User:Geogre/Civility

''Note: This page is meant to be a (long) analysis of why we have trouble with "civility" at Wikipedia, why we argue when we try to discuss it, and why there are only specific ways forward. When this is done, I will create a short (and maybe polemical-sounding) executive summary. Therefore, if you don't believe the summary, read this. If you don't have patience for this, read that.''

'Purpose: Perhaps no issue has been as longstanding a bone of contention as "civility." The term has bedeviled all electronic discourse, from the old days of BBS's, Usenet, and the like to the current days of web forums and cooperative editing environments. In few cases have people analyzed the conventions that they are seeking to uphold, detailing the world they seek to create, or think about the sorts of sanctions available in a non-physical world to create proper function. While I cannot claim to speak for anyone, and I am an "anti-civility" person in general, I hope to point the way for a better consideration and a more fruitful discussion.'

Civility and Politeness: a preamble
The term "civility" is built quite obviously on the Latin civitas, or "city/citizenship." The Romans talked quite a bit about "virtus" and what "manly" values made for the proper "citizen." Most of what they talked about, however, had absolutely nothing to do with "politeness." The term "Polite" comes just as obviously from the Greek πολίς (also "city"), but only indirectly. In English, it gets to us from "polity" and "politic." I.e. to be "polite" is to be "politic." Unlike the civil virtues, the politic virtues consist almost entirely of a method and manner of speech.

It may seem that I am making distinctions without difference, and perhaps I am being a little literary, but these distinctions, I believe, bear fruit. They can at least help to explain why different Wikipedians will claim to be for "civility" and disagree utterly about the propriety of insult and profanity. ''My hypothesis is that most of the arguments about "civility" are arguments between "civic virtues" and "politeness." Those who want civic strength may want impoliteness, and those who want politeness may not want civic strength.''

In the physical world, sociologists tell us that communities spend a vast amount of energy setting up codes of belonging. They enforce their own group identity by inclusive and exclusive strategies -- by rewards of belonging, statements of what it means to be a member, rites of passage for full membership, and by ostracism, shaming, outlawry, and coercion through police (see that "polis" in there again?). On a microsocial level, we have, again, inclusive and exclusive strategies. To analyze these, let's use two metaphors: the town and the cocktail party. (I say the town, because I think anything like Wikipedia with its thousands of members cannot be a village. The cocktail party is a microcosm within the larger group where rules of reward and coercion will be less certain, and we can think of it as being any given page's active discussion.  Wikipedia is like a city, but any discussion is like a cocktail party (only with more unwanted sexual advances, perhaps).)

Social correction
It is an enormous topic, how societies police themselves, and so I beg leave to look only at commonplaces and extremes and to examine only the large, legal community and the small, unregulated community (the town and the cocktail party).

Real towns
In a regulated society, laws emerge to address behaviors that have proven to be destructive to the welfare of the community, and they emerge to prevent behaviors that are foreseen as destructive. Laws provide both a positive and a negative enforcement mechanism. When you are "legal," you rise in the regulated society, and some will have honor days, civil awards, and the like, and when you break the law, you bring about penalty that is, like the law itself, impersonal and through a mechanism that is intentionally masked. Western legal societies distinguish between "civil" and "criminal" law, the law of corrective behavior and of punitive behavior.

Think about, for example, the recent mortgage crisis. Prior to de-regulation, banking laws had forbidden the combination of stock trading and banking. This had arisen out of a prior age's experience with monopoly. The laws were stricken in the US in the 1980's, because the argument was that capital markets govern themselves best when they have no inhibitions or prohibitions on pure supply and demand. Now, legislators are considering re-establishing such regulations. This set of laws arises out of experience.

Cocktail parties
Cocktail parties are as regulated as towns, albeit with cultural codes. Before one attends a cocktail party, one attends a neighborhood birthday party as a pre-teen, then socials as a teen, then mixers in college, and, during all of this, a set of informal expectations and demands are passed on and inculturated. No one hands out a rule book saying, "Do not belch loudly," although etiquette books have always sold well. No one says, "Do not go off to a bedroom with a date and have sex." No one says, "Do not tell people all of your medical complaints," or "Do not spend your time complaining about ethnic groups." However, a person who is at a cocktail party has been invited. Therefore, the person is assumed to feel an honor or obligation. The host has offered a place of congregation, possibly food and drinks, and has extended invitations, and therefore participants will willingly obey codes of behavior that are not found at home (where all of the above behaviors may be allowed).

