publisher colophon


A New Medium

In an Internet cafe in Beijing, Huatong, a young female player, showed us her World of Warcraft character. She played a druid—a class that “shape-shifts,” or takes on different forms, including several beautifully rendered animals. She found these shifting forms stimulating, and presented to us, one by one, the druid’s shapes and abilities. Without being asked, she mentioned that she thought about the game when she was not playing because it was so “interesting.”

In this chapter, I argue that video games like World of Warcraft constitute a new digital medium. The fusion of immersive visual experience with intense, skilled performative activity, represents a significant evolution in the history of digital culture. Video games afford rich stimulation to visual sensibilities while at the same time developing complex spaces of performance with opportunities for mastery and active participation.

Each of these elements—visual experience and performance—is compelling on its own. Together they are intoxicating. In Huatong’s descriptions, both the visual and performative qualities of the druid’s varying forms were deeply interesting to her, and she elaborated on their manifestations in each form.

Visual and performative experience in World of Warcraft were entwined, feeding back into one another. Accessing new visual experience and advancing in the game were mutually efficacious; attaining a level of performative mastery was necessary for “seeing new content,” as players said, while at the same time experiencing new content opened possibilities for performative challenge.

The term performative refers here to activity that demands conscious attentiveness and skill. It is time delimited, entailing a specific instance of performance, and engages an audience, either others or oneself. I do not use “performance” as a metaphor for the reproduction of culture through certain actions (see, e.g., Butler 1990) in which people are unaware that they are “performing” but rather as it is used in sports or theater, where participants engage in specific, recognizable instances of performative activity.

Contemporary culture offers many activities that afford either a strong visual experience or a venue of performance. Film and theatergoing, for example, involve immersive visual experience, but the audience has no part in the performance. Bowling is an enjoyable performative activity, but bowling alleys are visually austere and uniform, as are the sites of most sporting activities.

Interesting visual-performative activities occur in the real world, and they seem to spark the same enthusiasm as video games. Dance clubs that feature square dancing, salsa, or ballroom dancing combine visual and performative experience. Participants wear colorful, stylized costumes and perform complex dances. Members of church choirs, alongside parishioners, sing in churches decorated with seasonal flowers and religiously themed accoutrements. Paintball pits teams against one another in competitive play in woodsy environments. Hunting and fishing take place outdoors, often in locales of beautiful open country or on picturesque waterways. Participants in U.S. Civil War reenactments recreate the extravagant clothing of the era and enact elaborately staged mock battles and other historically inspired activities. Fantasy masquerade balls invite guests to appear in costume to dance and revel, as do events such as Burning Man and Mardi Gras.

World of Warcraft is very much like these activities. But it is a packaged, self-contained, computer-based medium created and reproduced on a massive scale enabled by digital technology powering immersive graphics and complex game mechanics. WoW is more accessible than most real world visual-performative activities because the entry point was a cheap commodity—a networked personal computer. The genius of WoW, and other video games, lies in allowing millions of people to attain skilled performance in artistically designed spaces entered through an ordinary computer.

One of my guildmates succinctly captured the notion of WoW as a visual-performative medium when he said, “WoW is baseball in elf costumes.” The term “elf costumes” evoked the visuals so vital to players. Baseball was a metaphor for the competitive contests central to play in World of Warcraft: quests, raids, duels, battlegrounds, and arenas.

When I first started giving talks about WoW, I emphasized community and collaboration. An audience member who played the game said, “Well, that’s interesting, but you’ve left out the most important thing—mastery of the game.” The comment indicated the need to attend to the development of skilled performance as critical to player experience. The emphasis in the interviews on leveling, constant player chat about attaining better equipment, and the seriousness of the collective organization undertaken to support raiding argued for mastery as crucial to game experience.

A notion of mastery complexifies an account of WoW as primarily a series of Skinnerian rewards. Yes, players wanted the loot and the experience points. But mere acquisition was not the sole source of satisfaction; the loot and points accumulated, over time, toward the player’s object of becoming a better player—of “improving yourself,” as Mark said. Focus on the object of activity proscribes reductive accounts such as Skinner’s as the end point of analysis. Attention to higher levels of activity moves analysis beyond the moment of reward to the larger historical trajectory of the activity, recalibrating the time horizon in which human activity is intelligible and for which analytical logics must be devised.

In the following sections I discuss the contours of performance in video games, the means by which visual/performative activity is encapsulated in digital rules, and the place of video games in mass culture.


Let us discuss WoW raiding as an example of the centrality of performance.1 (We might also investigate ordinary quests, 5-man instances, battlegrounds, and arena play, all of which entail performative challenge.)2 In a raid, as in a sporting contest, the outcome can turn on the smallest mistake or advantage. A spell cast a moment late can kill a player or wipe a raid. A player hanging on by a few points of health can still battle to victory. Raid action is in constant dynamic change. Players spoke of the importance of “situational awareness” (a military term), denoting the ability to mentally process rapid multidimensional environmental changes. The same ability is mandatory in team sports, where the positions of the ball (puck/Frisbee/ . . .) and the positions of other players, as well as their specific movements, must be tracked and responded to rapidly.

Golub (2007) provided a detailed account of the 25-man Magtheridon raid that nicely captured its performative demands:

Because the party will wipe if they attempt to fight both Magtheridon and the magicians, the magicians must be killed in under two minutes . . . Magtheridon will release a dangerous “blast nova” that will also kill the entire party. In order to prevent the blast, players must simultaneously click on the [magicians’] cubes to delay the blast for one minute. However, because a . . . player can only click a cube every two minutes . . . two teams of five players each must take turns . . . Finally when Magtheridon is at [a certain stage] he . . . will begin collapsing the walls of the room. Players must avoid the pieces of falling rock which will kill most of them instantly.

In this encounter, players are required to achieve precise timing in clicking the cubes. They must be vigilant of the magicians’ imminent demise and the point at which Magtheridon will issue the blast nova. Toward the end of the encounter, players perform the duties of their class while simultaneously moving their characters around Magtheridon’s Lair to avoid the collapsing walls. As Golub remarked, “The fights . . . can be quite intense.”

In my own raiding experiences I was acutely aware of the need to perform. During one encounter Scarlet Raven was just learning, the raid leader admonished the raid, “Innikka had demons three times and she survived and she’s a healer!” The comment was a rebuke to more powerful players. Though many had died, I had squeaked by even though healers have reduced power against mobs (demons here) which must be fought, in this encounter, by individual raid members.

Such scrutiny of player performance was deployed to direct and improve performance. Bardzell and Bardzell (2008) observed that a character in a virtual world is known to others through the performance of its actions. A character is “a subjectivity constituted by actions in-world.” (Their use of the word subjectivity is the same as the use of the word subject in activity theory.) A character is not merely an image or static representation but a performance. In Magtheridon’s Lair, either you can click the cube at the right moment or you cannot. There is no way to disguise an inability to perform this action precisely and accurately. Bardzell and Bardzell said, “A subjectivity . . . cannot lie; it is as it does.”

The centrality of performative mastery in video games was noted by Kennedy (2005) who reported that the female Quake players she studied used terms such as athleticism, balance, coordination, and taking risks to describe why they liked Quake, a first-person shooter game:

Although it is the avatar that performs these feats of athleticism or coordination within the game space, it is the players’ skill in controlling the interface that shapes this performance. The sense of agency the players experience is doubled; the player experiences a freedom of movement and sense of authority and mastery within the game, alongside a sense of empowerment through their skill in mastering the technology.

Kennedy pointed to the same issues of performance, mastery, agency, and empowerment I discovered among WoW players (see also Turkle 1984). Drawing a further connection between games and performance in sports, Kennedy (2005) reported a comment from a woman who ran a gaming site:

“Women are starting to realize that they have the same abilities in sports—and things like sport—as men,” said Leann Pomaville, a 38 year old former school teacher who runs the girl gaming sites Da Valkyries and Quake Women’s Forum. “Quake is a game where your own personal skill makes all the difference, like in a sport.”

In the same vein, Taylor (2003) observed of EverQuest players that “the actual fight is as much an opportunity to demonstrate the valued qualities of game mastery as anything.” WoW players were deeply attracted to “the competitive atmosphere,” as one player put it, noting that she was fascinated with “learning what strengths and weaknesses your character has against another’s” (Choontanom 2008). Wine (2008) emphasized the “drive to be a good player” in her analysis of discourse on a WoW priest forum.

