Community, Part II – social behavior in the online world

Community, Part II – social behavior in the online world – includes a related article on establishing online rules

Jerry Michalski

When you enter a party, you pick up an amazing amount of information. You can see how people are arranged: in small, loose clusters; in conspiratorial huddles, heads averted or bowed down; in rapt groups, which you might want to join; negotiating something (is it a merger, or simply a golf date?); or paired off in wooing rituals. You can hear laughter coming from some groups, nothing from others and loud debate from a few. You can rely on most people to represent themselves honestly, since others who know them are likely close at hand. It’s hard to mask one’s identity at a party.

Usually, you can deduce rank and status. Hosts accompany VIPs or special guests. Certain behavior stands out: You would notice if someone talked to himself in the middle of the room, or if most of the participants slipped away to private rooms. All of this information helps you decide where to go, whom to approach and how to approach them. Do you want to argue? Laugh? Are you expected to bow or use an honorrific? Hey, there’s Zoe! Did that person just wink at you?

You know how to behave: how to introduce yourself, how long to stay in a conversation to be polite, when to interrupt a conversation. You can stand in one group and listen to another. These things hold true for formal business meetings, too. Who’s talking with whom matters.

The online world offers few such cues. As we noted last month, a textual environment is a powerful least common denominator, masking cues that might bias users against interacting with someone and allowing equal access to all. But those cues can be useful. They are a shorthand that gets people feeling comfortable quickly and helps them make important decisions, (politically) correctly or not.

The online world need not be a perfect mirror of real life, but it does need a similar set of subtle signals if newcomers (newbies, in online argot) are to feel comfortable quickly (and want to come back) and if people are to be more productive.

The June issue of Release 1.0 framed the market for electronically supported communities, outlined their significance and offered some ideas of what functions people would find worth buying. This month we examine three aspects of real-world texture that such worlds should provide:

* Context: what is special about this place? what is in it? what protocols do I need to observe? what can’t I do? who governs this place? can I affect the place or the rules? how can I make the context work for me?

* Presence: are others here? are they who they say they are? are they listening? how do they feel about what I just said? are they being snide or loving? how are online and physical presence different?

* Interaction: how can I connect with others? what can we do together? what tools or protocols would help? how can I create and maintain feelings of trust and communion in virtual space?

We consider these features in the context of the race to find what consumers will use and pay for. As we explore, we keep three questions in mind: How can service providers foster or create these features to distinguish themselves and build a stable following? Can new features scale to support audiences large enough to be profitable? And what is the relationship between virtual communities and real-life activity?

Faites vos jeux

Media companies, cable tv, production houses, carriers, software, computer hardware, communication and companies are all placing their next-generation service bets now through technology alliances and infrastructure investments. How carriers design their piece of the information highway (especially its on- and offramps) directly affects what services they will be able to provide, and who will create them. If they bet too low, they won’t be able to offer appealing services later on; if too high, they will be unprofitable.

For example, some phone companies are counting on ADSL technology (Asymmetric Digital Subscriber Loop, aka video dial tone) to leverage existing copper connections to subscribers. While ADSL is probably satisfactory for movies on demand, it is not responsive enough for real-time interaction, which consumers will probably want. That means coax and high-speed switches– and more alliances between phone and cable tv companies.

But what are the driving applications? Online multi-player action games are a start, but we believe a key motivator is the desire to connect with others in some meaningful way. Many little companies are doing imaginative work in this area; many will need funding from the big ones. This issue of Release 1.0 investigates some playful offerings and some serious ones, emphasizing how they can help people connect with other people.

Systems you might like

Today, most inter-personal online services are structurally similar to each other, and very different from cable tv or the phone system. Eventually, the spaces in between will be filled by new services. Imagine the following choices in the year 2000:

* YourWorld allows subscribers to design their space and the rules they live by, within limits: the space is generally realistic (no bug-eyed please). No business is allowed in YourWorld, which has no memory. That is, the system encourages real-time communications, and wipes the slate clean every three months. People come to YourWorld to learn things firsthand. They want to discover legal systems and invent languages and social norms. Text spoken here.

