Bulletin, August/September 2006

We Live Here: Games, Third Places and the Information Architecture of the Future

by Andrew Hinton

Andrew Hinton is a senior information architect at Vanguard. He participated in the founding of the IA Institute and wrote its "manifesto." He keeps a home on the web at www.inkblurt.com.

I first heard the phrase in 1997. I was in a vast, strangely decorated chamber filled with catwalks, ledges and a red flag rustling on a dais surrounded by a pool of water.
I was on the Internet, logged into a game server running a version of Quake called Capture the Flag. By now the screen, keyboard and mouse had disappeared and I was "moving," not mousing; "talking," not typing.

Through my headphones came the sounds of two other players running, jumping and harpoon-grappling through the hallways. I was new to this particular map and clumsily learning its intricacies – shortcuts, hiding spots, hidden health and ammo packs, while they were doing virtual wind-sprints, practicing for a tournament. We were chatting over a public channel within the game, and I was joking how they seemed to know the map so much better than I. One of them replied: "Ha-ha, yeah... we live here!"

I was a recovering philosophy major who tended to overthink everything, so the words struck me as deeply resonant. And in the ensuing years, many of which I've spent practicing information architecture, the resonance of this phrase has not decreased, but has come to me often in the midst of my work. The fact is, it isn't hyperbole anymore to say "we live here" about an online environment. Our daily contact with simultaneously shared digital environments is only increasing and deepening with time. And I have come to realize that my early experiences in this game, and in the online community that emerged around it, have been a significant touchstone for how I understand my work as a practitioner of information architecture. 

This realization causes me to ask myself certain questions: Do game environments prefigure where we're headed with conventional software and networked experiences? Do they represent in some essential and prototypical way how the Web functions culturally? Are there lessons to learn there?

The Web/Game Hybrid
In 1996, when Quake arrived, there were already many multiplayer games on the Internet – digital environments with game characteristics that allowed many players to interact in them simultaneously. There were MUDs and MOOs, as well as some graphically rendered role-playing games. (For further definition and information on MUDs and MOOs see the Wikipedia articles about them on www.wikipedia.org.) But Quake was the first major three-dimensional computer game to use a set of open standards that allowed anyone to host a game server and create content and modifications for the game engine. In many ways, it gave birth to a sort of game/Web hybrid. 

When I say “Web” here, I mean the Internet’s “Web-ness” – that is, the characteristics envisioned by Tim Berners-Lee when he proposed the World Wide Web: hyperlinked, peer-to-peer content creation and publishing using open standards. It’s strange how easily we forget that when Lee created the first Web browser, it could publish pages as well as read them. He also specifically believed in hypertext as a democratic antidote to the hegemony of hierarchy, and he purposefully released his invention as an open architecture upon which others could easily build without impediment. 

Quake was Web-like because, although its core engine was proprietary, its designers released it with a set of tools and open standards that allowed the user community to make it into something infinitely bigger than original game itself. The Capture the Flag map, for example, had been beautifully designed and rendered by a regular user, and its virtues outstripped those of many of the maps that came from the publisher. Much like the meritocracy of the Web, this map was made by a user, published by that user and made a favorite through the collective consensus of use. Even the version we were in was a modification of the original game code. Everything from the game rules to the graphics, sound and physics were customized by a community of game coders. Modifications have spawned a huge, mainly non-profit, cottage industry of game and map design and communities of designers. And this is in addition to the many thousands of players who didn't build anything, but spent many hours playing and socializing in the game and outside it in chat rooms and discussion boards. 

As a result, the vast majority of actual user activity around the game of Quake wasn’t even playing the game itself – it was in the massive web of relationships, conversations, teamwork and collective creativity that happened in its orbit. 

The Game (and Game Culture) as Designed Experience
This phenomenon may sound like accidental anarchy but it wasn't. Because of what they'd seen happen with their previous hit game, Doom, the folks at id Software knew that their game would spread if they just created the right conditions and then got out of the way. The thousands of game servers that sprouted up almost overnight proved them right. 

Some of the conditions they created: 

  • A game with open standards that allowed anyone to create new maps and game modifications

  • An open language (called QuakeC) that turned anybody with rudimentary C programming knowledge into an immediate game hacker 

  • A function that – at no charge – allowed users’ game servers to automatically update location and status at id Software’s central directory, so that any Quake player on the Internet could browse for a quick pickup game. 