The cocktail party is an artificial social setting and therefore carries with it artificial behavioral constraints. It is "in society." (Without stealing my own thunder, online life takes place at home. People do not feel the invitation, do not feel the honor extended, and therefore do not accept the rules of etiquette.) The positive reinforcement for proper behavior is only conversation, flirtation, laughter at jokes, and agreement. The cocktail party rewards its players with friendship. (This, of course, is also at play in the online world.) However, what we care about here is what happens when a person fails the cocktail party.


 * A person can be labeled a "boor" and therefore lose future invitations
 * A person will see the faces of the other people showing displeasure
 * People will move away from the bounder
 * Some members of the community may eject or physically assault the cad.

Notice that most of the weapons party-goers have are exclusion. There is the withholding of rewards (instead of laughing at your blind-guy joke, I make an ugly face), the withholding of conversation (I walk away from you, as you begin to talk about your recent colonoscopy), the withholding of opportunity (I let it be known that you will never be invited ever again), and, finally, the withholding of presence (I manhandle you toward the door and give you a shove). Notice also, though, that the things that will trigger these coercive behaviors are unstated and, most importantly, communally negotiated. No one person decides for the party that you're a "drag." Instead, you have to annoy quite a bit of people to do that. A slip here or there does not get you thrown out. You do not have specific laws to obey, nor are there specific enforcements. The party doesn't have a group in it going around ensuring that all interactions are good, doesn't have bouncers. Furthermore, it doesn't have a Party Leader who will listen to all conversations or be the appellate for disagreements.

This leads us to two simple conclusions about the town and the party and their regulation:
 * 1) The town functions by law, and law is made without any knowledge of a particular individual, and it is enforced blindly (at least optimally).
 * 2) The party functions by group negotiation, and its rules are unstated; when there is a doubtful case, parties defer to others, and there is no particular set of penalties and sanctions.

Many people think of us as being like a cocktail party, online, and not many like a regulated society, and yet it's obvious that neither model is appropriate, because we're online. We're no particular place. I cannot see your face, and I am not showing mine. This brings up the next step in analysis, I think: "How do social negotiations break down online?"

How online worlds "break" real world models of civility and politeness
I apologize for the length of that header.

Invisible Cities and impossible ones

 * 1) Don't worry: Be happy
 * A regulated society online is possible, but it is unlikely. Would a community online set up a legislature?  Well, Wikipedia has something akin to one in its policy process.  Other sites that have laws have top-down designs, where some official body (moderator, owner, dude-who-founded-it, domain owner) sets out rules by fiat.  Wikipedia set out with a flat hierarchy and a slow (or paralyzed) policy process.




 * A good many users feel that the flat hierarchy is a problem. While few will knowingly or explicitly call for a dictator (or King Log), in any arbitration case talk page comments will indicate a desire that there be clear rules, that ArbCom, for example, rule.  Additionally, multiple communities ("cocktail parties," below) have felt this frustration or desire, noted that there is no ruling group, and have appointed themselves the "clueful," etc.


 * 1) Don't smile: This is serious
 * However, there is actually no need to build a regulated society at Wikipedia, because users only need laws in the breech. I.e. they only call for rules when there are violations.  No one calls out for a rule against writing articles about family pets until a user writes a page about a family pet.  For the sorts of civil law and behavioral rule that worldly societies have to legislate, Wikipedia has either no need (no bodies are present) or has an assumption.  I.e. Wikipedians come from regulated societies, and they carry those regulations with them when they go online.  Each person is bound, in a way, by the jurisdiction in which she or he lives.  Because of this, most people who call for rules are calling for them about speech rather than action, and this allows them to miscast their concerns as concerns with politeness and well regulated speech.