As in contemporary sports, performance in WoW was expressed in a series of publicly reported metrics (see Taylor 2008). Performance was measured and displayed in tables and logs. The “combat log” was a history of game actions, reporting in minute detail what transpired during a particular encounter, including metrics such as who delivered a “killing blow” and how much damage or healing resulted from a player action. Players seeking to improve their performance, and raid leaders critiquing an encounter, studied these logs carefully.3

Players embraced performance measures. Going far beyond the tools provided by Blizzard, player-created software modifications computed and reported performance metrics. (“Mods” were available for free download on the Internet and placed in a folder provided by Blizzard.) One of the most popular categories of mods was “damage meters” (see Taylor 2008). Damage meters gathered game data, computed metrics, and output the results in a window. The results could be reported in chat. Meters measured not only damage, but healing and many other quantities related to game performance. The illustration shows a window from Recount, a player-created software modification.

Raiding guilds studied these oracles religiously. Players kept an eye on their own performance, striving to be near the top of the meters. After each raid, Scarlet Raven posted detailed statistical results generated by a program called WoW Web Stats (WWS). These metrics, archived on the guild website, became a permanent part of guild history.

Players’ obsession with gear derived from their interest in performance. WoW required skill, but if two players of equal skill competed the better geared player won. The impact of gear on performance was made plain in reports provided by meters; players could see the effects of the gear they were accumulating. Players often posted questions in guild chat of the form “Is X or Y better?” where X and Y were similar pieces of gear. Guildmates would discuss why a particular piece of gear might be better for the particular player asking the question.

Another series of player metrics ranked performance in terms of guild “progression” through more and more difficult raiding dungeons. Websites such as ranked guilds according to progression measured as dungeons completed and bosses downed. Players dedicated to raiding eagerly read these websites. Sean, an undergraduate player I interviewed and got to know as a student in several of my classes, was excited about the “leap in rankings” his guild experienced after completing the Serpentshrine Cavern and Tempest Keep dungeons. Progressing through dungeons “felt fantastic,” he said. WoW was “not a Harry Potter fantasy game,” as he put it, but a site of serious competitive play and performative challenge.

A 10-Man Raid on the Kilrogg server

A member of an internationally known guild, Nihilum, wrote on their website:

Well, storming through BT [an advanced dungeon] and basically putting the smack down like we did killing Illidan [a difficult boss] way before many others would see him dead was probably a better experience than AQ40 [another difficult dungeon] was. God we owned BT! When Illidan died it was an amazing feeling.

Like Sean, who said it felt fantastic when his guild progressed, the Nihilum guild member expressed a release of positive emotion—“God we owned BT! When Illidan died it was an amazing feeling.” “First kills” in World of Warcraft were famously moments of performative ecstasy. Players savored them for days, they posted accounts and pictures on guild websites (see also Golub 2009).

At the other end of the spectrum of performative experience, when a boss or dungeon was under control, players having mastered it, the encounter was said to be “on farm status.” (Farming will be taken up in the next chapter; it referred to game activities lacking challenge.) Players linguistically marked the quality of dungeon experience according to the level of challenge, accenting the importance of performance. When encounters could be handled with aplomb, players engaged them not for performative experience but for loot—with loot, of course, always feeding forward to the next opportunity for performance and improving a player’s numbers on the charts and meters.

A further indication of the salience of performance in WoW was player humor. Players made fun of each other’s abilities, joked about how opponents cheated when a duel was lost, or facetiously suggested raiding low-level dungeons. For Alliance players, a running gag was mock discussion of organizing a raid to kill “Hogger,” a level-11 “elite” mob (i.e., difficult-at-level mob). When a player in my guild would ask on a non-raid night “we doing anything?” someone might reply, “get ready for Hogger.” The player-created website contained comical accounts of fictional attempts to defeat Hogger. Players drove the satire home by invoking WoW’s performative constructs and lingo. Below is the first Hogger post and a few others. The first brief post inspired a cascade of ever more preposterous Hogger stories, some quite lengthy and one with original artwork:


by Zvert

Tried 2 solo him with my 70 hunter, got him 2 65% and i wiped. He hits like a truck.

Re: Hogger

by Jasmela

My guilds current progression [ratio of bosses downed in a dungeon]:


12/12 - Karazahn

6/6 - Zul’Aman

2/2 - Gruul’s Lair

1/1 - Magtheridon’s Lair

6/6 - SSC

4/4 - TK

5/5 - Hyjal

9/9 - BT

6/6 - Sunwell

0/1 - Hogger

Re: Hogger

by belegamarth


Even with your guide, we had 2 40-man raids (lvl 80s with tier 12 gear) there and only brought him down to 64%. When he got to 70% he summoned the boars and most of our healers went down. Then he uber pwned us with his paw. One of these days we are going to get that Huge Gnoll Paw . . . As for the water, that sweet succulent spring water, I’ll keep grinding till I get it.



The jokes played off WoW metrics—“[we] brought him down to 64%” and “when he got to 70% [of his health]”—as well as character and gear levels and progression ratios such as 0/1 Hogger.

The Software Artifact

This section turns to the implications of the visual-performative medium as a digital entity encoded in rules. Among games, video games uniquely digitize rules of play, encoding and containing them in a software artifact. As Fron et al. (2007b) observed, “Video games . . . dictate and enforce rules automatically through software.” Taylor (2004) noted, “. . . we cannot overlook the role software and design play in shaping online life.”

In WoW, play was nearly perfectly reduced to performance and mastery of digital rules. Mechanical encapsulation (Kallinikos 2006) established and enforced rules automatically through the agency of the “black box” of software closed to user inspection and alteration. The kind of interpretive flexibility that, for example, a referee or umpire brings to a sport was absent. Mechanical enforcement of rules removed a source of social authority with whom to negotiate and rethink rules of play as they were followed (or not followed) during actual performances of play.

Digital encapsulation curtails (but does not eliminate; see Consalvo 2007) the possibility to engage rules at their margins to gain advantage. In sporting contests, by contrast, players are constantly aware of rules they might usefully bend and manipulations players on the opposing team might be planning to leverage. The software artifact’s mechanics permit constant, reliable vigilance and enforcement that give little quarter to ambiguity or interpretation; they eliminate the vagaries of the different personalities and proclivities of human agents such as referees and the impact they might have on the interpretation of the rules.4 As inscriptions within a machine, digital rules are not established and reestablished in interactions between human agents meeting in a shared space of performance; rules are removed from such influences.

WoW players could not critique or question human judgments about particular instances of play during play, but they could air their concerns and grievances about rules of play on official forums maintained by Blizzard. A different mechanism, then, for moving the game in new directions was in place.5 Rather than actual performances instigating change, player discourse about performance was encouraged and analyzed. As each new patch was issued, it was obvious that players’ opinions had been considered and their concerns rewoven into the software artifact. The black box was thus breached, although indirectly, through conversations about the game that occurred outside the game.6

The dialectic between players and the corporate provider was an important means by which World of Warcraft was altered and refined. However, this set of interactions, and their useful consequences, should not cause us to be inattentive to the force of the software as a designed entity. Activity theory and actor-network theory posit that technology embodies a powerful agency not strictly under human control (Callon 1991; Latour 1994; Kaptelinin and Nardi 2006; see also Mumford 1934; Ellul 1964; Kallinikos 2004, 2006, 2009). Technology forcefully shapes human activity and invariably entails unintended consequences (Winner 1977). In formulating WoW as a performative medium, I want to draw attention to the importance of the design of the game itself, its potent agency as a particular kind of medium that engaged players in certain kinds of regulated performances.

A revealing instance of the powerful shaping of the game’s design occurred in the wake of the first software expansion of World of Warcraft. The expansion, for which players purchased and installed new software, extended the game with significant content. “The Burning Crusade” added new geographies, quests, and dungeons. It was released in North America and Europe in January 2007 and in China in September 2007. Players spoke of “TBC,” as it was known, and “pre-TBC,” indicating its importance as a critical juncture in the history of the game.

The expansion brought the addition of ten levels of play. Players could level their characters from level 60 to 70. The 20- and 40-man dungeons became obsolete. They still existed, but new, superior 10- and 25-man dungeons were introduced. The loot in these dungeons was so amazingly better than the old loot that players abandoned the pre-TBC dungeons (although after awhile some players started conducting nostalgia runs).

TBC dungeons were exciting venues for fresh performative challenges in new visual surroundings. They contained clever storylines, imaginative graphics, and crafty bosses.

All this new content sounds like a great thing. And, in many ways, it was. It enabled players to improve their characters through enhanced loot and to enjoy novel visual-performative events—exactly what I am arguing WoW provides as its core experience. However, amid the happiness was a decidedly negative outcome. For many guilds, TBC entailed a transmogrification of guild social dynamics. Guilds such as Scarlet Raven, in which players had been playing peaceably together for months or years, suddenly fractured into two groups: players who advanced quickly through the new levels and acquired desirable new loot, and players who progressed more slowly and could not keep up with the swift progress of their peers.

