Monday, June 3, 2013

New location

Hi!  My actual blog is at Eponymous Fliponymous, this is just an archive these days.  Come see me over there, or at Huffington Post, or at BiNet USA!

Sunday, July 29, 2012

Article #12 (and last): Strange Attractors and Suppleness
being a look at the importance of inital conditions, and why administrators should be able to take a punch in the nose gracefully now and then
An attractor is a point or points on a graph that affects the other points plotted. It's a sort of mathematical gravitation. An attractor perturbs your lines, and can be located and defined even if it cannot be seen, much as the planets beyond Saturn were discovered (as well as how Obi-Wan found Camino). There are three types of attractors. A fixed-point attractor is like the vanishing point in perspective drawing, the point all lines converge at. Another attractor is a set of points that everything eventually circles through, a limit cycle. Strange attractors come from the science of Chaos.
Meta issues often resemble strange attractors. Activity cycles around them in infinite pattern, but never actually reach them. In math, the Lorenz Butterfly is one of the most famous. It resembles a mask, with lines in a complex lemniscate, the "eyes" never passed through, and the orbits switching unpredictably from one side of the graph to the other.1 One of the features of the butterfly is that, if you start two lines at slightly different points, the resulting lines will be different. This is the sensitivity of initial conditions, the reason weather is unpredictable in the long term, and the source of the Butterfly effect, the idea that a butterfly flapping its wings in Brazil can spawn a monsoon in Bangladesh.2 Both of these butterflies have meaning to the Metacore.
Meta issues (as opposed to freewheeling flamewars) are interactions between management and the population that have gone out of control and into a potentially infinite iterative cycle. One key part of this cycle is the history of the administrator in question. A neophyte host is bound to make mistakes -- we all do. If they do not recover from these mistakes, meta can become a spin cycle. One example is the administrator of Cafe Utne. Having been embroiled in several unresolved (and in some cases, unresolvable) issues, he is now in a position where even a small issue can blow up. Admittedly, he has a crew of very vocal critics who watch his every word and action -- but if he had recovered better from earlier incidents, it's possible that those critics would be more willing to let some of the smaller things slide, or might not have formed at all. Although it's the inconsistencies in the policies that cause most of the Meta, the history of the personalities involved has a significant effect.
Hypothesis: Personality conflicts between individuals are one of the drivers of Meta.
Stepping briefly into the analog world, I've observed that groups like clubs and companies are subject to the impersonal rules of group dynamics, which take as a given that all groups share characteristics regardless of the individuals involved. This works when you are discussing an entire industry, or a group of factories in the same town. However, when you start looking at individual groups, the personalities of individuals within the group, especially at the top, make every one different.
A perfect example of this comes from the PBS show Frontline's program on McWane and ACIPCO. They make exactly the same product, cast iron pipe, but McWane is recognized as one of the worst companies in the country to work for. The founder set policies that bring in high revenues, but lead to high turnover (near 100%, in one plant), apallingly high injury and death rates, and a very dissatisfied work force. ACIPCO, an employee owned company, has a waiting list, less than .5% turnover, and is recognized as one of the best American employers. J.R. McWane, the founder of McWane, was president of the company that became ACIPCO -- he quit to start his own company when the founder of ACIPCO willed the company to the employees.
Back to the digital world. Two otherwise identical virtual communities, with the same software and purpose, can be radically different. If one is run by an autocrat, or as an aristocracy, it will develop a different culture than a democratic, egalitarian community. Decisions made early in the life of the community will have a snowball effect later on, and that effect will not be linear, but quite possible exponential.
The administrators of a virtual community have a certain level of control over the culture and life of that community. If they lose that control by not acknowledging mistakes, by favoritism, by allowing the appearance of capriciousness or conspiracy, they will often end up in a position where they lose the respect of the people underneath them in the hierarchy. Once that happens, they will often leave the position3, or, to quote Mittens at Neopoiea, "just keep tightening up the screws, until the only ones left are the toadies and the Yes folks who need it that way. A lot has to do with playing fair with the people you're involved with." Playing fair is one of the most important ways to prevent meta spin cycles. Respect is two-way -- if the upper levels of a hierarchy allow the appearance of contempt for the lower, soon it will go both ways. An example from E2 is "April Troll's Day", when the administration, in an admittedly ill-advised April Fools prank, spent a large part of the day trolling the site and hoaxing h4x0ring. Regardless of your opinion of the administration's right to do so, regardless of your opinion of the humor (or lack thereof) of the prank, it had an effect on the site that did not go away on April 2nd, spawning four nodes about the events of the day, and leaving a lasting impact on some people.
At Cafe Utne, there was an event with an even broader impact. Known as "the Well-Engaged Crisis" in some circles, it spins (quietly, for the most part) around a proposed software change (that never happened). The reasons that it was a crisis rather than simply a nonevent is the perception that the Cafe Patrons were not properly treated, were not given any respect by management. The lack of resolution to this meta matter serves to increase the repercussions of later issues, in a small but not insignificant way. (According to the administrator of Cafe Utne, the Well-Engaged Crisis has not had as much of an impact as the actions of the then-administrator on that and other issues.) It's not uncommon for a mid-level admin to take heat for something done by another person at the same level.
Hypothesis: The administration reaps not only what it has sown, but the fields of their predecessors and contemporaries as well.
Axiom: In the eyes of the lower levels of the hierarchy, the higher members are frequently interchangable. This is increased by the attitudes of the higher levels. Although it's certainly true that particular admins will get reputations as exceptionally fair or trustworthy, one bad apple does, indeed, often spoil the barrel.4 If there is a policy of always backing the mid-level admins in public, is it any suprise that the actions of one host are attributed to the subculture of subadmins?
That explains the butterfly effect, but what of the strange attractor? The strange attractor, in Meta, is the event or idea that no one talks about, usually for fear (justified or otherwise) that to speak of it will bring down sanctions from above, or from your peers. Often it's something that the management has deemed Not To Be Spoken Of, formally or informally. Formally would be, to use April Troll's Day as an example, the note at the end of the nodes stating "There is nothing more to say. Contact an editor if you really think there's more."(paraphrase) A less formal but more meta-inspiring case was at Utne, where a poster was asking questions of the publisher if the magazine in an "Ask The Publisher" topic. When he persisted in asking valid if uncomfortable questions (that the publisher never saw fit to answer), he was sanctioned by the administration. This had a chilling effect on substantive dialog on his issue (one of copyright regarding his own words vs. corporate copyright over the entire site, an interesting YOYOW issue worthy of yet another, possibly too lengthy article ;) ).
In a lot of virtual communities, it is considered a sin to publicly speak out against management. (I use sin to mean "an action that, although not against formal rules, violates the norms and goes against the folkways5 of the community.) This forms the heart of many Meta issues.
Axiom: The upper levels of the hierarchy must be resilient and confident enough to absorb a higher level of criticism from the lower levels than vice-versa.
At the higher levels of a formal, software based hierarchy as is found at nearly all virtual community sites, the admin and sub-admin levels, there are specific software tools that can be interpreted as having extra power over others. At the lower levels, such as the general membership, the only power you have is your words. If you have the power to hide, erase, or edit those words, you need to, as well, have a thick enough skin to handle criticism both mild and sharp. Although it can get out of hand and become outright pillory or abuse, you can't be in a position of power and not expect to upset someone, sometime.
Axiom: The amount of verbal abuse directed at an administrator or group of administrators is directly related to how they handle such criticism.
Blaming the victim? Absolutely not. For one thing, being specially empowered is incompatable with victimhood. For an admin to say, "I can't take this, leave me alone" and click the button that silences their critics is victimising others. "If you can't take the heat, get out of the kitchen" is more than a Truman quote -- it's a law of nature.
Part of the reason for having mid-level administrators is to take the workload off of the top admin. Another part is to take some of the heat off of the top admin. If the actions of mid-level administrators leads to an increase of Meta activity, if someone's fragile ego or sense of superiority leads to an escalation of a situation that could have been handled with fairness and tact, they are a detriment to the community. At a site that has a genuine interest in building community one or two such incidents would be grounds for the host to learn how it feels to be hosted. At a site viewing the community as a commodity, where hits or revenues are a primary driver, such incidents are simply slid under the rug.
It's easy to sit back in the comfort of hindsight and objectivity and say, "don't take it personally." When Don Corleone was shot by a competing gang, he told his son, "It's just business". But when it turns from a dispute into a conflict, when it becomes a thrash, it's very personal, even if it's being driven by rules of group dynamics. A different group would treat the hypothetical Suomynope thrash in a different way, as it's driven by the conflict of personalities as well as the culture of the community. But that same group that thought Suomynope got only her fair due might consider changing conferencing software from unthreaded to threaded a major issue, where Suomynope and her peers might not care at all as long as they still had the same people to talk to at the same URL.
The other part of the strange attractor is when the conflict is not actually about what it seems to be about. Very often, what looks like a straightforward conflict about a simple issue is actually a hierarchial conflict. This contributes to the circular nature of thrashes, because no matter how successfully you resolve the issue at hand, if the true nature of the conflict is never addressed it will never go away.
Often, the actual root issue goes unrecognized. Equally often, it simply goes unmentioned because, for one reason or another, those that recognize it decide not to talk about it. If their silence is motivated by fear of retribution, the hierarchy is not supporting the community, and the hierarchy is allowing itself to damage the community, as well as guaranteeing the next iterations will be more severe. If their silence is motivated by a desire not to delve into the deeper issues at that time, it's neutral as far as the community is concerned -- the issue will come up again, inevitably, but may be easier to handle at that time, and the next few iterations may be equally or less stressful. If their silence is motivated by a knowledge that the deeper issue is unresolvable, and damage would be done by pursueing it, the impact is positive, and as the knowledge spreads that there's really nothing that can be done, further iterations are damped.
A goal for a Metacore group would be the conversion of strange attractors to limit cycles or fixed points, by identifying and attempting to resolve the deeper issues, ending the thrash not by judging winners and losers, but helping turn issues into authentic non-issues (as opposed to simply drawing a curtain over it). The Metacore group, if it functions this way, is an asset to a community, by ending thrashes and letting everyone else get on with the reasons they belong to the community in the first place.

