copyright claims were targeted at the characters and storyline of the game, while the trademark claims were targeted at the title, Romero's name, and a trademark on the game's packaging.¹⁵¹ Likewise, a court allowed both copyright and trademark claims to go forward where the defendant used excerpts from a novel in its promotional materials (copyright) as well as the photograph and name of the character as its corporate logo (trademark).¹⁵² And some courts have limited Dastar's holding to claims for false designation of origin, letting claims for false endorsement proceed.¹⁵³ But for many creators, it is likely that Dastar will bar straightforward attribution claims under the Lanham Act and that copyright law will not be a vehicle for attribution claims for creators who are not also copyright owners.

C. Attribution as a Matter of the First Amendment

There is one additional doctrine to consider that affects the nature of attribution: whether an individual can communicate anonymously or pseudonymously without threat of legal reprisal, which implicates important issues of how individuals and audiences construct and interpret authorial identity. Although authors may have interests in communicating under their given names, they may also desire to communicate free from the bias or prejudgment that may accompany attribution. The determination of such rights requires a balancing of interests: the author's desire to communicate without disclosing their identity or to keep modes of communication separate versus the audience's desire to use identity information to evaluate the communication being presented.

In 1988, Margaret McIntyre handed out leaflets regarding a referendum on a proposed school tax levy in Westerville, Ohio.¹⁵⁴ McIntyre opposed the levy, and her leaflets encouraged members of the community to vote against the proposal.¹⁵⁵ On some of the leaflets, McIntyre included her name; other leaflets were signed "Concerned Parents and Tax Payers," even though McIntyre wrote the leaflets alone.¹⁵⁶ Five months after the levy passed, a school district official filed a complaint with the Ohio Elections Commission, alleging that by failing to include her name on some of the leaflets, McIntyre violated an Ohio statute prohibiting the general distribution of any publication designed to influence voters or engage in issue advocacy without the name and address of the person responsible for the publication.¹⁵⁷ The Commission fined McIntyre $100, and the Ohio Supreme Court upheld the fine, finding the law to be permissible under the First Amendment.¹⁵⁸ The U.S. Supreme Court, however, reversed, concluding that “an author’s decision to remain anonymous, like other decisions concerning omissions or additions to the content of a publication, is an aspect of the freedom of speech protected by the First Amendment,”¹⁵⁹ and holding the Ohio statute unenforceable.¹⁶⁰

The act of pseudonymity in the case implicated two ostensibly opposing interests. From the author’s perspective, as the Court noted,¹⁶¹ pseudonymity might encourage speech that would otherwise not be produced. An author might choose to communicate pseudonymously or anonymously in order to avoid retaliation or the judgment of one’s community, to separate one’s various authorial endeavors, or to make one’s arguments or creative efforts more persuasive or palatable.¹⁶² More perniciously, authors may have an interest in pseudonymity to wreak havoc, harass or defame another individual, or otherwise avoid responsibility for the effects of their communications.

From the audience’s perspective, however, pseudonymity may sometimes be a hindrance rather than a help. Audiences may desire to impose judgment on individuals for their views or assess the validity or worth of a communication by putting it in the context of the individual’s other writings or credentials. One can more easily assess the validity of a scientific claim, for example, when one knows whether its author is a scientist or a layperson. To the extent pseudonyms cannot be easily connected to a known individual, audiences may justifiably claim that they are deprived of information that makes interpretation of communications more difficult.¹⁶³

Deciding McIntyre's case, therefore, required the Court to elevate one set of interests over the other. Individuals like McIntyre could have claimed that there is individual and societal value in being able to participate in discussions on important issues of the day without the attendant risk of being prejudged by the community's view of the speaker or being retaliated against afterward. The individuals purportedly benefited by the Ohio election statute, by contrast, would have argued that it is too easy for individuals who communicate pseudonymously to propagate untruths or other misleading statements. By requiring individuals to sign their names to statements they make in the context of election-related communications, the law encourages thoughtfulness and accuracy, as well as providing an easy way of identifying the statement's author for informal or formal means of regulating or responding to such statements. These views are directly opposed to each other: valuable speech may not take place if undesirable retaliation occurs, and harmful speech might gain undeserved credence if audiences are deprived of information that allows them to discount its worth. By invalidating the Ohio statute, the Court elevated the possibility of better speech over the possibility of better evaluation of that speech.¹⁶⁴

The Court's opinion thus embodied a normative view of the author-audience relationship. Because the First Amendment aims to encourage more and diverse speech, audiences should expect to have to work harder to evaluate at least some communications. And the Court arrived at this point by treating the author's name not as something extraneous to the communication but as part of the communication itself. So characterized, it is difficult to distinguish a law requiring writers to include their name on a communication from a law requiring writers to include any other kind of content. We would balk at a law that required writers to present their opponent's views in a favorable light, or a law that required writers to state their allegiance to the government in any communication. We would, presumably, similarly reject a law that required writers to include a mention of their age, occupation, or educational background in a communication, even though that information might be useful to audiences in assessing or interpreting the communication. In the Court’s view, once a name is seen as information-providing content, it is easy to contemplate a law requiring a writer to include it as anathema to the free speech principles the First Amendment is supposed to represent.¹⁶⁵

And so McIntyre, who purported to be acting on behalf of “Concerned Parents and Tax Payers,”¹⁶⁶ even though she likely composed the text of the leaflet entirely on her own, had a First Amendment right to include that content and exclude other content, even though some might have characterized “Concerned Parents and Tax Payers” as a false statement of organizational identity.¹⁶⁷ McIntyre’s act was no different from that of a corporation that renames itself after experiencing an unfortunate event¹⁶⁸ or an interest group that chooses a name that is designed to communicate a viewpoint rather than accurate organizational information:

“Citizens for Better Medicare,” for instance, was not a grassroots organization of citizens, as its name might suggest, but was instead a platform for an association of drug manufacturers. And “Republicans for Clean Air,” which ran ads in the 2000 Republican Presidential primary, was actually an organization consisting of just two individuals—brothers who together spent $25 million on ads supporting their favored candidate.¹⁶⁹

This point was also illustrated by Buckley v. American Constitutional Law Foundation, in which the U.S. Supreme Court invalidated a Colorado state law requiring circulators of voter initiative petitions to wear badges stating their names, finding that such a requirement, unlike a prior registration requirement, would put the speaker at risk of harassment in the heat of the moment.¹⁷⁰ The same was true in Justice For All v. Faulkner, a 2005 case in which the U.S. Court of Appeals for the Fifth Circuit invalidated a policy at the University of Texas at Austin, which required all printed publications distributed on campus to contain the name of the university-affiliated person or organization responsible for the publication.¹⁷¹ Given the university's claimed rationale for the policy of preventing nonaffiliated individuals or groups from distributing publications on campus, the court suggested that while asking suspected nonstudents for identification might be permissible under the First Amendment—in other words, identifying themselves only to campus officials—a requirement that speakers identify themselves to the public through putting their names on each publication “sacrifices far more anonymity than is necessary to effectively preserve the campus forum for its intended beneficiaries.”¹⁷²

A more cautious approach was taken in Majors v. Abell, a 2004 decision from the U.S. Court of Appeals for the Seventh Circuit.¹⁷³ At issue was an Indiana statute that, like the Ohio statute at issue in McIntyre, required political advertising that expressly advocated on behalf of or against an identified candidate to contain clear and conspicuous notice of the identity of the individual who paid for the advertisement.¹⁷⁴ Reviewing the legal landscape, the court noted that such cases presented a difficult constitutional issue, requiring the balancing of the speaker's right to remain anonymous against the usefulness of identity information to the audience in evaluating the nature of the statement made.¹⁷⁵ It noted that the Supreme Court had both lauded the importance of broad disclosure while also distinguishing cases in which disclosure had to be made only to an agency as opposed to the community at large.¹⁷⁶ But it concluded that it was reluctant, “without clearer guidance from the Court” to invalidate the Indiana statute.¹⁷⁷ Judge Easterbrook, writing dubitante, noted:

Doubtless “a speaker’s credibility often depends crucially on who he is.” But how does this support obligatory disclosure? Speakers who prefer concealment in order to reduce their personal risks, and who accept the discount that readers attach to advocacy from unnamed sources, do not impose burdens on strangers. What then is the justification for regulation?¹⁷⁸

The First Amendment does not, of course, stand in the way of a state’s prohibiting false statements about another under the common law of defamation. But it might stand as a hurdle to a state’s ability to prohibit a different kind of allegedly “false” statement: the use of a pseudonym in the context of advocacy, even if that use deprives audiences of information that would be helpful to their evaluation of the communication. The choice of attributional signal thus becomes part of the content that is conveyed;¹⁷⁹ to say otherwise would mean invalidating all pen names, trademarks, screen names, and other monikers.¹⁸⁰

This privileging of authorial speech over audience interests can be seen in other First Amendment-related areas of the law.¹⁸¹ For example, the doctrine arising from the Court’s 1964 opinion in New York Times v. Sullivan, which allows public figures to bring a claim for defamation only if they can show the false statement was made with malice—“knowledge that it was false or with reckless disregard of whether it was false or not”¹⁸²—means that some false statements about public officials will circulate without a remedy. Beyond the concept that speech about public officials requires more breathing room, the doctrine also assumes, as commentators have noted, a certain ability on the part of individuals to separate wheat from chaff in such communications—to consider the nature of the communication, the identity of the speaker (if available), and the other information circulating about the public figure and decide whether the statement is worth believing or not.¹⁸³

Similarly, in various cases decided by the Supreme Court concerning the regulation of advertising, the Court has suggested it is not the government's place to save consumers from unwise decisions, so long as fraud or deception are not at issue. For example, in a 1976 case in which the Court invalidated the Commonwealth of Virginia's ban on the advertising of prescription drug prices by pharmacists, the Court highlighted the presumed ability of consumers to engage with the information they are provided:

It appears to be feared that if the pharmacist who wishes to provide low cost, and assertedly low quality, services is permitted to advertise, he will be taken up on his offer by too many unwitting customers. . . . They will respond only to costly and excessive advertising, and end up paying the price. . . . All this is not in their best interests, and all this can be avoided if they are not permitted to know who is charging what.

There is, of course, an alternative to this highly paternalistic approach. That alternative is to assume that this information is not in itself harmful, that people will perceive their own best interests only if they are well enough informed, and that the best means to that end is to open the channels of communication rather than to close them. If they are truly open, nothing prevents the "professional" pharmacist from marketing his own assertedly superior product, and contrasting it with that of the low-cost, high-volume prescription drug retailer. But the choice among these alternative approaches is not ours to make or the Virginia General Assembly's. It is precisely this kind of choice, between the dangers of suppressing information, and the dangers of its misuse if it is freely available, that the First Amendment makes for us.¹⁸⁴

Although the details of an individual case might matter, the overall theme of this jurisprudence is that the government cannot always, consistent with the First Amendment, require authors to disclose information, like accurate attribution, that audiences might find valuable. Attribution is content, and speakers' interests in choosing that content have First Amendment implications.

D. Putting It Together

The idea that authorial attribution is content, with the attendant speech and autonomy implications, perhaps helps to explain attribution's general absence from IP law. Post-Dastar, attributional interests are largely the province of copyright law, which itself leaves attributional interests largely to private ordering. (Attributional interests are distinct from copyright law's core concern of authorial control of the use of protected works.) Likewise, the First Amendment interests inherent in choosing a form of attribution, or none at all, align with a hands-off role for regulators. The law does not require any creative output to bear attribution; it generally does not require attribution where there is none; and it does not require an attribution signal to be identical to any other name a creator or company uses.