The single biggest source of the fracture was a new 10-man dungeon named Karazhan. Acquiring equipment from Karazhan was necessary in order to advance to 25-man content. For guilds going from, say, raiding Molten Core, an old 40-man, to a sudden focus on 10-man was jarring. Only 10 people got a chance to raid. And they all had to perform to a high standard. In good old MC, a few players not really doing their jobs would not wipe a raid. In Karazhan, every player had to know what they were doing and be reasonably geared. The composition of the raid involved a precise balance of classes. On raid nights often there was a single group going to “Kara,” leaving behind many players who had previously been raiding.

An apparently simple solution would have been to start another Kara raid with an additional 10 people. But often the right combination of player classes was not available. Players grew discouraged at not getting a raid invite and would sign up for raids less frequently, exacerbating the problem.

Within Scarlet Raven, considerable disturbance ensued. Incipient guild cliques became more visible. Players advancing quickly wanted to play with others doing the same. They expressed irritation at slower or less skilled or geared players. The advanced players did not consider it their obligation to help the slower players, who were described in Scarlet Raven website posts as “less dedicated.” The design change from 20- and 40-man raids with some latitude for error, and openings for nearly all who wanted to raid, to a 10-man raid requiring better performance and precise class composition, generated a situation in which the rich got richer and the poor got poorer. Skilled, geared players preferred raiding with skilled, geared players. They did so, getting better and better equipment. Others, squeezed out of the raids, progressed even more slowly.

Not surprisingly, when the “dedicated” were geared and ready for 25-man, they found that having left others behind, they were impeded in their efforts to form a large raid. One solution was to recruit from other guilds. Scarlet Raven recruited successfully, although this action created fresh tensions among existing members. The 25-man raids came onboard much more slowly than everyone thought they would.

A little over a year after the launch of TBC, frustration among the ambitious raiders reached a tipping point. Raid progress had been too slow. Some of the most serious, best-geared players defected to hardcore guilds where they would be assured a rigorous schedule of disciplined raiding. Scarlet Raven tried raiding with those available, but gear limitations made it impossible to defeat new bosses. Within three months of the departure of key raiders, the guild closed down 25-man raiding. Scarlet Raven’s rapid descent from the server’s most successful casual raiding guild (according to rankings) to a guild with limited 10-man raiding was wrenching.

Many of the Scarlet Raven players who left had been with the guild a long time and were popular, visible members. They had hung in, waiting for others to catch up. But by refusing to institute a systematic program of helping the less geared, their departure was inevitable—if we understand WoW as a visual-performative medium. The social glue was insufficient to sustain the community building necessary to properly equip enough players to perform in the difficult dungeons. Nor was the simple fact of community adequate to retain geared players. It is a cliché of multiplayer games that players “come for the game and stay for the people.” But relations between game and people must be understood in more complex terms. WoW was engineered to require sociability for the most challenging content, but it only weakly developed community as shared commitments and durable social bonds.

It was players’ desire to grab the TBC opportunities to enhance performance and experience new visual content that led more advanced players to push forward with their own progress, altering social dynamics. Hypothetically, it would have been possible for Scarlet Raven to devise a social solution for a challenge imposed by a technological design decision. The guild could have slowly accommodated the less geared or skilled members. But such a solution would have abrogated the logic of the game, which was to perform to one’s fullest ability. WoW was a voluntary space of performative play, not a place in which players wanted to spend hours helping the less fortunate. Efforts to mitigate the disruptions entailed by TBC were difficult to institute because they violated the spirit of free play and the very reason to play the game, which was to perform and see new content—to engage the visual-performative medium. Once the new content was locked up in the 10-man dungeon with its limitations and constraints, and once this dungeon became a bottleneck to other content, guild dynamics altered dramatically.

This powerful design dynamic created what activity theory calls a contradiction. A contradiction is a systemic inconsistency or discrepancy (Engeström 1987; Groleau et al. 2009). Raiding guilds were formed for the purpose of raiding. At the same time, guilds were supposed to be friendly, supportive groups that a player could count on. These two purposes stood in relations of tension with the changes brought by The Burning Crusade; it was difficult for guilds to accommodate both.

The contradiction between raid progression and group support led some well-geared players to depart. On the other end of the gear spectrum, players who identified themselves as “nonelite” also became frustrated and /gquit. For example, one player wrote on the guild website in her good-bye message:

I came into the guild about 7 months ago and was quite aware of the elites versus the non, and I persevered to make my character good enough. I built some great friends and wish them all nothing but good fortune however in all my time here my class captain did not address me even once and even ignored me one day I sent him 3 whispers asking for a good time to help me. I read all the forums and did research on my own however sometimes you just need some advice from those who have gone through it. I made myself available for raiding at least 4 times per week and often was called in to fill a vacant spot. I believe I was respectful and tried to help all my guildies when asked.

She ended somewhat bitterly:

I am sorry this was not the type of mature player you needed and again wish you the best.

Thus, one change in dungeon design—a change that had some positive effects—also had a significant negative impact on many guilds.

Sean described how he left his guild because he desired “an environment that I want to play in” which he defined as

players that are same skill level as me or better, people who share the same goals of progression, a guild with my friends that I select and who respect me.

He said that in TBC, “I could not do my job”; in other words, he could not perform adequately because less skilled players were unable to raid effectively. He invoked the social: “a guild with my friends that I select and who respect me.” However, the nuances of “social” in this discourse were linked to the performative; Sean wanted to be surrounded by other performers of similar skill level who shared mutual respect for the “jobs” they did.

A Scarlet Raven player wrote sadly on the guild website:

i remember the old guild, the ones who would always be up for Scholomance or ubrs even strat [dungeons] to help people get there gear. this guild was built on helping . . . but tbc has changed all that.

I have spoken with European players who told the same story of post-TBC disruption. (At the time I was in China they were still waiting for TBC.) Scarlet Raven was irreversibly altered after The Burning Crusade. The relaxed atmosphere in which guildmates engaged each other in playful player-designed activities and a lot of wacky humor was gone. Pre-TBC, for example, Scarlet Raven had “gnome races” in which guild members created level-1 gnome characters and undertook various challenges (of a very silly, very delightful nature). The guild used to conduct competitive timed runs of one of the old dungeons, Upper Black Rock Spire. (I have fond memories of being on the first winning team in which we did UBRS in 53 minutes.) Such activities, and the guildwide camaraderie that generated them, fell away from the guild.

I left Scarlet Raven in June 2008 for a new raiding guild. My research in Scarlet Raven was extensive enough for the analysis I wanted to do, and I wished to continue to learn from raiding both as a researcher and a player. I kept in touch with friends from Scarlet Raven, and found the new guild through a former Scarlet Raven player. I often checked to see who from Scarlet Raven was online (using a simple WoW command). Usually only a few people were logged in; the guild was on life support awaiting the next expansion. Eventually the guild was dissolved, and players split into two small guilds.

Why did Blizzard institute such an abrupt change? Karazhan was created with the best of intentions—to open raiding to small guilds that could not muster enough players for larger raids. This was, in some respects, a good thing. The Derelict would never have been able to experience 25-man content unless it partnered with another guild, a difficult exercise in coordination undertaken by few guilds. However, the Derelict enjoyed Karazhan. The problem was in positioning a difficult 10-man dungeon as an obstacle for 25-man and in instituting such a change in a context in which many guilds had been raiding with large groups which could no longer easily be accommodated.

The design of a software artifact, then, may powerfully shape human activity. The pervasiveness of the disruption of guild dynamics following the TBC changes, across guilds, countries, and cultures, suggests the enormous directive influence inherent in rules encoded in digital technology. In the next section, I consider the disquiet often aroused by the power of digital rules and ways in which they have been analyzed in the literature. I argue for seeing rules as a potential resource rather than a hindrance to positive human activity.

Rules Rule

The story of TBC indicates that we pay special attention to the influences of artifacts on human activity. While it is clear that player experience refigured World of Warcraft in significant, meaningful ways (chapter 7 will address this topic), on balance I argue that the software artifact was uniquely powerful in creating the world experienced by players.

In examining such software artifacts, we turn to their encoded rules which embody directive force, giving designers the tools with which to transmit designed experience. While all artifacts mediate our relation to reality in important ways (Vygotsky 1986), the expressive flexibility of software rules entails not merely mediation—as a hammer mediates our relation to a nail—but the remarkable ability to conjure worlds of participatory aesthetic activity which can continually develop and change, generating historical depth and context. Karazhan was an important moment in World of Warcraft that occurred at a point in time. Through new content embodied in rules, World of Warcraft has moved on and will continue to do so as long as it is a viable commercial product. WoW is thus in some ways like life itself. But it is very much unlike the ineffable processes of ordinary life in that software artifacts can be deliberately constructed and implemented according to strict, articulated design goals authorized by designers. Rules are the vessel in which designers’ intentions and creativity are expressed in a machine-readable format and delivered to a large public.