1) For a great example of Lorenz's strange attractor, visit It's worth your time for its beauty and strangeness.2) There is no direct connection between Lorenz's Butterfly and the mythical storm-spawning Monarch, even though Lorenz's equation is a meteorological model. It's a beautiful coincidence, according to every credible source I've read.
3) The higher in the hierarchy, the less likely an admin will quit over a meta issue, of course.
4) It wouldn't have become a cliche if there wasn't some truth to it.
5) Mores are the rules followed by a community, often codified into law. Folkways are the unwritted principles, not as severely sanctioned if broken. A Neopoeian example: posting someone's meatspace address is verboten both by the formal rules and the culture. Calling your opponent in a debate a Nazi is a violation of Godwin's Law, is generally frowned upon, but is not specifically punishable unless it's accompanied by other rule violations. "According to the American sociologist William Graham Sumner , who coined the term, folkways are social conventions that are not considered to be of moral significance by members of the group (e.g., customary behaviour for use of the telephone). " -- "folkway" Encyclopædia Britannica
Article #11: Thrashing About About Thrashing
being a glance at recursion, thrash fatigue, and some notes toward possible solutions
One "feature" of online conflict is a relentless, numbing circularity. Easily solvable conflicts do not become thrashes, by definition, so a thrash is a conflict rarely resolved in a way that satisfies all parties, especially when the conflict crosses hierarchial lines or goes on across two or more communities, and the side that perceives itself as the losers are loathe to let it die. It's possible the winner will gloat and thus spark a new round, but generally it's those who think they're on the short end of the stick who keep thrashing.
For thrashing it is. Seemingly endless rounds of point-counterpoint-point-Jane, you ignorant slut, dragging in and debating points further and further away from the original dispute, fighting about how you're fighting, cutpaste word-by-word rebuttals. Furious flurries of post and riposte that seem to accomplish nothing, floundering in the murky water of Meta and crying "Damned be he that first says 'Hold, enough'!"
There are two primary reasons thrashes seem endless. One is that aggrieved parties will bring up their case as often as seems neccessary. The other is the addition of new people with similar grievances. If a community's policies (or the administration's way of carrying out those policies) are Meta prone, there will always be an influx of fresh thrash, leading to a recursive cycle of arguments with new cases and old being used as ammunition.
In your first few thrashes, it seems fresh and visceral and new and important. While it certainly is important, few if any Meta issues are brand new. Thrashes feed on new people with old problems. This can lead to a sort of thrash fatigue (not my term). "Oh, that again. Didn't we go through all this with wossname at The WELL, wasn't it all beaten to death by user:NobodyCaresAnymore at DotDash back in the 20th?"
There are three approaches to thrash fatigue. One is to ignore it, keep your meta-lance sharp and charge full tilt into any windmill which appears on the horizon. Another is to grow jaded and tune out thrashes as unimportant or unsolvable. For all but the most dedicated metaheads (and the person whose oxen have been freshly gored) this is an easy and appealing pseudo-solution. Filter out the unpleasantness and make your own way, after all, it's not your fight.
The third approach is the MetaCore. Metaheads love a good thrash, and what is a MetaCore group but a bunch of metaheads? The distinction between MetaCore and a bunch of thrash junkies is, MetaCore people don't start thrashes or thrash for the sake of thrashing. Instead of just lurching to the attack, or walking away, find solutions. Many of the more common sorts of conflicts can be resolved. Finding the root cause of the problem, rather than circling endlessly around the specific details, can help point to satisfactory resolution.
To illustrate a metacore approach, I'm going to pose a hypothetical situation at E2, in early 2001. {Suomynope} has several writeups deleted with no explanation by an anonymous editor. She complains about this in the chat module, and is silenced for a brief period. She comes up fighting every time, more and more combative, and ends up silenced for several hours. This leads her to post several anti-editor writeups, which are promptly nuked with XP loss, eventually costing Suomynope to temporarily lose a level. Some other users, seeing this happening, join in, some cautiously, some full-steam. They are also silenced in the chat module. These users find some of their own writeups being deleted. The editors doing the deletions claim that the writeups in question are weak or flawed and would not have stood the test of time (and, for purposes of this example, we'll assume this is true. I'm just looking at the mechanics of a possible thrash, not starting one). The users complain, claiming that it was a punitive measure, since the deletions may not have occurred at that time had they not called themselves to the administrative echelon's attention. At this point, the thrash can begin to spread, involving more users and spinning out of control for weeks (at approximately 6 hour intervals, as that's the approximate cycle time of chat module conversations), or a clampdown can ensue. Either way, neither side will be satisfied with the results.
From the perspective of Suomynope, it all started because she was singled out for unfair treatment. From the editorial perspective, it can easily be blamed on immature whining. From the perspective of the 5 or 50 new users who registered while this was going on, E2 is unfriendly to New Order, and they may well decide not to bother putting time and energy in.
Obviously from the metacore perspective, the problem was accountability. It was not the deletions or the whining, but the lack of accountability that started the spinning. Although this sort of thrash could still happen at E2, some accountability measures (such as a message bot informing users of deletions, which encourages editors to identify themselves and the reason for the deletion) have been put into place, making it less likely for this particular thrash to spin out of control or require draconian measures that would lead to the community voting with their feet.
Many of the thrashes at (and about) Utne Cafe are caused, in part, by a perception of closed ranks, by the idea that no matter what a host does, it will be backed unquestioningly by the administration, and by a perception of a double standard, where "ordinary" patrons are [hosted] for words and actions that are tolerated when said and done by members of the administration.
Two axioms: 1)Double standards are inevitably damaging to the health of the community, although it may not significantly affect the website's hits and revenues. 2)The perception of a double standard, if widely enough held, is nearly as damaging as a proven double standard.
Some thrashes are unsolvable because the participants have mutually incompatable terms. If a patron sincerely insists the only way to satisfy them is the resignation of the site admin, they cannot win, and the thrash will go on until everyone tunes it out. Although there may very well be unfairness and censorship going on, it's not realistic to expect someone to quit a paying or otherwise rewarding job over it. If the admin would be willing to step down over these sorts of issues, it's likely they would never have occurred.
Other thrashes will never end (although they may fade away) because the administration demands things the aggrieved party can or will not do. I have noted before the fist, the idiolect, the unique writing and posting style that can serve as an indicator of identity. If someone's style is considered disruptive (a very subjective judgement, especially if the system includes a filter or ignore user feature), and they are hosted for it, it's improbable the thrash will ever resolve. Your voice is your voice, and if you're kicked out of the party because you sound like Gilbert Gottfried or Fran Drescher, it's not likely you'll go in for speech therapy.
The "carrot and stick" system of administrative and built-in rewards and punishments (or, if you prefer more neutral language, positive and negative reinforcement) at E2 bear some examination here as a contrast with Motet and other similar systems, which have only sticks. Quantitative levels and reputations, C!ings and downvotes, blessings and cursings, allow positive reinforcement for patterns of behavior supported by the culture and the administration. Although it's not perfect (because no system is), it's more flexible and supple than a system like Motet, which has as its only formal carrot elevation by the administration to host status, and the informal carrot of favors from hosts such as images and banners in topic headers, or background colors or images. (One of the differences between Motet and eCore is Motet does not allow custom sitewide themes, but does allow hosts to customise the appearance of individual topics.)
I believe accountability of the administration to the community to be a key of thrash prevention and resolution. Transparency is another. One piece of evidence I have for this is an experimental conference I host at Neopoeia, called Plaza.ME. It's a zone with specific rules that give as much power to individuals as possible within the software. Each registered member of Neopoeia is allowed to have one topic, which they can set the appearance of; have open to all, open to some, or read-only; set the speech guidelines and hide or erase (with YOYOW protections built into the charter) others posts by contacting a caretaker host (an acknowledged sock puppet named "Igor") in a special administrative topic. The only opinion Igor holds is a commitment to service and free speech.
People have used their topics as intra- and inter-site archives, light chat, journals, political free-fire zones, semi-private personal conversation nooks, home bases, and angry soapboxes. Thrashes have been self-correcting and limited for the most part, with minimal hostly intervention. It's not a democracy (more a benevolent dictatorship), but so far it has been a successful experiment, in the sense of providing the maximum benefit to the largest group with the minimum unfairness.
A few disagreements are not bad for a virtual community. They serve to identify the weak spots, the places and policies that need fine-tuning to better serve the community. Continuous thrashes are not in themselves corrosive, but are an indicator that there may be a non-obvious problem within the structure of the community, and the only way to end the thrashing (as opposed to just stopping it, because it will come back) is to locate, define, and resolve (or prove the unresolvability of) the underlying issue.
Article #10: Visual Metaphors for Virtual Community
being a brief survey of ways to look at the places we live
As human beings, it can be hard to wrap our minds around this new medium. Understanding a communiy as a series of inputs and database calls (which, in the end, is the fundamental software basis for any virtual community more sophisticated than a chatroom) is difficult for most.
Many visual metaphors have sprung up to try to explain the structure and mechanics of virtual communities, and your choice of metaphor forms a filter that influences your experiance and perception.
An early metaphor that comes down to us from the days of MUDs is virtual community as a roleplaying game. It's not suprising that it's easy to draw an equivalency between sitting at a computer and competing (or cooperating) with other people and software entities for status, level, eXperience Points, powers, spells, and abilities under the aegis of an administrative Aesir, and sitting around a table rolling dice and exploring imaginary dungeons and towns with other people and non-player constructs controlled by a Dungeon Master. Many early MUDs were specifically dungeons, and the idea of magic is strong online. Since it really all takes place in our heads, why wouldn't magic work? At E2, [the Everything2 Role Playing Game] is a great (and witty) example. Role-playing is very common across many virtual communities -- there are legions of furries, and I play the role of a genetically engineered dolphin living in a mobile tank of filtered seawater in a couple of communities. People know I'm human, but it is fun and a good icebreaker.
To the lurker, the observer who chooses not to interact directly with the community, it can be a giant and intimate soap opera. People are creatures of habit and will ususally act in predictable ways, just like Erica Kane or Bo and Hope Brady will rarely venture far out of character for long. If you are a dedicated, long-term lurker, it's easy to see story arcs, lead and supporting characters, subplots, and even commercials in the form of spam and ad banners.
A subset of this, but more active, is virtual community as theater, especially absurdist theater. This can be a healthy way of keeping a sense of humor, but can be taken so far as to distance people from the real effects of their interaction with others. If you are playing a role, it's easy to assume others are too, when they may be totally sincere.
A common metaphor seen in larger communities is website-as-city, with busy sqaures, intimate coffeehouses, and soapbox-filled parks, connected by broad avenues of general interest and dark alleys of Meta. The vision of cities of various sizes and atmospheres connected by superhighways of the Internet is an appealing and supple one.
Another way to look at virtual communities is as a magazine or newspaper. This is especially apt for sites like Slashdot or E2 that have a large population of contributors, and a large group of non-participant observers. There are news articles, top stories, extensive op-ed pages, and cartoons/jokes. The community is managed by a cadre of editors who have significant input into the daily look, feel, and content of the site.
Another visual metaphor is a house or houses, with the administrators as owners. This is especially apt in discussions of hosting. A common argument is, "You have the right to free speech, but I have the right to kick you out if you kick my dog, spil red wine on my white carpet, and set the draperies on fire." Although this is certainly a valid argument, it has an equally valid counter-argument, "If you invite a group of essentially strangers into your home for a party, you have to expect a few spilled ashtrays and a fistfight or two. It's not a party until somebody gets punched." In a large community with a lot of vertical stratification (like Motet boards) this is quite useful, as several houses on the same cul-de-sac can be quite independent and discrete, but it's easy to walk from one house to another, and especially loud or violent parties can significantly impact their neighbors.
One very prevalent metaphor comes from history, and is frequently quoted in Meta issues. "The Tragedy of the Commons" is the principle that an area owned and used by all will be exploited and destroyed by a few. In relatively unmoderated communities, the "Wild Wests" of the Internet, this can be true. The September That Never Ended is a USENet metaphor for this phenomenon, the flooding of an existing community by unsophisticates with little or no committment to the health of the community. (It's no coincidence that Comittment and community are such similar words. They seem to be related in direct proportion.) Defenders of free speech often find themselves in conflict with the defenders of the commons. Too strong a defense of either side can be detrimental, leaving on one hand a site looking like Yasger's Farm the day after Woodstock, but on the other extreme, a prairie criss-crossed with barbed wire and guarded by a militia of riflemen, where only a select few are allowed to graze, docile and sheeplike, on nomimally public land that still bears the marks of bison tramping through.
Finally, it's possible to view a virtual community as an orchard, with nuts and fruits growing on topic trees, the rotten ones discarded by administrators who patrol the rows pruning unhealthy branches, the larger trees cutting off light and resources from the smaller ones. "There is unrest in the Forest/There is trouble with the trees/For the Maples want more sunlight/And the Oaks ignore their pleas..." -- from Neil Peart's "The Trees"
There's a similarity to the Commons issue and conflicts here, with Metaheads seeing themselves as checks and balances on chainsaw-happy, overzealous pruners, and admins seeing argumentative strangler figs choking off conversations with off-topic chatter and flames.
(There is another metaphor that I have been introduced to recently that is supple, robust, and very applicable to virtual community. This is web community as a feudal society, with owners and admins as lords of the manor, and end users as serfs. In order to fully appreciate this metaphor, you must remember that peasant was once an honorable profession, perhaps the most honored, and noblesse oblige must apply for a feudal community to be healthy.)
No one of these visual metaphors is universally applicable, but these are common ones, and easily accessable.
Article #9: Anonymity and Secrecy in Open and Closed Virtual Communities