A very important point: each of these conditions was a design decision. This approach wasn’t just about marketing – the designers assumed that making the game that way was essential to making it a good game. 

Because of its open architecture and its applicability across the whole Internet, what Quake represented was a highly specialized instantiation of the Web. Granted, Quake was a product sold for profit, but, beyond the proprietary core of the game engine, everything else about it was chock full of Web-ness. And as a result it grew right along with the burgeoning World Wide Web of the late 1990s. 

The game creators realized that, given the right conditions, the power of open communities easily eclipses the planned efforts of any single organization. The challenge is to create structures that encourage and channel that power without hindering its collective energy and creativity. Just as the Web became much larger than the Berners-Lee’s first server and browser at CERN, the collective experience of Quake far outstripped the boundaries of the original design. 

What Quake (by which I mean, at this point, Quake-as-cultural-experience) became, then, was what urban sociologist Ray Oldenberg called “the third place” in his book The Great Good Place. Where our “first place” is home, and our “second place” is work, our “third place” is “determined most of all by its regular clientele and is marked by a playful mood, which contrasts with people's more serious involvement in other spheres.” Oldenberg is describing physical places, like the corner pub or town commons, and, with the increase of suburban sprawl, he grieves for the loss of these places in American life. (Ray Oldenburg, The Great Good Place, 1990)

Upon a moment’s reflection, it’s obvious that people are now filling this need for a “third place” online. Do game environments, being richly designed to feel more literally real than the regular Web, fulfill the need for place in a particularly satisfying way? If true, this conjecture may help explain the incredible rise in multiplayer online games that specialize in being a persistent, realistically rendered place. 

The Rise of the MMOG
Digital games are a significant cultural and economic force that is often overlooked in conventional media. Among adults, 75% of heads of household play some kind of computer or video game and sales of these games reached $7 billion in 2005 (www.theesa.com/facts/top_10_facts.php). 

The truly amazing phenomenon, however, is the rapid growth of Massively Multiplayer Online Games (MMOGs). These are games like EverQuest, World of Warcraft and Second Life, where, unlike Quake and its first-person-shooter progeny, you don't hop from server to server playing pickup games and tournaments, but situate yourself in a single, massively shared world as a digital representation of yourself or a role-played character. 

At the time when I was playing Quake in 1997 there were only about 100,000 registered users playing in MMOGs (Ultima Online was a pioneer), but that number shot up to over 2.5 million by 2005 and, incredibly, it has nearly tripled in the last year (http://mmogchart.com).

World of Warcraft (WoW), by Blizzard Entertainment, now has over six million subscribers – partly due to its expansion into Asia. In order to insure a continuous experience, the world itself is centrally controlled and managed, unlike the decentralized user-based game servers in Quake. 

While WoW is a goal-oriented game where players increase their character’s level and wealth through competition and quests, Blizzard knew that this framework alone wasn’t enough to keep people coming back. So the game provides tools and avenues for social networking and community building in-game. There are many quests requiring organized teams and plenty of ways to customize one’s game experience.

It turns out that a player’s level and in-game wealth end up being secondary to the personal character traits of the person behind the character in the social milieu of the game world. Like any neighborhood, being rich and strong doesn’t mean you’re a good leader or well liked. Underneath all the fantasy trappings, it’s the real human connections between players that keep them invested. Even for many players who prefer single-player action, WoW gives them a social context for their exploits. While the original MMOGs didn’t cater to these needs as much, over time developers realized that any successful MMOG folds agile social networking tools into its game environment (www.fineartforum.org/Backissues/Vol_17/faf_v17_n08/reviews/
and http://terranova.blogs.com/terra_nova/2006/02/alone_together_.html).