In the context of this discussion, “rules” are taken as structures in a software program. In digital games, software rules also embody game rules. We might use the term algorithm to refer to the software structure, but for our purposes the emphasis on control connoted by rules is needed (rather than sequencing of instructions).

In his analysis of enterprise resource planning systems,7 Kallinikos (2004) noted:

The study of technology and its social impact cannot be exhausted at the very interface upon which humans encounter technology. Essential strips of reality are not observable or even describable at the level of contextual encounters.

Kallinikos urged that we examine the qualities of software itself, in particular its capacity to direct activity through encoded rules. In studying only “contextual encounters,” we miss the stuff of the artifacts themselves—the powers flowing from their form and function. In essence, Kallinikos says there is something to learn about hammering not from “contextual encounters” involving watching people hammer but from addressing the hammer itself. How much does it weigh? What is its shape? Of which materials is it constructed and why? How does the hammer fit into a human hand? (An analysis certainly more easily done when the hammering is not taking place.)

This line of inquiry would appear to contradict the emphasis in Dewey and activity theory on studying the actual activity of human subjects. However, what Kallinikos noted was that we seem to find it difficult to attend to both the artifacts themselves—with their capacities for mediating activity—and the lived experience of engaging them. Kallinikos observed that in studies of software artifacts examination of the artifact itself is often sidestepped in favor of asserting the importance of user resourcefulness in dealing with the artifact. Studies center on managing software in practice. Analysis often addresses the ways in which people work around the perceived ill effects of the execution of software rules in human activity or how rules are manipulated. Orlikowski (1992), for example, conceived a notion of interpretive flexibility through which workers managed problems created by computer-aided software engineering tools in their organization. Other research celebrates resistance to the power of rules—for example, that of hackers (see, e.g., Consalvo 2007; Jordan 2007)—or ways in which human actors such as game masters (employees who provide in-game customer support) and guild leaders smooth the bumps that invariably issue from the inflexibilities of rule-driven systems (Ryan 208).

Kallinikos argued that such analyses of “contextual encounters” construct a user control that is largely illusory, downplaying or even denying the astonishing capacity of software to call worlds of human activity into being and to sustain, within narrow limits, specific actions in those worlds. While user actions such as devising workarounds or hacking are important, Kallinikos (2004) suggested that the user contribution in a rule-generated world (such as a video game or enterprise resource planning system) “must be attributed its right proportion.” The world created by the rules is huge; user response is relatively small, reactionary in nature, and sited within the framing established by the rules. It is the very rules that give meaning to the activities of hackers trying to break them, to employees finding local solutions to the inflexibilities of large software systems, and to game masters and guild leaders patching over the rough spots. Rules direct and motivate the activity of these actors, not the other way around. As Kallinikos (2009) observed:

Emerging use of artifacts and workarounds are efforts to reinvent or resist the expectations tied to a technology, and in this regard are inescapably framed by the functions and forms embodied in the artifacts.

To a considerable degree, we are products of our technologies. Kallinikos provides a strong version of this claim:

The agent that acts upon, interprets, or reshapes technology has, to a significant degree, been made an agent by, among other things, technology itself.

At a workshop on “productive play” sponsored by the National Science Foundation and hosted by Jason Ellis, Celia Pearce, and me in May 2008, virtual worlds pioneers Chip Morningstar and Randy Farmer touched on the debate, remarking on what they called the “tyranny of emergence.” They observed that we all hope that online communities will emerge in bottom-up fashion from participatory activity, and we approve as participants take whatever technical affordances they are offered and appropriate them to their own ends. But at the same time these veterans cautioned that we cannot forget the “hand of God” that is the software artifact and the power of designers to shape activity (see Nardi et al. 2009). As Kallinikos said, it is the “right proportions” of the relative contributions of users and software we seek to apprehend.

Pickering’s (1995) notion of the “mangle” is a distinctive formulation of the relative contributions of user practice and technology. Unlike Farmer, Morningstar, and Kallinikos, who urged attention to the primacy of software rules and their directive force, Pickering proposed that a mangle—a mélange of user practices, socioeconomic conditions, and technologies—produces experience. This vivid metaphor suggests the entanglement of diverse elements; taking the metaphor at full face value, the conceptual mangle “mutilates” the varied elements it passes over, pressing them into one another.

Engaging Pickering’s notion, Steinkuehler (2006) discussed ways in which player practice modulated undesirable effects of the rules in the video game Lineage. She described how high-level players traveled to areas with new, low-level players and repeatedly killed them. As upsetting as this was for the low-level players, such activity was permitted by the rules. In response, Lineage clans (i.e., guilds) organized themselves to periodically sweep through low-level areas to establish order by killing offending players.

Clan action was a resourceful response to a perceived problem of the rules. Following Pickering, Steinkuehler argued that a mangle of player practice and game rules “interactively stabilized” the game world, relieving new players of the predations of advanced players. She observed:

This “interactively stabilized” system (Pickering, 1995) emerged from the mangle of both designed-in technologically instantiated rules and human intentions.

However, it would have been trivial—a matter of a few lines of code—for the game’s designers to change a rule to make it impossible for a high-level player to kill a low-level player in designated areas (as in World of Warcraft). Lineage’s rules were in place to promote a certain style of play. The story of the rescuing clans presumes that we cheer them on as they scour the countryside meting out justice to errant players. But we must also acknowledge the fun high-level players were having pwning nubs—actions Lineage’s designers chose to preserve through the rules. A world in which ganking is possible was precisely the world intended by designers, and it was that world they encoded into the rules.

While “interactive stabilization” is one possible outcome of the function of any system, the mangle suggests that equilibrium is the expected outcome of dynamics within a set of interacting variables or factors. The events surrounding Karazhan did not result in stabilization but destabilization. Notions such as mangles appear to be a small subset of general systems theory, which elaborates varying system dynamics (beyond stabilization) (see Ashby 1956; von Bertalanffy 1968).

The mangle, it seems to me, underplays not only relational dynamics but also temporality. In the account of the Lineage clans, the temporal flow of events is clear. Clans acted in response to realities entrained by game rules. The necessary accommodations, reactionary in nature, followed events engendered by the rules. Rules established conditions that made killing low-level players possible, which in turn led to clan posses. I see less a mangled hash of interacting elements in which rules lose their directive force, surviving simply as one element in the mix, and more a temporal progression from rules to reaction.8

Paradoxically, the clans’ (re)actions served to reinscribe the very rules whose outcomes distressed them. Lineage’s designers had no need to exercise their power to change rules because the clans instituted a workaround. Clan actions, rather than asserting dominion over the rules, redoubled the rules’ power.

Metaphors such as mangles are useful in calling attention to a heterogeneous mix of practices, economic conditions, and technologies that shape experience. But in the context of digital artifacts, such metaphors may camouflage the outsize contribution of software shapings, imputing, indirectly, a sort of equality between elements. Notions such as stabilization mask asymmetrical accommodation to the power of digital rules. Steinkuehler (2006) described the lengths to which she went as clan leader to ensure that her guild participated in policing low-level areas, including an elaborate rewards system. Rather than a natural merge of practice and rules, practice was forced to stretch and extend itself to accommodate the rules—which remained untouched.

In assessing user response to rules in enterprise resource planning systems, Kallinikos (2004) observed:

Interpretation and local reshaping of the package can take place only within those narrow limits allowed for by the constitution of technology . . . To navigate effectively in the engraved routes of large-scale information platforms entertains the hope (or illusion) of gradually acquiring control over them. (emphasis added)

Attention to the “constitution of technology” is necessary for understanding how we encounter digital technologies. Kallinikos’s argument that we attend to the precise ways in which digital rules shape experience suggests that theories that itemize elements of experience without assessing their relative contributions and particular means of influence (such as the narrow range of possible customizations in enterprise resource planning systems), cannot account for the capacity of digital technology to direct activity (see also Taylor 2006a; Fron et al. 2007b). A kind of figure-ground reversal seems to occur in which actions such as the heroism of the Lineage clans, which was a relatively small adjustment to the world shaped by the rules, become emblematic, defocusing the larger shapings entailed by the execution of rules. Kallinikos’s argument indicates a need to rise analytically to the level of the world of human activity created by the rules of software systems, examining, in the large, the quality of experience produced by the rules in dynamic interplay with human action.

