Virtual Worlds and Moral Choices

Raph Koster had a link on his blog about a week ago to a post about a family of monsters slain by adventurers in World of Warcraft (http://duocenti.blogspot.com/2008/07/murlocs-family.html). It’s very interesting, and led me to read an insightful, though much older post by Raph, “The Evil We Pretend to Do,” and ultimately the Terra Nova post, “The Horde is Evil” (http://terranova.blogs.com/terra_nova/2005/12/the_horde_is_ev.html), which claims that the choice to play an orc or undead in world of warcraft is a morally significant decision. This series of posts together have led me to consider the following question: what does it mean to make a moral choice within a virtual world? Rather than considering the specifics of World of Warcraft, I’ll try to discuss this in more broadly applicable terms.

There are a variety of ways we might answer this question, the first of which is presented by “The Horde is Evil.” We might think that there is moral significance to a choice of character type in an MMO. By choosing to be represented in the world by a character associated with evil, you are presenting an implicit approval of that evil. This doesn’t seem right to me. Choosing a character type in an MMO, especially if by that we mean a character’s race, is a matter of choosing a setting. You’re not choosing any actions your character has performed, but choosing the cultural and historical background of that character. If that counts as a morally weighted (and negative) choice then you could just as easily say that choosing to play World of Warcraft is a morally negative choice to begin with, because you’re choosing to play a game in a world that includes not only evil, but rampant warfare. You could instead have chosen to play a game, or even an MMO, that includes no war, no combat, and no evil. Insisting that a choice of setting in a virtual world is morally weighted is tantamount to declaring that MMORPGs are, on the whole, morally evil, which beyond being simplistic isn’t an argument that’s going to have any traction among gamers.

So if I disagree with the thesis of “The Horde is Evil” does that mean I don’t believe that decisions in a virtual world can have moral weight? No it does not, but we might mean something else entirely by that. Many people will agree that a decision that affects the happiness of others carries implicit (or explicit) moral weight, and online virtual worlds allow users to affect each others’ game experience. If I stick around a spot where a rare treasure appears on a timed basis and continually gather it, I might be preventing other players from gathering it; if I use a macro to collect massive quantities of a valuable resource, I devalue that resource that other plays have earned legitimately. Beyond these unintentionally side-effects of my actions, there is the phenomenon of “grieving.” Grievers are bored or insecure players (or both) who intentionally go out of their way to ruin the game experience for other players. This could entail verbal harassment, repeatedly killing another user’s character, joining a party and then refusing to help when they run into trouble, or a variety of other activities intended to provide amusement by frustrating others. As grieving is an activity that directly affects other real, living people it’s clearly a moral decision, but is it a very weighty one? In the majority of cases the worst consequence of grieving is temporary annoyance and frustration. What’s more, this behavior isn’t really specific to virtual worlds. Even if it takes a particular form within a virtual world, the behavior is essentially harassment, and the virtual world in this case is simply a medium through which the harassment occurs. The virtual world is playing the same role that a phone does in a prank phone call: the content of a prank phone call does tell us much about the moral nature of telephones.

There might be a third thing that we mean by saying that a decision in a virtual world can be a moral decision – we might think it’s moral or immoral not because of the impact it has on players, but because of the way it affects the virtual world itself. This strikes me as the interesting concept, because it’s not immediately clear what to make of it. On the one hand a virtual world is fictional, so it seems that the events that occur within it shouldn’t be any more weighty than they are in a book or film. However, the nature of a virtual world is that it is participatory: many people are invested in the virtual world (emotionally and financially), and they can shape it through their actions.

We may need a few more conceptual tools before we evaluate this idea. First, I will suggest that what is fundamentally capable of being moral or immoral is choice. When we say that a person is moral, we mean that they are a person who makes good moral choices. Part of the reason I reject the thesis of “The Horde is Evil” is because of this. I don’t think it’s coherent to suggest that a race (such as orcs) can be inherently evil – they can only be evil by virtue of their choices. The second idea that I am going to put forward as useful for our purposes here is that morality, at some root level, depends upon value. What I mean is that whether you are a deontologist, a consequentialist, or virtue theorist, you should think that what it means for a decision to be moral is that it promotes something of value. The exact nature of this question can be debated, but I don’t think it makes sense to talk about a decision having moral import without suggesting that something of value is affected by it.

Now let’s see how these concepts fit in with our thoughts about virtual worlds. If we’re going to judge something as moral, we’re judging a choice to be, so the actions taken by the computer can never be considered moral or immoral (though the decisions of the designers and programmers that resulted in the computer performing that action might). Rather, we’re concerned about player choices, as manifested in the actions of their avatars. What’s more, we’re concerned with how these choices affect things of value within the virtual world. Given that the world is, as its name implies, virtual, and exists primarily in the minds of the players (as mediated by the game), it seems to me that the only kind of value to be found within it is instilled in it by humans. This kind of value is easy enough to find, however: virtual worlds are largely about identifying with the interests of your character or avatar, and acquiring possessions, experience, and the like for them. A player might also identify with the interests of their guild, faction, or with the virtual world as a whole.