being thoughts on anonymity, with a discussion on open and closed communities, and closed areas inside open communities
In order to have a community, you have to know who you are talking to. One of the great differences between previous paradigms and the Internet Era is the ability to have a solid, consistent level of anonymity and continuous, meaningful social interaction. The only absolutely vital identifier in a virtual community is an individual, unique, secure login (or series of logins). A username like [eponymous] tells nothing more or less than the name Kevin Smith. The information that someone gives about themselves can help fix an identity, but --
Axiom: Online, nothing a person says about themselves can be taken at face value without independent verification. This is just as true in meatspace, of course, but not as obvious. If you meet a vaguely familiar-looking stocky bearded man in a black trenchcoat hanging around outside a convenience store and he says "Hi, I'm Kevin", he may be pulling your leg. The presence of a camera crew and Ben Affleck provides external verification. If you meet [eponymous] in a virtual community, the only way to prove that it's the same Epo is to develop trust in the identity.
One of the primary ties to an ID is email. If user:KSmith has an email address of, he's much less reliably identified than If Epo's email is known to you, it's not unreasonable to have doubts that 1337Hax0r@hotmail is the same person (although it's certainly not impossible, as multiple web-based email addresses is not uncommon, especially among people who value their anonymity. I have five or six valid email accounts with nothing in common save the person reading them, and I doubt that's unique.)
Another ID is syntax, punctuation, and writing style over time. An analogy is the "fist" of cold-war era clandestine shortwave radio operators -- an individual could be identified by the pattern and cadence of their keypunches, and it's reported that the fist could be identified even if the operator was aware of this and tried to change it by switching hands or other alterations. Likewise, a person who consistently uses or avoids certain words -- or has a special affinity for an unusual punctuation or nested lists or germanic sentence structure -- more easily identifiable is. A known entity with a very distinctive style (a great example is a poster I know who always uses short lines, no punctuation, and ends all posts with the word "out" alone on a line, usually with a humorous or ironic twist) is easily spoofed in the short term -- but only a genius could do it for six months without a slip. Anthony refers to this as idiolect, the formal term for the individual dialect we all have -- the concept that no two speakers of a language actually speak the same exact language. He also pointed out a meatspace example of this, the computer analysis of _Primary Colors_ which linked the book to its previously anonymous author. Thanks, Anthony!
Each community will have a different view of anonymity and its acceptability within the community. Many sites do not allow people to post personal identifying information about others. TrueNames, phone numbers, home addresses, I.P. numbers, children's names and schools are verboten, and for good reason. Interestingly, except for I.P. number, in meatspace this is exactly the sort of information you are frequently asked for in relatively casual conversation, and in smaller communities, demurral results in strange looks and people comparing your face with pictures on the post office wall.
Anonymity in virtual community is a continuum. At one end are closed, highly regulated communities where people are required to use real identities, through looser sites where your words are your credentials, to communities where the right to anonymity is a basic premise and founding philosophy. Neopoeia is at one end of the spectrum, where the owner and administrator is so committed to anonymity that no one on the site knows his TrueName. Even people he trusts, when they need to contact him outside the board, get a web-based email, a P.O. Box, or a cell phone. This does not decrease, for the most part, other's trust in him, because the only thing he demands of others is a respect for his (and other's) anonymity and one login per person (this rule does have a certain amount of flex, as there are a few acknowledged sock puppets, with specific administrative roles and known puppeteers. They are transparent.)
Open, when discussing virtual community, has three meanings -- welcoming, accessible, and transparent. In the second (and simplest) meaning, a community can be viewed without registration, and has a simple registration procedure. Cafe Utne and E2 are examples of this. Less accessible are places like Neopoeia, where the only way in is to email the admin or have someone inside request your registration. I'm certain there are closed communities out there where only the invited are accepted.
Closed areas exist in most virtual communities, and have many purposes. One of the most common is an administrative area, such as E2's Valhalla chatroom or Cafe Utne's Hosts-Only conference. These can range from sleepy, functional areas to lively, smoke-filled backroom discussions. One of the common (but not universal) requirements for participarion in these sorts of admin zones is a vow of secrecy, a "What happens in here, stays in here" ethos, a formal or tacit agreement not to air the family's dirty laundry in public. As a host at Neopoeia, I honor that committment, but I can say this about it -- the subcommunity of software-empowered hosts is not that different from the community of Neopoeia at large. This is, I believe, not unusual in a reasonably egalitarian culture, but it may be at least partially a function of the egalitarian ethos of Neopoeia itself.
Other closed administrative cultures can gather reputations less savory, deserved or otherwise. Hypothesis: The perception of the administrative culture will be based on the worst actions of a minority of admins. The higher in the hierarchy the person offended is, the worse the offense will be considered to be. Since people are unable to see behind the administrative veil, the visible actions of the admins will be watched closely. [Anthony] has suggested there is an iceberg effect, where people will multiply public capricious-appearing actions by a sort of "skullduggery coefficient" to get an idea of how prevalent secret machinations are.
I can only directly report on one open administrative area. It was in a subcommunity of Utne Cafe, conference MetaMatters, and it was, in part, an experiment in online democracy, a "patron-run space". It failed, for several reasons. One was an inability to come up with a consensus for how to manage the conference. Another was the structure of Motet has no support for any sort of polling or voting, making democracy a technical challenge, with multiple cut-and-paste ballots on competing issues going on simultaneously. One of the problems with voting was the inability to verify identities, which led to the possibility of sock puppets voting and real people being accused of being puppets. I spent significant time and energy in my first few months of MM defending myself as an individual and not someone else's sock puppet. One patron (who has since been booted from Cafe Utne) wrote scripts to simplify voting, but they never caught on. Aside: The booting raises the point that people with value to the community who are idiosyncratic, creative, and outspoken are often booted when the criteria for booting are too subjective. Whether the administrative view that this poster was too disruptive, abrasive, and voluble is true or not, it doesn't mean that they were not valuable or legitimate members of the community the website purports to promote.
The greatest strength of MetaMatters was the fishbowl. Since the subcommunity was administered by a rotating group of three patrons, there was no closed area -- simply a topic, universally readable, writeable only by the "unhosts", with a companion "gallery" topic open to all, and another gallery that was not supposed to be posted in by the unhosts. (Neopoeia has a similar area, a conference that no hosts are permitted to read. I know nothing about it, other than its existence.) The fishbowl increased trust in the unhosts. Although it would have been possible to conspire against MM patrons through email and other backchannels, it was less likely than if a private space had ben provided for it.
Axioms: If an admin makes any action against a community member, someone will think it's an abuse of power. If two or more admins make the same action against a community member, someone will think it's a conspiracy. As long as there is any possibility that the action is an abuse of power, a conspiracy cannot be disproved. Without access to the closed administrative areas (or a credible confession), a conspiracy cannot be proved.
Closed comunities are difficult to know about, by definition.
Howard Rheingold is the acknowledged expert on virtual community, to the layman. Although his community BrainStorms specifically prohibits any sort of research, I do not believe that prohibits a passing secondhand opinion. I have heard it is very quiet in there. That does not mean there isn't a vital, close-knit community inside the gates. It does mean that to the outsider, it is closed in all three senses -- not welcoming, inaccessable, and opaque. Since Rheingold is a researcher, it is my belief at least part of the prohibition on research is a desire to prevent the Worm OroboroUs effect.
Cafe Utne, Neopoeia, and E2 have non-administrative closed areas as well. Utne allows anyone interested to purchase one, for a small fee. For practical economic reasons these are rarely open to the general public. In MetaMatters, there was discussion of having free, open subconferences, but the discussion never led anywhere, and was mooted whan paid subs became avalable. A dedicated conspiracy theorist could claim the interest in and discussion of subconferences led Utne to a new source of revenue, but, axiomatically, this can neither be proven nor disproven. Neopoeian private conferences are free, and assigned at the discretion of the administrator. His policy is based on compelling argument rather than profit. I have full or partial ownership of several. One is an archive, and another is simply a private space I use for conversations I prefer to have there -- I have used it to help mentor an Everythingian who was more comfortable with Motet software than eCore, for example, and used it to set up and preview these topics. E2 offers private chatspaces as a reward for leveling up in the formal hierarchy.
Closed, non-administrative areas have different impacts on the life of a virtual community. A common closed area is gender-designated. Utne and Neopoeia have women-only conferences, for example, and Utne has a mens-only conference as well. At the other end of the spectrum, I have reports of a private area at one community that was used "conspiratorially", in an attempt to influence a member or members of another community. Since the source is someone I consider credible, I cannot believe it is the only time it's ever happened online.
Backchannel is another form of closed area, and many believe it's where conspiracies happen. If there is a private mailing list among some members of a community, and it is used to build or justify a case to boot someone from that community, it is, by definition, a conspiracy, regardless of its success or the validity of the complaints of the conspirators. Of course, if three friends in a car ditch the fourth and go to another party instead of going home as they claimed, technically that's also a conspiracy. The legal definition of conspiracy requires the contemplation or execution of an illegal act (In a case during the Seattle WTO riots, a man was charged with "felony conspiracy to commit a misdemeanor" -- try to wrap your mind around that one for a while.) and booting is not against any laws. People planning to rob a bank is a conspiracy, people planning to throw someone out of a party is not, whether it happens in public or private.
Actions planned in private, however justified, can appear capricious or unjustified, especially administrative actions. This can have significant effects on the health of a virtual community.