Rules May Nurture

It is just that quality of rule-based worlds to which I turn now. Notions such as mangles and interpretive flexibility suggest that, with some effort, the ill effects of digital rules can be managed. A different response to the power of rules is to denounce them as “totalizing”—figuring rules as entities that brook no interpretative activity and restrict the moves a subject can make (see, e.g., Jordan 2007). Based (somewhat loosely) on the work of theorists such as Foucault and Lyotard (1984), the notion of totalizing rules in computer games draws attention to the foreclosure of “metagaming” in which players alter the rules to suit themselves, as we all (putatively) did with Monopoly and other board games (Jordan 2007).9 From this perspective, the fact that we cannot alter rules within digital games is troubling. Jordan pointed to the problem, proposing hacking as a solution:

Academic programs in game studies are in the unique position to offer institutional support for hobbyist emulation and ROM-hacking communities, and it is our obligation to utilize the resources of thousands of programmers and gamers in our common goal of reverse-engineering and documenting, recontextualizing, and reworking these products beyond mere objects for consumption.

It is through hacking, Jordan argued, that we may “transform, adapt, and distribute the systems of rules themselves.”

Whether viewing rules as manageable because they can be flexibly gotten around, or declaiming rules locked down and in need of being broken into, both arguments cast digital rules in a bad light. Analyzing World of Warcraft invites us to examine rules from a different angle, considering them as resources preserving good design. In WoW, the artistry of the graphics, the excitement and imaginativeness of the quests and raids, the careful progression through increasingly challenging levels of play, the interdependencies of character classes were undeniably grounded in the creativity and ideations of World of Warcraft’s designers. Their labors were caught up in the inscriptions encoded in the software and made available to players in replicable form through execution of the rules.

Anyone who has played World of Warcraft knows how vibrant and present player interaction and negotiation are in the experience of play. I will give an example of such interaction and negotiation, indicative, I hope, of the texture of everyday play. At the same time, I argue that such texturing is largely (though not fully) engineered by the rules, as it was in Lineage.

As noted, when players are grouped and a mob drops a valuable item, players must decide who will receive it. The decision can be highly charged, as a desirable new item invariably and inevitably enhances the player’s abilities (themselves a product of the rules of the game encoded in the software).

In group play, it was sometimes unclear which player was most deserving of an item. In a raid, perhaps a member of an appropriate class rolled for an item but another player of the same class was equipped with inferior gear. It would benefit the guild as a whole were she to get the item. The raid leader might step in and allocate the item to the player who lost the roll but appeared to need it more. Even when guilds used systems in which players bid for items based on points accumulated through raid participation, raid leaders might dictate or suggest that a player who had bid fewer points receive the gear. The point systems, intended to remove ambiguity, were, in practice, deployed flexibly in response to judgments about guild and player needs rendered in the thick of play. Guild leaders sometimes negotiated these matters among themselves in voice and chat channels reserved for officers. It was not simply game rules that decided the crucial matter of who attained prized gear; complex, occasionally contentious assessments, shaped the process. These assessments, and the negotiations that generated them, loomed large in player consciousness, producing an ambience of constant human interplay.10

Player practice shaped loot distribution in another way; under some circumstances, in particular in pickup groups, but also 5-man guild runs, players had to decide whether to cheat. One of the game’s mechanisms for allocating loot was a roll of virtual dice by typing the /roll command into the chat window. The odd thing was, players could cheat by using an addon that mimicked the output of the roll:

Zeke rolls 75 (1 -- 100)

The player who won picked the item up from the mob.

Cheating could not be detected by other players unless someone recorded a series of rolls and came forward to say that so many rolls in the nineties were statistically impossible. Some players did not even roll; they simply walked over to the mob and picked up an item, a practice known as ninja-ing. Player practice materially figured into loot distribution; it was trivial for a player to defeat the virtual dice or ignore them altogether.

Players could also choose a system of loot distribution in which they selected “need” or “greed” for an item. If the item was appropriate to their class, need would be the correct choice. If more than one player chose need, the system rolled the dice and the player who won would find the item automatically placed in his or her bag. But sometimes players chose need even though another player clearly deserved the item more—another type of ninja-ing. When a ninja struck, a good deal of ill feeling ensued.

One of the most important aspects of play, then—loot distribution—was subject to player intervention; crucial decisions about whether to cheat, and assessments regarding who should receive loot, were shaped by player practice and interaction.11

Without denying the centrality of such practices and interactions, we must regard them as materially constituted through the engineering of the rules. World of Warcraft might easily have been designed differently so that a high roll would automatically place gear in bags, or so that classes could not roll on gear they could not use. The design of a game dictates where opportunities for human intervention shall be offered; a hand of God is embroidered into the software. It was not simply player inventiveness that gave rise to the negotiation and sociality that molded game experience; latitude for such player activity depended on specific rules of play. Just as fractious Lineage players were permitted to rampage, so were WoW players afforded means by which to cheat, as well as means by which to share wealth if they so chose.12

Hunter and Lastowka (2004) observed of an early virtual world, LambdaMoo, that its founder, Pavel Curtis,

. . . and the other four [programmers] who worked with him could quite literally reshape LambdaMoo’s heaven and earth, which was nothing more than a database of textual representations and coded rules that governed the objects within that represented space.

Hunter and Lastowka recounted that when Curtis and the programming team came to a point at which they wanted to abdicate responsibility for running LambdaMoo (which became a burdensome responsibility), they realized they could not because the software, which they controlled, was so definitive of participant experience.

The design of a software artifact dominates experience while not completely determining it. This asymmetry of player and software shapings is desirable when good design is encapsulated in digital rules with their capacity to reliably reproduce experience. In the case of video games, invariant execution of rules constitutes a resource for preserving and propagating vision and artistic imagination. World of Warcraft, whose software encoded elegant artwork, clever game mechanics, and support for specific forms of social activity, gave rise to play experience that found appeal to millions of people from diverse national, cultural, and socioeconomic backgrounds. I suggest that we examine outcomes of rules as situated in particular artifacts rather than as a monolithic category, conceiving rules of well-designed software artifacts as neither inflexibly totalizing nor calling out for user remedy but as nurturing, protective, caring.

In WoW, the enforcement of game rules conserved a rich, complex play experience. The design was hardly perfect, and the experiences surrounding Karazhan indicate the troubles that may issue from design decisions, but, taken in the large, World of Warcraft was an out of the park, bases-loaded home run, becoming a de facto standard for multiplayer games according to most any metric we might consult, including sustained player base, formal awards, and (presumably) profitability.

Let us now consider a case in which participants in a virtual world are responsible for its creation and content is not pregiven, locked down in rules, by designers working for the corporate entity. A point of comparison with the highly designed universe of World of Warcraft comes ready to hand. We will inspect a virtual world devoid of its own content, dedicated instead to participants’ creations: Second Life.

In Second Life, participants construct content using tools provided by Linden Lab, Second Life’s purveyor (Malaby 2006, 2009; Boellstorff 2008). Malaby conducted extensive field research at Linden Lab, reporting that the design of Second Life was guided by a philosophy of emancipation endorsing whatever creativity might spring from the hands and minds of Second Life “residents,” as they are called. From the beginning, Linden Lab’s ideology centered on the importance of furnishing participants tools to build whatever they wanted (Malaby 2006, 2009).

Launched in 2003, Second Life has had considerable time (in an Internet sense at least) to establish its trajectory. The practical results of the radical freedom envisioned by Second Life’s designers, have been, I would say, mixed. While Second Life contains a rich assortment of diverse activities (Boellstorff 2008), on the whole, participants have gravitated toward creating content devoted primarily to two activities: shopping and sex.

In a world in which people can do whatever they want, the reproduction of consumption as a primary activity is, in my view, a somewhat disappointing turn (though consistent with the larger culture). The prevalence of sexually themed activity is a draw for some but uncongenial to others. Hansen and his colleagues (2009) analyzed the written responses of senior business executives who spent time in Second Life as part of an assignment for an MBA class:

Anonymity is considered a likely contributing factor to the often alarming pornographic activity within Second Life. In the words of a program management executive, “In my 40 years of life I don’t think I have ever run into so many sexually-motivated characters.” Stories of sexually-oriented activity and rudeness abounded in the data. Executives reported a multitude of situations where they were harassed, stalked, and on one occasion, sexually violated. One executive indicated that he cannot see a way to control such activity, which greatly limited any significant conceivable role for the medium in organizations.

Another executive observed:

In addition, even if you filter out mature content, the risqué clothing, skin, and parts business is unbelievable, so you are never safe from a flash of some body part while walking through a Herman Miller studio or Coldwell Banker’s HQ.