* RedLife blends personal and business life. It has advertising. and an underlying market-making and funds-transfer system. Entire companies subscribe to ReaLife, and individuals can run their whole lives through the online service, which offers plentiful mass storage. The system features an integrated media environment, which allows users to switch effortlessly between typing, speaking and video. Portable handheld terminals and real-life notification of online events tie the system to real life. ReaLife relies on personalities as guides and attractors.

* The Rubber Room has few rules, and those are regularly broken. Creatures of all sorts roam the halls/parapets/marshes. Some are automatons, some are people in disguise. Games and fantasies abound. Video is the principal means of communication. Rubber Room subscribers prize great special-effects jocks, and 3D-jays (we-already have DJs and VeeJays) have huge cult followings.

* TrueNet has trusted communications. People are definitely who they say they are. Pseudonyms and avatars are permitted (in a separate zone). but a flag indicates which people are behind a mask., so you can act accordingly. Another zone is safe for kids. TrueNet groups subscribers into cohorts (by choice, or merely chronological) of a finite. manageable size. TrueNet has special sections where people can meditate together or just hang out, without worrying about connect time.

Clearly these thought experiments open many questions. Can one service have many places with different characteristics, or is one “place” the basic unit of measure? Can people move easily from one service to the next, and have their bills consolidated? How are people billed — by time online, transaction or number of objects used? Does it cost extra to talk to single women, lawyers or Bill Clinton? How are these features made clear to newbies?

Profiting from your persona: ‘zines

The services most likely to thrive are those that entertain, and those that maximize salience (interest and relevance) and communion (the feeling of connectedness that sometimes occurs in communities). That means trusted sources and trusted spaces. Referrals and recommendations are more useful and personal than surveys, statistics and scientific tests (as in the real world). As the number of participants grows and the volume of information becomes unmanageable, moderated forums will increase in importance. People will sell their attitudes, their personalities and their attention (see Release 1.0, 3-92). They will do it by hosting a ‘zine (a term that started with low-circulation, science fiction fanzines).


Subscribers who don’t like the feeling of a particular online service or its rules may keep quiet, go through channels to change them — or leave. One of our premises is that human behavior is more interesting than any mechanized information feed, and more so when it’s about stuff that matters. And what can matter more than the rules that govern everyday (online) existence and the objects and features that users encounter? This section focuses on services that offer that kind of power and flexibility, either in software that users control directly, or through jointly developed rules of governance.

Mirroring reality

One way to make the online experience more comfortable and comprehensible is to model it on some known (or imaginable) reality. For example, in 1981, when Dave Hughes started the Old Colorado City BBS, an early bulletin-board system in Colorado Springs, he named sections of it after places one might find in a community: the Old Town Bank for business; Roger’s Bar (with Hughes as the electronic bartender) for discussion; the Little Red Schoolhouse for education; and so on.

There actually is a Roger’s Bar in Colorado Springs, but the only high-tech element in it is a phone Jack that the owners installed at Hughes’ request. Most of the other “places” in the Old Colorado City BBS are invented, even though many outsiders believe it is a faithful map of Colorado Springs. In fact, the BBS’s popularity led to some sprucing up by the city, which has now taken over the BBS and calls it CityLink. So in a strange, roundabout way, the city that Hughes envisioned and instantiated electronically has affected the physical community it represents.

Build your own: the MediaMOO

One of the lessons of early online systems is the importance of participatory design and construction of the shared online space. The Internet is home to some such online environments, most notably MOOs: Multi-User Dungeons (MUDs) that are Object-Oriented, minus the swords and sorcery.

Allowing participants to construct personally meaningful artifacts is an essential element of MOOs. They can do this by either defining so-called contributory objects (simple instantiations of existing object classes, such as a costume or greeting card) or programming new objects or places (usually by tweaking existing code). Instantiate a dog, and it will automatically follow you and wag its tail. If you examine the code and learn the language, you can make the dog guard your house, or maybe do tricks or exhibit other dog behavior — or have it talk.

