The Yale Law Journal

VOLUME
118
2008-2015
Forum

Virtual World Feudalism

19 Jan 2009
James Grimmelmann
Land in Second Life

Second Life is divided into 256-meter by 256-meter “Regions,” which can be subdivided into smaller rectangular plots. Second Life’s developer, Linden Labs, regularly auctions plots to its residents. Linden also provides tools enabling users to market and securely transfer plots among themselves. Prices range from under $100 for a small plot to over $3,000 for a full Region. Linden assesses landowners a monthly “tier fee,” ranging from $5 for 512 square meters to $195 for a Region.

This description might make Linden sound like a modern lo-cal government that auctions public lands, maintains title records, and collects taxes, except for one crucial fact: if you do not pay your tier fee—or if you break any of many other rules of conduct—Linden will seize your land. What offline governments can do only after lengthy legal proceedings, Linden does unilaterally, just by changing an entry in a database. What’s more, Second Life’s Terms of Service give Linden the right to do so for any reason whatsoever.

Scholars, considering this imbalance of power unfair, have called for explicit property rights for virtual world users. They have a point, but full-blown property rights have their own problems. Stronger rights in virtual land would override express contracts between developers and users. These rights would also threaten developers’ ability to run their worlds effectively and could leave them powerless against abusive users who spoil the experience for others.

Linden Labs as Lord

We can resolve this tension by describing a user’s interest as seisin rather than as ownership. A tenant seised of land had sworn homage to the lord from whom he held. In exchange, the lord symbolically delivered the tenant into possession. Thereafter, the tenant owed the lord various services and feudal incidents, and in return the lord was obliged to defend his possession against outsiders to the relationship. Every element of this system maps cleanly onto Second Life. A user swears homage by clicking “I agree” to Linden’s terms and conditions; Linden delivers her into possession by changing an appropriate database entry. She owes tier fees in place of feudal incidents; Linden defends her possession via software-based access controls.

Crucially, seisin intertwined substance and procedure: the tenant’s remedy for disseisin was to appeal to his lord to set matters right. As S.F.C. Milsom writes, “[T]he lord is both the grantor who makes the grant and the law which protects it.” This unity applies with even more force to virtual property, which can have no existence apart from the developer’s actions to maintain it. The developer both “makes the grant” of virtual property and is in the best position to provide the “law which protects it.” The same database entry that makes one user an “owner” also triggers the automatic access controls that keep trespassers out.

Second Life even sports a parallel to feudalism’s hierarchical chains of subinfeudation. Several large commercial operations purchase entire Regions from Linden, landscape and subdivide them, and then rent or sell plots to users. Like feudal lords, these “land barons” play a major role in dispensing justice related to landownership. Many of them impose “covenants” on their land, such as a prohibition against running businesses from virtual homes. Tellingly, users upset at a neighbor’s violation of the covenant must look to the land baron for recourse; Linden Labs has no involvement in these local disputes. Similarly, Linden stays out of seignorial disputes between these (land)lords and their tenants. Whereas offline landlords are expected to rely on the state when evicting recalcitrant tenants rather than self-help, Second Life land barons have no recourse but self-help.

Implications for Law

This historical parallel between feudalism and Second Life provides useful guidance for offline courts asked to decide questions of virtual property. After the civil war of 1135-53 disrupted relationships of homage and seisin so badly that custom became unreliable, Henry II provided royal remedies for many tenants whose lords unjustly withheld seisin. Although eventually the king’s law and the king’s courts would control the ownership of land, these early interventions were modest; the king was not hearing land disputes in his own court in the first instance, but merely correcting “failure[s] of seignorial justice.”

Just as Henry II’s reforms led to some supervision of lords’ courts to protect tenants’ rights, so offline courts have a role in defending users’ interests in virtual worlds against developers. But that role does not extend to protecting absolute in rem “property” rights. Instead, a world’s own internal processes should ordinarily be the first and last stop for most questions of virtual possession. For now, offline courts should review only Linden’s adherence to fair procedures in important individual cases, rather than ruling on the legitimacy of its land practices in general or trying to hear virtual land disputes directly. (Of course, just as the king punished breaches of his peace criminally, offline courts can also appropriately hear cases when users do things with serious offline consequences, like laundering money, uploading malware or sending death threats.)

Thus, a workable system of virtual property would have some notably feudal features. Users would have possessory interests, secure against dispossession by other users, and enforced in the first instance by the developer. In return, users would have obligations of payment and loyalty to the developer’s rules of conduct. The developer could not dispossess users without reason, but offline inquests to test the developer’s reasons would be reserved for instances of clear-cut abuse. Similarly, to the extent that the developer allows these “feudal” relationships to develop among users, offline courts should ordinarily acquiesce.

The recent case of Bragg v. Linden provides a good example of a matter that should have been left to the developer’s discretion. Although plaintiff Marc Bragg’s allegations that Linden expropriated his land were explosive, Linden answered them with credible evidence that Bragg had taken unfair advantage of a bug in the land transaction system. That fact alone makes Linden’s suspension of his account sensible. The case settled, but had it reached a decision on the merits, the law should have treated Linden’s response as presumptively legitimate.

Conclusion

This analysis of the feudal dimensions of Second Life should make us optimistic about the legal future of vir-tual worlds. After all, for all its flaws, feudalism was a functional organization of society—indeed a better one than some of the alternatives. Virtual worlds built purely for play and where the stakes are not too large should remain mostly untouched by the offline legal system. Part of their value is as an imaginative alternative to “real” life.

As for the worlds with greater connection to offline values, we should recognize that they often follow the relational logic of feudal land holding rather than the in rem logic of modern land holding. By doing so, we can propose reforms that do less violence to the communities we’re trying to protect. In time—as indeed happened his-torically—the natural evolution and growing economic importance of virtual-world communities are likely to lead them to modern property rights. But until then, we should supplement in-world feudal justice, not supplant it.

Author Information: James Grimmelmann is Associate Professor of Law at New York Law School.

Preferred Citation: James Grimmelmann, Virtual World Feudalism, 118 Yale L.J. Pocket Part 126 (2009), http://yalelawjournal.org/forum/virtual-world-feudalism.