One executive highlighted the impact this could have on business:

If we recommended our customers to use this site, and they subsequently were propositioned or mistreated in any way, I believe that our reputation would suffer irreparable damage.

In practice, the freedoms cherished by Second Life’s designers resulted in the dominance of certain activities, pushing away other activities into which potential participants, disaffected by the dominant activities, might have entered.13 This is not to say that sex and shopping are bad, or should be dismissed, but that they constitute a narrow range of experience. Sexually themed activity has a powerful dynamic; its presence in Second Life draws sharp lines between those who will enter a world where it is pervasive and those who will not.

A partial solution to the problem of off-putting content is to pay for private spaces within Second Life that are inaccessible to the general public. Many businesses and organizations maintain them. They serve an important purpose and are integral to the success of Second Life. But at the same time we may view such spaces as a retreat to walled gardens, defeating the ideal of an agora of diverse activities in which we might browse and learn and be enchanted. If participants enter only into restricted areas in Second Life and must stay off the streets, so to speak, the impact of the creativity Linden Lab’s founders envisioned is diminished in its capacity to generate abundant, varied experience.

Malaby (2006) noted another sequela of residents’ activities. Visually, the virtual world evolved into something of a junk heap. Malaby reported that Linden Lab employees complained that Second Life was “ugly, trashy, and junk-filled.” Malaby remarked that

the Second Life built landscape is remarkably variant, with giant advertising signs next to enormous modernist skyscrapers next to medieval castles next to five story dinosaurs next to perfect recreations of art deco gas stations. With no gravitational limits on building, the skyline is correspondingly unconstrained, and the word that springs to mind . . . is, indeed, “trash.”

Without the gifts of artists to compose the landscape and rules of governance to channel participants’ creations, the ensuing hodgepodge dismayed those who worked at Linden Lab, and even sympathetic observers such as Malaby commented on the infelicitousness of the visual prospect in Second Life.

That some Second Life residents rejected its “trashy” aspect was evident in the practice of residents moving to private spaces in which elements of order and unity were exactly those reintroduced. Malaby reported that one of the first things owners of private spaces did was establish rules. Private spaces asserted “sovereignty . . . to build . . . in a particular style . . . with restrictions . . . and covenant[s]” (2006).

It is perhaps ironic to locate finer sensibilities in a video game centered on killing cartoon monsters than in a virtual world devoted to the creativity of the citizenry. But there is no substitute for good design; a well-designed digital artifact such as World of Warcraft— with its rules and constraints and limits—can be constitutive of a “work of art . . . not remote from common life,” as Dewey said. The design of Second Life—its bid for tools-without-rules—and the pursuant disorder and disunity, made its public spaces less works of art for common life and more environments that provoked the “lamentations” of its own designers that it was not more beautiful (Malaby 2006).

Second Life is, of course, a digital object with its own software rules; rules per se guarantee nothing. My argument is that digital rules provide a special kind of resource with which good design can be preserved and protected through encapsulation in the black box. In this sense, rules may nurture by providing a safe haven for cultural objects of integrity and excellence. I see the design encapsulated in the rules of World of Warcraft as a work of art—one that gives rise to participatory aesthetic experience, to the remaking of experience and community, as Dewey formulated.

Thus, I view with caution proposals to instigate more player control and “community participation.” Taylor (2006b) suggested that participatory design techniques could be utilized in multiplayer video games for the purpose of “according [players] some power and responsibility to govern their own community and world.” While there is appeal in this idea, I believe it is quite difficult to determine where the design of a game such as WoW leaves off and “governance” begins. Obviously players shape game experience, as Taylor (2006b) and others have pointed out and as I have discussed. But I wonder if allowing players latitude in matters such as character design is a workable idea. Taylor (2006b) described a protest that took place within World of Warcraft instigated by players displeased with changes in their characters. She noted that multiplayer games have few spaces for such protests and that corporate entities such as Blizzard do not encourage them.

I am skeptical that those gathering to protest were representative of the millions of people playing World of Warcraft or that there is any way to develop representative political processes within a game such as WoW. I am skeptical that the protesters were better equipped than Blizzard designers to alter character design. I have discussed ways in which Blizzard consulted players. They had their own methods for doing so; it is hardly the case that players were silenced by the corporation.

Methods of participatory design are intended to balance power among stakeholders. Participatory design is deliberately democratic. I cannot speak too highly of democracy as a political system, but artistic production is, to me, another matter; it is inherently singular, anomalous, moving on the edges of culture. It has no interest in balancing competing claims through fairness or compromise, although it is, of course, not immune to influences outside itself.

When I attended the “UI and Mods” session at BlizzCon, I was cheered to hear one of the panelists declare that World of Warcraft was intended to be an immersive experience of a lively visual world in which the mobs are moving, the scenery interesting, buildings and landscape features designed to be woven into strategies for fights. In other words, the gestalt of the designed world was the basis of play experience. The designer said:

Don’t play the user interface, play the game . . . [We want you] to look at the world not the bars.

Another panelist said that World of Warcraft should be “organic and fun.”

In contrast, many player-created software modifications redirected the game, moving it toward a focus on numbers. The interface to these numbers became a central focus of player attention. As with much quantification, numbers rarely tell the whole story, but they become the salient point of contact between actor and world.

I confess I loved Recount. But at the same time, it frustrated me that certain aspects of my healing activity were not represented in its calculations. Recount computed only direct heals, ignoring the mitigation of damage provided by the “shields” a priest can cast or the fact that priests have an ability to increase the armor of the player being healed, also preventing damage. Many guilds chastised players whose WWS numbers were not as high as a class or raid leader thought they should be (see Taylor 2008; Wine 2008). At one point, a (misguided) priest captain in Scarlet Raven declared that I was casting too many “Prayer of Mending” spells, which he had determined by studying WWS reports. This efficient spell did not reveal all the healing I was actually doing because, in the modification that measured heals, healing activity from Prayer of Mending was attributed to the players who received the healing (due to a technical complexity in the design of the modification). The class leader wanted me to move up in the rankings by reducing my use of the spell, even though the spell was beneficial for the raid.

I have seen more and more use of player-created software modifications in World of Warcraft. Mods often push the game away from the gestalt identified by Blizzard as core to WoW play experience toward a world of meters, alarms, shortcuts, and annoying announcements cluttering the chat window. I was forced to adopt the mods Clique and Grid in order to speed my healing to maintain my position on the meters. The use of Clique did nothing but allow me to mouseover to heal more quickly. WoW encounters were designed to allow players enough time to click on a target and then select a spell. (Blizzard did not design a game impossible to play.) My use of Clique was a defensive maneuver in response to the power of player-created meters. Likewise, I was pressured into using Grid, which I grudgingly admit had some very useful features not found in the standard WoW UI, but which also lacked some features (such as the ability to move individual raid group windows).

Golub (2009) argued that mods help players in difficult raid encounters. This is certainly true, but many of the mods players used simply were not needed once an encounter was learned. For example, Deadly Boss Mods (DBM) warned players of impending events, allowing them to prepare and take action. I used Deadly Boss Mods until it stopped working (I tried everything to restore it, and eventually did). But for a time I got used to not having it. Golub observed that players used mods like DBM for mobs such as Akil’zon the Eagle Lord who casts a devastating electrical storm that can only be avoided if players run to him. But if a player’s situational awareness is switched on, there is a warning from the game and time to run.

Golub reported that players in his guild used an audio alert mod when battling the dragon Sartharion to anticipate when lava walls rise up and must be negotiated by running to a gap between them. I died several times in the lava walls but became an ace at them in pretty short order. Sartharion was one of my favorite encounters in part because it was so visually stimulating. Lacking mods, I was forced to maintain visual attentiveness.14

So, while mods were helpful without question and I see them as critical to player experience, as I will discuss in chapter 7, some mods served to hasten progress in the game rather than to make it “organic and fun.” I would not trade the fun of having gained command of the lava walls for quicker progress in the game, and, indeed, some of my guildmates who used mods got the hang of the lava walls after I did. It is not always the case, then, that “the community” is invariably right; there are competing perspectives on what makes for good play. Although she is in favor of strong player participation, Taylor (2008) observed that

[T]hrough [modifications’] rationalization and quantification of action, they . . . strongly inform (and potentially limit) what is seen as “good play” or what is viewed as reasonable.

In analyzing discussion on a priest forum, Wine (2008) remarked:

The mistakes healers make are some of the most public, and posts in the forums try to head this off by letting priests know the “right way” will be expected of them in a raid.