For example, there is the MediaMOO run by Amy Bruckman, a graduate student in Seymour Papert’s Epistemology and Learning Research Group at the HIT Media Lab. MediaMOO is designed to support the professional community of media researchers. Built with MOO software developed by Pavel Curtis at Xerox PARC, the core of MediaMOO resembles the Media Lab: Rooms and their occupants are situated similarly to the real thing (there are no graphics; this is all described only in text). In addition, individuals create their own rooms with special features and creatures. They can also add to or comment on others’ creations.

Lately Bruckman has been exploring how children behave online and how systems such as MOOs can affect their learning experience. Bruckman and her colleagues are designing a text-based MOO to be called Moose (MOO Scripting Environment) Crossing, which will feature a multiple-window client program and a Logo-like scripting language.

Governance: the Well and Habitat

Giving up control and allowing participants to co-create the rules and the space itself seems to be more effective (and maybe less stressful) than trying to design and manage the perfect environment centrally. The self-organizing process is as much a part of forming community as any aspect of a place. The participants may be thinking, “Let’s design a place we love and find a way to coexist with others who want to be there” and not, “Let’s build a successful online service.” But the results may be the same, or indeed better for the provider. The Well: tools not rules

The Well is a good case study of the evolution of governance in an online service. It had a unique start in that its designers had no commercial charter: They Just wanted to create an online place to hang out together, as many of them had done in real life at the Farm, a large commune in Tennessee. Since these developers didn’t consider themselves to be great authorities on governance and were not above listening to their subscribers, early members had much to say about the Well’s rules of conduct. On commercial services, those decisions are typically made by Management, with little debate among members. That can lead to uprisings, as Prodigy experienced. Habitat, described below, offers more lessons on governance.

Over time, through much online discussion, the Well’s management and users co-developed community standards. This path is not easy: The Well has had its share of uprisings and staff turnover. One general principle is a preference for tools, not rules. For example, rather than create a specific policy on who is being a nuisance and should be punished, some Well users created Bozo Filters, which are invisible to the people being filtered out and offer a shield to users being bothered. A Bozo Filter skips postings and messages from people on a user-created exclusion list. Of course, there are further complexities: How do you participate in a multi-party online conversation, if you are filtering out one or more participants’ contributions? What if the nuisance people begin to say things you’d like to hear?

Habitat: ToonTown online

Habitat is more than a nifty online multi-user game: It is a social laboratory whose participants grow to care. Habitat users can design and inhabit entire characters, which Habitat calls avatars. Players can customize their avatars with different heads,(1) colors and accoutrements. With a joystick, they can make the avatars gesture and move (e.g., point, pick up and drop things, turn and walk). In fact, one group of users scheduled and choreographed online dances, which required calculating data-network latencies so they could send their commands at the right times to achieve synchrony. Habitat characters speak to each other in cartoon-style speech balloons.

To see the sysop, participants have to go to the fountain and invoke the oracle. Habitat has an economy, with Tokens as its medium of exchange. There are vending machines, ATMs (automatic token machines) and goods for sale; there are pawn shops, and even arbitrage (at one point, some players discovered that a pawn shop was offering more for an item than a certain rending machine’s discounted prices).

During its original trial run in 1987, Habitat’s developers experimented broadly, with fascinating results.(2) An important breakthrough occurred when they let go of their need to plan everything centrally and said they would build the structures and features that users wanted. For example, one participant (a Greek Orthodox priest in real life) wanted to start a church, so Habitat’s developers created a church, home of the Order of the Holy Walnut, which preached nonviolence and became quite influential in Habitat.

Moral choices in a Toon world?

Death matters in Habitat. Avatars who are killed drop whatever they are carrying, lose what is in their pockets, and then get teleported home, head in hands (literally). Death, coupled with the system designers’ willingness to experiment, led to unpredictable (and very engaging) phenomena.

To provide drama and raise moral issues, Habitat’s developers offered weapons for sale. which rapidly became a problem. Half of the members saw Habitat as a game, and the avatars as expendable puppets, to be killed over and over; the other half saw Habitat much more seriously, and considered life important. One Habitat member wandered around randomly, killing others. Some of the victims were upset. As a compromise, the sysops made Populopolis (the major city in Habitat) a weapon-free zone.