This activity isn’t just kid stuff. The median age of WoW players is about 29, which means for every teenage user, there’s a 30-something (www.nickyee.com/daedalus/archives/001365.php). Increasingly, for adult players, WoW is becoming “The New Golf,” where people make business connections and talk shop (http://joi.ito.com/archives/2006/02/10/world_of_warcraft_the_new_golf.html). John Seely Brown and Douglas Thomas write that, far from being a shallow diversion, “the process of becoming an effective World of Warcraft guild master amounts to a total-immersion course in leadership.” People are even starting to put such credentials on their resumes (www.wired.com/wired/archive/14.04/learn.html and www.wired.com/wired/archive/14.06/warcraft.html). According to entrepreneur and digital activist Joi Ito, WoW "represents the future of real-time collaborative teams and leadership in an always-on, diversity-intensive, real-time environment. [The game] is a glimpse into our future."

While WoW has become immensely popular, there are limits to what people can do in the game world. Unlike the Web, where users collectively determine the shape of what is created, the main content within WoW is still created by the publisher, and they control (sometimes controversially) the way in which players can express themselves in the game. What about a game that doesn’t have these restrictions?

Enter the game Second Life. In this MMOG launched by Linden Labs, the players are responsible for creating almost everything. While there are some competitive games or quests within Second Life, it’s principally a social environment that encourages community and creativity. Users have in-game tools for creating intricately designed avatars that can look like just about anything or anyone, and through activities in the game they can earn currency to purchase clothing, gadgets and points used for building items themselves. 

There are amazing feats of architecture going on in Second Life (many of them by actual architects), and if you spend real money you can have “land” there to build upon – but you can play as a non-landholder for free. Real business happens there, and "real-life" business is taking notice. Second Life was recently the cover story in Business Week, and last year Wells Fargo, one of the oldest financial institutions in North America, spent actual time and money building a place in Second Life called Stagecoach Island (Figures 1-3: Second Life Places Built by Wells Fargo; also http://news.com.com/Wells+Fargo+launches+game+inside+
). Visitors can learn about mutual funds and mortgages from live personnel – or they can go skydiving. It's a serious game inside a non-serious game, in a virtual world with virtual currency, where players learn about finance and the real economy.

Stagecoach Island  Stagecoach Coffee Shop  Stagecoach Mall

Figures 1-3 Above, left, Stagecoach Island; above middle, Stagecoach Coffee Shop; above right, Stagecoach Mall.

All images created by Wells Fargo and used with their permission.

One notable result of Second Life’s openness and success is that it’s become an even more conducive environment for people to make real-life connections in business or creativity or even (especially) romance. And it’s attracted quite a user base: 15 years ago, the place where all the cutting-edge Internet thinkers and writers would hang out was “The Well” – a bulletin board service (BBS) hosted in California. But now, the new generation of Internet literati is gravitating toward Second Life and other places like it. Part of the draw is that Linden Labs operates under the Creative Commons licensing standard so that anything a user writes, designs or builds in Second Life belongs to the user. 

As with any new medium, people worry that a game like this will usurp the users’ lives and make them more isolated from flesh and blood people, but the opposite is the case. For most players, Second Life is a complement to their regular lives, truly a “third place” in the Oldenburg sense. And the game designers have gladly enabled this kind of permeability. For example, from your avatar in the game you can communicate with the outside world – it’s not unheard of for someone’s avatar to have an extended message chat with a friend in RL ("real life"). Linden Labs has also developed a map API so that, for example, a Second Life player with a blog can post “I had a conversation on a park bench in the game today” and offer a link that another player could click in the browser to launch Second Life and take them straight to that park bench – much like links work in Google Earth. 

Players have even gone so far as to model real environments in Second Life. The Electric Sheep Company, an actual company of designers who make a living creating objects and environments for other players in Second Life, recently hosted an event they called The Happening – an arts and culture benefit for Washington, DC – at a local favorite coffee shop. They also created a model of the coffee shop in Second Life, and had the party there as well, simultaneously. And last year, when SF author and online activist/writer Cory Doctorow’s new novel was released, friends created an e-book version of it within Second Life, and Cory had a book party there and autographed virtual copies of the novel for friends. While these events may have started out as occasional stunts, the blurring of boundaries is only becoming more common (www.electricsheepcompany.com/happening.php and www.dragonscoveherald.com/blog/index.php?p=864). 