The limitations entrained by modifications and narrow concepts of the “right way” do not open games to a wider variety of experiences and personal preferences but move to rationalized systems of control. With respect to player-centered governance, we may need to be careful what we wish for.15

I am doubtful of the feasibility of participatory design methods in large virtual worlds. At the time of this writing, World of Warcraft had 11 million players. These players lived in countries as different as Denmark and Dubai. They spoke different languages and came from diverse cultures. How are we to engage them in exercises of participatory design? While I am sympathetic to the aims of participatory design, its techniques were developed in the context of a specific local culture (Scandinavian trade unionism), and the techniques it devised are accountable to particularities of that culture (see Kensing and Blomberg 1998). There is little evidence that participatory design techniques scale to large multinational, multicultural venues.16

The question, then, is how best to sustain and develop communities that spring up in and around video games and virtual worlds. My discussion of design and governance is intended to problematize participant input, to call into question a propensity to see all participant input as an a priori good.

Indeed, the very notion of community must itself be reexamined. It is perhaps not so simple as community = players, when surely community is a complex assembly including, in the context of World of Warcraft, Blizzard and the corporate entities behind websites such as, from which players derive critical information, as well as advertisers who support these sites.

Developing nuanced understandings of the varying roles of participants and corporations in online worlds is a daunting task. Devising ways to attain dialogue between interested parties remains a challenge, one we will struggle with for some time to come. Productive interchange must reckon with the diversity and complexity of authentic participant experience, the visions of artists who design games, and corporate realities.

I mention a final testament, in the context of World of Warcraft, to the power of a well-designed software artifact to preserve and propagate vision and inspired imagination. The potency of WoW’s artistry gave rise to the player practice of conducting “nostalgia runs” in old, pre-TBC dungeons. Having gained a foothold in TBC, some players journeyed back to old content they had not seen. Dungeons such as the Temple of Ahn’Qiraj and Blackwing Lair, extremely difficult at level, offered imaginative settings and contests players wished to experience. Players whose characters were not geared enough when TBC arrived, or who had not had time for advanced content, went back to see what they had missed. These players could not obtain a single piece of equipment better than that of the new TBC dungeons; acquiring gear was moot. Doubling back to the old dungeons was purely for the pleasure of novel visual-performative experience.

Since players were equipped with powerful new gear, the old dungeons provided visual experience for the most part, although players were sometimes surprised at their performative difficulty. They still had to figure out what the mobs were up to and how to defeat them.17 Some players upped the ante by returning with small groups, refiguring performative challenge by attempting, say, the 40-man Blackwing Lair with a smaller group.

Players’ sustained drive to “see the content,” as they put it, whether in high-end dungeons or quotidian quests, owed much to the care and nurturance afforded by the black box, sheltering, as it did, code in which was registered the desires and visions of talented designers. Empowered through digital technology to call forth complex worlds of human activity, creators of the virtual world preserved their artistry, ensuring its continuance through inscription in digitally encoded rules.18

Performance and Participation

I have proposed that video games constitute a unique visual-performative medium, affording—on a massive scale, by way of cheap, commodity technology—visual-performative experience available in the real world in more limited ways. The accessibility of desirable participatory experience has enabled a significant evolution in digital culture with global impact.

Dewey’s notion of participatory aesthetic activity, on which I have relied, is one of many concepts of participation in the literature on video games. Raessens (2005) observed:

Many authors refer to concepts such as . . . “participation” to characterize the distinctiveness of computer games and the media culture that has developed around them . . . [But] these terms are used in various and sometimes contradictory ways, a situation that leads to confusion.

Huisman and Marckmann (2005), for example, suggested that participation must involve “an open dialogue between user and designer.” Unless players are extending a game through their “own imagination,” play activity is merely consumption. Huisman and Marckmann observed, “[I]t turns out that [participation] is interesting only when the number of users is limited . . . The smaller the number of users, the greater the influence each one can have.” This formulation excludes World of Warcraft and other popular multiplayer games and seems limiting.

Without exhaustively analyzing notions of participation (see Raessens 2005), let us examine Raessens’s thoughtful construction of participation in video gaming. I will suggest that it has some limitations, while recognizing that it identifies certain critical aspects of participation.

Raessens’s notion of participation is theorized at the action level of the activity hierarchy. He formulated video game participation as comprised of three elements: interpretation, reconfiguration, and construction. Interpretation is figuring out how game rules work. Reconfiguration builds the player’s game world by selecting objects and actions from a fixed set offered by the game. Construction adds new game elements such as player-created software modifications.

These elements of participation were present in World of Warcraft, and they faithfully capture a good deal of the texture of everyday WoW activity. Structural decomposition of participation into its actions, as Raessens has done, is useful and necessary. What is missing for me in Raessens’s depiction is the passion that animated participatory activity in World of Warcraft—the object of activity that imbued actions with intensity and interest.

I began my investigation wondering about the undergraduates’ excitement over multiplayer video games. The actions of interpretation, reconfiguration, and construction, while pleasurable in themselves, were deeply absorbing, not merely momentarily diverting, in building toward something bigger, specifically, the development of performative mastery. Dewey noted that the flow of aesthetic activity moves from “from something to something.” Actions such as interpretation and reconfiguration satisfy in themselves, but at the same time they advance toward an object—a horizon of fulfillment in a larger trajectory (see Kuutti 1998).

In analyzing participation, it is useful to attend to both short and long time frames, as activity theory and Dewey’s conceptualization of aesthetic experience indicate. Ethnographic methods play well with this strategy, mindful as they are to nuances of small moments that embody larger themes, while sticking, over time, with the subject (in both senses of the word) to trace paths visible only as they emerge in their particular temporality. For analyses grounded in notions of texts (such as Raessens’s), temporal flows of human activity may be less salient as structural characteristics of artifacts predominate the analytical field of view.

Game design that provides pleasurable means building toward the possibility of a kind of greatness in a player’s personal history produces a compelling form of participation. Sean’s discussion of his choice of guildmates, frustrations with certain players, and elation at progression were indicative of a larger object of continually improving his performance in World of Warcraft. For him, knowing rules, configuring objects, and reconstruction were organized and motivated by the object of becoming a better player. By themselves these actions cannot explain Sean’s attachment to World of Warcraft or the meanings of the game for him. The object of performative mastery imbued his actions with significance and interest.

The Spectacle of Images

I want to end this chapter by looping back to Huatong’s fascination with the druid character type and her contemplation of its varied renderings. Performative activity in World of Warcraft took place in a rich visual matrix. Analogizing video games and sports takes us only so far, lacking, as it does, acknowledgment of the brilliant visual spectacles that constitute contemporary video games.19 Metrics and competition suggesting sportslike activity tell half the story, but instead of the literal uniformity of sporting uniforms and the plainness and predictability of, say, basketball courts or soccer fields, video games conjure striking visual worlds remarkable in their vivid realizations of unique imagined universes.

In the interviews, players often discussed and evaluated WoW’s visuals. They talked about the game artwork, mentioning colors, character images, animations, buildings, and game geography as important aspects of play experience.

An American player was touched by the animation of a leaf falling from a tree.

Sheryl: I love the scenery. It’s just beautiful, the animals, everything. You know the little leaves that they’ll have falling. Or, like a feather on the ground. It’s just like, out of all the things they had to do in this game, they remembered to do a little leaf.

Chinese players commented:

Chen: It seems the game is done with a lot of devotion and heart by its producers and every [visual] detail is elaborate.

Bao: I like the characters, the pictures, the production. And the characters have a vivid performance.

A player in Shanghai said:

Liu: My guild members and I fight the monsters and then we rest and look at the area together. We explore and wander through different areas . . . Looking at the scenery is recreational.

Players conversed about how gear looked:

Herold: Not all AQ40 [a dungeon] sets are ugly, personally I like the leather ones for rogues n drood [druids].

They expressed delight when a new piece of gear (or an enchant) was sparkly, showy, unusual (in a cool way), evocative of the class, or otherwise well designed.

Dimminix: Priest T6 [a category of gear] is HAWT20

Players registered disappointment in gear that copied old designs, was clumsy looking, or ugly. It was especially regretful when a piece of gear had “good stats,” that is, powerful attributes, but poor design:

Rigg: man, I hate it when ugly sets have good stats.

The notion of seeing new content alluded to both visual and performative experience, but new content had tremendous visual impact when first encountered, bringing forth complex visual worlds no one had ever seen before. Juul (2005) observed:

Most video games . . . project a fictional world: The player controls a character; the game takes place in a city, in a jungle, or anywhere else.

Juul theorized that fictional worlds are, in part, imagined by players who “fill in any gaps.” While agreeing that video games bring forth imagined worlds, my data suggest that these worlds are less a fiction in which players fill in gaps and more a powerful visual experience like viewing a striking landscape—the world is fully realized, and one need only gaze at it.