Disclaimer: Of all the virtual communities I have observed or been a member of, I have been there because I want to be there, not for research purposes. Anything I have learned has been incidental, from experience. Although I am finding the birth of C2's community fascinating, I'm not here to study it, but to be a part of it, and hopefully to do what I can to improve it. Improving it does not mean forcing it into some theory-driven mold, some image of a perfect community that lives only in my head. It is my belief such attempts can be deadly to community unless it is a shared vision.
Article #8: The Dilution of Censorship

Being an overview of censorship and accusations of censorship in online communites, and a meditation on what freedom of the press means on the Internet
Axiom: One of the largest sources of conflict between the power structure of online communities and the registered members who make them up is the perception of censorship. This varies in direct proportion to the stated "Free Speech" policies of the community.
At Bare Knuckles Politics, for example, it's clearly stated that ideas and opinions counter to the site's limited political purpose will be deleted. This means that few, if any, will complain about censorship, and even if they do, their voices will not be heard. It's similar to partisan newspapers of the 19th century. If a journalist's story does not have the editorial slant required, the reporter simply goes to the presses next door.
At sites where a Freedom Of Speech ethic is in place, the complaints of censorship will increase (on a per-case basis, not necessarily in raw numbers). The stronger the freedom of speech at, the more likely any editorial action is going to be seen as abrogation of rights. For example, the ongoing thrashes of Cafe Utne are in most cases due to hosts or the administrators hiding and deleting posts, closing topics (making them read-only), and removing user's ability to post. The reason this garners such a vehement reaction by the "offender" (and frequently by a whole crew of past bootees, metahistorians, and interested onlookers, both at the Cafe and another Motet) is not the simple act of removing one or more person's voice. It's removing their voice from a website that is attached to and sponsored by a magazine that champions the freedom of speech and alternative viewpoints. In Neopoeian vernacular, deleting and booting is called "hosting" or "being hosted". It's a useful turn of phrase, based on the mid-level administrators being termed "conference hosts" on Motet systems, and I use it interchangably with "deletion of words and/or account revocation".
I bring up the newspapers of the 1800's to drive home an important point about censorship -- every website with even a hope of developing a virtual community is going to have to host someone, at some time, for some reason. However, it's not censorship in a legal sense (although it can be nasty, partisan, personal, and ugly). Freedom of speech is freedom of the press, and anyone who can scrape up the money and the hardware to put together a website is the owner of the press. Deleting a person's post from a website, or even locking down their account, is not censorship. The web is a supersized Wal-Mart of soapboxes, and if you can't find one there that fits you, there's lumber and paint out back in the USEnet aisle. The web is a million partisan yellowsheets, and the difference between an ideologue and a muckraker is in the eyes of the beholder.
This is, of course, according to the legal definition of censorship. There is a grand scale of difference between throwing someone in jail for publishing an opinion and deleting a post from a website that is owned by a private corporation or an individual.
Legally, what's frequently called censorship online isn't.
But, and it's an important but, the effects of hosting on a virtual community are very similar to the effects of censorship in meatspace. It is counter-community, if the motives are capricious or the justification unclear. In the event of a booting, where the user's words may remain on the site but are uneditable by the bootee, it represents both disenfranchisement and denial of ownership (YOYOW). At least one major Utne thrash centered around a booted user's inability to selectively delete their historic posts.
Wal-Mart refuses to sell music that it deems offensive. Although this is not censorship in the legal sense, it does deprive the artists of a major market and prevents them from reaching a significant segment of the population, and prevents that target market from ready access to the material. If a private site like Rheingold's Brainstorms deletes your words or revokes your membership, you have lost access to a small, if influential, audience. If a public site deletes your words or revokes your membership, you have lost access to a larger market, especially if the site is indexed by search engines. The size and vitality of the community you are presenting your words in lends cachet to your work, and this is lost if they are deleted. An identical article at Slashdot or at will reach a different audience and have a different (and likely much smaller, at the second) impact. Although this does not fit legal definitions of censorship, there is no other word in current use to describe the phenomenon.
I have mentioned before that a dichotomy between a community's stated or understood philosophy and the actual actions of the administration is a key source of online conflict. Conflicts over censorship, the appearance of censorship, and charges of censorship are the primary source of major hierarchial conflict in virtual communities. (Flamewars between members on the same hierarchial level is a different subject altogether. Hosting is not subject to Godwin's Law, but it certainly as been known to lead to applications of it.) The effect of on the virtual community varies, but the following general observations apply.
  • The higher the hosted is in the formal or informal hierarchy, the larger the ripples. Booting an account created for spamming, or someone who has not been around long enough to have garnered a lot of respect, or who brings with them a bad reputation, or booting a secondary "sockpuppet" account when it's against policy, will cause little if any disruption of the community. Booting someone who has been around for years, or who has achieved a high enough level to have a significant following, or who has a good reputation in the community, without clear cause, will strain the relationships that make up a VC.
  • Although it is unlikely that one or even a series of bootings will cause a site to go out of operation, the booting of anyone who is a genuine member of the community will cause a change in the culture of that community. Frequently, this is the intent of the administration, a warning to others not to commit the offenses that led to the booting. However, the repercussions of booting are often not considered or unpredictable. The ghosts of famous bootees haunt sites for years.
  • Changes in the culture are directly related to the reason for the hosting, but not always in the direction intended. The amount of the precession is directly proportional to the appearance of caprice. If someone is hosted for using foul language and linking to porn, people will be less likely to use the same language and links. If someone is hosted for making comments contrary to management, this may lead people to be kinder to the administration. It is far more likely to lead to the formation of a vocal opposition group (for example, the Meta Brigade at Utne), self-censorship (in that people, afraid to cross an invisible line, may not only avoid knocking the upper levels, but avoid any discussion of the event at all), or take other actions (see The Worm OroboroUs for a partial list), especially in a site where management has claimed to be open.
  • Actions in one community can affect others, and it depends on the overlap between membership. In the world of virtual community, there is overlap between communities. In meatspace, few people are members of several communities. The virtual world is much smaller. The meta conference at Neopoeia, for example, is frequented almost exclusively by current or ex-members of Cafe Utne and SpeakEasy.
  • In nearly all cases, those accused of censorship have the best motives. In a long-standing community with a strong enough Free Speech ethic, "evil administrators" are vanishingly rare, because easily verifiable cases will lead to the community either not surviving, the administrator leaving, or the stated ethic changing. This does not mean that well-intentioned administrators never make mistakes. It's the way the administration reacts when mistakes are pointed out that can damp or intensify the thrash.
  • Deletions and bootings to prevent copyright violation or obvious trolling contribute to the health of the community. Hosting, as a reaction to derogatory comments directed at management, do not, and may harm the community. There is a time and place for criticising management, but there is no reason for it to be kept secret. If the criticism is clearly unwarranted, it will be obvious to all members of the community. Alex Delacruse, pseudonymous Admin of Neopoiea. -- "Calling hosts idiots and the management style flawed should be the last reason people get hosted. If people in positions of authority can't tolerate that much, their egos are far to fragile to trust with oversight. If the culture of authority actively supports fragile egos, [the culture is flawed]."
  • The long-term chilling effect on the community when popular or well-known members are hosted capriciously cannot be overestimated. It is the easiest way to put skeletons in the closet of a virtual community. The effects of any administrative action in this medium are magnified. "The tighter the lid on the box, the more people try to pry it up. And the more people try, the more nails the admin hammers in." -- Mittens, Neopoeia, re SpeakEasy.
In meatspace, if an authority figure tells you to shut up, you are likely to grumble, but not shout, unless you feel deeply about the issue. If they tell you to leave, you will not usually need to be maced and led out in cuffs. But in a virtual community, the mace and cuffs are easily accessed, because of the perception that nothing that happens here can cause permanent or even temporary harm.It is true that being censored in or booted from a VC causes no physical harm, and usually no financial harm. Psychological harm is well outside my ability to speculate on, as it is an individual issue, and my concern is groups.
Censoring or booting for anything other than clear violations of stated policy alter the social fabric of the community, and this effect is greatly magnified when the policies are not clear, or overly dependent on subjective judgement. One of the functions of a MetaCore group is to determine the level of justification for censorship, and to help clarify policies, when possible. Since metacore groups are not a formal part of a virtual community's administration, their recommendations and suggestions are not binding policy. However, a fair and reasonably non-partisan metacore group is an asset to a website that values the virtual community that exists there. Axiom: MetaCore groups will often have the appearance of an anti-administrative slant, because they are a balancing factor between the administrative (software empowered) and non-administrative levels of the hierarchy. There is a difference between a metacore group and "a bunch of whingers", and it is this -- metacore groups are motivated by the health of the community, and justice for individuals is a significant part of this goal, so the identity of the individual is not as important as the acual facts of the case. If a metacore group only defends speech and individuals it approves of, it is not a true metacore group. The litmus test is this: "I do not like what you say. I will defend your right to say it."
Hate speech is both the murkiest metacore issue and the most likely to lead to damage to the community, and it will be discussed in a seperate article.
I am certainly guilty of diluting the term censorship in this article. It seems to be impossible to avoid, as I do not feel I'm in a position to invent a new term.
Article #7: Hierarchy of Peers