 * There are a few places where issues of action come into play, but these are rarely ambiguous. The problems of pornography, age of consent, pedophilia, stalking, and real world harassment do emerge, but in most cases Wikipedia sorts these matters very quickly.  (The exception is the issue of pornography, which remains open on Wikipedia, just as it does in the regulated societies that Wikipedians come from.)  When it comes to breaching such regulations, Wikipedia has only one response: deletion and total exclusion.  There is no rehabilitation process possible where there is no full person present, and there is no chance of a fine or penalty.  All Wikipedia can do if a user harasses another in real life or posts pictures of explicit sex or underage sex is delete the account and the media and police the user's reappearance.


 * 1) The boat sailed without you
 * The most important failure of the invisible city is that it has no reward, no citizenship, no public spirit. It has no rules of permission.  Because of the sluggishness of regulation and the coercive nature of it at Wikipedia, there are few declarations of rights, few positions of freedom, and no encouragement of those things that build a society.


 * What happens when the people in your town begin blaming immigrants? What happens when they want to lynch someone?  What happens when they want to try the school teacher for teaching evolution?  In actual life, there are civic virtues that are propounded, and there are lessons of "good citizenship."  The invisible city has nothing like that.  So, on the one hand, Wikipedia can't be a regulated society, doesn't need to be, but, on the other, by only responding with prohibition and censure, it has no strength.

"That ain't no way to have fun" (the Bad Party)
There are all sorts of terms for the extralinguistic communication that exists in any utterance. There are several additional signifying layers to any spoken line.
 * 1) Parties rely on phatic communication, which is absent online
 * [Scene: Doorway, in June 2008; two young men in white, short-sleeved shirts are on the threshold of a house, while a slovenly, frumpy, pseudo-intellectual stands inside.]
 * Man: "No thank you. I believe I am already well informed about the end of the world."
 * Proselyte #1: "But, Sir, our founder had a message of universal significance about what we need to do to ready ourselves."
 * Man: "No, no. I know what I have to do to get ready."
 * Proselyte #2: "Sir?"
 * Man: "Well, I need to find a nuclear trigger, and then I'll finally be ready to end the world."
 * Proselytes leave.

Now, in that little psycho-drama, you are missing the facial expressions of both "Man" and "Proselytes." Because you can't see them, you imagine them. You also cannot hear the inflection of the language. Was Proselyte #2 sarcastic? You imagine the most likely tone for the words, based upon your experience. You cannot see inside the house. Was it filled with red flags, or were the walls covered in pictures of Jodie Foster? You cannot see the hands of the people. Was one crossing fingers, or was another holding up a single finger extended? Were the proselytes hip to the joke and leaving because of that, or were they terrified?

In an actual conversation, physical presence colors everything. If the person is beautiful, respondents will report positive interactions and will be nice. If the person is physically imposing, he or she might be accustomed to intimidation. If the person is disabled in some way, people will speak differently to the person. If a person is old or very young, the conversation changes. If a person is in a palatial manor or a tarpaper shack, the conversation changes.

Online, "Nobody knows you're a dog," but nobody knows you're the lost Tsar of the Russias, either. The phatic and social and physical portions of communication can be a burden in actual life that is shed by those online, and they can be vital components that are lost, and this can account for the psychological appeal and revulsion of online communication. However, what we need to care about here is that these channels, and there are at least four major channels, of communication are simply absent.

The usual tactic of social negotiation -- warning gesture, then withdrawal of presence -- simply does not work online. I can't get away from you, no matter how much of a creep I am, if I'm not actually "there" in the first place. Also, you can't rub me off on a nearby cluster of idiots, if my attachment is always direct.
 * 1) Get away from me, Creep!

In a conversation online, we have the illusion that many people are present, and yet each conversation is one on one. I post to BoboTheClown, in the midst of Bobo's argument with TimmyDGnome, and it seems that I am "in the middle," but I have directly contacted one person. If Bobo doesn't want to respond to me, he can't actually move away. He goes to a different talk page to debate JimmyCrickets, and I post to BoboTheClown again. In other words, every interaction is direct. Whenever we have problems with each other, we have no way to shake each other off. We are never in a crowd, always directly accessible.

This is a very important point, and I hope it's not lost in the clutter: We cannot get away from someone who disagrees with us. If e-mail is enabled, we can't even get away fully by leaving Wikipedia. Because we cannot use our customary social negations, we often see "snapping" and insult.