This emphasis on the visual is not to say that textual representations were not important in World of Warcraft. They were very important—in chat, quest descriptions, character statistics, and other game elements. We might call video games such as WoW visual-performative-textual media, if it weren’t so inelegant. But I don’t think we need do that. “Baseball in elf costumes” crystallizes what brings video games into the realm of the deeply compelling for a vast audience; the marriage of performance and stimulating visual experience impels players to spend long, dedicated hours engaged in activity in game worlds.

We have a handy comparison with which to argue for the primacy of the visual. Text-based role-playing games (such as Arctic or Avalon), while similar in certain ways to their image-rich cousins, never attained the wide appeal of video games. Text-based games utilized similar storylines, offered opportunities for performative excellence, and engaged players with challenges. But despite a core audience of enthusiasts, text-based games have always been, and remain, a niche. The performative element, sans visuals, seems insufficient to enable these games to break into the big time, even though they go back to the 1970s (Jerz 2007). Players today could choose such games, which are cheaper to produce and require less bandwidth, but they overwhelmingly select games with graphics.

The notion that video games are themselves texts, i.e., a form of narrative (see Murray 1997), is another sense in which we may consider the question of textual representation. Games such as A Tale in the Desert engage true narrative; game play changes based on a clever mix of player activity and designer storytelling (see Fujimoto 2005). However, the blockbuster games, including multiplayer role-playing games and first-person shooters, are only weakly narrativized as far as play goes. One can play without knowing a shred of what players call “lore.” Contemporary video games utilize narrative, but they are not essentially stories in the way that “narratologists” have argued (see Aarseth 1997).

World of Warcraft was indeed based on a rich textual backstory. In the forests of Tirisfal Glades, the Scarlet Monastery, once devoted to the Light, was conquered by fanatical zealots. The rocky environs of Shadowmoon Valley sheltered the powerful warlock Gul’dan, who transformed the Black Temple into headquarters of the Shadow Council, an organization bent on destruction. Silithus, an insect-infested desert, was home to the fallen Aquiri Empire, whose capital, Ahn’Qiraj, housed a temple sealed off by Night Elves in an effort to prevent complete infestation.

This is beautiful stuff, but I had to look it up on WoWWiki. The lore does not come through seamlessly in play experience as the surface of the visual world does. Some players loved WoW lore and would have known what I looked up. But they were a minority; guild chat almost never mentioned lore while players constantly talked about how their equipment looked and discussed their desires to see new content.

WoW’s visual world borrowed elements of theater; characters accumulated wardrobes of colorful, fanciful costumes and props and participated in encounters in stagy settings. A cast of fantasy beasts, such as the druid forms Huatong loved, were on display, as well as demons, ogres, trolls, giants, dwarves, and dragons. The costumes, props, and characters afforded direct visual experience that did not require the mediation of narrative or the need to “fill gaps” in an incomplete fictionalized world. Visual elements were intact and complete in themselves. As Andy Warhol said of the importance of surfaces:

If you want to know all about Andy Warhol, just look at the surface of my paintings and films and me, and there I am.

The impact of WoW’s visual surface was evident in consistent player comments about animations, landscape, gear, and so on—players were sensitive to, and liked the way things looked.

But the visual surface of the game had other important work to do. Players interacted with visual elements with more than gaze; much of what they saw was intended to be interpreted for purposes of play. The visual environment contained quest items, friendly and unfriendly NPCs and players, flowers to be picked, ores to be mined, bodies of water to dive into, forges at which to smith weapons and armor. In battle, “line of sight” was critical; players carefully arranged their characters around walls, buildings, and hills so that their spells could reach those they were healing or fighting or to prevent an onslaught of mobs which would be riled if approached “in the open” rather than from behind the safety of a door or wall. Players knew well what the fantasy world looked like from a performative standpoint: where a flag carrier might be hiding in the Warsong Gulch battleground, how to jump from a ledge to a balcony in Blackrock Spire to get to the entrance of one of its dungeons, ways to use the pillars in the Ring of Trials arena to competitive advantage.

Players not only interpreted visual elements; their actions altered the visual world—unlike other visual media such as television or film. One of the most satisfying aspects of play was the impact of action on the world—watching an enemy’s (usually rather comically enacted) death, seeing a raid member appear beside one through the magic of a warlock’s summoning ability, calling forth a “pet” such as a baby wolf or tiny, psychedelically colored bat.

WoW’s design, then, was a kind of theatre in which audience and performers were one.21 There was plenty to look at,22 but at the same time, players themselves were onstage. In boss fights, players performed in spaces very much like stages; in some, bosses stood on literal platforms exactly like stages. In others, a confined space such as a cave or library delimited a stage in size and orientation. The Zul’Gurub dungeon, for example, was gaudily theatrical—amid Mesoamerican ruins decaying in a jungle, crocodiles, panthers, tigers, serpents, and giant spiders guarded bosses who, during battle, transformed to huge, powerful animals. The biggest boss was Hakkar the Soulflayer, visible on a high platform as players moved through the instance killing lesser bosses.

It was an exciting moment when players assembled on the virtual stage in anticipation of a challenging fight. The relation of player to play was given literal form in the Opera encounter in Karazhan. A stage manager, Barnes, invites the “audience” of players to watch a play. But when the curtain rises, the characters onstage transform from actors to mobs, attacking the players—who are required to perform to avoid death. The flip from viewer to actor dramatizes players as agents who do not merely watch but themselves act. The encounter subverts pervasive cultural instructions to sit quietly, passively, one’s activity constrained to viewing (recall Dewey’s impatience with high-culture art). By abruptly inverting audience and actor, the directive authorizes, indeed demands, participatory activity.

Barnes invites the audience to. . . .

. . . a performance of Little Red Riding Hood.

WoW’s visual surface, then, did double duty; players could gaze appreciatively at their surroundings, but, simultaneously, the world’s visual features invited players to participatory activity.

Within commodity culture such as television, visual elements lack the capacity to instigate participation. Baudrillard (1983) asserted that mass media produce a “narcotized,” “mesmerized” consciousness of passive immersion in a spectacle of simulated images. Turkle, a pioneering media theorist, noted that with television, “the body of the television spectator is not in the picture” (1984). While television may have its social side (see Jenkins 1992; Mankekar 1999), I believe Turkle’s powerful observation, and the arguments of theorists such as Baudrillard, are generally descriptive of the medium.23

In the early 1980s, Turkle began to report the engaged, active experiences of computer users, including gamers, arguing that computers afford an experience in which the (virtual) body is in the picture. Even before Turkle’s work, in the late 1970s an immersive text-based game, Colossal Cave Adventure, generated a cult following. Colossal Cave Adventure established a digital space that described a game geography in words, offering players interesting performative opportunities. The game began:

Somewhere nearby is Colossal Cave, where others have found fortunes in treasure and gold, though it is rumored that some who enter are never seen again. Magic is said to work in the cave . . . You are standing at the end of a road before a small brick building. Around you is a forest. A small stream flows out of the building and down a gully. (Jerz 2007)

Players remembered:

The game spread like wildfire across the Internet, inspiring such obsessive efforts to solve the game that it is rumored numerous college seniors did not graduate that year as a result.

The entire computer industry was set back by a week. (Rickadams n.d.)

Though tongue-in-cheek exaggerations, these statements indicate the excitement and intensity players experienced in response to Colossal Cave Adventure’s performative challenges—the same excitement that suffuses today’s video games, which reach a much wider audience through their image-rich design.

New text-based games, worlds, and communities continued to develop after Colossal Cave Adventure such as LambdaMoo (Cherny 1991; Mnookin 1996; Damer 1998, 2009), Habitat (Morningstar and Farmer 1991), and the Well (Rheingold 1993), as well as console and arcade games (Hunter and Lastowka 2004; Huhtamo 2005; Malliet and de Meyer 2005). Mass culture, with its accumulation of images, seems quite capable of generating participatory experience; video games afford arenas of activity in which visual experience is unified with active performance. Baudrillard’s fear of the “simulacrum,” i.e., simulation that devolves to ersatz experience, and his rejection of the virtual appear to have been based on too narrow a sampling of virtual media, one that neglected the rapid development of interactive networked media.

Participation in virtual worlds is not simulation but performance. There is no faking performance; it is brutally honest. The software enforcing the rules and the players watching to see whether you click the cube at the right moment compel honesty. Postmodern theory asserted the delusional quality of mass-produced images, but even as those images were proliferating, new means of authentic expressive performance, embedded in vivid visual spaces, were emerging as forms of mass culture.

Additional Information

Related ISBN
MARC Record
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.