Governing the Magic Circle panel 2

Mark Lemley –
Obama administration seems to be interested in directing ISPs to cut off parts of the internet they don’t like—an internet kill switch, which worked so well in Egypt. Lemley could only think they’d been there before, in 1994, when people argued that we should break the internet into little pieces that could only interact with each other through regulated portals. Rather than talking about gaming, he wants to be the curmudgeon: he was around for the first generation of internet law, with an eerie similarity to what’s going on now in law of virtual worlds.

In the early days, the internet was a thing unto itself, small and relatively insular subculture creating its own norms outside the view (and care) of the world at large. The internet was populated by second class citizens; what we did didn’t really matter.

Virtual worlds share many of these characteristics—you hear the same sorts of things from non-gamers—it’s just bits, ones and zeros. Had an argument with a prominent and savvy federal judge about whether one could defame an avatar; his answer was “no” because it’s not real, it doesn’t mean anything.

In various respects it sucks to be a second-class citizen. You can defame an avatar. Those who invest ourselves in such things understand that there can be real world consequences. One of the benefits, though, is that you were in fact left alone.

That didn’t last on the internet. As more money started to flow, more important people started to think attention had to be paid. As the internet started to substitute for things people bought offline—television, music—then we attracted realworld interest, in the form of being seen to cause harm. More and more people in the mainstream started to spend time on the internet, and the real harms that one could suffer became things people thought should be remedied.

Virtual worlds: becoming more and more real—in imagery, POV, interactivity, complexity, and also in how much money is at stake. As they become more real, we change the nature of those worlds, not just psychologically—it’s different to shoot someone who looks like you. As we invest more of ourselves in persistent personality, we subject ourselves to more harm. As we spend more money, the world outside sits up and takes notice.

Lemley lived through the declaration of independence of cyberspace, John Perry Barlow in 1994. That looks more like the revolution in Iran than the revolution in Egypt. Government rushed in and continues to do so. Magic circle is similar to the idea of the independence of cyberspace. This analogy should give us pause.

If you read books, watch TV, or play games: one ironclad law of entertainment is that if there is a magic circle, an insulation from surrounding harm, it will be breached. (The holodeck!) People find the ability to break through and harm shows up. (He says the other ironclad law is: a fishtank in an action movie will be shattered by the end.) We’re already seeing copyright and TM suits in Second Life; he defends Linden. While the traditional attitude of the outside world has been “who cares, it’s just a game,” that attitude is not going to last.

The first lesson of the internet independence movement is not particularly optimistic for staying unconnected to the law of the real world.

Other lessons: Code is law. Physical architecture of the world we create is at least as powerful a determinant of what people will do as legal rules/social norms. Having three internets would be as good as having none. We needed a set of choices trading off certain interests against others—we had to make it everywhere, consistently. The advantage of virtual worlds over the internet is that our games don’t need to interoperate and can be labs for experimentation with different rulesets and different levels of interaction between inside and outside worlds. Loss of one game’s independence from legal world is not necessarily catastrophic the way the loss of an internet architectural feature would be.

Law, which works by precedent, is backward-looking and tends to flail around when handed something new. Law wants an analogy—it’s not really completely new, it’s kind of like X. New spaces open up room to question things established in other, longer-existing spaces. Lots of resolved legal questions, so resolved we don’t think about them, turn out to be hard online, depending on what we think the internet is “like.” Orin Kerr: internal v. external perspectives. Is the internet the way it works? Or is it the way it appears? Are we hitting someone because it looks like we’re hitting someone? Keyword ads: No direct analogy to real world, but that didn’t stop courts from trying—this is just like hijacking, or this is like putting things next to each other on a grocery shelf. Opens up the possibility for us to think legal rules, maybe in ways that feed back to other spaces. If we rethink proximity online for TM, we might rethink it more generally for TM. We might see that with virtual worlds.

Opportunity/danger: a lot of legal rules are going to bump up against virtual worlds in unpredictable ways.

Next lesson: as a general matter, openness breeds creativity. One of the things we got right in internet was opening it to development. Other idea: if we figure out how to make ideas act just like land, we’ll make money. The fact that not doing that led to a powerful surge in creativity is important. Also, law can support creativity. Mostly by accident, §230 gave us free speech on the internet, intending to do the exact opposite.