All of this suggests that to further codify attributional rights beyond what can already be found in the law would not only be practically challenging but would also risk providing suboptimal solutions. As the U.S. Copyright Office concluded in its 2019 report, despite attribution’s significance to many creators, regulators need to consider

how a U.S. moral rights regime would work across industries and types of works [given the existence of] sectors where attribution interests tend to be governed by sometimes-written, sometimes just generally understood norms, rather than by contract or statutory law. . . . A one-size-fits-all moral rights regime would risk obliterating these context-specific and largely self-governing differences between types of works and industries.¹⁸⁵

In short, the flexibility inherent in attributional choices and the varying interests of authors and audiences naturally lead to disuniformity across creative spaces and suggest a role for those spaces to play in creating attribution mechanisms.

III. The Norms and Practices of Attribution

Whether in response to law’s deficiencies or regardless of them, various communities have established norms and practices around attribution.¹⁸⁶ As Catherine Fisk has noted, this is at least in part what authorship means—not only to be the beneficiary of legal rights but also to be “constituted by social and legal processes of recognition” that accord one status in a particular community.¹⁸⁷

Nonlegal enforcement mechanisms for attribution can resemble legal structures. In the academic community, for example, attribution is enforced through rules around plagiarism, such as by strong community norms, a school Honor Code, or disciplinary review boards. In the art world, attribution can be determined both by the market and by artist authentication boards, which sometimes reach opinions that are at odds with one another.¹⁸⁸ Many scholars have written about what has now become known as “IP without IP”—a topic that includes community norms around attribution and credit.¹⁸⁹ Attribution may also function differently in spaces in which participants already have established relationships. For example, if members of a community already know one another from interpersonal, face-to-face interactions, one might see a trend toward requiring participants to use their given names in the online version of that community.¹⁹⁰ And online, various attempts have been made—some more successful than others—to develop rules or technologies that allow participants to attribute more easily or disclaim authorship.¹⁹¹

The implementation of attributional choices can, of course, become easier online, where platforms can develop systems to facilitate not only the choice of one’s own attributional signals but also the attribution of content to others, whether directly (by the user of content) or indirectly (by others who use technologies such as reverse image searches to discover and then highlight original sources).¹⁹² The choice between these kinds of organic systems and systems implemented by institutions, whether legal or platform, tracks the typical standard-rules tradeoff. In the legal literature, this tradeoff notes that legal rules (for example, laws that set the maximum speed limit on certain roads) provide clear guidance about what the law is but do not easily allow exceptions (the person rushing an ill friend to the hospital). By contrast, laws that set regulation by standards (people must drive in a reasonable manner) allow for flexibility but risk uncertainty and capriciousness in enforcement.¹⁹³ In this context, then, systems that create attribution requirements need advance rules about the various contexts in which people might or might not want to attribute. Does a parody require attribution of the material being parodied? How does one attribute in a mash-up? But such systems also have to recognize that one size may not fit all.

Thus, organizations that are constructed more formally around membership may have rules around whether one can freely choose one's authorial or professional name or whether such choices must defer to the previous choices of others. Roller derby participants, as described by David Fagundes in 2012, maintain a "Master Roster" that keeps track of skaters' names and generally prohibits duplication of names even if two skaters compete in distant leagues and so are unlikely to be in a situation where viewers might be confused about which skater is which.¹⁹⁴ Actors' Equity Association, the actors' union, prohibits members from using a name that is the same as or confusingly similar to another member's name, unless the other member consents or the organization finds that there is no likelihood of confusion.¹⁹⁵ (Actors who have to change their name cannot simply add a middle initial; they must change the first or last name completely or add a full middle name.¹⁹⁶)

By contrast, communities that have more fluid and dynamic membership may develop attribution norms through technological means that allow for adaptation and changes in approach. As readers are no doubt deeply familiar, the social media network formerly known as Twitter is a platform that allows users to broadcast short messages either to the public or to one's "followers" (depending on the user's settings). Because, as one researcher has noted, Twitter enables a form of social communication and thus "sit[s] between mere status updates to any and all followers, and the direct (private) messages between two users which the Twitter platform also supports,” users developed certain conventions that provided points of reference in a conversation long before the platform was modified to accommodate those conventions.¹⁹⁷ One of these conventions was the “retweet,” which enabled a user to repost another’s message and give credit to the originator.¹⁹⁸ At first, this was simply a matter of manual convention. A user would type “RT” followed by the other’s username, before the content that was being redistributed: “RT @JohnDoe: Cogent commentary on today’s news out of the White House. http://bit.ly/XYZabc.” The “at” symbol served to indicate that the characters that directly followed were another’s username, and the text that followed the colon represented user JohnDoe’s original message. The user who retweeted then could also append their own text before the retweet: “I agree completely. RT @JohnDoe: Cogent commentary on today’s news out of the White House. http://bit.ly/XYZabc.”¹⁹⁹ This convention thus highlighted the various dimensions of attribution—here, not only giving credit to the originator of content or an idea but also indicating one’s role in a conversation, particularly when the retweeter appended their own text to the retweet.²⁰⁰

In response to this user innovation, Twitter developed its own retweeting functionality in the form of a “retweet” button: the functionality allowed users to retweet without depleting character limits and included the original message verbatim, maintaining the integrity of the original tweet.²⁰¹ Despite the ease this provided, one researcher characterized Twitter’s response as “based on an incomplete understanding of the practice of retweeting.”²⁰² In regularizing the process, it “removed any direct opportunity for users to indicate the context in which they sought to make such retweets.”²⁰³ Here, then, we have the typical standard-rule tradeoff: a systematic operation is easier to implement and monitor for compliance, but it may not be flexible enough to be appropriate for every situation.²⁰⁴

TikTok, by contrast, was slower to arrive at a similar result.²⁰⁵ TikTok’s use of automatic content recognition occasionally resulted in what users deemed to be misattribution to sources other than the original.²⁰⁶ As a result, participants on the platform engaged in manual efforts to provide credit, such as including the names of others in on-screen text, manipulating metadata, and other ways of circumventing the technology.²⁰⁷ These efforts might have reflected frustration with the capabilities of the platform but also signaled a community desire for a more baked-in approach to attribution. As one group of scholars explained at the time:

TikTok’s laissez-faire approach to attribution creates economic value for the platform but may do little to motivate creators, if they feel their creative labor is going unnoticed. Yet, the additional effort required from other creators, to acknowledge original authorship through (mis)attribution practices, can build “emotional value” that contributes to community building and incentivizes creativity.²⁰⁸

No doubt in response to these concerns, TikTok introduced “new tools to better enable creator credit and equitable attribution for our creator community and content originators” in May 2022 so as to “support a culture of credit, which is central to ensuring TikTok remains a home for creative expression.”²⁰⁹

Without platform control, not every attempt at a systemic solution has been successful. For example, in 2012, Maria Popova and Kelli Anderson developed what they called the Curator’s Code. The Code supplied symbols for blogs and other websites to use to designate either a link of “direct discovery” (known as a “via”)—meaning content that came directly from another source—or a link of “indirect discovery” (known as a “hat tip”)—meaning that the idea, lead, or inspiration came from another source.²¹⁰ Popova told the New York Times that “[d]iscovery of information is a form of intellectual labor . . . When we don’t honor discovery, we are robbing somebody’s time and labor.”²¹¹ Unlike with Twitter, where the technology developed organically in response to community practices, this system was ultimately unsuccessful, likely because it required wholesale buy-in and adoption.²¹² A similar reaction attended the Council on Ethical Blogging and Aggregation, which aimed to develop a set of guidelines for aggregation and attribution online²¹³ and whose efforts were criticized by some on the grounds that the guidelines merely codified what “decent online writers already know” and thus added an unnecessary level of bureaucracy.²¹⁴

Contrast these efforts, however, with the much greater success of Creative Commons. The organization, founded by Lawrence Lessig in 2001, created a set of licenses enabling copyright holders to depart from the default choices that U.S. copyright law would otherwise provide.²¹⁵ In particular, these licenses allowed copyright holders to forgo certain rights (or, put differently, cede those rights to the public) while retaining others.²¹⁶ Most notably, one of those rights—not generally part of U.S. copyright law—is the right to attribution. Initially, attribution was simply one of a number of options that copyright holders could choose for their license, but it became so popular (with over ninety-seven percent of those using the licenses in 2004 selecting this option) that attribution became a default component of every Creative Commons license.²¹⁷ Two further moderately successful success stories include (1) the development and adoption of the ORCID identification system by the academic community as a way of providing unique identifiers and avoiding the problem of name disambiguation²¹⁸ and (2) the requirement among some free and open-source software licenses that condition use of code on the provision of a notice file that provides attribution to the original source.²¹⁹ Unlike the systems described above, each of these systems found success by engaging a broad range of community members in the system’s development.

Despite these successes, automated tools or other standardized practices will not always satisfy a community’s desires. Because attribution is an important aspect of community norms, it may well require individual, manual effort as a signal of compliance with those norms, as opposed to standardization. This was the conclusion of a 2011 study, which found that “credit-giving, done manually, is associated with more positive reactions but that automatic attribution by the system is not associated with a similar effect.”²²⁰ The study’s scholars thus proposed that “designers should create affordances that make it easier for users to credit each other, rather than to merely pursue automated means of acknowledgment.”²²¹ A commentator and participant in the fandom community Archive of Our Own offered a similar critique of a “kudos” button adopted by the community; in the commentator’s view, the button “provides the illusion of giving feedback and minimizes reader engagement with the text,” changing the community from “[one] of equal exchange into one of consumption.”²²² This likewise suggests the value of a community norms-based approach to attribution over positive law. While norms can prove to be a barrier for new entrants to a community, they can more easily adapt to changing understandings and needs than statutory law, which imposes rules from the top down.

These are lessons that the many generative AI model developers must grapple with as they try to determine whether and how to incorporate attribution norms into their models.²²³ Beyond the Dastar challenge of whom to attribute when the contribution of any one source is indeterminate, it may well be that any automated solution will not carry the same meaning in every community.²²⁴ To what extent will users of AI expect an attributional signal indicating that text or an image is generated by AI or indicating the source material from which the model learned? To what extent will automating the production of such symbols interfere with the creative intent of those prompting the model?²²⁵ What is the appropriate tradeoff between the rule-like efficiency of top-down automation and the standard-like satisfaction of more individual interventions? The examples in this Article provide lessons for lawmakers, technologists, and researchers—not only about the human desire for attribution as a fundamental matter but also about the limits of both positive law and technology to fully satisfy that desire.

Footnotes

*James G. Cutler Professor of Law, William & Mary Law School. Many thanks to the Jurimetrics editorial team for their careful reads and many thoughtful suggestions. Any remaining errors should be attributed to me alone.

1. For recent treatments, see Dan L. Burk, Cheap Creativity and What It Will Do, 57 GA. L. REV. 1669, 1673 (2023) (contending that artificial intelligence “shifts our concern away from the problems of unauthorized reproduction and falling marginal costs toward concerns over the value of authenticity”); Stefan Bechtold & Christopher Jon Sprigman, Intellectual Property and the Manufacture of Aura, 36 HARV. J. L. TECH. 291, 295 (2023) (describing “an environment in which artifacts are promiscuously reproduced but where aura persists or is even manufactured”). ↩︎

2. The belief that an NFT is imbued with the concept of authenticity is furthered by artists’ claims that their permission is needed to create an NFT associated with their work. See, e.g., Katarina Feder, Know Your Rights: Can I Make an NFT of Someone Else’s Artwork? + More Artists-Rights’ Questions, Answered, ARTNET NEWS (Mar. 25, 2021), https://news.artnet.com/art-world/know-your-rights-march-nft-1954800 [https://perma.cc/5DHM-VSZU] (“[Y]ou can make an NFT of your own work free and clear. If you want to make an NFT of an artwork that’s not your own, you need to go to the source for permission.”). Neither the technology nor the law supports that claim where the NFT does not embody a copy of the work but merely points to the work elsewhere on the Internet.