The drunk at the party is ejected, but the Wikipedian who is legitimately insane can't be shuttled away. We have various "do not feed the trolls" platitudes that are efficient, if followed, but when someone is actually nutso, there is little to do about it. The informal and formal social actions that eject someone who is unfit for a given society do not work online.
 * 1) "I think you've had enough. I'm calling a cab."

Additionally, there are different settings and definitions of fittingness. We all know that the NRA member who crashes a Brady Bill discussion group is not fit. That person is going to start arguments and try to dismantle the conversation. Neither the Brady Committee persons nor the NRA member is illegal, nor generally unfit, but the presence of that person in that party is bad. If the conversation itself is licit, then there is no way for people to keep out the gainsayers online.

I used to moderate a forum on a BBS network on Feminism. Guess who the new members were every day? Yep: the "women should shut the hell up and get back in the kitchen" crowd. Who were the new members to the Christianity group? Atheists. Because we had top-down structures, we ejected those persons very quickly, but not without side effects. Online, we had no way to call a cab for the drunks except coercive exclusion (and then those people would rejoin under other names and be even angrier).

On Wikipedia, you can look at any highly charged topic for an occasion of "incivility" because there is no way to eject a participant. If a person comes to the Michael Moore page wanting to call him names all day, that person can get by on the talk page with the most ridiculous stuff, and the people who try to keep the page mainstream and neutral will have to fight every day against that person. There is no mechanism on Wikipedia to say, "You may be fine editing generally, but you shouldn't be at hot button pages where you hate the way the mainstream views the subject." We don't have a way to say, "You can't come into this party, but you are welcome down the street at the frat house kegger." (N.b. I do not advocate any method of ejection, either.) In fact, things have to get bad, stay bad, and then get worse before an RfAr starts, and then, after much suffering, the person who hates the subject might endure a block or ban. The failure of the "I'm calling you a cab" feature has resulted in much frustration.

Clarifying terms
At Wikipedia, we are only words, of course. Therefore, our interactions are verbal, for the most part. We have to distinguish, I think, between the desire to be civil and the desire to have pleasant experiences. These are not synonymous, although they are often allied.


 * 1) To be civil is to be building community, defending community, and acting as a function of the civil group
 * 2) To be pleasant is to avoid anything that might give disquiet to another user and to provide interactions that primarily give to others.

Civil virtues
As I said well above, there is a long discussion of "civil virtue" that we can look at. Heck, in the US, youngsters even take a class called "Civics," which, unfortunately, focuses too much on government and not enough on community. The problem with talking about civic virtues at Wikipedia is that the shedding of real life burdens of personhood and the building of online presences means that Wikipedians are, to a greater or lesser degree, antisocial by preferring an online, manufactured society over the duties and functions of the actual. I know that I'm antisocial enough for any three people. Nevertheless, let's set that aside and focus on the fact that, for some reason, people are wanting to create a society here, alternative or not.

The eBook of Virtues would include:
 * 1) Sacrifice for the wellbeing of the group
 * 2) Promulgation of justice
 * 3) Passing up private gain, if it is to anyone else's loss
 * 4) Fostering the young and inexperienced
 * 5) Respecting differences and practicing pluralism
 * 6) Excluding voices and parties that would negate pluralism
 * 7) Protecting the weak from depredations of the strong
 * 8) Blowing the whistle on wrong doing, either through satire or policy
 * 9) Admitting mistakes and being prepared to pay for them
 * 10) Not requiring others to pay a price for their mistakes, if they are truly contrite.

These virtues are sometimes directly at odds with social graces, or the arts of conversation. For example, blowing the whistle is often the same as ringing a klaxon. Respecting differences can often mean not joining friends in the condemnation of some. Protecting the weak from the strong means being exposed to the strong, yourself. However, only when we have these virtues as unambiguous goals can we have the second layer of goods, the "pleasantness" of conversation.

I would argue, and insist, that it is only possible to demand that all people speak politely if we are first being virtuous. If a person is not being civil, per above, then polite behavior is, in fact, corrosive to the group, to the society, and, in fact, "incivil."