Powers that be have every incentive to kill that openness. Internet revolution as separate space failed because people in real world fearful of losing business models are the ones who by and large end up having the power to pass legislation. Internet is still under threat, but nobody today would say what lots of scholars said 15-20 years ago, that it should be its own jurisdiction. Maybe so, but that’s not going to happen.

Odds are always good that big power and big money will find ways to control new communications media. It’s still possible to keep it open and not seized and sold back to us. Maybe virtual worlds will be different.

Laura Heymann - A Name I Call Myself: Creativity and Naming in Virtual Worlds

Offline example: people upset that other people “stole” the wonderful name for a newborn. The usual response is “get over it,” which is probably true, but there’s still a feeling of ownership/authorship with respect to a name. What’s at the root of that anguish? Due to TM-like concern about dilution. But it’s also about the creativity that went into finding the name and choosing the name—attributional concern with being recognized as the “author” of that name. Common metaphor of theft used to describe that duplication.

Naming systems/practices involve both TM and copyright-related interests. In the act of personal naming, the law doesn’t play much of a role despite the very common feelings/tacit agreements. Disconnect between how things operate and what the law provides/participants’ understanding of what the law provides.

Use virtual worlds as a laboratory.

Names aren’t subject to copyright protection. They aren’t long enough to contain the modicum of creativity required. There are of course many creative acts of naming. Justin Hughes also talks about social facts: names are a matter of social agreement, and once we’ve all agreed to call a person Robert, that name takes on the status of a fact. Relatedly, Dave McGowan suggested that copyright law shouldn’t operate in namespace. With copyright, we want to encourage variation/expression, whereas with names we want consistency and not workarounds. Also, to the extent copyright is about incentives, there would appear to be no need. People are more than incentivized to come up with names. In virtual worlds it’s mostly a condition of entry.

Same questions come up in TM. SCt noted that TM isn’t about creativity but denotation. Jeanne Fromer emphasizes that TM doesn’t go to the agency that creates the mark, but the entity that uses it in commerce. Yet TM is not devoid of creativity—Abercrombie spectrum gives a nod to creativity in that it gives more protection to coined/fanciful marks. Clever new TM diversifies both the marketplace and the marketplace of ideas, making them stronger/more valuable. Initial conception of dilution was targeted at coined/fanciful marks, giving copyright-like interests some play—you can’t copy that famous unique mark.

Same efficiency concerns about functionality apply, thus giving us nominative fair use. New Kids says you can use the team’s name to talk about it.

Yet: naming and creativity are intertwined in a way with significant social meaning. Particular ethnic/cultural groups name children in various ways. African-Americans may be more likely than white Americans to give distinctive names; both sets are more likely to play more with names for their daughters than names for their sons. Graffiti taggers: they use their names to express creativity. Writing over someone else’s tag is an act akin to assassination. Hip-hop performers rarely perform under original names; development of a name is an important part of hip-hop’s performativity. Name might connote things; engage in commentary/criticism/parody of existing structure; communicate directly with one or more individuals—the Roxanne wars of the 80s, battling over the right to the MC name Roxanne.

Naming as expressive: artist Kristin Sue Lucas, who in 2007 persuaded California court to allow her to change her name from Kristin Sue Lucas to Kristin Sue Lucas—a reboot. Court allowed her to do it to “refresh” herself as though she were a webpage.

Copyright doesn’t ignore relevance of names—the fact that a character has been named, as in Gaiman v. McFarlane, was important to whether the characters had been sufficiently delineated to be protected. (I’m not sure that “Medieval Spawn” is a sufficient name to make this persuasive; see also later cases finding that Tiffany and Domina were derivative works of Angela.)

Internet names express identity: ilovecats@aol.com. Dave Fagundes explores the world of roller derby pseudonyms—represent cleverness and ingenuity; creative names generate “name envy.” Community doesn’t allow duplication.  In Second Life, duplication is allowed—micro-creativity. Display names are now allowed: you can change your display name, which can differ from username; allows greater creative expression, including ideograms. Second Life has only three rules for display names: can’t use (1) celebrity’s name/TM, (2) engage in deception, (3) use vulgar, offensive, hateful, or harassing display name; but you can duplicate another’s display name. Led to some protest about diminished individuality/stolen creativity.