In a 2022 study, researchers discovered the uniqueness of NFTs provided a psychic value in addition to its functional value. In other words, the fact that the owner of an NFT could justifiably claim to be the only one to own that particular NFT conferred value in and of itself. Tanusree Sharma et al., “It’s a Blessing and a Curse”: Unpacking Creators’ Practices with Non-Fungible Tokens (NFTs) and Their Communities, ARXIV 1, 3, 9 (Feb. 2022), https://arxiv.org/pdf/2201.13233 [https://web.archive.org/web/20250609180127/https://arxiv.org/pdf/2201.13233]. ↩︎

3. See, e.g., Carys J. Craig, The AI-Copyright Trap, 100 CHI.-KENT L. REV. (forthcoming 2025), https://papers.ssrn.com/sol3/papers.cfm?abstract\_id=4905118; Letter from Robert J. Kasunic, Assoc. Reg. of Copyrights & Dir. of Registration Pol’y & Prac., U.S. Copyright Off., to Van Lindberg, Taylor English Duma LLP (Feb. 21, 2023), https://www.copyright.gov/docs/zarya-of-the-dawn.pdf [https://perma.cc/Z4CS-QG5T] (analyzing copyrightability for content generated by Midjourney). ↩︎

4. Taylor Lorenz, The Original Renegade, N.Y. TIMES (Feb. 13, 2020), https://www.nytimes.com/2020/02/13/style/the-original-renegade.html [https://perma.cc/6VSL-CJJL]; Jewel Wicker, Renegade Creator Jalaiah Harmon on Reclaiming the Viral Dance, TEEN VOGUE (Apr. 15, 2020), https://www.teenvogue.com/story/jalaiah-harmon-renegade-creator-viral-dance [https://perma.cc/6GZ2-VDWH]. The incident, of course, involves considerations of race and class, among others. ↩︎

5. Lorenz, supra note 4; see also Morgan Sung, Black TikTok Creators Are “Striking” to Protest Uncredited Viral Dance Trends, MASHABLE (June 23, 2021), https://mashable.com/article/black-tiktok-strike-dance-megan-thee-stallion-thot-shit [https://perma.cc/E89N-H8AG]. ↩︎

6. Scholars differ on their assessment of this state of affairs. Compare Rebecca Tushnet, Naming Rights: Attribution and Law, 2007 UTAH L. REV. 789, 822 (2007) (“Sometimes, however, law and morality should be left to diverge, when law’s tools are too crude to make the fine distinctions that prevail in ethics.”) with Jane C. Ginsburg, The Most Moral of Rights: The Right to Be Recognized as the Author of One’s Work, 8 GEO. MASON J. INT’L COM. L. 44, 45 (2016) (“Of all the many counter-intuitive features of US copyright law—and they abound—the lack of an attribution right may present the greatest gap between perceived justice and reality.”); id. at 48 (rejecting as “overstated” the claim that “it’s simply too difficult to implement an attribution right in practice”). ↩︎

7. 17 U.S.C. § 106A. ↩︎

8. 17 U.S.C. § 1202. ↩︎

9. Laura A. Heymann, The Birth of the Authornym: Authorship, Pseudonymity, and Trademark Law, 80 NOTRE DAME L. REV. 1377 (2005) [hereinafter Heymann, Birth of the Authornym]. ↩︎

10. See Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23, 37–38 (2003). ↩︎

11. See, e.g., Jennifer E. Rothman, The Questionable Use of Custom in Intellectual Property, 93 VA. L. REV. 1899, 1925–26 (2007) (describing communities where the norm is to expect attribution for use of another’s material); id. at 1926 (“This attribution norm stands in stark contrast to the governing legal regimes which do not generally require, and certainly do not highlight, acknowledgement of sources or influences.”); cf. Mark A. Lemley, The Law and Economics of Internet Norms, 73 CHI.-KENT L. REV. 1257, 1260 (1998) (taking a “skeptical look” at “the idea that law should give deference to private norms on the Net”); id. at 1267 (noting that “virtually all of the empirical work on norms has taken place in small, close-knit communities with little change in membership over time” and that “[n]orms develop most clearly and most easily in a static community”). ↩︎

12. Nathan Wintersgill, Trevor Stalnaker, Laura A. Heymann, Oscar Chaparro & Denys Poshyvanyk, “The Law Doesn’t Work Like a Computer”: Exploring Software Licensing Issues Faced by Legal Practitioners, in 1 PROCEEDINGS OF THE ACM ON SOFTWARE ENGINEERING 40:1, 40:15 (2024), https://dl.acm.org/doi/pdf/10.1145/3643766; cf. Tushnet, supra note 6, at 801 (“English students and law professors aside, people rarely read footnotes, read through the credits of a film, or pay attention to disclosures in general.”). ↩︎

13. About CC Licenses, CREATIVE COMMONS, https://creativecommons.org/share-your-work/cclicenses/ [https://perma.cc/68DZ-YZV2]. ↩︎

14. Casey Fiesler & Amy S. Bruckman, Creativity, Copyright, and Close-Knit Communities: A Case Study of Social Norm Formation and Enforcement, in 3 PROCEEDINGS OF THE ACM ON HUMAN-COMPUTER INTERACTION 241:1, 241:7 (2019), https://dl.acm.org/doi/pdf/10.1145/3361122; Aram Sinnreich et al., Ethics Reconfigured: How Today’s Media Consumers Evaluate the Role of Creative Reappropriation, 12 INFO. COMMC’N & SOC’Y 1242, 1253 (2009) (noting that for many respondents in the study’s survey, “authenticity and legitimacy are premised on the explicit acknowledgment of the source materials or ‘original creator.’”); see also Jeanne C. Fromer, Expressive Incentives in Intellectual Property, 98 VA. L. REV. 1745 (2012). ↩︎

15. Consider, for example, the meme community, where the entire point is to make meaning collectively without an expectation of attribution to the originator of the meme. Amy Adler & Jeanne C. Fromer, Memes on Memes and the New Creativity, 97 N.Y.U. L. REV. 453, 532 (2022). ↩︎

16. See infra notes 197–204 and accompanying text. ↩︎

17. See How It Works, CONTENT AUTHENTICITY INITIATIVE, https://contentauthenticity.org/how-it-works (last visited Apr. 6, 2025); Scott Lowenstein, Using Secure Sourcing to Combat Misinformation, N.Y. TIMES (May 5, 2021), https://rd.nytimes.com/projects/using-secure-sourcing-to-combat-misinformation [https://perma.cc/NAC7-DQLG]. ↩︎

18. For thinking along these lines, see, for example, Joey DeBruin, The Creator Attribution Ratio, FLYING PENGUINS (May 13, 2021), https://www.flyingpenguins.io/p/the-creator-attribution-ratio [https://perma.cc/SCN6-XJVY]. ↩︎

19. See Louis Kaplow, Rules versus Standards: An Economic Analysis, 42 DUKE L.J. 557, 561–62 (1992) (“One can think of the choice between rules and standards as involving the extent to which a given aspect of a legal command should be resolved in advance or left to an enforcement authority to consider.”). ↩︎

20. This Part builds on previous work in Laura A. Heymann, The Law of Reputation and the Interest of the Audience, 52 B.C. L. REV. 1341 (2011), [hereinafter Heymann, The Law of Reputation], and Heymann, Birth of the Authornym, supra note 9. ↩︎

21. Christopher Jon Sprigman et al., What’s a Name Worth?: Experimental Tests of the Value of Attribution in Intellectual Property, 93 B.U. L. REV. 1389, 1401 (2013) (characterizing the different values of attribution as “extrinsic value” (helping a producer gain more in the future), “intrinsic value” (the psychic benefit of attribution to the creator), “moral value” (social recognition), and “social value” (the value to consumers)). For an analysis of scholarly proposals around attribution, see Tushnet, supra note 6. ↩︎

22. Andrés Monroy-Hernández et al., Computers Can’t Give Credit: How Automatic Attribution Falls Short in an Online Remixing Community, in CHI ’11: PROCEEDINGS OF THE SIGCHI CONFERENCE ON HUMAN FACTORS IN COMPUTING SYSTEMS 3421, 3422 (2011), https://dl.acm.org/doi/pdf/10.1145/1978942.1979452. ↩︎

23. See, e.g., James Meese & Jennifer Hagedorn, Mundane Content on Social Media: Creation, Circulation, and the Copyright Problem, SOC. MEDIA & SOC’Y, Apr.–June 2019, at 1. ↩︎

24. Id. at 1, 6 (reporting respondents’ views that attribution matters differently when commercial versus noncommercial uses are at issue, as well as the belief that expectations of attribution depend on the amount of labor put into the work); Trevor Stalnaker et al., Developer Perspectives on Licensing and Copyright Issues Arising from Generative AI for Software Development, ACM TRANSACTIONS ON SOFTWARE ENG’G & METHODOLOGY (forthcoming 2025) (manuscript at 18), https://dl.acm.org/doi/pdf/10.1145/3743133 (reporting the view of a survey respondent that if a LLM copied a large amount of code, the developers of the code “deserve attribution”). ↩︎

25. Adam M. Grant & Francesca Gino, A Little Thanks Goes a Long Way: Explaining Why Gratitude Expressions Motivate Prosocial Behavior, 98 J. PERSONALITY & SOC. PSYCH. 946, 954 (2010) (“Small expressions of gratitude can motivate prosocial behaviors by leading helpers to feel socially valued.”). ↩︎

26. Monroy-Hernández et al., supra note 22, at 3422. ↩︎

27. See, e.g., Nate Matias, Gratitude and Its Dangers in Social Technologies, CIVIC MEDIA (Aug. 5, 2014), https://civic.mit.edu/blog/natematias/gratitude-and-its-dangers-in-social-technologies.html [https://perma.cc/9HDV-VZ92] (“I see acknowledgment as something focused on relationships and community, while attribution is more focused on a person’s moral rights and legal relationships with the things they create, as they are discussed and shared.”); Nate Matias, Researching Love and Thanks on Wikipedia: CrowdCamp Hackathon Report, CIVIC MEDIA, https://civic.mit.edu/blog/natematias/researching-love-and-thanks-on-wikipedia-crowdcamp-hackathon-report.html [https://perma.cc/5HE8-J73P]. ↩︎

28. One study suggests that even young children have an understanding that it is wrong to copy another’s idea. Kristina R. Olson & Alex Shaw, ‘No Fair, Copycat!’: What Children’s Response to Plagiarism Tells Us About Their Understanding of Ideas, 14 DEVELOPMENTAL SCI. 431, 438 (2011) (“[T]his research suggests that children as young as 5 years old differentially evaluate strangers who produce original work versus those who plagiarize.”). ↩︎

29. JESSICA SILBEY, THE EUREKA MYTH: CREATORS, INNOVATORS, AND EVERYDAY INTELLECTUAL PROPERTY 166–69 (2014); see also Jessica Silbey, We’re All Pirates Now: Making Do in a Precarious IP Ecosystem, 39 CARDOZO ARTS & ENT. L.J. 691, 698 (2021) (noting that most creators and innovators in her study “describe proper attribution and credit as central to ethical behavior and fairness in their field”). Sometimes, of course, the desire for attribution is motivated by economic interests. Melvin Gibbs et al., The Importance of Moral Rights to Authors, 8 GEO. MASON J. INT’L COM. L. 87, 89 (2016) (“For us, attribution—that is our currency. I don’t exist if people don’t know who I am. I mean that in the most literal sense of ‘I don’t eat.’”) (comments of musician/composer Melvin Gibbs). ↩︎

