Our ongoing series on law and gaming continues this week with some thoughts about guilds and government.
As you may have gathered, I'm very interested in questions of democracy and governance in online games. And as many of you have pointed out, so far I've talked only about governance and politics at the level of the game's designers. I've been acting as though the designers' decisions were the only one that mattered, and all that players do is lobby the designers for the changes they want.
But that picture is painfully incomplete, in the same way that a description of democracy and government in the United States would be painfully incomplete if it mentioned only the President, Congress, and the Supreme Court. There are a zillion other institutions between the Big Three and ordinary citizens. States, counties, cities, and towns have governments too; administrative agencies, from the FAA to the FCC, do as much to affect the lives of Americans, as does Congress.
More fundamentally, democracy involves more institutions than those written into law as "government." Political parties, political action committees, and news agencies are all woven into our political process. So too are companies: Wal-Mart's decision about where to open up a store can do more to affect a town's future than any decisions made by its official government. And then there are the zillions of other civic groups that Americans take part in: churches, PTAs, bowling leagues, professional associations, and neighborhood watches, to name just a few. (Robert Putnam's famous book, Bowling Alone, details the contributions these groups make to public life, and bemoans their decline.)
So I'd like to start filling in our picture of public life in online games by talking about their closest equivalent to this universe of intermediate institutions: guilds.
The Power of Players
Now, last week,
I dwelt on the difficulties game designers face when trying to get players to behave themselves. For various reasons, mostly stemming from players' ability to flip the switch and log off, games themselves have limited ability to punish anti-social players. Most of what designers can do is give the players rewards for playing nicely and help them build up rich relationships with other players. In doing so, it seems, designers may be giving other players tools to keep evildoers in line. The argument I'll be developing over the next couple of weeks is that such enforcement by other players works.
My favorite example of the incredible (and incredibly positive) power player institutions can develop is EverQuest's raid calendars. When I first heard about them, they blew my mind. Many quite lucrative areas in EverQuest are safe only for large groups of players working in close cooperation. Without that cooperation, monsters will wind up attacking the "wrong" members of a raiding party, leading quite rapidly to mass death. For obvious reasons, then, awful things tend to happen if two separate (or worse, two mutually-hostile) groups are raiding in the same place.
Economists would recognize this situation as a classic "coordination problem." So long as people can make a rational collective decision about who goes raiding where and when and stick to that decision, everyone does extremely well. If the coordination breaks down, everyone's unhappy. Complicating matters, the actual allocation of raiding slots has the potential to be a highly contentious decision: everyone would like to get the plum slots and stick the others with the lousy off-hours low-profit-margin high-risk ones.
But, in large part, the rotation of raids has become a startlingly orderly process. Last I checked, on most EverQuest servers, access to the prime raiding zones is rationed according to raid calendars. These calendars are maintained on web sites outside of the game which let parties sign up and "claim" a particular zone for particular days. As long as everyone follows the calendar's sign-up rules and then goes raiding only when the calendar says it's okay, the coordination problem is automagically solved.
And that's pretty remarkable, when you think about it. These calendars are not self-enforcing. EverQuest and its servers know nothing of them. Nor, for that matter, can players themselves physically prevent groups not on the calendar from raiding a particular zone or coming and deliberately trashing a raid. People play along because they've learned it's in their interest to play along. A group which disregards the calendar will earn the instant enmity of other groups on its server and may be unable to do any raiding without interference until it has sufficiently atoned in the eyes of others. This threat works: players, by and large, listen to the calendar.
And -- I assert -- it's all thanks to guilds. EverQuest's guilds -- its more permanent form of player alliance -- are the units with which the calendar works. When you sign up for a slot, you sign up a guild, saying that you and your friends in the guild with you are going to use that slot.
It's the guilds who are punished if they disobey the calendar's demands; it's other guilds who mete out that punishment. It's through their guilds that players express their choices about where and when to go raiding; it's guilds that hold players to their decisions. In short, the guilds are the focal point of the raid calendar system; it would almost certainly not exist in their absence.