None of this is to say that we should extend the law to cover names/noncommercial activity, but that assumptions that underlie IP laws don’t always map very well to individual behavior. These interests are interests that IP takes care of in other areas—creativity, effort, credit, uniqueness. Virtual worlds can develop norms that either reflect those interests—derby girls—or decide to embody rules and norms that don’t map onto those concerns. Consider using these as laboratories: is there a reason to adjust the law.

Architecture also shapes creativity. Move to display names opened up creativity—allowed more characters (foreign languages), move from firstname-lastname. But constraints may also enhance creativity: 140-character limit in Twitte.

Ask whether participants’ experience shapes their expectations of what the law provides—whether a participant in a virtual world that doesn’t allow duplication of names comes to believe that the law should work in the same way. Where does a person upset about copying of her baby’s name get the idea that she has the right to be upset?

Rebecca Tushnet - Pseudonymity and Performance as Good Faith

I talked about the Archive of Our Own and users’ desires for play, privacy, and performance.

It’s a cliché that the internet allows people to try on new identities, liberating themselves—but potentially harming others. Anonymity is the usual target of critiques of such identity play, with theorists contending that anonymity is often used to harass and to perpetuate vicious sexual, racial, and other stereotypes, suppressing the very liberatory potential that disconnection from physical identity was supposed to offer. These analyses, while certainly offering powerful examples of the misuse of expression disconnected from identifiable people, often set up a binary opposition between a persistent identity linked to legal names and pure anonymity.

These are not, however, the only alternatives. I want to discuss how persistent pseudonymity and role-playing work in certain online spaces, where using a name that most people with whom you interact understand is not your “legal” name can create rich communities and interactive works of art. Talking about such spaces may help elucidate why proposals to require “real names” in online interactions can seem so threatening not just to “free speech” in the abstract, and not just to people living under oppressive regimes who fear being thrown in jail or worse for speaking out, but also to the project of the creation of an online self.

My discussion will focus on how pseudonyms, including role-playing pseudonyms, work in online spaces, using the example of a fan-run archive, the Archive of Our Own, which was designed to host fanworks—creative works based on existing media, such as Star Trek, and “real person fiction” or RPF, creative works starring actual celebrities, more or less realistically portrayed. The Archive has Terms of Service addressing specific concerns about pseudonymity, connecting pseudonyms with real identities, and playful use of “real” names. Users want to be able to distinguish between online and offline identities, maintain persistent identities that aren’t linked to their legal names, change those identities under certain circumstances, role-play characters (including celebrities), and deal with harassment. These goals require a complicated and context-sensitive set of rules recognizing that the celebrity Daniel Radcliffe and a poster using the name “Daniel Radcliffe” are unlikely to be the same person or have the same interests.

So: our users don’t use their government names; they do nonetheless value persistent identities, both as authors and as readers interacting with authors. (Side note: we make it easier to say nice things than nasty things by making it easy to leave kudos with a click of a button; you can leave any kind of comment you want, but you can’t do a thumbs-down with a click of a button!) They distinguish between RPF and reality. They are concerned to maintain a separation between legal names and fan identities, and we implement that in the ToS, including by allowing people to orphan works—detach any identifiers from them—and prohibiting the reassocation of the person with the work. Our users want ways to avoid harassment, which requires context specific decisions. Writing a story about someone is not generally going to be harassment.

The system is both open in a lot of ways and constrained in other ways—not like many of the other systems we study canonically in academia.

Betsy Rosenblatt: Fagundes characterizes derby as emergent/Ellicksonian: couldn’t that be in Second Life? That is, derby has a no-copying norm because the community decided to do it; the Second Life community could do the same, except that it’s bigger.

Heymann: there is a master roster controlled by derby volunteers. People do seem excited about the ability to vary display names weekly, not simply as denoting/differentiating but creative expression. Lots of people are concerned about copying, though.

Lemley: idea of name uniqueness is of recent and contingent vintage. TMs are almost never unique in history—lots of Americans, Federateds, Dells. People have contextual clues other than the name itself. Same is true with personal names. Internet gave more impetus to uniqueness, because domain names were unique—we started to take uniqueness as a necessary feature, but it’s really just an artifact of one particular system. If people in Second Life are different/distinguishable, they can have the same name. We may have to worry about impersonation/confusion, but we shouldn’t assume chaos will result.