30. See Gibbs et al., supra note 29, at 91 (“Also, the fact is, is when you’re a performer, you built up a certain brand name on the basis of your previous albums. So even before your album comes out, it gets a value, an economic value, because it’s going to be based on the popularity of the songs before that, more people are going to listen to it, maybe buy it, get licensed in films.”) (comments of songwriter/recording artist David Lowery). ↩︎

31. See Rebecca Tushnet, Payment in Credit: Copyright Law and Subcultural Creativity, LAW & CONTEMP. PROBS., Spring 2007, at 135, 152 (discussing how fan creators often receive attribution, or credit, as a form of compensation rather than monetary payment). ↩︎

32. WILLIAM SHAKESPEARE, OTHELLO act 3, sc. 3, l. 188. ↩︎

33. Heymann, The Law of Reputation, supra note 20, at 1342. ↩︎

34. The late Columbia University professor Carolyn Heilbrun, for example, published mystery novels under the name Amanda Cross because she did not want her tenure committee to know that she was writing novels on the side. Robert D. McFadden, Carolyn Heilbrun, Pioneering Feminist Scholar, Dies at 77, N.Y. TIMES (Oct. 11, 2003), https://www.nytimes.com/2003/10/11/arts/carolyn-heilbrun-pioneering-feminist-scholar-dies-at-77.html [https://perma.cc/88WJ-HL8X]. Likewise, J.K. Rowling, the author of the Harry Potter series, published crime novels under the name Robert Galbraith because, in her words, "I really wanted to go back to the beginning of a writing career in this new genre, to work without hype or expectation and to receive totally unvarnished feedback. . . . Since my cover has been blown, I continue to write as Robert to keep the distinction from other writing and because I rather enjoy having another persona." About Robert Galbraith, ROBERT GALBRAITH, https://robert-galbraith.com/about/ [https://perma.cc/P3S6-SGGL]. ↩︎

35. See 1 J. THOMAS MCCARTHY, MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 3:12, Westlaw (database updated Nov. 2024). ↩︎

36. HAROLD LOVE, ATTRIBUTING AUTHORSHIP: AN INTRODUCTION 16 (2002); see also A Letter from Jerome (403), EPISTOLAE, https://epistolae.ctl.columbia.edu/letter/424.html [https://perma.cc/5JJD-XS6N]. ↩︎

37. 61 DESIDERIUS ERASMUS, VOLUME II PART 1 (1516), reprinted in COLLECTED WORKS OF ERASMUS: PATRISTIC SCHOLARSHIP 74 (James F. Brady & John C. Olin eds., 1992). ↩︎

38. Id. at 75. Interestingly, Erasmus’s discussion presages a parallel with trademarks. He continues, wondering why readers would accept an attribution on faith, rather than based on their own (or another’s) assessment of the work’s stylistic characteristics:

You are not satisfied with the label on a physician’s pillbox, but you sniff, handle, and taste the contents; nor do you at once believe the unguent to be balsam if the unguent jar is so labelled. Instead you take every precaution not to be fooled by a label and receive poison in place of medicine and mud in place of ointment. Not content with the inscription, you test a coin. Not trusting the price-tag, you examine closely a piece of cloth. You rub gold on a touchstone when uneasy about its colour. There are characteristics by which we distinguish a real gem from a counterfeit, and it is not enough simply to call it an emerald or an amber. In the case of books alone will any name at all assigned without good reason be enough for acceptance?

Id. ↩︎

39. See Multiple-Use Name, WIKIPEDIA, https://en.wikipedia.org/wiki/Multiple-use\_name [https://perma.cc/G8RL-CPXQ] (May 27, 2023, 1:46 PM); see also Mariano Longo & Stefano Magnolo, The Author and Authorship in the Internet Society: New Perspectives for Scientific Communication, 57 CURRENT SOCIO. 829, 847 (2009) (discussing use of multiple-use names and sites like Wikipedia, which do not depend on authorial identity for authority). ↩︎

40. Although it is certainly not a foolproof method, for environments that facilitate creation, such as virtual worlds, blog comments sections, social media platforms, and other sites of online engagement, an emphasis on stable pseudonymity can foster a desire to maintain the reputation attached to that pseudonym and avoid intemperate and misguided content. JULIAN DIBBELL, MY TINY LIFE: CRIME AND PASSION IN A VIRTUAL WORLD 23 (1998) (noting that “guest” visitors to a virtual world acted more brazenly than those who had made “the critical passage from anonymity to pseudonymity” that reflected a concern for reputation in that environment); Rebecca Tushnet, The Yes Men and The Women Men Don’t See, in A WORLD WITHOUT PRIVACY: WHAT LAW CAN AND SHOULD DO? 83, 107 (Austin Sarat ed., 2015) (noting that “the empirical evidence on real name policies as deterrents of misbehavior is unpromising”); id. at 119–22 (describing the value of pseudonymity in supporting community building). ↩︎

41. See Christopher Buccafusco & David Fagundes, The Moral Psychology of Copyright Infringement, 100 MINN. L. REV. 2433, 2457–62 (2016) (using moral foundations theory to explore the psychology behind why creators feel a dignitary harm when their work is used in unauthorized ways). ↩︎

42. See, e.g., Werner Güth et al., An Experimental Analysis of Ultimatum Bargaining, 3 J. ECON. BEHAVIOR & ORG. 367 (1982). ↩︎

43. Id. at 371. ↩︎

44. Id. at 373–74; see also Martin A. Nowak et al., Fairness Versus Reason in the Ultimatum Game, 289 SCIENCE 1773, 1773 (2000) (describing past studies and suggesting, in addition, that “fairness . . . is linked to reputation”). ↩︎

45. Adler & Fromer, supra note 15, at 532. ↩︎

46. Seth F. Kreimer, Sunlight, Secrets, and Scarlet Letters: The Tension Between Privacy and Disclosure in Constitutional Law, 140 U. PA. L. REV. 1, 85 (1991). ↩︎

47. As the literary theorist Gérard Genette has described, there are multiple ways in which an author’s name could be presented (or not): (1) complete omission of a name (anonymity); (2) false attribution to another with permission; (3) false attribution to another without permission; (4) false attribution to oneself with permission (ghostwriting); (5) false attribution to oneself without permission (plagiarism); (6) attribution to a fictional author; and (7) attribution to a fictional name (pseudonymity). GÉRARD GENETTE, PARATEXTS: THRESHOLDS OF INTERPRETATION 47–48 (Jane E. Lewin trans., 1997). ↩︎

48. Nathaniel Popper, Decoding the Enigma of Satoshi Nakamoto and the Birth of Bitcoin, N.Y. TIMES (May 15, 2015), https://www.nytimes.com/2015/05/17/business/decoding-the-enigma-of-satoshi-nakamoto-and-the-birth-of-bitcoin.html [https://perma.cc/RM3Z-XMPV]; see also SATOSHI NAKAMOTO, BITCOIN: A PEER-TO-PEER ELECTRONIC CASH SYSTEM (2008), https://bitcoin.org/bitcoin.pdf [https://perma.cc/FKV6-P6CJ]. ↩︎

49. See, e.g., Who Is Banksy? A Guide to the Street Artist, Printmaker and Anti-Establishment Provocateur, CHRISTIE’S (Sept. 12, 2024), https://www.christies.com/en/stories/collecting-guide-banksy-street-artist-3352f6af3e3441548909161320dc618a [https://perma.cc/FA8E-6APK]; Who Is Banksy? Everything We Know About the Anonymous Artist, BBC (Aug. 13, 2024), https://www.bbc.com/news/uk-england-bristol-50249349 [https://perma.cc/JR9W-TX53]. ↩︎

50. Of course, this accountability could also come in the form of a criminal or civil suit. And it can be engaged in perniciously, in what has become known as “doxing” or “doxxing.” E.g., A.B. 1979, 2024 State Assemb., Reg. Sess. (Cal. 2024). ↩︎

51. LOVE, supra note 36, at 7. ↩︎

52. Cf. Paul Ohm, Branding Privacy, 97 MINN. L. REV. 907, 911–12 (2013) (proposing that companies that handle customer information be required to choose different brand names for different privacy commitments). ↩︎

53. These are all questions now arising in the case of authorship and generative AI. See, e.g., Letter from Robert J. Kasunic to Van Lindberg, supra note 3. ↩︎

54. For additional discussion along these lines, see Laura Heymann, Authorship, Attribution, and Audience, JOTWELL (Dec. 7, 2015), https://ip.jotwell.com/authorship-attribution-and-audience/ [https://perma.cc/UH6U-Q8AB] (reviewing Kirsty Robertson, The Art of the Copy: Labor, Originality, and Value in the Contemporary Art Market, in PUTTING INTELLECTUAL PROPERTY IN ITS PLACE 158 (Laura J. Murray et al. eds., 2014) and Lionel Bently & Laura Biron, Discontinuities Between Legal Conceptions of Authorship and Social Practices: What, If Anything, Is to Be Done?, in THE WORK OF AUTHORSHIP 237 (Mireille van Eechoud ed., 2014)). ↩︎

55. For a discussion of discursive communities in the consideration of fair use, see Laura A. Heymann, Everything Is Transformative: Fair Use and Reader Response, 31 COLUM. J.L. & ARTS 445 (2008). ↩︎

56. See Bridgeman Art Libr. v. Corel Corp., 36 F. Supp. 2d 191 (S.D.N.Y. 1999). ↩︎

57. Id. at 196–97 (concluding that "slavish copying" does not manifest the originality required for copyright protection). ↩︎

58. Of course, conclusions about the value of the work may differ. Compare Peter Schjeldahl, The Joker: Richard Prince at the Guggenheim, NEW YORKER, Oct. 15, 2007, at 90, 92 (writing that Prince's prints of Marlboro cowboy advertisements "stick us with the fact that those pictures are beautiful" and that "[a]ny opinions we may have about advertising, cigarettes, and the West founder in our visual bliss"), with Robert Hughes, Mucking with Media: The Whitney Offers a Long Trek Through the Alien Goo, TIME, Dec. 25, 1989, at 93 (decrying that Richard Prince's appropriations "are credited with value as art"). As Andrew Gilden and Timothy Greene have noted in their discussion of fair use, whether such an individual is viewed as an artist or as an infringer may well depend on the individual's fame or wealth. Andrew Gilden & Timothy Greene, Fair Use for the Rich and Fabulous?, 80 U. CHI. L. REV. ONLINE 88, 99 (2013). ↩︎

59. Peter Jaszi & Martha Woodmansee, Introduction to THE CONSTRUCTION OF AUTHORSHIP: TEXTUAL APPROPRIATION IN LAW AND LITERATURE 1, 3 (Martha Woodmansee & Peter Jaszi eds., 1994). ↩︎

60. Catherine L. Fisk, The Modern Author at Work on Madison Avenue, in MODERNISM & COPYRIGHT 173, 179 (Paul K. Saint-Amour ed., 2011). Indeed, Fisk notes, because the copyright in work created by employees typically belongs to the employer (either under the work for hire doctrine or by contract), attribution becomes the only thing that a creator can hope for to maintain the connection to their work, even if it is unearthed long after the work is distributed to the public. Id. at 176 (“Once we realize that much modern creativity is exercised in an employment setting where salaried creators sign away their rights in their work as a condition of hire—sign away, in effect, their very status as authors—we can see that the attribution of work, rather than ownership of the intellectual property represented in it, defines the modern connection between many creators and work of all kinds.”). ↩︎

61. See Laura A. Heymann, Book Note, 85 CAL. L. REV. 761, 763 (1997) (reviewing WILLIAM DOMNARSKI, IN THE OPINION OF THE COURT (1996)); cf. Jeffrey S. Rosenthal & Albert H. Yoon, Judicial Ghostwriting: Authorship on the Supreme Court, 96 CORNELL L. REV. 1307 (2011) (analyzing Supreme Court opinions to determine reliance by justices on law clerks for opinion writing). ↩︎

62. Robert K. Merton, The Matthew Effect in Science, 159 SCIENCE 56, 62 (1968) [hereinafter Merton, Matthew Effect in Science I]; Robert K. Merton, The Matthew Effect in Science II: Cumulative Advantage and the Symbolism of Intellectual Property, 79 ISIS 606, 609 (1988); id. at 621 (discussing the relationship between attribution and the advancement of knowledge). ↩︎

63. Merton, Matthew Effect in Science I, supra note 62, at 59. ↩︎

64. Here, too, norms vary depending on the field. See, e.g., Don Riesenberg & George D. Lundberg, The Order of Authorship: Who's on First?, 264 J. AM. MED. ASS'N 1857 (1990) (proposing guidelines for the order of authors in medical articles).