But what about the Web-ness mentioned earlier – the peer-to-peer distributed democracy of Web information? Doesn’t the necessity of having the game on a centralized server take away from that ability? To some degree, yes. It just means that the game population is at the mercy of whoever is running the server. In the community Quake experience, having a distributed collective of game servers meant that, even if one server went down or its administrator decided to be a jerk, there were still many other places to go and play – plus the bulk of the community happened outside the game, on the Web, anyway. What if something as immersive and persistent as Second Life could exist independent of a central infrastructure? 

It turns out there is such a thing, though it’s in very early stages of development. It’s called Croquet: an immersive 3D multi-user platform that is open source and peer-to-peer. It’s the brainchild of a company calling itself Qwaq, which includes the likes of Alan Kay, one of the creators of object-oriented computing and the modern windowed computer interface, known for the statement "the best way to predict the future is to invent it." 

Qwaq is pitching Croquet as a paradigm for a future operating system (OS) – one that doesn’t require a client-server architecture. What this means is that one’s OS would be part of a distributed, organic, immersive environment that makes use of some of the interface conventions found in a place like Second Life to foster collaboration and cross-platform interoperability. The early demos feel like alien worlds at the moment; but for anyone who remembers the first time you tried using a mouse on a computer, it has a similar feel of the strange but inevitable future. Essentially, the MMOG paradigm has broken the boundaries of entertainment and entered the imaginations of serious futurists pushing evolution of next-generation software (www.opencroquet.org).

Gamespace and Ubicomp
For a lot of people, though, the future is less about the interfaces themselves than about how computing is spilling into our everyday objects and lives: “ubiquitous computing.” Mark Weiser, the Xerox scientist who coined the term, describes ubiquitous computing (ubicomp) as “roughly the opposite of virtual reality. Where virtual reality puts people inside a computer-generated world, ubiquitous computing forces the computer to live out here in the world with people” (www.ubiq.com/hypertext/weiser/UbiHome.html).

This description makes it sound as if virtual reality and ubicomp are somehow mutually exclusive, but they’re not. They’re complementary sides of the same thing. The more useful virtual reality becomes, and the more widespread its use, the more it inevitably interweaves itself into our daily lives. An example is the one given earlier, of the Second Life player chatting with an out-of-game friend, avatar-to-cell phone. 

But you don’t need a game environment to experience ubicomp now. Anyone with a Tivo can schedule their home television recording from their office or their web-capable PDA or rent home videos from Netflix the same way and engage in sharing favorites lists with friends in a social network as well. But even more significantly, the ubicomp world is one where common devices and objects are networked both in cyberspace and in real space – their physical locations and IP addresses coequally defining their existence and, essentially, their relevance. The fact that users in Second Life can model a real space and have an event in both at the same time only serves to prefigure the blurring between map and landscape that is to come.

That’s why some researchers are studying it to see how human beings behave in a truly saturated ubicomp environment. One problem: We don’t yet have these environments, either because they’re too expensive to create or the technology simply hasn’t quite reached the degree of refinement it will have five to ten years from now. So how do you study something that doesn’t exist yet? You model it in a game environment. 

The TATUS project, conducted in 2004 at Trinity College, Dublin, “aimed at overcoming [the] cost and logistical issues. Based on a 3D games engine, the simulator [was] designed to maximize usability and flexibility in the experimentation of adaptive ubiquitous computing systems.” Using the Half-Life game engine, the research team created the tool to model “a true representation of the world so that a person’s natural surroundings become in essence a user interface. The natural movements and gestures of people occupying the space become the input commands to the application controlling the environment.” Essentially, they used virtual reality (VR) to model the future of real life (RL). Over time, VR and RL are inevitably spilling over into one another (www.cs.tcd.ie/Dave.Lewis/files/05a.pdf).

As you can see, game environments have changed drastically over the last 15 years; they’ve gone from being a computer-nerd pastime to being a mainstream activity that’s relatively integrated into the regular lives of professional adults. And even beyond that, the immersive-yet-permeable game environment has become a compelling paradigm for those who are thinking hard about the future of networked computing. 

Why Game Environments Are Significant for Information Architecture
I contend that information architecture isn’t really about “information.” Information is really just an artifact of human discourse. It’s a thing we can touch to know that someone has spoken. It’s what tells us we aren’t alone and allows us to reflect, to create shared stories and meaning. IA practitioners often obsess that we’re responsible for investing the spaces we design with meaning. But this is like saying a game designer’s responsibility is to invest the game environment with fun. That’s not our job; our job is to create the right conditions that allow users to bring the machinery to life.