(From here on out, I'll be using "guilds" as a generic term for a wide variety of institutions. Ultima Online has "guilds;" Star Wars Galaxies has "groups," "factions," and "associations;" The Sims Online has "neighborhoods" and "clubs." These institutions, in general, have some level of "official" recognition: the game engine blesses them and gives them some moderate powers. At the same time, the engine is typically very neutral about them: it will bless any group of players who ask to be a guild and meet certain prerequisites (e.g., not belonging to too many other guilds already, picking someone as guild leader)).
So, what I'd like to do is to think about how to think about guilds. It's a large topic ("That's not a topic, that's a career!"), so I'm splitting it across this week and next. This week will be how not to think about guilds; next week will cover what we can learn from them.
Guilds Are Not Sovereign Governments
"Government" is perhaps the most basic category of political science; at a formal level, guilds seem to fit that category fairly well. They have members (citizens!) over whom they exercise power (sovereignty!); they have owners (leaders!) and officers (bureaucrats!); they make decisions in consultation with their members (democracy!). Not surprisingly, some people, writing about games, have spoken of guilds as "sovereigns"
The obvious reply to this is that games already have sovereigns, called "designers." Guilds are just as subject to the rules of a game as are players. The designers could easily destroy all guilds with a tweak to the code; there is no decision a guild could make that could not be immediately be overridden by the designers.
A couple hundred years ago, in a little case called McCulloch v. Maryland, Chief Justice John Marshall put this line of reasoning in his usual pungent language: "[T]here is a plain repugnance in conferring on one government a power to control the constitutional measures of another . . . ." In a game, every measure of a guild "government" is subject to the total control of the "government" of the designers. Your guild "government" is repugnant to the nature of government itself. Q.E.D.
This reply, by itself, is weak -- sovereignty, like jurisdiction, is not some absolute binary quality that admits of no compromise –- but it does highlight the important observation that the purportedly "sovereign" powers of guilds come from the game’s code; these powers are held at the sufferance of its designers. Even more damagingly to the guilds-as-sovereign hypothesis, these "delegated" powers are also incredibly weak.
What can guilds do in code? They can control their membership (sometimes), they can communicate on a guild-specific chat channel, they can put items and money into a shared guild treasury, and they can set and unset a few honorific status bits. From the perspective of the game’s code, guilds are no more able to carry out sentences of imprisonment, confiscation, or corporal punishment than any random player would be; nor can they hand out rewards over and above those freely offered by their other members. Any "sovereign" in real life claiming such paltry powers would be invaded -- or laughed -- out of existence in short order.
Guilds Are Not the General Will
Of course, at least since the demise of the divine right of kings, real-world governments have claimed that their power is not inherent in their existence, but instead is granted them (more or less willingly) by their subjects. In keeping with "modern" political theory, then, one might look to guilds as the embodiment of Hobbes's Leviathan --
For by art is created that great LEVIATHAN called a COMMONWEALTH, or STATE (in Latin, CIVITAS), which is but an artificial man, though of greater stature and strength than the natural, for whose protection and defence it was intended; and in which the sovereignty is an artificial soul, as giving life and motion to the whole body; . . . Lastly, the pacts and covenants, by which the parts of this body politic were at first made, set together, and united, resemble that fiat, or the Let us make man, pronounced by God in the Creation.
-- or Rousseau's General Will --
"Each of us puts his person and all his power in common under the supreme direction of the general will, and, in our corporate capacity, we receive each member as an indivisible part of the whole."
At once, in place of the individual personality of each contracting party, this act of association creates a moral and collective body, composed of as many members as the assembly contains votes, and receiving from this act its unity, its common identity, its life and its will.
On such a view, the members of a guild lend it their personal power in exchange for its protections and some role in shaping its decisions. This view is obviously correct, at least in a narrowly functional sense, but many who subscribe to it seem to take it just a few steps too far.
