“No, I meant punishment. Enforcement of rules can’t happen without the threat of some punishment hanging over the heads of the people who are supposed to be following the rules; that’s what enforcement is. Granted, it is possible to restrain people from ever breaking a rule in the first place, but in that case, presumably there is no need for the rule in the first place; and in any case, the restraint makes rules enforcement unnecessary. As long as rules enforcement is necessary, it requires the threat of punishment.”

So, when enforcement isn’t needed, the rule isn’t needed? That’s another false assumption.

Restraint isn’t a type of enforcement?

I only need a simple counterexample to prove your line of thinking unsound:

Rule: Subjects of comments may not be more than 256 characters

Enforcement: don’t accept a posted comment with a subject of more than 256 characters. Nicely inform the user of the rule.

Rule: The Firewall should not accept connections to port 25

Enforcement: drop connections to port 25

I think you are confusing two definitions of enforcement:

1) the use of punishment to make people follow rules

2) the things that make rules followed

You use sense 2 to complete your argument, but when I point out that enforcement (sense 2) includes restraint which is not punishment, you say that sense 1 is the definition of enforcement, so I’m wrong — non sequitur.

Only rules that require identity in their definition need identity for enforcement. That demolishes your second argument. I suspect you’ve fallen into the trap of anthropomorphizing the philosophy of law by defining polity in terms that must link back to punishment. The main problem is that polity has a typical context of human affairs. Ultimately what we want to improve is the store of “objective knowledge”, which is not a human affair or behavior, since it is by definition (in adjective) objective. No matter what humans do, the objective world we’re trying to catalogue will not change fundamentally. Furthermore, a democratic polity doesn’t make the information any better — then plurality or majority wins the day. I don’t think you want to intermingle expertise and democracy — they are diametrically opposed at some points, so I consider this secondary appeal to democracy a red-herring to making a better expert-driven Citizendium.

BTW, I don’t disagree with identity validation, I just think the argument is unsound. You and wikipedia have done a Hegelian show: no identity trust to complete identity trust. I suspect the happy medium is anonymous authors, identified editors, and nothing you’ve said here shows why that’s not valid.

Seth

“Even if they are imperfectly enforced, rules against speeding are nevertheless enforceable.” But how imperfectly? There’s a difference between “a small number of criminals occasionally get off” and “almost everyone breaks the law habitually and only a small fraction of them are caught.”

In this case, the law has not had its intended effect: the flow of traffic on American freeways is not 55mph. Meanwhile, we have everything from wasted fuel to accidents caused by people slamming on the brakes every time they see a cop. We have a thriving trade in radar detectors and other devices designed specifically to circumvent the law. We have made the vast majority of the country into habitual lawbreakers, which has an obvious effect on people’s respect for the law. We waste valuable police effort. And finally, being so haphazardly enforced makes these laws easily open to abuse–and, in fact, minorities, people in flashy cars, drivers in certain jurisdictions, etc. are generally much more likely to get pulled over for the exact same crime.

“… unlike, say, the old laws against certain sexual practices, after modern mores made it politically impossible to enforce them.” Is this really how we want laws to be changed? Look at the drug laws, which are currently on the borderline of the public will. The majority of people are criminals, there are jurisdictional battles around issues like San Francisco’s marijuana growing clubs, the laws are enforced unfairly, etc. Is this the model we want to follow?

“But we didn’t decide that, yes, it really was necessary to ban people sometimes in order to enforce rules, until relatively late in the game (toward the end of 2001). Consequently, as I explained in my memoir, it was as if we didn’t have any rules at all. That’s partly my fault, I’m afraid…” I understand wanting to make up for past mistakes–just make sure you don’t let the pendulum swing too far in the other direction and make mistakes just as big this time around.

Also, this seems to support, rather than contradict, the possibility that Wikipedia’s current policies might actually be ok, and they only fail because they were instituted far too late (and too piecemeal), so an unworkable culture has already hardened that’s difficult to fix. If we start with overly restrictive policies and only loosen them after we’ve already built up a culture, might we not face the same kind of fate?