Similar instincts shape views on celebrity endorsements. The theory behind such practices is that there is a transfer of meaning from the celebrity to the product for which consumers expect some level of authenticity. What, after all, is an endorsement but a loan of one's reputation to the product (or person) being endorsed through an association between a name and the subject of the endorsement? Lawrence Lessig, The Regulation of Social Meaning, 62 U. CHI. L. REV. 943, 1009 (1995) ("[Michael Jordan] endorses Nike shoes. Some of his social capital is transferred to the product endorsed, and the meaning of wearing Nike shoes changes."); GRANT MCCRACKEN, Who Is the Celebrity Endorser? Cultural Foundations of the Endorsement Process, in CULTURE AND CONSUMPTION II: MARKETS, MEANING, AND BRAND MANAGEMENT 97, 97 (2005) (setting forth a "meaning transfer" theory of celebrity endorsements). ↩︎

65. See, e.g., Saul Levmore, The Anonymity Tool, 144 U. PA. L. REV. 2191, 2208–09 (1996); cf. Simon Stern, Sentimental Frauds, 36 LAW & SOC. INQUIRY 83, 92 (2011) (discussing the concept of "natural fraud"—using falsehoods "aimed at making the victim credit the speaker with a kind of moral authority—which the victim relies on to define his own character"). ↩︎

66. For example, several writers noted the strong reaction to what they saw as an outing of the author who wrote under the name Elena Ferrante, when an article in the New York Review of Books linked that identity to the author's "true" identity. See, e.g., Nilanjana Roy, The Unmasking of Elena Ferrante: Just Who Is It For?, FIN. TIMES (Oct. 3, 2016), https://www.ft.com/content/a6cf402a8962-11e6-8aa5-f79f5696c731 ("Before Gatti's intrusive reporting, [Ferrante] had made a successful stab at splitting her authorial self and her personal self: a way of carrying out her supposed responsibilities to her readers, while retaining some freedom. . . . But Gatti's revelations have changed everything: almost every story on Ferrante today, even those highly critical of Gatti's article, include the name of the subject of his investigation. Soon, you will not be able to search for 'Elena Ferrante' without the other surfacing, the real-life author chained to the pseudonym. It is no longer possible to go back to not knowing."). ↩︎

67. Louis Menand, Literary Hoaxes and the Ethics of Authorship, NEW YORKER (Dec. 3, 2018), https://www.newyorker.com/magazine/2018/12/10/literary-hoaxes-and-the-ethics-of-authorship [https://perma.cc/8A4D-NXSS]; John Gregory Dunne, The Secret of Danny Santiago, N.Y. REV. BOOKS, Aug. 16, 1984, at 17, 25–27 (reviewing DANNY SANTIAGO, FAMOUS ALL OVER TOWN (1983)). ↩︎

68. For an overview of U.S. legislation in this area, see U.S. COPYRIGHT OFF., AUTHORS, ATTRIBUTION, AND INTEGRITY: EXAMINING MORAL RIGHTS IN THE UNITED STATES (2019), https://www.copyright.gov/policy/moralrights/full-report.pdf [https://perma.cc/Q5YQ-96TM]; see also Patrick R. Goold, The Lost Tort of Moral Rights Invasion, 51 AKRON L. REV. 1093 (2017). ↩︎

69. 17 U.S.C. § 102; 17 U.S.C. § 101 (definition of “fixed”). ↩︎

70. See Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 345 (1991) (noting that “original” means “only that the work was independently created by the other” and that “it possesses at least some minimal degree of creativity”). ↩︎

71. Time Inc. v. Bernard Geis Assocs., 293 F. Supp. 130, 143 (S.D.N.Y. 1968). ↩︎

72. Id. By contrast, the same court held more than thirty years later that photographs of public domain artwork (as would be used in an art history book) were not copyrightable given that the photographs involved “slavish copying” and thus no originality. Bridgeman Art Libr. v. Corel Corp., 36 F. Supp. 2d 191, 196–97 (S.D.N.Y. 1999). ↩︎

73. Alfred Bell & Co. v. Catalda Fine Arts, Inc., 191 F.2d 99, 105 (2d Cir. 1951). But see SUSAN SONTAG, ON PHOTOGRAPHY 117 (6th prtg. 1978). Sontag quotes Ansel Adams as saying, “A photograph is not an accident—it is a concept,” and she further notes, “To take a good photograph, runs the common claim, one must already see it.” Id. ↩︎

74. Cf. Jeanne C. Fromer, Claiming Intellectual Property, 76 U. CHI. L. REV. 719 (2009). ↩︎

75. 17 U.S.C. § 101 (defining “work made for hire”); id. § 201(b) (noting “the employer or other person for whom the work [made for hire] was prepared is considered the author for purposes of this title”). ↩︎

76. This was true even of the Copyright Act’s British progenitor, the Statute of Anne. David Saunders & Ian Hunter, Lessons from the ‘Literatory’: How to Historicise Authorship, 17 CRITICAL INQUIRY 479, 493 (1991) (noting the Statute of Anne did not “assume or require a necessary equivalence between the person of the copyright owner and the aesthetic persona of the writer”). ↩︎

77. 17 U.S.C. § 101. ↩︎

78. Id. ↩︎

79. See, e.g., Lindsay v. Wrecked & Abandoned Vessel R.M.S. Titanic, 52 U.S.P.Q.2d 1609, 1613 (S.D.N.Y. 1999) (“All else being equal, where a plaintiff alleges that he exercised such a high degree of control over a film operation—including the type and amount of lighting used, the specific camera angles to be employed, and other detail-intensive artistic elements of a film—such that the final product duplicates his conceptions and visions of what the film should look like, the plaintiff may be said to be an ‘author’ within the meaning of the Copyright Act.”). ↩︎

80. 17 U.S.C. § 106A. ↩︎

81. 17 U.S.C. § 106A(e)(1). ↩︎

82. 17 U.S.C. § 1202(c). ↩︎

83. 17 U.S.C. §§ 1202(a), (b). ↩︎

84. E.g., IQ Grp., Ltd. v. Wiesner Publ’g, LLC, 409 F. Supp. 2d 587, 598 (D.N.J. 2006). ↩︎

85. Agence Fr. Presse v. Morel, 769 F. Supp. 2d 295, 305–06 (S.D.N.Y. 2011). As the U.S. Copyright Office has noted, however, even given this interpretive breadth, section 1202 likely does not align with at least some core attributional instincts. For example, “[a]lthough it does establish a cause of action for attachment of false CMI as well as CMI alteration or removal in some situations, it does not create the right to be credited in the first place—something that many creators view as a shortcoming.” U.S. COPYRIGHT OFF., supra note 68, at 90. ↩︎

86. John Tehranian, Toward a New Fair Use Standard: Attributive Use and the Closing of Copyright’s Crediting Gap, 96 S. CAL. L. REV. 1, 34–35 (2022). ↩︎

87. See infra Part III. To the extent that a use is a fair use, of course, there will (appropriately) be no opportunity for negotiations. In a 2007 article, Greg Lastowka proposed including a consideration of attribution as an additional element in a fair use determination. Greg Lastowka, Digital Attribution: Copyright and the Right to Credit, 87 B.U. L. REV. 41, 84 (2007); see also Tehranian, supra note 86, at 58 ("draw[ing] and expand[ing] upon, with some important alterations," Lastowka's proposal). ↩︎

88. For useful information on trademark history, see Sidney A. Diamond, The Historical Development of Trademarks, 65 TRADEMARK REP. 265 (1975); Daniel M. McClure, Trademarks and Unfair Competition: A Critical History of Legal Thought, 69 TRADEMARK REP. 305 (1979); Benjamin G. Paster, Trademarks—Their Early History, 59 TRADEMARK REP. 551 (1969). ↩︎

89. Diamond, supra note 88, at 289. ↩︎

90. See id. at 280. ↩︎

91. The use of "Ask for" or "Look for" language in advertising enables this activity, as does "Accept no substitutes." ↩︎

92. See, e.g., FreecycleSunnyvale v. Freecycle Network, 626 F.3d 509 (9th Cir. 2010) (finding that plaintiff's trademark had been abandoned through failure to control use by authorized party). ↩︎

93. James Brooke, Factory Jobs Move Overseas as Japan's Troubles Deepen, N.Y. TIMES, Aug 31, 2001, at A1 (quoting an economist who described, "[Y]ou do the brand management in Seattle and the manufacturing in Indonesia"). ↩︎

94. It is not surprising, however, that companies like McDonald's may seek to establish different types of brand identities in different geographic areas, including internationally, in an attempt to respond to local markets. For a discussion on McDonald's relationship to the Black community, see MARCIA CHATELAIN, FRANCHISE: THE GOLDEN ARCHES IN BLACK AMERICA (2020); see also Laura A. Heymann, The Meaning of McDonald's [®], JOTWELL (Sept. 17, 2020), https://ip.jotwell.com/the-meaning-of-mcdonalds/ [https://perma.cc/E7LQ-FK2P] (reviewing CHATELAIN, supra). ↩︎

95. See Heymann, Birth of the Authornym, supra note 9; Laura A. Heymann, Naming, Identity, and Trademark Law, 86 IND. L.J. 381, 381–82 (2011). ↩︎

96. See, e.g., In re McUlta, 189 F. 250 (M.D. Pa. 1911) (holding that individual did not engage in deception by filing for bankruptcy after name change despite claim by creditors that failure to inform them of original name was fraudulent) (adopting order of referee); id. at 252 (“A name is used merely to designate a person or thing. It is the mark or indicia to distinguish him from other persons, and that is as far as the law looks.”). But see, e.g., In re Verrill, 660 N.E.2d 697, 699 (Mass. App. Ct. 1996) (concluding that granting petition of incarcerated individual for a name change would “cause significant confusion in the criminal justice system if he were ever released” because the individual might be able “later to elude criminal prosecution and conceal his identity”). ↩︎

97. 15 U.S.C. § 1125(a). An author or artist who has registered their name (or stage name) as a trademark for goods or services can take advantage of the statutory provisions governing infringement of registered trademarks to enjoin or receive damages from another individual or entity who has unlawfully used the name. See, e.g., LADY GAGA, Registration No. 3,695,038 (for entertainment services). ↩︎