That is, having observed that guilds derive power from the consent of their "governed," people equate these powers with those enjoyed by real-world governments. But players in games don't have very much direct individual control over each other; nothing special happens to magnify these powers when they are pooled. Guilds enjoy a fair amount of effective control over their members, for reasons I'll discuss next week, but if they carry out the General Will, it is a weak one indeed.
The best way to see this fact is to focus on the "backbone" of code required to make guilds function at all. No game that does not implement guilds in code has ever seen them evolve spontaneously. The flat and basic features of most games are too weak to enable successful social organizations on such a large scale. A chat channel, membership controls, and a guild treasury may not seem like much, but empirically, they make the difference between a functioning guild and a dysfunctional one.
The flip side of this observation is that these features fail to provide many things that one would expect an organic will of the people to demand. Do you want your guild to have elections for its leaders? Well, you can certainly hold informal elections, but in EverQuest, unless the old leader is willing to step down, these elections amount to nothing.
Another classic example turns up in games that implement conflict using their in-code political systems; these conflicts are typically designed to exist in near-balance.
When one faction comes close to "defeating" another, the expected consequences -- surrender, pillaging, the sowing of fields with salt -- turn out not be available, because the game’s code has no provision for one side actually winning. The game is modeling military power, rather than actually letting players wield it.
A Tale in the Desert lets players propose and vote on binding referenda on game "laws" which will be written into its code; the last few months have seen the game’s designers "clarifying" this feature with one statement after another limiting the topics on which referenda may speak.
Guilds Are Not Just Tools of Game Designers
Thus far, I've noted that guilds are given very few powers by game designers, but that they are highly dependent on those powers. One might reasonably extrapolate to the conclusion that guilds are therefore instruments of game designers; when designers tell guilds to jump, guilds will ask "how high?" One would be wrong, not least because, in practice, designers don’t ask guilds to do anything. Having created guilds, designers occasionally tweak their mechanics or give them new features, but otherwise tend mostly to leave them alone.
The features we've mentioned as characteristic of guilds –- membership control, shared treasuries, and chat channels –- are remarkably neutral. One could use them to do all sorts of things, and guilds do, but the features themselves hardly speak to any particular uses. EverQuest's designers probably never even imagined raid calendars. Nor do administrators use their other powers of persuasion ("Make Argaven sleep with the fishes and we'll dump 10,000 platinum pieces in your treasury.") to push guilds into any particular course of action. Administrators seem to treat guild decisions in the same way they treat individual player decisions, as being largely beyond their control. They will frob a game’s features with the goal of nudging overall behavior towards greater happiness for all, but are, if anything, more deferential to guild choices than to player choices.
By way of contrast, designers do give player guides particular functions to carry out –- for example, to get players unstuck from the scenery and to warn players who curse at others -- and it is instructive to note that designers use every technique available to them to confine guides' discretion and shape their decisions. Guides are given powers tailored to specific circumstances, are punished for abusing those powers, and are in regular contact with designers for handing difficult cases and receiving further instructions.
None of these statements are true of guilds.
The idea that guilds are a guided "part" of the bureaucracy of the game administration attributes too much intent to the administration. In particular, it assumes that the designers much care about the details of their game.
They do not, in the sense that matters for political purposes. They care deeply about their games in the sense that they want players to enjoy playing them (and to spend money pursuing that enjoyment); they often care in the sense that they want to be able to enjoy playing the games themselves. But the designers are not of the game world; their relation to it is very different from that of players who live within the confines of its code.
To summarize, guilds depend both on code-based features given them by game designers and on the cooperation of players. Their decisions are taken largely in accordance with the wishes of their members, but within the structural context -- and the limits -- provided them by the designers.
And yet (or so I assert while deferring proof) they are able to fill important institutional voids in multiplayer games; their intermediate nature allows them to solve problems that may be intractable for players or for designers acting alone. Why might that be?
Next week: guilds as corporations, social clubs, close-knit communities, and representatives . . .