All web communities have some sort of user verification and some sort of hierarchy among the users. Most of the sites I've participated in have both a formal and an informal hierarchy, a pecking order of sorts. I'm comparing, contrasting, and discussing a variety of systems and their meta implications, not advocating a particular system or suggesting a direction for any of the communities I'm looking at.
The formal hierarchy is easy to discern, as it will be documented as fully as possible, for the utility of all. Informal hierarchies are harder to discern and, in some cases, can be impossible to pin down with any degree of accuracy.
[Community2]'s system is not yet fixed as far as I know, but one of the suggestions I've seen is for a four-level hierarchy consisting of unlogged guests with read access, provisional users with write access, peer-verified users with full write and vote access, and coder/administrators with behind-the-scenes access. The public finger utility sorts alphabetically, eliminating some of the obvious hierarchy. The ethics and aesthetics of the site appear to put coders and administrators on the same tier as verified users. Within the ranks of verified users, things are designed to be as level as possible, with no specific official levels or ranks. The formal policies, culture, and informal hierarchies are as yet unformed, I simply bring it up as a point of comparison.
The magic mountains of [|Pyroto] are entirely different. A rigid hierarchial system that requires no peer validation (although there is a rule against multiple IDs), with a sharp level gradient. From the front page -- Pyroto mountain is a web site interactive game where people log in under a "wizard" identity, answer increasingly difficult trivia questions for power and status, post messages to each other, and occasionally wield "spells" that can increase or decrease others' power and status. These systems are one of the primary reasons for the prohibitions against multiple IDs, because a user with many names could use them to boost their own abilities and level, thus distorting the game aspect in ways that can adversely affect other users. Pyroto as a community is a cross between a [MUD] and message board. In order to gain levels, you have to answer trivia questions -- but you must also make a certain number of posts to the message boards. The actual content of the posts is not measured, but people who can't find anything to talk about wash out. The levels are complex, with low level users starting in a small part of the site but able to expand their view to other "mountains" by use of mana (which regenerates over time) and spells (which are gained by leveling). The people at the highest levels don't seem to notice the teeming masses below them, in part because the message boards are keyed to level. This encourages the bulk of your conversation to be with people at or near your own level, unless you choose to drop down to a lower level board. This formal horizontal stratification means that you're unlikely to meet high rankers, although you can message them if you wish. But why? You haven't had the chance to get to know them, and there is little reason to msg someone unless you know them (or at least something about them, their work. Contact between higher and lower ranks is most likely to be initiated from above.) The formal hierarchy is so rigid and defined that I was unable to determine any sort of informal status-among-peers, although I never operated on the highest levels. It's possible that there is some form of status at the tops of the mountains, but unprovable without getting there yourself.
The level system of [Everything2], based on a combination of number of posts and numerical reputation (XP, based primarily but not entirely on peer voting), is less stratified than the Pyroto system. Higher level users receive software perks such as supervotes (Chings), bio images, ability to create chatspaces and lurk without displaying their login in the public finger utility (which is sorted by XP). Since all users share the same public space, all members of the hierarchy are accessable to lower-ranking members, for example, it's entirely possible for a level 2 user to get to know and interchange messages with a level 10 user. Levels are not the only system in place, as levels have a voluntary aspect -- any high-level user wishing to lower their rank has an easy method to do so by voluntarily giving up reputation. Another, less formal hierarchial system is [User Since:]. This is also not rigid, as it's not unknown for a user to leave and come back under a different name.1 The informal hierarchy-among-peers seems to be pretty egalitarian and based on the same information that the levels are based on, with the addition of pleasantness/politeness, accessability, and membership of subgroups, which are formal, and opt-in rather than opt-out type groups2. Some user subgoups at E2 include a couple of "database improvement" teams, parents, GLBT, and a variety of "expert" groups (video games, for example). The level system discourages multiple logins, because to gain enough rank in more than one login to have a substanive effect on a user's standing, they would be able to achieve that standing anyway. Multiple logins do exist, but generally for benign purposes (some people have multiple logs that they use for chatspace fun, I have access to two logins but one never nodes, etc), identity obfuscation or changes (often with administrative approval), and (rarely succesful) trolling.
On Motet systems, there is no formal hierarchy other than Admin, Host, and User (or Poster). Since the status of posters has no formal system at all, any conclusions about the hierarchy must be tempered with IMHO and YMMV. Since the level of a poster is informal and subjective, it is unquantifiable. On moderated boards multiple logins are generally not allowed except in special cases, where the administration openly or tacitly approves of "sock puppetry". Boards with auto-registration's admins can check IP logs and determine most, if not all, cases of multiple IDs. Within the User rank the hierarchial elements include longevity on the system and, more subtly, longevity and rank on related systems, although since it's informal, is inconsistent for example, some users view rank on their system inversely to rank on another, so an admin at another board may well be considered anathema, whereas someone booted from that board has instant status. Another measure is facility. If someone is unable to communicate coherently (due to spelling errors and misuse of words indicating that they have little or no knowledge of vocabulary and grammar, and never seems to add anything to conversation, they are liable to be discounted. Volubility is another -- an unusually voluble poster may be discounted because they may be seen as watering down their credibility, and someone who rarely posts may also find themselves in a "what the hell do you know" category. An important note on this, however. Some who post in high volume, if they are good writers with interesting views, are likely to be respected, as are people who post rarely but whose few posts are pithy, witty, and well-written.
Another part of the informal hierarchy are informal usergroups, sometimes called cliques. Some of these groups are indeed cliques, tightly knit exclusive groups with a united front aspect. They tend to "take over" a conference (or a gropup of topics within a conference) and allow little or no other participation, or do everything thay can to run off people they don't accord status to. An example of a clique that isn't really a clique is the group of users who frequent a specific conference. The Motet system compartmentalizes by interest rather than rank, so people who are interested in free-for-all discussions of politics may never cross paths with a group who are interested in writing and commenting on poetry. The thing that makes these groups not fit the definition of a clique is they are willing to accept new people in their conferences, regardless of their opinions, with no eye to rank. The status of a poster who may be held in high esteem in Politics doesn't matter to the people in TextileArts, and vice versa. It's possible for two people to be bitter opponents in one topic and be cordial and deferential in another. In the Motet hierarchy, the board as a whole is a common pool (like the eCore Everything) but there are vertical strata. A host in one conference may be regarded as just another poster in another. Also, users can use the software to ignore another user or a complete topic. Someone who is being "filtered" by a large number of people tends to have a lower level of overall status, although that is not 100%.
The degree of stratification is a function of the administration, who determine exactly what conferences/major divisions exist. Also, on a per-conference level, the hosts affect the status of the posters. A popular host can elevate or denegrate the status of a poster by their reactions and attitude towards them. An unpopular host can do the same, although their impact tends to be inverted, as someone censored by a host generally seen as an autocrat tends to gain status among those who consider themselves egalitarians.