98. Gilliam v. Am. Broad. Co., 538 F.2d 14, 18 (2d Cir. 1976). ↩︎

99. Id. at 24–25. ↩︎

100. Id. at 25. Notably, section 43 was available even though Monty Python also had a copyright infringement claim that was likely to succeed, given that it held the copyright to the script on which the unauthorized broadcast was ultimately based. One judge on the panel concurred to note his view that, given this, it was unnecessary, and perhaps unwise, for the court to consider an additional remedy under the Lanham Act. Id. at 26 (Gurfein, J., concurring). ↩︎

101. See King v. Innovation Books, 976 F.2d 824 (2d Cir. 1992). ↩︎

102. Id. at 826. ↩︎

103. Id. ↩︎

104. Id. ↩︎

105. Id. ↩︎

106. Id. at 827. ↩︎

107. Id. at 828. ↩︎

108. Id. at 833. ↩︎

109. Id. at 829. ↩︎

110. Id. at 833. ↩︎

111. Id. at 830. ↩︎

112. Id. at 831. ↩︎

113. King v. Allied Vision, Ltd., 155 F.R.D. 440, 444 (S.D.N.Y. 1994), aff’d in part, vacated in part, 65 F.3d 1051 (2d Cir. 1995). ↩︎

114. Waits v. Frito-Lay, 978 F.2d 1093 (9th Cir. 1992). ↩︎

115. Id. at 1097. ↩︎

116. Id. ↩︎

117. Id. ↩︎

118. Id. at 1097–98. ↩︎

119. Id. at 1098. ↩︎

120. Id. at 1103. ↩︎

121. Id. at 1106. ↩︎

122. Id. at 1111. ↩︎

123. Smith v. Montoro, 648 F.2d 602 (9th Cir. 1981). ↩︎

124. Id. at 603. ↩︎

125. Id. ↩︎

126. Id. ↩︎

127. Id. at 605. ↩︎

128. Id. at 603. ↩︎

129. Id. at 607. ↩︎

130. 539 U.S. 23 (2003). ↩︎

131. Heymann, Birth of the Authornym, supra note 9, at 1437–41. ↩︎

132. Dastar, 539 U.S. at 25. ↩︎

133. Id. at 25–26. ↩︎

134. Id. at 26. ↩︎

135. Id. ↩︎

136. Id. ↩︎

137. Id. ↩︎

138. Id. ↩︎

139. Id. at 27. ↩︎

140. Id. at 26–28. ↩︎

141. Id. at 27. ↩︎

142. Id. at 27–28, 38. ↩︎

143. Id. at 29. ↩︎

144. Id. at 37. ↩︎

145. Id. at 34. (I have discussed in Heymann, Birth of the Authornym, supra note 9, why precedent did not necessarily compel this conclusion.) The Court also noted that Congress had enacted a limited attribution right in VARA, so to recognize a broader right via the Lanham Act would render VARA superfluous. Dastar, 539 U.S. at 34–35. ↩︎

146. Id. at 35–36. The Court did suggest that even if copyright claims turned out not to be available in this case, the false advertising provision of the Lanham Act might provide some relief:

If, moreover, the producer of a video that substantially copied the Crusade series were, in advertising or promotion, to give purchasers the impression that the video was quite different from that series, then one or more of the respondents might have a cause of action—not for reverse passing off under the "confusion . . . as to the origin" provision of § 43(a)(1)(A), but for misrepresentation under the "misrepresents the nature, characteristics [or] qualities" provision of § 43(a)(1)(B).

Id. at 38. Note, however, that the basis for such a claim would rely on the nature of the assertions made in the advertising, not necessarily on the presence or absence of an attributional signal. ↩︎

147. Flaherty v. Filardi, No. 03 Civ. 2167, 2009 U.S. Dist. LEXIS 22641, at *25 (S.D.N.Y. Mar. 20, 2009) (suggesting, in a dispute over a screenplay, a Lanham Act claim might be available if the defendant had removed the physical title page from the plaintiff’s script and substituted a page listing his name as the screenwriter); Barefoot Architect v. Bunge, No. 04-99, 2009 U.S. Dist. LEXIS 99861, at *45 (D.V.I. Oct. 26, 2009), aff’d in part and vacated in part on other grounds, 632 F.3d 833 (3d Cir. 2011) (“Outside of a situation where the good offered is merely an unrevised carbon copy of an already existing product, use of the contents of such a product is not prohibited by the Lanham Act.”). ↩︎

148. Compare Moser Pilon Nelson Architects, LLC v. HNTB Corp., No. 05CV422, 2006 U.S. Dist. LEXIS 58334, at *34 (Aug. 7, 2006) (distinguishing case in which the defendant passed off plaintiff’s architectural design as its own on its website (not actionable) from case in which an architect took the plaintiff’s physical drawings, removed the plaintiff’s name and seal, and replaced them with his own (potentially actionable)), with Gensler v. Strabala, 764 F.3d 735, 737 (7th Cir. 2014) (“A false claim of authorship, without the making of copies (or some other act covered by 17 U.S.C. § 106), is outside the scope of copyright law [and so not precluded by Dastar].”). ↩︎

149. See, e.g., Zyla v. Wadsworth, 360 F.3d 243 (1st Cir. 2004); Steele v. Turner Broad. Sys., Inc., 607 F. Supp. 2d 258 (D. Mass. 2009); Marvel Worldwide, Inc. v. Kirby, 756 F. Supp. 2d 461 (S.D.N.Y. 2010); Williams v. UMG, 281 F. Supp. 2d 1177 (C.D. Cal. 2003), aff’d, Nos. 04-56314, 04-56398, 04-56399, 2006 U.S. App. LEXIS 12358 (9th Cir. May 12, 2006); Richard Feiner & Co. v. N.Y. Times Co., No. 07 Civ. 11218, 2008 U.S. Dist. LEXIS 58454, at *10–11 (S.D.N.Y. Aug. 1, 2008) (finding no Lanham Act violation under Dastar where newspaper was alleged to have unlawfully used plaintiff’s photograph in edition of newspaper, since defendant accurately identified itself as the producer of the edition of the paper). As noted above, the Dastar Court left open the possibility of a false advertising claim pursuant to section 43(a)(1)(B), but courts have rejected such claims when they are premised on the failure to attribute authorship, on the ground that if authorship were a quality or characteristic, Dastar’s holding would be eviscerated. See, e.g., Lapine v. Seinfeld, No. 08 Civ. 128, 2009 U.S. Dist. LEXIS 82304, at *45–46 (S.D.N.Y. Sept. 10, 2009). ↩︎

150. See, e.g., Michael Grecco Photography, Inc. v. Everett Collection, Inc., 589 F. Supp. 2d 375, 387–88 (S.D.N.Y. 2008) (holding that Dastar is not limited to works in the public domain); Pellegrino v. Epic Games, Inc, 451 F. Supp. 3d 373, 385 (E.D. Pa. 2020) (“Under Dastar, a claim that concerns the origin of an idea embodied in a tangible good is governed by copyright law, not the Lanham Act.”). But see id. at 385–86 (noting that Dastar does not bar claims for false endorsement, as opposed to false designation of origin). ↩︎

151. Capcom Co. v. MKR Grp., Inc., No. C 08-0904 RS, 2008 U.S. Dist. LEXIS 83836, at *36–37 (N.D. Cal. Oct. 10, 2008). Both claims ultimately failed. ↩︎

152. Bach v. Forever Living Prods. U.S., Inc., 473 F. Supp. 2d 1110, 1118 (W.D. Wash. 2007) (regarding the book Jonathan Livingston Seagull). ↩︎

153. See, e.g., Pellegrino, 451 F. Supp. 3d at 385–86 (citing cases). ↩︎

154. McIntyre v. Ohio Elections Comm'n, 514 U.S. 334, 337 (1995). I have previously discussed McIntyre and its implications for authorship in Heymann, Birth of the Authornym, supra note 9, at 1427–31. ↩︎

155. McIntyre, 514 U.S. at 337. ↩︎

156. Id. ↩︎

157. Id. at 338 n.3 (citing OHIO REV. CODE ANN. § 3599.09(A)). ↩︎

158. Id. at 338–40. ↩︎

159. Id. at 342. McIntyre had passed away, but her husband, as executor of her estate, pursued the case in the U.S. Supreme Court. Id. at 340. The opinion, written by Justice Stevens, garnered the votes of Justices O’Connor, Kennedy, Souter, Ginsburg, and Breyer. Justice Ginsburg wrote a concurring opinion that noted that narrower laws might well be found to be permissible. Id. at 358 (Ginsburg, J., concurring). Justice Thomas concurred in the judgment, writing separately to state his view that an originalist approach would yield the conclusion that the First Amendment was contemplated to apply to anonymous speech like McIntyre’s. Id. at 358–71 (Thomas, J., concurring in the judgment). Justice Scalia, joined by Justice Rehnquist, dissented, contending that the long-standing practice among states of prohibiting anonymous speech in connection with elections militated in favor of upholding the Ohio statute and that, in any event, the Ohio statute was a permissible attempt to reduce fraud and increase accountability in election-related communications. Id. at 371–85 (Scalia, J. dissenting). ↩︎

160. Id. at 346. ↩︎

161. Id. at 341–42. ↩︎

162. Id. at 342 (“Anonymity thereby provides a way for a writer who may be personally unpopular to ensure that readers will not prejudge her message simply because they do not like its proponent.”). ↩︎

163. See discussion supra Section I.B. Thus, as Alfred Moore has written, the debate involves “two contradictory normative principles. One is that anonymity is valuable because it enables expression free from fear of repercussions. The other is that anonymity is destructive because it enables expression free from fear of repercussions.” Alfred Moore, Anonymity, Pseudonymity, and Deliberation: Why Not Everything Should Be Connected, 26 J. POL. PHIL. 169, 169 (2018). ↩︎

164. The Court held similarly in Watchtower Bible & Tract Society of New York v. Village of Stratton, a 2002 opinion invalidating a local ordinance requiring door-to-door canvassers to first register their names with the mayor's office and obtain a permit. Even though such individuals would be presenting themselves in person to homeowners, the Court held that their names might not be known to those residents; preserving that anonymity implicated significant First Amendment-related interests. Watchtower Bible & Tract Society of N.Y., Inc. v. Vill. of Stratton, 536 U.S. 150, 167 (2002); see also Serv. Emps. Int'l Union, Local 3 v. Mun. of Mt. Lebanon, 446 F. 3d 419, 427 (3d Cir. 2006) (relying on Village of Stratton to invalidate part of municipal code that required registration with police department prior to door-to-door canvassing, noting that the court did "not regard . . . as a material distinction" the fact the ordinance required only registration and not a permit); Peterson v. Nat'l Telecoms. & Info. Admin., 478 F.3d 626, 632 (4th Cir. 2007) ("[T]he First Amendment protects anonymity where it serves as a catalyst for speech."); id. at 633 ("By distinguishing a person's physical appearance from his or her legal identity (name, address, etc.), Watchtower Bible merely establishes that individuals may speak in public without otherwise forfeiting their right to conceal the personal information necessary to locate and harm them in retaliation for engaging in unpopular speech.").