Social graces
Well, we know full well what etiquette is. I gave examples, above, from dinner parties and cocktail parties of boorish behavior. As much as we know these forbidden actions, we know the usual actions that are smiled upon. Ask the other person to talk about himself, hold a door for another person, smile, shake hands, etc. At Wikipedia, we are words and conversations, though.

We know that there are social registers in language, I hope. Most languages and dialects have at least three registers: official, familiar, and slang. Basically, a person speaks one way to a judge in court, another way to wife/husband, and yet another way to the closest friends. These three registers each have their own syntax, their own vocabulary, and their own lists of taboo and licit language. Additionally, languages themselves, lexically, will list terms as "formal," "informal," and "profane." What happens when we mess up in conversation is that we use one social register for a different social setting. Speaking in formal language to one's close friends may be "snooty," and speaking with profanity to one's doctor may be gauche. Additionally, we violate the social graces when we are intentionally mean or vicious.

So, the social graces would include at least these qualities:
 * 1) Deflect criticism from oneself, rather than combat it
 * 2) Criticize another's actions, but always allow the other to "save face"
 * 3) When a person is "down," be supportive
 * 4) Do not make personal reflections about another's mentality or physicality
 * 5) Do not make sexual suggestions to another
 * 6) When another betrays a secret or an inappropriate detail, turn the conversation away from that
 * 7) Praise the efforts of others.

Note that profanity is only profanity when it is, literally, profane. "Profane" and "vulgar" are assessments of difference rather than inherent qualities of words. A term is "vulgar" when it should not be used in preference to another term or when there is a striking difference between the actual and anticipated audience (social register). When I was a teenager, I might say "pussy" and refer to pudenda with my similarly testosterone-intoxicated friends, and I would never think of doing so now, in my forties, even speaking to the very same people. I might now say "bitch" of someone I dislike who is female (not likely, but possible) with my close friends and rely on the difference to indicate the vehemence of my disgust, while, for another person speaking to friends, that might be a normal slang term for any female. The point is that no word or expression or phrase is profane without a consideration of audience. Therefore, it is a social grace to not be profane, but that only means knowing one's audience and pitching the language appropriately in terms of culture, age, familiarity, and expectation.

Online, we have a devil of a time understanding audience. My audience for this essay could be anyone from my worst "enemies" at Wikipedia to someone doing a Google search. How do I know what is profane to my audience? On the other hand, how do I know that what you said is profane? Remember the point, above, about how all communication at Wikipedia is one-on-one. If I see that you have written, "This is bullshit, d0de!" it's odd for me to announce that you were being profane, if you were writing to someone who expects, understands, and traffics in that language. If, on the other hand, you have addressed your comments to a general audience, I can fairly easily say that your words are vulgar, because, at least in English, all general address is presumed to be at the upper social register.

Synthesis (Civility in practice)



 * 1) Civility applies to those qualities that make for a strong social unit. These include politeness and deference in conversation, but they are not tantamount to them.  Civility is that which makes the site's community function well.
 * 2) Polite speech is only possible if people are behaving in ways that are in accord with social goals.
 * 3) If a person is behaving in a way that harms the community, or if a person is being anti-social, then we have two different methods of response:
 * 4) We can respond by law and policy
 * 5) We can respond by social action.

It is vital that no one invert these priorities, that we not ever substitute politeness for civility, that we never commit the double atrocity and absurdity of thinking that words, by themselves, have a quality of being allowed or disallowed and that the use of "bad" words is "incivility."

To know which type of response is appropriate, we have to discern the two types of breach of civility.
 * 1) A breach of the rules of behavior at Wikipedia will cross the lines of policy
 * 2) A breach of the rules of conversation at Wikipedia will not be against any policy.

Saying, "You suck, loser!" is boorish, childish, and a violation of the rules of conversation, but those rules are entirely derived from context, audience, and community, and therefore there is no way to put that statement on a pan in a balance to see just how bad it is. On the other hand, deleting every image I upload with a summary of "rm troll" will break the deletion guideline, and we can easily compare that action to existing policies. The action exists without context.