On a Motet-style board with no moderation whatsoever, the hierarchy among equals skews. Voluble posters with firm opinions are seen as stronger and louder, and the loudest voices can dominate a conference, disrupting any conversations they have no stake in nurturing. Also, if there is no user verification by the administration, people can log in under as many names as they can supply email addresses for, and literally have conversations with themselves. A skillful puppeteer can be highly respected, whereas a clumsy one can be seen as a low-status nuisance, or vice versa. The lack of a formal hierarchy, either admin or peer-driven, removes any objective measure of status and leaves any attempt to measure user status murky and incomplete.

    On the strengths of my observations, I'd like to set down drafts of some laws that virtual communities seem to follow. As always, these are my opinions, and I'm happy to discuss these or alternate views. I do not claim perfect objectivity nor omniscience.
  • Virtual communities have a mechanism to somehow verify and classify members, whether by administrative means (from hand-registration to coded level systems). A website with no mechanism whatsoever to tie a person to a login is unlikely to develop genuine community on a sitewide level.
  • Virtual communities have formal and informal hierarchies among the non-software-empowered members. The more comprehensive and rigid the formal hierarchy, the less important the informal hierarchy.
  • An important component of the informal hierarchy is the longevity of the user on a site. If you have been somewhere for five years, you will be accorded a higher status than someone who has been there five months. It is more likely that more people will listen to what you have to say, although that depends on your reputation among your peers.
  • Personal reputation is part of the informal hierarchy. Someone perceived as honest will have a higher status than a proven liar, if honesty is esteemed on the site. In a culture of tellers of tall tales, the accomplished liar may be more respected than the boring truthspinner.
  • Status in a virtual community does not have to equal power. The actual power will always reside with the administration. The administration may give power to high-status members by granting perks to high-level users, or inviting them to leave the ranks of the users and become members of the administration.
There is, of course, application of these laws to meatspace groups. It can be clearer in virtuality, because in meatspace, you can't review the code.
1) personal note: I've seen this happen, where I offered unsolicited advice to a newbie who seemed to be floundering, and promptly got my head bitten off for assuming that a user since:this afternoon was actually new to the system, when actually they claimed to have been there longer than my 2.5 years. My only defense is the writeup needed work, and as a mentor in the E2 system, I felt I was trying to help a new user succeed, not criticising someone above me in the hierarchy. If I had known the noder had been there that long, I would not have offered to help, I would have simply downvoted and moved on, because according to the systems and culture of E2, a higher-ranked user "ought to know better" than short, poorly written nodes.2)Opt-in vs Opt-out -- Due to the nature of the internet and virtual community, I am unaware of any opt-out groups (that you are placed in without your conscious action, but you can leave on request without leaving the community entirely without having to satisfy any prerequisites) unless you wish to count administrative groups such as Conference Hosts or Editors. Since those groups have a higher measure of power than the general populace, they aren't really relevant to this particular discussion, although they are a keystone of MetaCore  {Facebook now offers opt-out groups by default.  They suck.}
Article #6: Chagnon, Heisenberg, and Metacore  {Note: some of this article has been hotly refuted because of problematic aspects of Chagnon's research that I was unaware of at the time I wrote the article.  Assume that at the time, I was unaware of the true scope of the problems around him.)

Being both a caution to scientists and a statement that I am not a scientist, merely opinionated
You cannot understand the level of community at a website without being a part of that community. (If your information comes from members of that community, in formal or informal interviews, than your information is skewed by your source's interpretations that they formed as part of the community.)
Napoleon Chagnon's groundbreaking study of the Yanomamo Indians, fabled in story and song as the Fierce People of the Amazon, is a seminal anthropological work. It has become a source work for others, such as Marvin Harris's Cows, Pigs, Wars, and Witches, which has been standard fodder in some college-level anthropology and sociology courses. In recent years, it has been partially discredited in some circles, as it's coming out now that the people being studied merely acted as they thought the anthropologist wanted them to. The Yanomamo will never be free of the baggage of Chagnon.) Note: Chagnon has not been totally discredited, and many of the most egregious charges against him have been proven unfounded. However, what I am concerned with here are not deliberate mishandlings (such as the accusations of biowar against the Yanomamo or the "genetic agression" theory), but the Hiesenberg effect, as applied to the soft sciences, that observation of an event changes that event. Thanks to Jeremy for making me reexamine and clarify this point.
Trying to "study" an online community as an outsider will always fail, because: the interactions you have with the community will make you a part of it, compromising your objectivity; if you are a true outsider, a dedicated lurker, the community will have depths you will be unable to understand (examples, you can't lurk at an E2 Nodermeet or a WELL face-to-face); you are going to seriously piss people off.
Real people do not like to be research subjects without their informed consent. Internet communication is sometimes more intimate than meatspace relationships for a lot of reasons, including anonymity and a presumption of safety (For example, I'd be very hesitant to hang a rainbow flag with a pink triangle outside my apartment door, as it's my place of business, and I could suffer anything from no ill effects at all to an egging to being fired or (until Lawrence V. Texas, anyway) prosecuted. I have no such qualms about participating in a GLBT usergroup or website or openly posting it on my bio, because I am effectively anonymous in my name, specific location, and gender. Anyone with sufficient time and commitment could find out my real name, adress, and phone number. In meatspace, it would take less effort to ask a couple questions and divine the answer by my body language and dodging of the question.).  {This has changed, as at the time I wrote the article I was not yet out of the closet in meatspace.  I've been flying my bi pride flag high since 2009.}
Commenting on community, how it develops and grows or balks and mutates, is an art, not a science. There is no quantitative analysis possible, not computer model you can plug all the elements into and predict if it will grow and thrive, or what form the community will take. The most atractive, supple, and well designed site with the best intentions can become a Levittown of tiny parcels "Where the Lowells speak only to Cabots And the Cabots speak only to God", or a community can develop in spite of the intentions and actions of the administration. The San Fransisco Gate ran an essentially unmoderated Motet site for years. It became, undeniably, a community -- until the newspaper decided it was too expensive to maintain, and shut the entire site down. It did not destroy the community, merely fragmented it. Many members went to other sites with the same software, joined other communities, while bringing with them the friends, attitudes, habits, and customs that made the Gate a special experience for them while it existed. The Gaters are a community-in-exile, a (dare I say) metacommunity.
You cannot understand a community without being a part of it. One of the features of eCore websites is a real-time chat utility with messaging. This chat, open to everyone, is one of the carriers of memes, ideas, and relationships that can only be experienced by being there. You can read the archives and get an idea of events, but a lot of subtlety is lost. When there is a lull of minutes or hours, it's not felt, even if it's recorded. It's ephemeral in actuality, but not in effect. The real-time chat is a community builder, a place where people can interact in real time, without restrictions of topicality, where relationships can start and be nourished.
Another community builder is backchannel. Private messages, private topics, private chatrooms, telephone calls, snailmail letters, all are places where the individual relationships that both form the seed of and can be the ultimate expression of community happen1. In most cases, the upper levels of the hierarchy of a site have such private spaces designated. On a site where community is encouraged or just happens, the administration forms a de facto subcommunity that may or may not have anything to do with the expressions of power of this group.
When reading topics (on a non-threaded system) or threads, there is a difference between reviewing them and being there as they happen. The best way to show this is with a diagram. For the purposes of argument, A is a topic about foo, B is foo', C is commentary about foo, D is commentary about foo', and E is a metathread talking about the four previous dicussions. This is by no means an unusual event.