Anyway, if this sounds adversarial, I didn’t intend that. My position still stands: I don’t think there is an obviously right answer, but I think CZ, with its proposed policies, is at very worst a good experiment to help determine the right answer, and may turn out to be close enough to right to succeed; at any rate, I can’t see any reason not to support it.

I agree that we need separate arguments for making real names public — but we’ve got them.

Andrew went on to say: “In fact, I’m not sure that you’ve established (or, for that matter, that it’s true) that perfect verifiability is necessary. A system that has a “pretty good” chance of detecting sockpuppetry and a “pretty good” chance of assigning appropriate punishment may be a “pretty good” system in practice. The fact that a small number of sock puppets will escape verification and therefore subvert the system is not necessarily a fatal flaw…”

I am not sure that my arguments require “perfect verifiability,” either, which is a good thing, because CZ doesn’t have perfect verifiability, just “pretty good for practical purposes” verifiability. What the existence of a democratic polity does require, however, is that we confirm identities reasonably well.

Andrew also said: “What’s the percentage of American citizens who have parked 61 minutes in a 60-minute zone? Driven 66mph on a 65 freeway? Smoked marijuana? Miscalculated (or intentionally cheated or questionably-intentionally fudged) their tax returns? I’d guess it’s at least a majority in all of these cases.”

Point well taken, but the point is that rules, to exist, have to be enforceable. Even if they are imperfectly enforced, rules against speeding are nevertheless enforceable–unlike, say, the old laws against certain sexual practices, after modern mores made it politically impossible to enforce them.

Andrew once more: “Finally, I certainly understand the attempt to find out what’s wrong with Wikipedia, but are you certain that it’s a technical/procedural problem in the first place? Does their checkuser system work poorly because it’s a bad idea? Or it could it be a good idea, unfortunately undercut by Wikipedia’s interface with the outside world (the open invititation to edit anonymously) actively inviting abuse? Or by internal cultural flaws (the insularity, cult of personality, ad-hoc but deep hierarchy, etc.)? Or maybe just because it was added too late, after years or admins judging identity questions based on individual intuitions?”

Well, I wasn’t trying to explain everything that’s wrong with Wikipedia, but only its governance problems (and doing that, by the way, wasn’t my main purpose in any case), and I admit that I gave only a partial explanation of those. After all, back when I was involved, there were several people who broke the rules on a very regular basis and did so using their own names–or, we knew who they were, anyway. But we didn’t decide that, yes, it really was necessary to ban people sometimes in order to enforce rules, until relatively late in the game (toward the end of 2001). Consequently, as I explained in my memoir, it was as if we didn’t have any rules at all. That’s partly my fault, I’m afraid…

Gee, I hope not. I think identity is necessary for democratic polity, but that doesn’t mean I’m not in favor of anonymity online. Those communities just can’t be democratic polities, that’s all! Anonymity online is itself a positive innovation of modern society, on my view.

“‘Rules cannot be enforced without effective punishment. (From the definition of “enforcement.”)’

“This doesn’t follow — I think you mean to say “restraint” instead of “punishment” (also in the rule above it).”

No, I meant punishment. Enforcement of rules can’t happen without the threat of some punishment hanging over the heads of the people who are supposed to be following the rules; that’s what enforcement is. Granted, it is possible to restrain people from ever breaking a rule in the first place, but in that case, presumably there is no need for the rule in the first place; and in any case, the restraint makes rules enforcement unnecessary. As long as rules enforcement is necessary, it requires the threat of punishment.

“There are a number of ways to make sure rules are followed without reference to identity, including screening by a set of people who are trusted.”

There is a difference between a commonly-used practice or standard, and a rule. Sure, there are ways to make an anonymous group follow certain practices voluntarily–but not consistently. As anyone who has ever managed an ordinary mailing list, forum, wiki, or other similar social software knows, eventually people will come along and make a big nuisance of themselves, often deliberately. Unless there is a regular, justifiable way to put such people in check–in other words, a rule and a way to enforce it–you’ll just have to undo the abuse (if your software allows it), and in any case suffer the consequences. And one of the consequences is that you can’t have a polity, because the fundamental policies defining your polity can’t be enforced.

“That you have identified editors is enough to create rules upon your content. Identifying authors is not really necessary unless you want to have specific privileges that become eliminated based on identity.”