Heymann: that’s why she’s interested in the ownership/authorship impulse.

Q: Levmore argues that new fields bring in divergent models, then they converge, at which point people try to steal each other’s money/commercial value. Archive of Our Own and Second Life are anomalous in not asserting platform’s ownership over names/characters/identities. Virtual worlds have built-in play-for-hire problem.

Lemley: it’s not clear to him that there has to be convergence. It’s great that copyright/TM policies all differ across worlds because they’re trying to do different things. As to internal creativity, things built within the game, there’s more free rein to set a policy. Marvel v. NCSoft (mentioned in question)—some numbers of users take advantage of their freedom to use things from outside the game. Not clear Marvel should be able to control player’s ability to make character look like a particular Marvel hero. But there’s a difference between copyright and TM—as long as it’s clear that this isn’t the person at issue, there’s a perfectly good argument that I can call myself Daniel Radcliffe. Copyright is thicker; attribution doesn’t protect you. Comic characters are troubling because they sit at the intersection.

Heymann: if EULA says the platform owns the name, how does that affect expectations?

RT: be aware that right of publicity laws requiring written consent may well trump the EULA.

Q: what about services like Livejournal that allow people to pay for name changes? How does that affect thinking about names?

Heymann: maybe it reflects or encourages thinking in terms of monetary value/ownership, or maybe it reflects that there are administrative costs associated, like paying for a new driver’s license. (I note that driver’s license is also identity/verification of identity; different cultures treat it differently and this is contingent.)

Lemley: his sense isn’t that he’s paying for value, but that he’s demonstrating that flexibility to change name has value. Interesting interaction between that and overlapping names. WoW: I can’t pay to change my name to imitation of another. More interesting: where you can pay to change to be the same as another’s name.

Q: Second Life formerly allowed you to secure your name for $50/year; they put a premium not just on the process of getting the name, but the maintenance of identity.

Lemley: on the internet we have very odd instincts about this. It doesn’t seem bothersome that if I want marklemley.com I have to pay for it to be registered. But if an arbitrageur shows up to register marklemley.com to sell it to me, that person is a cybersquatter subject to the full force of the law. Once we move away from the government as name allocator, have to decide who’s an authorized/unauthorized name provider.

Heymann: early internet names—“use” was use as identifier even if they weren’t using the domain to sell stuff—change in what the name represents.

Eric Goldman: keyword ads—TM owners are irritated that they have to bid on their names for searchers. But many TM owners do it and find high conversion rates, good investment v. other keywords they might use.

Lemley: they’re trying to buy exclusivity—they don’t have to buy their names but they don’t want others to be the top result.

Goldman: some would argue organic and paid aren’t direct substitutes.

Lemley: sure. But you can imagine Second Life saying: you can choose any display name, but for $50/year we’ll make it exclusive.

Q: questions of boundaries. Our language for thinking about boundaries is impoverished. Magic circle is a dead horse. Not all virtual worlds are games. Distinction between magic circle/outside is not the same as virtual/physical. You can play a game in Second Life or hang out in WoW. Bad effect: idea that seems to occur that now money is involved—assumption that moving back and forth across the boundary makes the boundary go away. No, it can make the boundary stronger. How to think about overlay/simultaneity instead of spheres? How do we think about jurisdiction in such circumstances?

Lemley: he loves the ability to blur what we normally think of as boundaries. Our ability to hop back and forth between different personas. Can it strengthen the boundary? Yes, but can constitutes it as different thing v. closed boundary. Real risk: each state claims to control the whole thing. Identification of boundaries/conditions in a clearer sense, whether in game or rigorous enforcement of norms might help; it can’t be a EULA. WoW has a PvP section and a non-PvP section, and you find out where those boundaries are. Skeptical of whether people outside the community will respect those boundaries, though.

RT: Viviana Zelizer is all about how a society goes about maintaining separation while also constantly crossing the boundary, e.g., between public/private, economics and love. Some relationships allow you to merge sex/affection and financial benefit (marriage) while others will get you thrown in jail. We have many tools to do this, but the question is their justice/suitability. (In terms of how crossing boundaries can reinforce them: girl on the internet v. girl in real life: blurring the lines of identity really can reinscribe the boundaries between identities—faking being a girl wouldn’t make any sense if there weren’t a difference between girl and boy.)