topic/thread     A          B         C          D           E
day v  post >
Mon             1,2         1         1         1,2,3        
Tue             3,4         2        2,3        4,5    
Wed           5,6,7,8,9     3         4         6,7,8        1
Thu              10       4,5,6                  9         2,3,4
Fri          11,12,13,14  7,8,9,10    5        10,11,12      5
Looking at this the way that you would see it if you left the site on Sunday Looking at this the way that you would see it if you left the site on Sunday and returned Friday evening, if you accessed these threads in this order, you would read A, B, C, D, E. If these are all topics or threads on a related subject, part of one multi-thread discussion, you're going to see five parallel conversations. But, if you are reading this a few times a day, you might see the posts appearing something like this --
topic/thread     A          B         C            D              E
day v  post >
Mon             1,2         3          7         4,5,6        
Tue             8,10        9        11,12      13,14,15    
Wed             21-26       19         16       17,18,20        27
Thu              28        30,33,35                29        31,33,34
Fri         36,43,44,47    37,38,39    40       41,42,46        45
Without trying to draw conclusions, it's pretty obvious that not only are the conversations influencing each other, but they are really one big, sprawling conversation. You are more aware of the back-and-forth, and you see immediately how a comment made in topic E affects the activity in topic A (just an example. You can't prove it by this diagram, I made it up, but it's representative of how things seem to happen.) Reading it as it happens, firing off messages and comments in chatspace and backchannels such as email, telephone, and private spaces, it's obvious. Reading it more asychronously, the tone and tenor of the conversation seems different. Humans are built for synchronous conversations. In asynchronous communication environments, they will develop conscious or unconscious simulations of synchronous communications. This is related to but not the same as the problem of interpreting posts and messages differently due to a lack of body language, vocal cadence, and tone of voice, the inadequacy of emoticons to convey the subliminals we depend on.The community is much more clearly seen in the daily back and forth, give and take, than in the sterile environment of last month's posts.
The difference between a member of the community doing some analysis and an external sociologist type trying to dissect the community to see how it works is the difference between a journalist recording the current events in Mayberry and Chagnon's work among the Yanomamo, which not only presented an inaccurate view of a culture, but may have damaged that same culture (by causing people to act differntly, handing out axes and canned foods which altered the traditional food-gathering and cultivation systems, etc). The local writer may or may not cause change in the community. It depends on their goal. They can bring down a corrupt justice-o-the-peace, or simply give everyone in town something to chuckle over at Floyd's Barber Shop.

1 -- Thank you to Ben, Anthony, Oolong, and SEF (among many others in different places) who helped me realize this simple truth that had been eluding me.
A Note -- this article has been refuted, quite well, in msg and in discussion on other sites. I plan to offer a rebuttal in another Metacore, Schroedinger, and Harris article, unless someone beats me to it.
Article #5, Quis Custodiet Ipsos Custodies?

Who watches the watchmen? -- Juvenal, Satires. In any hierarchial system, this is an, if not the crucial question. It has implications and applications online and in meatspace. On an international level, the 2003 conflict between the Pax Americana doctrine and the United Nations over the right of unilateral, pre-emptive wars (or, as they were called back in the height of the Cold War, "First Strikes") is an immediate example of dispute. On a national level, the McCarthy HUAC hearings come to mind as an element in a hierarchy operating without oversight with damage to the community as a result.
In online venues, especially sites that are virtual communities, the question tends to be much more immediate and intimate. One thing that all community sites requiring registration share is the ability for administrators to boot users. If there is a community where users cannot be censored for any reason, in any form, I've never run across it.
On a private site such as a personal web page with a forum, the golden rule (version 2) is unquestioned --"It's my site and I say what goes." The hierarchy is in essence autocratic, the author of the page has the final say regarding content, and no one expects them to give any reason or justification. A personal web page, no matter how popular or commented upon, no matter how sophisticated the forum, is unlikely (nearly, IMO, impossible) to foster community. There are no checks or balances other than voting with your feet.
On a very tightly monitored site that is designed for some level of community building, where the power lies with an individual or very small hand-picked group, such as Bare Knuckles Politics1, the hierarchy is unquestioned, and any dissenters are invited to dissent away. Preferably far away. The admin team is unquestioned, unwatched, because on the internet, you can always walk away clean, and, more importantly, the policy is clearly posted, right up front.
The waters get murkier on a site like Cafe Utne. The front page bans sock puppetry and calls for civility. The host staff (who can hide or erase posts, or restrict posters' ability to write but not to read) is, as you would expect, at times under attack for their actions. Although some forms of citizen oversight have been proposed or even experimented with, at this writing there is no real effective oversight. The Cafe has the advantage of a large patron pool (from the magazine Utne Reader, a sort of Reader's Digest of the "alternative media"). One of the reasons for the thrashes at and about Utne, IMO, is the essential dichotomy between the ideals of being alternative and liberal and greeny, and the realities of corporate ownership and trying to cater to market forces. Axiom: Inconsistency, real or perceived, in the actions and words of the upper levels of the hierarchy is a prime source of conflict online. One feature of the hierarchy on this software (Motet) is that the site is divided into autonomous conferences monitored by hosts, and the conferences are divided into user-generatable topics with up to 10,000 unthreaded posts permitted in each topic. Hosts can freeze topics in their own conferences, but have no power over other conferences. Sites with this or similar software (such as Conferencing On Web, or COW) do not have a built-in chat module.
I've been staying away from too much discussion of E2, but it does fit into the continuum I'm discussing here, as the example of eCore software I'm most familiar with so if there's something substantially different at, say, Perlmonks, I'd like to hear about it. The hierarchial system of E2 is undivided, with administrators and editors having more power over individual patrons than Motet, being able to delete or alter posts/writeups, or temporarily silence people in the built-in chat module. The user involvement in post moderation (voting, and the general but not absolute, in cases of contraversial nodes removal of posts/writeups with downvotes outnumbering upvotes by 5 or more) tends to limit (but not totally end) thrashing about erasures. More egalitarian in theory and practice than the systems above, it is still open to abuse and perceptions of abuse.2 The lack of differentiation between areas, with all writeups going into a common pool, and all editors/admins responsible in measure for the entire site instead of rigid partitioning changes and to some extent anonymises interaction between levels of the hierarchy, although the empowered do have the option of being non-anonymous in the exercise of power. Individuals taking personal responsibility for use of power, on any site, tends to reduce general levels of conflict, except in those cases where personal onus is perceived -- and even then the conflict is more likely to remain between individuals rather than overrunning an entire site.
When the hierarchial gradient is very high or very low, there is little conflict based on rank and power. To make an undoubtedly hyperbolic and inaccurate meatspace comparison, there was nearly no dissent in Stalin's CCCP (absolute power) nor in certain small Polynesian groups (where the power figure was very weak, and conflict was with other groups rather than within the group itself).
No WWW website will ever be perfectly egalitarian, because at some level, someone has to pay for it. On a so-called "free-hosted" site, where the admin can set up a site, assign all users equal power, and vanish, the webhost itself will retain the right to scuttle the site if it feels it is neccessary, due to copyright violations or complaints of egregious offensiveness. Abdication of responsibility does not produce instant freedom. At P&W magazine's SpeakEasy (another Motet)3 there was, for a while, a period of semi-anarchy, with one conference having no hosting or moderation whatsoever. Many of the topics there were dominated by people with strong personalities. They could, if they chose, dominate discussions, flood topics with multiple posts, disrupt conversations, and make people feel unwelcome, but they could not exercise any power over other posters that the others did not give them. This period was finite, as there is now an administrative presence and the board has become much more tightly moderated.
In a truly egalitarian system, anarchy, the answer is, "We all do, for we are all the watchmen." Although this state cannot be achieved online, it can be approximated as closely as the culture and administrator(s) choose to.
True community can be fragile or robust. The software has a lot to do with it, but the prevailing culture of the site is important as well. If the level of moderation is agreed to by all participants (which, really, can happen relatively often, via the "voting with your feet" principle) the conflict level is low. As communities grow in size and complexity, so does the need for consistent oversight and useful, honest feedback from the lower but more numerous levels of the hierarchy to the smaller but more powerful. The watchmen watch over the site, and make the decisions that protect the patrons from disruptive outside forces. The rest of the site watches the watchmen to make sure that they do not start protecting us from ourselves.