Some, of course, might say that a different evaluation and weighing of the balance should pertain today given the greater ability to distribute misinformation and disinformation through social media. For more recent, thoughtful overviews of the benefits and harms of online anonymity, see JOSH SMITH ET AL., WHAT'S IN A NAME?: A FORWARD VIEW OF ANONYMITY ONLINE (2020), https://demos.co.uk/wp-content/uploads/2023/02/Anonymity-FINAL-1.pdf [https://perma.cc/DPK7-J34G]; Moore, supra note 163; and Alfred Moore et al., Deliberation and Identity Rules: The Effect of Anonymity, Pseudonyms and Real-Name Requirements on the Cognitive Complexity of Online News Comments, 69 POL. STUD. 45 (2021). ↩︎

165. See, e.g., Brown v. Yost, 133 F.4th 725, 733–34 (6th Cir. 2025) (noting that the McIntyre Court considered the Ohio regulation at issue in the case to be a “direct regulation of the content of speech” because it required the supporters and opponents of a ballot initiative to include the names and addresses of the leaflets’ sponsors, thereby altering the content of their advocacy materials”). ↩︎

166. See Reply Brief of Petitioners, at 16, McIntyre v. Ohio Elections Comm’n, 514 U.S. 334 (No. 93-986) (1995), 1994 U.S. S. Ct. Briefs LEXIS 433 at *25 (noting that McIntyre did not intend the phrase to refer to an organization but, rather, to refer to the fact that she “believed that her leaflets expressed not only her views, but the views of other like-minded parents and taxpayers with whom she had communicated”). ↩︎

167. See Respondent’s Brief, at 1, McIntyre v. Ohio Elections Comm’n, 514 U.S. 334 (No. 93-986) (1995), 1994 U.S. S. Ct. Briefs LEXIS 322 at *5 (noting that the hearing before the Ohio Elections Commission “revealed that no such organization as ‘Concerned Parents and Tax Payers’ had ever existed”); Brief of the Council of State Governments et al. as Amici Curiae Supporting Respondents, at 15, McIntyre v. Ohio Elections Comm’n, 514 U.S. 334 (No. 93-986) (1995), 1994 U.S. S. Ct. Briefs LEXIS 514 at *25 (arguing that the Ohio statute promoted integrity in the election process by minimizing the potential “for deceptive presentation of a message so as to make it appear to speak for a larger or different constituency than it actually represents”); Brief for the Respondent, McIntyre v. Ohio Elections Comm’n, 1994 U.S. S. Ct. 433 (No. 93-986); id. at n.10 (“The fact that McIntyre represented on some of her flyers that they were authored by ‘Concerned Parents and Tax Payers’ could have misled voters in precisely this manner.”). ↩︎

168. See Note, Badwill, 116 HARV. L. REV. 1845 (2003). ↩︎

169. McConnell v. Fed. Election Comm’n, 540 U.S. 93, 128 (2003). Federal election law’s encouragement of corporate use of political action committees (PACs) may unintentionally encourage what one Justice called “compulsory ventriloquism.” Id. at 333 (Kennedy, J., concurring in part and dissenting in part); see also Citizens United v. Fed. Election Comm’n, 130 S. Ct. 876, 968 (2010) (Stevens, J., dissenting) (“[F]or-profit corporations associated with electioneering communications will often prefer to use nonprofit conduits with misleading names . . . to conceal their identity as the sponsor of those communications, thereby frustrating the utility of disclosure laws.”) (internal quotation marks omitted). ↩︎

170. Buckley v. Am. Const. L. Found., 525 U.S. 182, 198–99 (1999); id. at 199 (“The injury to speech is heightened for the petition circulator because the badge requirement compels personal name identification at the precise moment when the circulator's interest in anonymity is greatest.”). Colorado law also required circulators to file an affidavit with the circulator's name and address. The propriety of the affidavit requirement was not before the Court. ↩︎

171. 410 F.3d 760 (5th Cir. 2005). Because the university had designated the outdoor open areas of campus as public forums for student speech, its regulation was subject to strict scrutiny under the First Amendment. Id. at 769. ↩︎

172. Id. at 769–72. ↩︎

173. 361 F.3d 349 (7th Cir. 2004). ↩︎

174. Id. at 350. ↩︎

175. Id. at 351–52. ↩︎

176. Id. at 353. ↩︎

177. Id. at 355. ↩︎

178. Id. at 357 (Easterbrook, J., dubitante) (citations omitted). ↩︎

179. Laura A. Heymann, A Name I Call Myself: Creativity and Naming, 2 U.C. IRVINE L. REV. 585 (2012). Outliers do exist. See, e.g., Weiler v. Ritchie, 788 N.W. 2d 879 (Minn. 2010) (granting petition challenging appearance of candidate’s name on ballot under state statute which required names to be “the candidate’s true name or the name by which the candidate is commonly and generally known in the community”); Peter Nemerovski, You Can Call Me Al: Regulating How Candidates’ Names Appear on Ballots, 99 NEB. L. REV. 848 (2021). ↩︎

180. Notably, with respect to electioneering and campaign finance laws, the Court has taken a different approach. In McConnell v. Federal Election Commission, the Court rejected a challenge to a federal statutory provision requiring disclosure to the Federal Election Commission of the names of individuals who contributed over a certain amount to the direct costs of producing and airing “electioneering communications.” 540 U.S. 93, 199 (2003). The Court then treated this holding as precedent in its decision in Citizens United v. Federal Election Commission in rejecting a facial challenge to a related provision, which required disclosure of the person or entity responsible for a televised electioneering communication in the advertisement itself. 558 U.S. 310 (2010). Doe v. Reed followed, concluding (without extensive analysis) that disclosing the names of signatories to a referendum petition in response to a public records act request did not violate the First Amendment. 561 U.S. 186 (2010). The Court then returned to a more speech-protective approach in Americans for Prosperity Foundation v. Bonta in 2021, holding that a California regulation requiring tax-exempt charities to disclose their major donors to the state could not be sustained under the First Amendment, despite the government’s claim that the regulation prevented fraud. 594 U.S. 595 (2021). However, even the more restrictive of these cases seemingly leaves room for “Concerned Parents and Tax Payers” to be listed as the name of any entity related to advocacy communications, which arguably poses different concerns. In other words, the Court’s concern in these cases seems to be about disclosure vel non, not about verifiable attributional signals. ↩︎

181. See, e.g., Lyrissa Barnett Lidsky, Anonymity in Cyberspace: What Can We Learn from John Doe?, 50 B.C. L. REV. 1373 (2009); Lyrissa Barnett Lidsky, Nobody’s Fools: The Rational Audience as First Amendment Ideal, 2010 U. ILL. L. REV. 799 (2010) [hereinafter Lidsky, Nobody’s Fools]. ↩︎

182. 376 U.S. 254, 279–80 (1964). ↩︎

183. Cf. Lidsky, Nobody's Fools, supra note 181, at 825 ("First Amendment jurisprudence assumes, for most purposes, that the implied audience of core speech is comprised of rational information processors."). ↩︎

184. Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S. 748, 769–70 (1976). ↩︎

185. U.S. COPYRIGHT OFF., supra note 68, at 36–37. It’s worth quoting Rebecca Tushnet again: “Sometimes, however, law and morality should be left to diverge, when law’s tools are too crude to make the fine distinctions that prevail in ethics.” Tushnet, supra note 6, at 822. ↩︎

186. See, e.g., Gregory N. Mandel et al., Intellectual Property Law’s Plagiarism Fallacy, 2015 BYU L. REV. 915, 915 (“Across a wide variety of subject matters and contexts [in the authors’ studies], people tend to believe that simply providing proper attribution to the originator of a creative work or invention should enable the free copying of that work by others.”); id. at 948 (“There appears to be a significant disconnect between how intellectual property law treats attribution (or fails to treat it) and both creator and general public interest in the rights of creators to attribution.”); Fiesler & Bruckman, supra note 14, at 241:2 (“[W]hen people have imperfect knowledge of copyright law, norms can not only fill in the gaps, but they are often more strongly enforced than law.”). ↩︎

187. Fisk, supra note 60, at 176. ↩︎

188. A federal district court’s 1993 decision that a Calder mobile was authentic, to take but one example, lost any authority in the market when a recognized expert reached a contrary conclusion. See Patricia Cohen, Ruling on Artistic Authenticity: The Market vs. the Law, N.Y. TIMES (Aug. 5, 2012), https://www.nytimes.com/2012/08/06/arts/design/when-judging-arts-authenticity-the-law-vs-the-market.html; see also Amy Adler, Artificial Authenticity, 98 N.Y.U. L. REV. 706, 716–17 (2023). ↩︎

189. See KAL RAUSTIALA & CHRISTOPHER SPRIGMAN, THE KNOCKOFF ECONOMY: HOW IMITATION SPARKS INNOVATION 178–79, 185–88 (2012); Catherine L. Fisk, Credit Where It’s Due: The Law and Norms of Attribution, 95 GEO. L.J. 49 (2006); Jon M. Garon, Wiki Authorship, Social Media, and the Curatorial Audience, 1 HARV. J. SPORTS & ENT. L. 95, 100–02 (2010); Dotan Oliar & Christopher Sprigman, There’s No Free Laugh (Anymore): The Emergence of Intellectual Property Norms and the Transformation of Stand-Up Comedy, 94 VA. L. REV. 1787, 1809–31 (2008) (describing the norms governing attribution among stand-up comics); David Fagundes, Talk Derby to Me: Intellectual Property Norms Governing Roller Derby Pseudonyms, 90 TEX. L. REV. 1093 (2012) (roller derby names); Brian L. Frye, Plagiarism Is Not a Crime, 54 DUQ. L. REV. 133, 155 (2016) (“While some social groups adopt strong plagiarism norms, many adopt none at all. . . . If the members of a social group value attribution, then the group is likely to adopt internal plagiarism norms.”); see also ROBERT C. ELLICKSON, ORDER WITHOUT LAW: HOW NEIGHBORS SETTLE DISPUTES vii (1991) (“This book seeks to demonstrate that people frequently resolve their disputes in cooperative fashion without paying any attention to the laws that apply to those disputes.”). For a study testing the value given to attribution empirically, see Sprigman et al., supra note 21. ↩︎

190. David R. Millen & John F. Patterson, Identity Disclosure and the Creation of Social Capital, in CHI 2003: NEW HORIZONS 720, 720–21 (2003), https://dl.acm.org/doi/pdf/10.1145/765891.765950 (describing the “thick trust” that developed on an online site for residents of a small community in which users were required to identify themselves using their given names). ↩︎

191. Monroy-Hernández et al., supra note 22, at 3422 (noting in 2011 that “many designers of online communities pay little attention to issues of attribution in their designs—a fact that is reflected in user behavior”); Casey Fiesler, Note, Everything I Need to Know I Learned from Fandom: How Existing Social Norms Can Help Shape the Next Generation of User-Generated Content, 10 VAND. J. ENT. & TECH. L. 729, 746–754 (2008) (describing importance of social norms in online communities in regulating attribution, among other practices); Mark A. Lemley, Rights of Attribution and Integrity in Online Communications, J. ONLINE L., 1995, art. 2, at ¶ 52 (noting informal social norms against plagiarism can resolve some attribution issues), http://web.wm.edu/law/publications/jol/95\_96/lemley.html [http://web.archive.org/web/20100107023200/http://web.wm.edu/law/publications/jol/95\_96/lemley.html]. ↩︎

192. Xiyin Tang, Art After Warhol, 71 UCLA L. REV. 870, 923 (2024) (quoting an artist who noted "media is everywhere and you can have access to any information at any point" and "people will kind of catch you because they also have access to that information"); id. at 923–24 (quoting from an interview with an artist who "compared retweets and reposts on Instagram to a 'form of citation'—a way of acknowledging the original source material"). ↩︎