Therefore, dealing with uncivil people goes in two broad methods: The degree of violation often determines the severity of the response in both instances. A really dramatically illegal action will get a ban. A really dramatically impolite speech act will provoke a severe insult, a long winded derogation, a shunning forever, or a brutal ridicule. The important thing is that these are the only tools available for conversation to police conversation. They're "natural." They have implied rules (e.g. a person who really knocks the hell out of someone who had been only a little rude will have it "blow up" on him and find himself being shunned). They have measures of effectiveness.
 * 1) Meditation, mediation, and arbitration exist to deal with policy disputes and bad behaviors.
 * 2) Speakers deal with bad speech by
 * 3) Ignoring it
 * 4) Diverting it
 * 5) Being insulting
 * 6) Using invective
 * 7) Shunning
 * 8) Satire

What is the court, then, for determining if I have been right in calling someone a "bed wetting pissant with no more sense than a box of concrete?" What is the court for determining if that person had been right in saying that I am "a horse's ass" and a "fuckwad" (these are hypothetical examples!)? Have I just argued that we have nothing but a Wild West model? Have I just argued for utter permissiveness?

Conversation court
We know what to do with bad acts. What about bad words?


 * 1) All utterance is meaningful only in context of speaker, auditor, medium, and time (history)
 * 2) All improper speech is a failure to speak at the level expected by the audience
 * 3) All audiences are participants in a wider culture, as are all speakers.
 * 4) Language is the primary mortar and brick of culture, and it is the only material of online culture.

So, if you think that I have been "incivil" by calling you a turd, should you block me? No. You can testify that you felt insulted, but you cannot say whether or not the insult was a proper or improper response. This is the most important element of all:  the reader cannot say that an insult or compliment is warranted, because "warranted" depends upon community standards. Furthermore, and this is one of those important features of language in action and online life, all speech acts online are an answer to a question. Don't get hung up in the particularities of that. The point is that everything you say is in response. Therefore, you cannot judge whether or not the speech is out of bounds because you're only one speaker, and because that utterance was in response to something, and yours was in response to all of the speech in the exchange.

What we have to do, if we think that someone is being unjust, is appeal to a very, very wide audience. If you think that my exchange with you has been improper, you need a wide group to agree. You have to establish what the cultural norms are, here. The norms here are probably more restrictive than at SomethingAwful, but they're not Queen Victoria's. Essentially, if we want to determine whether or not a person has been too insulting, we have to have a wide, wide review.

Because only the whole community can determine what the community holds as its standards, and no single member of the community can speak for the rest, and because every utterance is in response, the worst possible way to assess whether comments were "incivil" is by diffs that isolate a comment. The whole community has to look at the whole context to determine whether a person's rhetorical strategy is a reasonable response or not.

Discipline and punishment
Suppose, therefore, we try to figure out the proper response, as a community, to the community's determination that a person's comments are against acceptable standards.
 * 1) Most violations of politeness will be mistakes
 * 2) Those that are not mistakes will be efforts at achieving a goal
 * 3) The goal will either be for the greater good (civil)
 * 4) or against the general good (uncivil)

First:
 * Any effort at cautioning would have to be based on a person who does more than one or two bad moves. Any effort to use a community sanction for a single comment or exchange would be disproportionate and really against our common goals (cooperative editing).
 * Any person making a mistake in comments has made a mistake. Mistakes are either going to be misprision or an inability to assess audience properly.  Therefore, it is proper and best to explain where the comments went wrong or where the reading went wrong.  (It's important that we explain to people who are being unnaturally sensitive that their readings are mistaken just as much as we explain to people being overly informal or vulgar that their assessment of the community is wrong.)
 * A person who consistently makes mistakes about the rhetorical environment or the speech community is either incapable of assessing it properly or culturally blind to the distinction. A person who is incapable (a sort of autism of verbal communication) may be unfit for a cooperative editing environment.  A person who is culturally blinded (some cultures have multiple registers of profanity, with some registers acceptable for mild annoyance and others only for starting fist fights) can learn the culture.