So what kind of effect can a player have on things of value? In truth, in most persistent worlds the answer is “very little.” In every persistent world I’ve ever seen, player characters never permanently die, NPCs cannot be killed, and monsters and NPC allies will eternally respawn. In most worlds it’s also not possible to take territory or possessions from other players, and all factions in the game will exist for as long as the world persists. These static virtual worlds actually disarm most of the moral weight of decisions within them, but if a world were designed to include more significant consequences of actions, then the idea of morally weight decisions by virtue of their influence on the world would become very real very quickly.

Let’s look at a concrete example. In the posts I referred to above, one topic of discussion is the fate of monsters called murlocs at the hands of adventurers who slaughter them in huge numbers in order to complete quests. There is something disturbing about how this is presented to the player – you are committing acts of genocide – but within persistent world, there are no lasting effects of a player completing the quest. The moral impact of killing murlocs is only the impact it has on the game experience of the player(s) killing them. Let’s consider an alternative possibility: the number of murlocs in the world is limited. Perhaps the murlocs reproduce at a certain rate, and if the rate at which players kill them exceed the rate at which they reproduce, their population will dwindle. Now we have the possibility not only that a player can have the visual experience of killing a murloc, but that their choice can have an effect on the world. If murlocs are slaughtered in large enough numbers, they might become extinct on that game server. What about a world in which player characters can die permanently? In these situations it seems like the possibility exists for in-game decisions to have a real impact on things of value to other players.

More generally, what I’m suggesting is that if players can shape the virtual world that they are in, they have the potential to promote things they value within that virtual world, whether that be immersive gameplay (in which case playing an villain could actually be a morally positive decision, by allowing for a more dynamic game world), the aesthetic features of the world, or murlocs. In this situation a destructive player can make a morally significant decision within the context of the virtual world, as can a constructive player. Given the nature of most worlds today, this is more an issue of the potential of virtual worlds, but it is still of immediate interest, at least to people like me.

After Raph Koster’s “The Other Side,” “The Evil We Pretend to Do,” Terra Nova’s “The Horde is Evil” and DuoCenti’s “The Murloc’s Family.”

2D and 3D Worlds

ROCKETON is a 2D virtual world that can layered onto any website. One may be tempted to question, then, how it intends to compete with the various 3D virtual worlds out there, such as IMVU or Google’s recently announced Lively. I’ll address this in two parts – first, dealing with the assumption that 3D graphics are better than 2D graphics by their nature, and second by talking about how a 2D format is specifically advantageous for what ROCKETON is trying to accomplish.

The advent of 3D game technology has opened up new possibilities for games. It has allowed for an unprecedented amount of graphical detail, allowed for shifts in perspective, and perhaps more importantly added a dimension in which objects can move in virtual space. There are, however, downsides to 3D graphics, including the problem hinted at in the post just below this one: the problem of the uncanny valley.

More generally, though, a key problem for 3D games is the inherent decrease in visual abstraction, which results in 3D graphics aging very quickly.

For example, Final Fantasy VII was lauded for it’s state-of-the-art graphics in 1997, as shown in the picture to the left. This was seen as a prime example of the amazing possibilities of 3D graphics when it was first released, but just a couple of years later looked dated.

This isn’t the fault of the artists working on the game, but is a natural consequence of the switch to polygon-based graphics. There is an added level of visual abstraction in the use of a 2D image to represent a 3D space, and it is a level of abstraction we’re very familiar and comfortable with – we make the same allowances whenever we draw pictures representing 3D objects in the world.

A 2D game also presents privileged perspectives of the objects in the world, which can make it easier to have a consistently appealing aesthetic presentation. A month after Final Fantasy VII was released, Konami released Castlevania: Symphony of the Night, a throwback to 2D sidescrolling that was able to build upon existing methods of making convincing 2D art and use layered backgrounds to create a game that is an aesthetic achievement even by today’s standards.

This isn’t to suggest that 2D games are inherently superior to 3D games – I’d be very sad not have played many of my favorite 3D games from the last 10 years – but that 2D and 3D games have comparative advantages, and in some cases 2D graphics will serve the vision of a project better than 3D graphics can.

ROCKETON is such a project that can make better use of 2D graphics than it could 3D graphics. The vision for the application is to allow users to bring their avatars to any site. Google’s Lively advertises that it can be put on any site – but what that means is that a point of access to the virtual world can be embedded on any site. There is no relationship between the virtual space and the site it is on. There is a French virtual environment called “Yoowalk” which tries to use websites as the backdrop for a 3D space, but the result strikes me as extremely awkward. The reason is pretty straightforward: websites are only 2-dimensional in nature. Most websites consist of text and pictures, both of which exist in two dimensions. In order to create a relationship between a virtual space and a website, the space has to accommodate the form of the website. What’s more, 3D environments take up more resources and take longer to load than a 2D environment does, which slows down the browsing experience and makes the environment less accessible. Finally, navigation in a 2D environment is more intuitive than it is in a 3D environment. A mouse only moves in two dimensions, and so can perfectly represent a range of possibilities in 2D but not in 3D space. 3D environment have shifting perspective, and objects in front of objects are common, which can make it hard to select the object you intended. A 2D environment offers a more fluid experience, in which a user can pick up and go where they want immediately, without having to wait for a long load time or figure out how to navigate a 3D space. Combine this with the fact that users know that they’re in-game property won’t be visually obsolete in two years time and you have all the components needed for a successful virtual world.

-Ellipsis

Beware of the Uncanny Valley!