To cope with theft, murder and other mayhem, members created a small-scale legal system. Habitat’s developers suggested electing a sheriff. Three people campaigned, and the sysops gave the elected sheriff a ten-gallon hat with a star on it. Despite having no special powers, he was surprisingly effective. Unfortunately, the Habitat pilot ended before participants could explore the implications of a legal system.

The elephant gun

One Habitat novice wanted to create a labyrinth and play the role of Death inside it. Thus was born the Dungeon of Death. To spice things up, the developers created a healing elixir and a special weapon they called the elephant gun: Normally, killing someone took 10 shots from a weapon; the elephant gun could do it in one. One day, sysop Randy Farmer inherited a Death character that was trapped in a corner of the labyrinth with several other avatars, and hadn’t been healed with the elixir. Unexpectedly, Death died, and the elephant gun ended up in an ordinary user’s hands (other characters can scavenge items dropped when an avatar dies).

Despite entreaties and polite requests, the user refused to return the gun to Farmer. Instead, the elephant gun became a prized possession; eventually, the sysops ransomed the gun for 10,000 Tokens and staged a melodramatic exchange, complete with intermediaries and news stories in the Habitat paper. Online time skyrocketed. Unknown to Farmer, the same thing had happened the day before to a different sysop, who had used his Sysopian powers to retake possession of the elephant gun, threatening to cancel the user’s account. This had caused a major ruckus among the participants, who felt that even the system’s overlords should have to play by the rules. (Of course, online time Jumps when people use the system to complain and commiserate, too.)

Wanna build an electric community?

Habitat was born over a lunch between Chip Morningstar and his Lucasfilm officemate Noah Falstein in 1985, inspired in part by the science-fiction novel True Names by Vernor Vinge. Operational testing began in December 1986 on QuantumLink, Quantum (3) Communication Services’ Commodore 64 offering. Despite the low-power Commodores and 300-baud moderns, Habitat delivered a surprisingly rich experience. In fact, much of what the developers learned was how to optimize a system for use by many unpredictable (human) entities, while preserving independence from underlying hardware with an object model.

Habitat was tested for about a year with 300 users. Six months later, a stripped-down version, now called Club Caribe and given a Club Med motif, was back on quantum/America Online. It runs to this day, but only on Commodore 64 front ends. Along the way, Fujitsu licensed the idea and wrote new software for its FM Towns multimedia machine. which features a CD-ROM drive; it still runs a Habitat system in Japan on NIFtyServe.

Homingstar spent much of the intervening time at Amix. and is now working with Randy Farmer (the principal Habitat sysop and developer, and the key front-end programmer at Amix) and Doug Crockford (a colleague from Lucasfilm) to shop their ideas around through their recent startup, called Electric Communities.

(1) New users have a choice among five or six relatively bland heads; they can purchase others, or might receive others as rewards for work well done. In the system’s test run, the heads became as much a medium of exchange as the Token, Habitat’s currency.

(2) Chronicled in Cyberspace: First Steps (see Resources, page 17).

(3) Quantum also ran a Macintosh (and later pc) service called America Online. When that business grew to dominate, the company changed its name to America Online.

Prometheus unbound

In the Disney movie “Who Framed Roger Rabbit,” real characters who go into ToonTown (the part of town inhabited by cartoons) are suddenly subject to a new set of physics, laws and behavioral rules, dictated by the nature of cartoons. They can expect to talk to cars and to survive if a safe falls on them from the building above. While the Disney studio could control its environment and its inhabitants’ reactions, Habitat’s creators found that any time they tried to make something happen or to circumvent the rules, it would backfire; if they left things alone and went along with the agreed-upon rules, users responded with intense usage and loyalty.

Although some users will be happy with authoritarian systems that offer guaranteed service and little autonomy, others — we suspect many others — will prefer services that walk the fine line between setting the rules and allowing participants to write their own. The rules needn’t be completely consistent. They can even be warped, but the participants have to play by the rules — including the owners.

COPYRIGHT 1993 EDventure Holdings, Inc.

COPYRIGHT 2004 Gale Group