Second:
 * A person whose goal is for the greater good may well be "incivil."
 * We must recognize that slang, profanity, and vulgarity are both bonding and exclusion language. Just as some groups will use their online slang (e.g. "cluebat" and "clueful" and "gnoming") to create a subculture of inclusion, so some will use online (or otherwise) slang to create an exclusion group (e.g. "drama," "wonkery").  Therefore it is absolutely appropriate to regard any use of exclusion slang as an intentional offense, and we have to look at the motivation and effect of it.
 * A person who is trying to protect the community or defend him/herself may well (almost certainly will) resort to insult or other disincentive language designed to push the other person away. If the person being "harmed" this way is engaged in harming the community, it is not "incivil" to use these tools.

Third:
 * Those uses of "bad words" that are for the private gain or for the personal detriment are actually uncivil.
 * Those uses of "bad words" that are designed to win an argument rather than to protect the community are uncivil.
 * Those uses of "bad words" (including exclusion slang) that are designed to create factions within Wikipedia rather than to invite participation are uncivil.

How do we determine the extent of the harm, and therefore the proportion of the response?

In any rhetorical situation we have the speaker's intention, the medium/culture, and the reader's perception. A good many societies have had trouble with verbal crime in real world situations, and they have had even more trouble with social crimes of speech. Remember the weird, hyperbolic arguments about political correctness?
 * [Scene: A busy office of cubicles. The comb-over, male supervisor lingers by the copying machine.  Two female workers in their 20's walk to the cubicle of a male co-worker, also in his 20's.  He is a frumpy pseudo-intellectual, and they are not.  One of the women is wearing a fashionably low-cut blouse and a mid-thigh skirt, while the other is wearing a fashionable pants suit.]
 * Pants Suit: "Hi. Are you going to the cafeteria for lunch, or are you going out today?"
 * Frumpy Pseudo-intellectual: "I was going to the cafeteria, but if you guys are going out, I'm going too."
 * Midi: "We thought we might go to the new tavern."
 * FPi: "Well, count me in. You look sexy as hell."
 * Pants Suit: "You know, on second thought, we probably don't have time to go out today."
 * Midi: "Yeah. Maybe another time."
 * Women walk off. Supervisor looks up, alarmed.

The PC problem was one where we had to decide if that guy had sexually harassed the woman who was dressed up. In the Harassment Awareness meeting and training session that followed, he protested vociferously that he had not meant to make her feel bad, that he had not meant to make her feel objectified, that he had meant a compliment and thought they were friendly enough to take it in the right spirit. He will go away convinced that the woman with the midi-skirt is a horrible person, even though it was the supervisor who had called the training session and made the report. The trainer at the session will tell the man that his remarks had been heard as harassment, that they had had the effect of belittling the women, that they had reinforced patriarchy. Between intention and reading, we have a dilemma.

If we only ask whether someone meant harm, we find that everyone is innocent. If we ask whether someone heard harm, we find that everyone is guilty. If we ask, absurdly, if someone could possibly have heard harm, then we go silent in fear.

In my second psycho-drama, you'll notice that the group had worked out its own negotiations. The women iced the guy. They told him what they thought. They might have been offended, but they were probably just pissed off. They might have been reduced to tears, but they were probably just thinking that the guy was clueless. He will either learn not to speak like that, or he won't be able to speak to them again. However, the supervisor has no way of knowing the degree of harm. He can determine that the comments pass a line proscribed in his rule book, but he can't tell whether that was harassment or just stupidity.

To know if something is or is not offensive, we have should question neither the speaker nor the reader, but, instead, the culture. Would a community determine his remarks stupid or criminal? To know the degree of offense, we have to look at the effect.


 * Therefore: the only way to know whether a person's "rude" or "bad" speech is actually uncivil is to look at the effects they have. Is the person coming back to fight (bad)?  Is the person editing less (worse)?  Is the person just avoiding the first (not bad)?  Is the person explaining the situation (good)?  The only way to know what to do with "incivil" remarks, after you have determined that they're not mistakes and not for the greater good, is to look at their effect.

By that standard, my calling a member of the ArbCom a moronic, corrupt, incoherent, self-serving, auto-sodomite is probably not going to be very bad. I doubt he will quit the board over it. If I call him those things because of some evidence of corruption or actual incoherence, I'm probably not even being uncivil at all.