1 -- the front page of BKP states, "Welcome to Bare-Knuckles Politics, a forum for the discussion of political matters from a liberal Democratic point of view. Republican lies, Greener whimpering, and other matters not consistent with the purpose of this forum will not be tolerated." I am not knocking this site's approach, it seems to work for them, and the people who post there seem to be happy with the system. They are protected from trolls and assured a forum where they can debate the finer points of their own views without having to deal with the opposition. If that's what you want, that's what you get. I make no claims or accusations regarding the fairness of this sytem, I am merely using it as an example.
2 -- The possibility of abuse does not mean there is proven corruption afoot. It simply means abuses of power are possible. Perceptions of abuse certainly exist -- who among E2 users has never been upset when an early, favored node was deleted? On the other hand, proven cases of false accusations of abuse of power do not mean that all such accusations are a priori without merit.
3 -- I write a lot about Motet sites because a lot of my experience with Meta over the last three years has been at or about Motet sites. Write what you know. If people have insights about other software community systems, I want to learn about them -- that is one of the many reasons I have for writing this metathread. I know next to nothing about the dynamics of mailing lists or USENet, to list two obvious examples. I am not an expert on virtual communities, merely a very interested observer. I'm no Howard Rheingold by any measure.
I am not obsessed. I am not obsessed. I am not obsessed.
Article #4, "A few definitions"

There are some terms that have proven to be useful shortcuts in Meta (and metacore) discussions. Community and its definition, of course, will still be being debated long after we are dust. Von Neumann machines will probably be thrashing about it as the universe sputters to a halt around them, at Tipler's Eschaton.
Thrashing is a major feature of non-threaded systems. (I do not have enough experiance with threaded systems to know if the system itself moderates thrashes, or if they are a feature.) A thrash is somewhere between reasoned discussion and a flamewar. The flames are usually directed at management of the site where the thrash is occuring or management of another site that forms the seed of the thrash. Example: Neopoeia. Created as, partially, an alternative to Cafe Utne, the debate about censorship and bootings at Utne has sustained a busy conference there since the board's creation, and continues apace, and will continue until one or both boards closes.
Booting is management's final recourse, the banning of a person from a site. Lesser penalties include restricting a person's ability to post. This can vary from the E2 system of borging for a few minutes, to only allowing a person to read articles and comments, taking away the ability to make new ones. Thrashing about bootings can go on for years, especially if the person wants to have access to the site. Imagine if DMan or Jay Stile tried to register at E2 every day under a different name, dropping barbs, stirring the pot. Although management has the tools to handle this, it can lead to innocent people who unwittingly choose a suspicious ID being banned from the site, or auto-registration being shut down altogether, or banning webmail addresses from registering (such as is done at The Mote), or requiring a valid credit card for identification. All of these can damage the community at a site, by decreasing the trust levels and stemming the flow of new users. Without an influx of fresh ideas and opinions, a site risks turning from an existing or nascent community into a small exclusive social club.
Another aspect of booting is when a user is booted from a site not for their poor actions at a site, but for stating unpopular or counter-management opinions. Sometime's it's hard to tell if someone is a troll or sincere -- the only way to tell is by their behavior over time. This does happen at a variety of sites, and I'm not trying to start an interboard thrash in this topic. Sometimes it's hard to discuss meta issues objectively when you feel that it's happened to you or someone you esteem, but for now I'm looking not at the individual issues but why they happen at all and how they can be prevented at any site with an interest in community building.
YOYOW is the doctrine that You Own Your Own Words. This has copyright aspects (for example, if the site states up-front that it will own the copyright to all user-contributed material, or if the site quotes a user's posts in another venue, atributing them to itself rather than the original poster), censorship aspects (for a deleted post may not be retrievable, thus denying the creator access; a booted user may claim the right to enter the site to redact or erase or archive their own posts, this depends on the copyright and access policies of the site), and responsibility issues. There are two ways to read YOYOW -- one is that you have legal right to your words, the other is that you have the responsibility to own what you say, to not present falsehoods as truth, to not flame others without expecting to be flamed back or otherwise face censure by your peers or management.
The Golden Rule. In any dispute with the upper levels of the hierarchy, it comes down to this. Treat others as you wish to be treated, and whoever has the gold, makes the rules. This does not make site management malevolent or benign, simply that the people who have invested money or time into the system will have the final say in what happens at their site. When the actions they take are inconsistent with the site's stated purpose (one of the largest non-cultural reasons a lot of people go there) or are clearly not fair, the fighting, the meta, begins. One of the functions of a Metacore group is to try to understand, illuminate, and, in a best case scenario, solve the dispute. I'm not talking about formal mediators or an advisor panel or a system, although one such may be suggested by the metacore. It's the people who are the most interested and who have no personal axe to grind who may be closest to finding a solution. If the parties involved are not interested in a solution, if the management insists that theirs is the right to make the decisions by fiat, or if the person who feels wronged is not sincerely interested in finding a solution, the metacore's influence stops there -- although solutions rejected by one site may apply at another.
Meatspace. Where your body is.
{reference removed}
Article #3, a short one, originally entitled "Are Virtual Communities Real?"  This was, by the way, long before the advent of social media as we know it.

The subject of the reality of virtual community is a source of near-infinite discussion and argument.
    Common arguments include:
  • Community exists wherever people are gathered together in a venue for linked purposes. E2, for example, has the purpose of the creation of a database of all human knowledge and experience. P&W SpeakEasy has the purpose of being a forum and chatspace and palimpsest for writers to network and polish their work. is devoted to hosted discussions on a variety of topics. To people enamored of the first view, these are all communities by definition, and duh. The interesting thing is that in real life, this would be a meeting or a committee. In meatspace, community has other definitions.
  • Virtual communities are people who know or get to know each other and communicate for linked purposes. The best meatspace analogue for this is the community of science fiction fandom. People isolated from each other in time and distance, who put their minds together and share features with traditional geographically based communities from outpourings of generosity for members in need to scandals and backstabbing.1 This argument is the most compelling for the existence of virtual communities as real communities, in my definition set.
  • Virtual Community is impossible. You can't argue productively with these people2. They are zealots. There are two subtypes -- those that believe community can only be applied to people who live in the same area or share a common profession, and therefore do not believe ideologically diverse non-geographical communities to be real communities, or they believe that when they shut off their computer, we all go away, and have no objective reality. "Christ what an imagination I've got/Welcome to the Matrix."3
  • Online Communities are by definition business "communities", and therefore cannot be real. communities. Similar to the above, this more subtle renunciation of virtual community says, "The people who hang out at coffeehouse A aren't a community but a mob of customers." No website fosters actual community because they are just trying to build a customer base.

1) There is an idea, rarely spoken but seemingly widely held, that since we are in (I hate to say it) cyberspace, only the best features of a community will develop. This is patently nonsense. Anywhere human nature functions, the bad side will be featured as well as the good. The ratio is going to be determined by the mores of the community, and by the administration's vision and actions. To draw an example partially from what I've seen in some of the planning spaces here and in action elsewhere -- if someone behaves extremely poorly, mistreating other people, they are likely to be filtered out. One of downsides to filtering as community discipline is that new users won't filter even the most blatant crap until they have read enough of it to decide, and people will resent being told who to filter. Their replies and comments to the crap can effectively unfilter. In an admin-filtered site, people consistently violating community norms (unrepentant sinners) are eventually kicked out. Once this happens, even one time, eternal vigilance becomes the only way to maintain the norms, and even if they are norms decided by the community, if the administration has to enforce, it will inevitably be seen as corruption, favoritism, and censorship.2) Although it's a great way to hone your rhetorical skills. Argueing with a brick wall teaches patience, the importance (and impossibility) of airtight arguments, and the joys of frustration.
3) addendum, from Mittens, a user at a motet site: "I think often a statement that 'we all disappear when the computer is turned off' is often a very real excuse for bad behavior of whatever stripe the user chooses. I've seen it happen, one person runs utterly amok, and truly is stunned that anyone would be upset by it at all. After all, they say, we are just made up could we possibly get hurt..."