My ultimate point is this: Game environments serve as useful prototypes for the future of conventional networked software design, both in the literal sense of learning from particular design patterns and in the more oblique sense of game architecture as archetype. If information architecture is about designing the structure of a new kind of place for human activity, one made of bits rather than atoms, surely it can learn from understanding this more literal, archetypal form of shared digital environment. 

Luckily for anyone wanting to learn more about games as social artifacts, an impressive growing body of research about them exists in the field called “Game Studies.” And the more one reads this work, the more it sounds like the sort of things those who do information architecture should be talking about. 
Constance Steinkuehler, who teaches game studies at the University of Wisconsin-Madison, explains that the study of multiplayer games “requires an understanding of the full range of human practices through which players actively inhabit those worlds of rules and texts and render them meaningful” (http://website.education.wisc.edu/steinkuehler/papers/

This powerful statement deserves some parsing. Players (users) are actively inhabiting (“we live here”) an environment that is essentially a framework of “rules and texts” – a world that carries little or no inherent meaningfulness, but that is “rendered meaningful” by the activity of the inhabitants. 

According to Hideo Kojima, creator of the well-regarded game Metal Gear Solid, good game design and good architecture are “essentially the same thing. … In both cases, the designer has to ask what service the structure will provide. From there you decide what raw materials you will use to help reproduce the sense of play and fun over and over. Looking at the different functions of a building can be a great study in game design" (www.wired.com/wired/archive/14.04/gurus.html). Essentially, just as in brick-and-mortar architecture, the success of the game is in how it shapes human experience. 

At this year’s South by Southwest Conference, Jimmy Wales of Wikipedia conducted the keynote interview with Craig Newmark, creator of Craigslist. A theme that arose again and again in the interview, stated by both Wales and Newmark, was that the key to the enormous popularity and effectiveness of their creations was that they didn’t try to micromanage or predict their users’ every need and behavior. They merely created the right conditions for achieving what their users needed to get done, and they “got out of the way” (http://2006.sxsw.com/interactive/programming/panels/?action=show&id=IAP060077).

I find these to be excellent expressions of what it means to do information architecture. We work to understand human needs and behaviors in a particular context and then design structures that facilitate, enhance or even redirect and modify those behaviors. What we do really is architecture, but in digital bits rather than atoms. We don’t make information into architecture for the sake of the information. We use information – rules and texts – as material to create digital architecture for the sake of its inhabitants. 

The Game Layer
Wikipedia and Craigslist are just two examples of third places that succeed in part because of qualities that also make an effective game environment – emergent spaces where user activity and interaction create meaning and relevance. Other places like Flickr, Digg, MySpace and Facebook have also become extremely popular in the last several years for similar reasons. They're frameworks – not predefined experiences – that provide relevant opportunities for discourse and play.

Usually when we say the word play we mean an amusing activity. But a secondary meaning is “latitude, leeway, elbow room.” A game-studies version of the term is spielraum – literally playroom in German. Game environments, and in a larger sense the Web, represent the inevitable triumph of spielraum. Even though the Web is made of rules, logic and predefined structures, people still find their own spaces, their own elbow-room, and they appropriate carefully carved Parthenons for use as fruit stands, clinics or singles bars – whatever they happen to need at the time. People are not only already taking these third places for granted, they expect all three places to be connected, interrelated and ever-present. How do we become better architects in a reality where architecture is so dynamic and deconstructed? We need a new perspective, perhaps even a new language, to meet that challenge.

As a start, I suggest we look to the game paradigm, and think of the idea of game in a broader sense – not unlike its use in game theory or Wittgenstein's use of "language game" – and recognize that, by understanding how a game is essentially a systemization of human behaviors toward some common enterprise, we can better understand the nature of what the Web is becoming.

There is a layer of the coming world that is spreading ever more widely into our lives – a Game Layer – that truly is a "world of rules and texts" – a layer that will surface the hidden, tacit subtleties of human experience as explicit, relational data, a layer that intermingles first, second and third places in ways we're only starting to comprehend. It is essential for those of us who claim to be architects of these places to better understand what we bring to this increasingly present future. After all, we live here.