Well, in order to enforce a rule, you must remove some privilege; if you don’t know who an author is, you essentially allow the author to run amuck, not necessarily under a given identity, but under new ones.

“Wikipedia insists that self-correction is done through peer review — and that the content of what is contributed is more important than who contributed it. By defining most if not the vast majority of their rules in terms of content and not identity, they really do have rules that require no knowledge of identity.”

This doesn’t make much sense: “self-correction is done through peer review”? If your peers are correcting you, it’s not self-correction. And in any case, I’ve always been a bit uncomfortable with this misuse of the phrase “peer review,” as if Wikipedians were academic journal reviewers. I wrote about this a long time ago on the Meta-Wiki (yet another Wikipedia component I started and named, by the way).

“By defining most if not the vast majority of their rules in terms of content and not identity”–this is confused. We might say that all rules have at least two components: a behavior to be ordered or forbidden, and a class of person said to be thus ordered or forbidden. The point of having a rule forbidding some action is precisely to forbid someone from taking the action. If you can’t identify who has done the action, you can ban the user account, you can undo the damage, and you might be able to change the software so that compliance is no longer voluntary, but what you can’t do is to prevent that same person from repeating the offense.

Wikipedia struggles mightily to adhere to rules, but it ultimately fails precisely because sockpuppetry allows uncorrectable people to stay on as long as they damn well please–as anyone familiar with Wikipedia knows very well. Wikipedia acts as an anarchical mob, not as a polity.

This doesn’t require true names, and in a democratic system it is often preferrable to use a pseudonym. The trick is to make sure that each person gets one and only one name.

By using pseudonymity, as opposed to anonymity, the tension between democracy and meritocracy is eliminated. Persons can demonstrate their experise under their pseudonym. The technology for securely transferring reputations from one name to another is also available. See the first reference here:

http://dss.secureid.org/stories/storyReader14

dss

I would go so far as to say, however, that by “democratization” people really *do* mean anarchy: their ideal of “rule by the people” is that each person can rule. Kind of like the race of Cyclops described in Homer’s Odyssey (each could make his own law). The problem of anarchy is that it facilitates bullying.

My “quibble” is that to limited extent, identities such as “Eloquence” or even “User:172″ can be built up and established by long familiarity. A person with a certain nickname which YOU can remember, and associate with what they’ve said and done, can become know to YOU even if you don’t know them as Charlie Parker, 123 Main St., Provo, Utah. If they get themselves banned, they could reincarnate themselves and go to the trouble of establishing a new identity. Only this time, they can profit from the lessons of the misdeeds that got them banned.

However, it is a rare dog who can learn new tricks like this. So my quibble is no more than a quibble.

For a massive online collaboration such as Wikipedia has been trying (and you are trying anew), in many cases we must demand more accountability. For anything but the most trivial or easily checked topics (such as geography and sports), we need more than just anonymity; we need a persona. For more important topics - and this is where Citizendium intends to focus I presume, nicknames are not enough. With a few exceptions that must be known to the editor in chief, the writing community needs to know who their fellow writers are.

Reputation carries weight. There’s no getting around that. If Michael Mann logs in and wants to write about the Medieval Warm Period, his reputation will precede him. He’s not an “authority” but actually someone accused of scientific fraud. He shouldn’t be allowed to contribute under a nickname, because we need to know that we can’t trust him to reveal data which contradicts his POV.

To make an objective article, or at least a balanced or neutral one, we need to be clear what POV each contributor espouses. I am completely open about siding with the “natural cycles” theory of Fred Singer. Mann and Connolley, if they log in, have a track record of bias against natural cycles.

Anyone who is a POV supporter needs an Editor above him to ensure that he “plays fair”. Readers would expect no less.

Indeed, we may have to make a special policy for scientific controversies - which goes beyond or is an extension of “NPOV”. Something like, Any claim to be “science” must adhere to the [[scientific method]] by (1) showing how the theory is related to actual observations and (2) specifying at least one derived hypothesis which could possibly be falsified by the discovery of facts which contradict it. Claims which do not fit this schema would simply be treated the same as political or historical claims: “he said vs. she said”.

Ed Poor