193. Cf. Kaplow, supra note 19, at 561–62. ↩︎

194. Fagundes, supra note 189, at 1117. ↩︎

195. Professional Name Protection, ACTORS' EQUITY ASS'N, https://www.actorsequity.org/join/WhyJoin/name-protection/ [https://perma.cc/92QK-MCFK] ("Actors' Equity Association shall not enroll an applicant under a name, nor shall a member use a name[,] which is the same as, or resembles (so closely as to tend to be confused with), the name of an existing enrolled member, except that an applicant may enroll under and use such name professionally, upon proof of consent by the existing member, or a finding by the National Council that under the circumstances there is no likelihood of confusion, or that there are extenuating circumstances."). ↩︎

196. Id. ↩︎

197. AXEL BRUNS, AD HOC INNOVATION BY USERS OF SOCIAL NETWORKS: THE CASE OF TWITTER § 2.1 (2011), <http://snurb.info/files/2011/Ad Hoc Innovation by Users of Social Networks.pdf> [https://perma.cc/ETU2-UD37]. ↩︎

198. Id. ↩︎

199. See, e.g., danah boyd et al., Tweet, Tweet, Retweet: Conversational Aspects of Retweeting on Twitter, in PROCEEDINGS OF THE 43RD HAWAII INTERNATIONAL CONFERENCE ON SYSTEM SCIENCES 1 (2010); id. at 3 (“There is no consistent syntax to indicate a retweet, attribution is inconsistent, the 140-character limitation and other factors prompt users to alter the original message, and adding commentary is prevalent. Furthermore, people use retweet language to reference content from other media and when paraphrasing others’ tweets. As a result, the text and meaning of messages often change as they are retweeted and the inconsistent syntax makes it difficult to track the spread of retweets.”). Users also developed variations on this theme including using “via” or HT (for “hat tip”) in place of the “RT” convention; for some users, these represented different types of attribution (for example, when the retweet did not involve a direct quote). Id. Questions also arose about how long of an attribution chain to acknowledge. Id. at 6 (noting “[s]ome users believe that it’s critical to attribute the chain of authors who passed along the message because this provides context and credit” while other users “chop the text such as to exclude multiple attributions,” believing that crediting only the most recent person in the chain was necessary). Here, we might recall the Dastar Court’s reference to a “search for the source of the Nile and all its tributaries.” See Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23, 35–36 (2003). ↩︎

200. boyd et al., supra note 199, at 1 (noting that retweeting “contributes to a conversational ecology in which conversations are composed of a public interplay of voices that give rise to an emotional sense of shared conversational context”). ↩︎

201. Biz Stone, Project Retweet: Phase One, X: BLOG (Aug. 13, 2009), https://blog.twitter.com/official/en\_us/a/2009/project-retweet-phase-one.html [https://web.archive.org/web/20241105060201/https://blog.x.com/official/en\_us/a/2009/project-retweet-phase-one.html] (“Retweeting is a great example of Twitter teaching us what it wants to be. . . . That’s why we’re planning to formalize retweeting by officially adding it to our platform and Twitter.com.”). ↩︎

202. BRUNS, supra note 197, § 2.3. ↩︎

203. Id. ↩︎

204. Bluesky, by contrast, included reposting and quote posting as core features of the platform. See Martin Kleppmann et al., Bluesky and the AT Protocol: Usable Decentralized Social Media, in DIN ‘24: PROCEEDINGS OF THE ACM CONEXT-2024 WORKSHOP ON THE DECENTRALIZATION OF THE INTERNET 1, 2, https://dl.acm.org/doi/pdf/10.1145/3694809.3700740. ↩︎

205. D. Bondy Valdovinos Kaye et al., You Made This? I Made This: Practices of Authorship and (Mis)Attribution on TikTok, 15 INT’L J. COMMC’N 3195, 3198 (2021). ↩︎

206. Id. at 3196–97. ↩︎

207. Id. at 3203. ↩︎

208. Id. at 3204 (internal citations omitted). ↩︎

209. Kudzi Chikumbu, Introducing More Ways to Tag and Credit Your Favorite Creators, TIKTOK (May 18, 2022), https://newsroom.tiktok.com/en-us/crediting-tools [https://perma.cc/FBT3-2MRY]. Part of this effort was no doubt in response to claims that Black artists’ contributions in particular were going uncredited. See also Chase DiBenedetto, Instagram Launches New Feature to Help Credit Black Creators, MASHABLE (Mar. 8, 2022), https://mashable.com/article/instagram-enhanced-tags-black-creators [https://perma.cc/JB3H-86ZR]. ↩︎

210. David Carr, A Code of Conduct for Web Aggregators, N.Y. TIMES, Mar. 12, 2012, at B1; Megan Garber, The Curator’s Guide to the Galaxy, ATLANTIC (Mar. 11, 2012), https://www.theatlantic.com/technology/archive/2012/03/the-curators-guide-to-the-galaxy/254294/ [https://web.archive.org/web/20221204233741/https://www.theatlantic.com/technology/archive/2012/03/the-curators-guide-to-the-galaxy/254294/]. ↩︎

211. Carr, supra note 210. ↩︎

212. See Ernie Smith, The Curator’s Code, Reconsidered, TEDIUM (Aug. 1, 2023), https://tedium.co/2023/08/01/blogging-curators-code-history/ [https://perma.cc/VX56-V5DL] (describing opposition from bloggers that made the proposal “dead in the water among its intended audience”). ↩︎

213. Carr, supra note 210. ↩︎

214. Hamilton Nolan, We Don’t Need No Stinking Seal of Approval from the Blog Police, GAWKER (Mar. 12, 2012, 10:00 AM), https://www.gawkerarchives.com/5892453/we-dont-need-no-stinking-seal-of-approval-from-the-blog-police [https://perma.cc/4EZZ-CY87]. ↩︎

215. See Share Your Work, CREATIVE COMMONS, https://creativecommons.org/share-your-work [https://perma.cc/8HPA-9WY8]. The organization reports that “Creative Commons licenses are used on over 2.5 billion works online across 9 million websites.” CREATIVE COMMONS, CREATIVE COMMONS CERTIFICATE FOR EDUCATORS, ACADEMIC LIBRARIANS, AND OPEN CULTURE § 1.1 (ebook), https://certificates.creativecommons.org/cccertedu/chapter/1-1-the-story-of-creative-commons/ [https://perma.cc/B67K-PCPQ]. ↩︎

216. As Greg Lastowka described it, “The Creative Commons licenses . . . operate, in part, to express ownership and make demands.” Lastowka, supra note 87, at 80. ↩︎

217. See Glenn Otis Brown, Announcing (and Explaining) Our New 2.0 Licenses, CREATIVE COMMONS (May 25, 2004), http://creativecommons.org/weblog/entry/4216 [https://perma.cc/N55B-5NPX] (“Our web stats indicate that 97–98% of you choose Attribution, so we decided to drop Attribution as a choice from our license menu—it’s now standard. . . . Important to remember: Attribution can always be disavowed upon licensor request, and pseudonymous and anonymous authorship are always options for a licensor, as before.”); Licenses List, CREATIVE COMMONS, https://creativecommons.org/licenses/ [https://perma.cc/2D8Q-UNRQ] (providing links to available licenses in various languages). ↩︎

218. ORCID (which stands for Open Researcher and Contributor Identifier) resulted from a meeting in November 2009 of publishers, scholarly organizations, and others aimed at developing a solution to the name disambiguation issue. Marcel Laflamme, On the Problem of the Namesake, 31 CULTURAL ANTHROPOLOGY 1, 2 (2016); ORCID, ORCID’S FIRST DECADE: FROM STARTUP TO SUSTAINABILITY 8 (2022), https://info.orcid.org/wp-content/uploads/2022/11/R2\_Orcid-10th-Ann-Booklet-FOR\_WEB.pdf [https://perma.cc/3TB2-QXHB]. ORCID’s report on its founding states that it “had already progressed further than earlier name identifier systems that had emerged within various communities, because of its engagement with all the stakeholders necessary to success.” ORCID, supra, at 10. Laflamme noted, on behalf of Cultural Anthropology, “Even as we have come to recognize [name disambiguation] as a problem in scholarly publishing, we also want to honor and engage with feminist efforts to challenge norms of individual authorship that involve writing under the sign of pseudonyms and collectives. . . . The sooner that anthropologists adopt this emerging standard, the more say we are likely to have about where, when, and how it is used in the future.” Laflamme, supra, at 2–3. ↩︎

219. Wintersgill et al., supra note 12, at 40:2 to 40:3. Similar systems can also be implemented by proprietary platforms. In 2014, Getty Images launched the ability to embed images for free for noncommercial purposes that include a watermark that provides attribution and links back to the Getty site. Gibbs et al., supra note 29, at 92 (“If it’s going to be shared, we want it to be shared with attribution, with a possibility and potential for people to find more images from these creators.”) (comments of Yoko Miyashita, Getty Images). ↩︎

220. Monroy-Hernández et al., supra note 22, at 3422. ↩︎

221. Id.; see also Kurt Luther et al., Edits and Credits: Exploring Integration and Attribution in Online Creative Collaboration, in CHI 2010: IMAGINE ALL THE PEOPLE, 2823, 2830–31 (2010) (suggesting systems be designed to “separate attribution and commendation” and suggesting that designers consider incorporating moral rights concepts in collaborative authoring environments). ↩︎

222. Ellen, What’s Wrong with the Kudos Button on Ao3 and What to Do About It, MEDIUM (Feb. 26, 2018), https://medium.com/@ellenannes/whats-wrong-with-the-kudos-button-on-ao3-and-what-to-do-about-it-203a9cc45cfd [https://perma.cc/WP7G-R2TH]. ↩︎

223. For examples of conversations on this topic, see Todd A. Carpenter, Ensuring Attribution Is Critical When Licensing Content to AI Developers, SCHOLARLY KITCHEN (Sept. 4, 2024), https://scholarlykitchen.sspnet.org/2024/09/04/make-attribution-mandatory-in-ai-licensing/ [https://perma.cc/32XV-9VCBJ] (“If AI tools do not have the capacity to reference back to the source material in their outputs, the reputation economy and the benefits that accrue to authors as a result will begin to fail.”); Stalnaker et al., supra note 24, at 23 (suggesting that developers may need to develop new licenses to control the use of their work by generative AI models). ↩︎

224. In this regard, consider an example from Reddit, where the subreddit “KarmaCourt” enables users to, in a light-hearted but also occasionally serious forum, seek justice for attributional and other transgressions on Reddit. Ned Donovan, Standing Trial in Reddit’s Karma Court, DAILY DOT, https://www.dailydot.com/irl/karma-court-reddit-doge/ [https://perma.cc/6LG4-DLNH] (May 31, 2021, 6:50 PM). Donovan writes that “the obvious satire that the subreddit offers is sometimes misinterpreted by its users, with bizarre results.” Id. For a slightly different take, see James Meese, “It Belongs to the Internet”: Animal Images, Attribution Norms and the Politics of Amateur Media Production, M/C J. (2014), https://www.journal.media-culture.org.au/index.php/mcjournal/article/view/782 [https://perma.cc/MT52-768C]. ↩︎

225. Among the many challenges in this context is the fact that generative AI is a technology and not a platform, so to the extent norms are developed by users, they may differ from platform to platform. The U.S. Federal Trade Commission released a report in December 2023 documenting views of creators. See U.S. FED. TRADE COMM’N, GENERATIVE ARTIFICIAL INTELLIGENCE AND THE CREATIVE ECONOMY STAFF REPORT: PERSPECTIVES AND TAKEAWAYS (2023), https://www.ftc.gov/system/files/ftc\_gov/pdf/12-15-2023AICEStaffReport.pdf [https://perma.cc/LDC5-9Z6Q]. ↩︎