I never thought I would see established Wikipedia editors refusing to allow the NYTimes as a source in an article, but that appears to be occuring in the Global warming article. This, in spite of the opinions I received by uninvolved editors at the Reliable sources noticeboard. Strange.

I can't believe you never thought that.

Has there ever been any indication that this would not happen?

(Didn't it already happen in some thing related to Moulton's problems?)

By rights, an encyclopedia article on the scientific data regarding Global Warming should solely use peer reviewed journals and statements by climate scientists. And there should be enough of that to fill scores of articles.

Virtually everything put out by the media on global warming has been in error. This has created mass confusion, especially in the US it seems where the data is subject to ludicrous media propaganda.

But on the whole, nice. The idea that newspaper reporting (even the best newspaper the world) should be given credence when it comes to science and technology topics that are independently peer-reviewed in other and better sources, is laughable. Mostly, that WP policy comes from stuff slipped into the WP:RS policies long ago by people who trust newspapers and don't know a damned thing about science.

Hopefully the policy will slip and be modified until it covers history, the arts, and so on, leaving the best newspapers to their proper role of being the most reliable sources we can find for many historical events in the time before news accounts can be properly balanced and mixed with other primary sources by academic historians.

Interestingly I note that among the few academic subjects on WP where the "big paper journalism is just as good a source as anything" does not sell, appear to be philosophy and political science. I call this the SlimVirgin effect. She may have worked as a journalist, and thinks journalistic droppings are not odiferous; but when it comes to politics she has strong opinons, and she majored in philosophy, so she knows journalistic and web sources aren't to be trusted, there. As for anybody else who has this expertise problem, however, Slim and her crowd have been raking them over the coals for years. New York Times cite or not.

It's still a case of "One Rule for the Cabal, another for Other People." The difference is that these days the old cabal is not what it was, and we may yet live to see some proper heirarchy of academic epistemology begin to approximately (in a pale way) control preferred sourcing on WP.

Color me optomistic!

I mainly write history articles, and newspapers are often useful because they record history. I believe that in this way they are useful for articles on other topics, including science, for the same reason. A newspaper reporter, as in this case, can look at scientific opinions and papers from all over and summarize their general opinion, discuss how things have changed or stayed the same regarding the topic over the years, and then quote opinions on where things are going.

As far as the actual science is concerned, peer reviewed scientific journals and reports are probably better. As Milton points out, however, WP policy is fairly clear that mass media are reliable sources for pretty much all topics, and for the GW regulars to insist that the NYTimes cannot be used in their article is the same as saying that their particular article is a special case that doesn't have to follow the same rules that all other articles (except BLPs) in WP have to follow.

Basically what he said. It isn't like we don't have the peer-reviewed, scientific info to write such an article. If we were to have "media perceptions of global warming" or "public perceptions of global warming" (and I can see the endless-spinoff-machine that is Wikipedia producing one) then this would be useful. Best to stick to hard facts, particularly when the media seem to have a blind spot for checking their sources. My job entails working on the Treaty of Lisbon, and some of the media claims about it make me want to kill one journalist a day until they collectively promise to speak to lawyers before hitting "publish".

But, don't you see where media reporting is useful? If the meteorlogical agencies of five different countries produce reports on climate change, wouldn't it be helpful to use a major newspaper reporter's synthesis of all five reports? In the article on Wikipedia, you would only need to make it expressly clear, "The NYTimes reports that, in general, the five studies found that temperatures have been steady or have increased slightly since 1999. The scientists involved, however, have stated that they still expect temperatures to increase in the long run unless appropriate measures are taken" or something like that. What's the matter with that?

This ends up turning to the basic problems of organizational structure, and a solution, and I thought of cutting it up, putting the org structure stuff in its own topic, but .... the shift was gradual and I didn't quickly see a place to cut it, so I'm leaving it here for now, not enough time today to do more with it.

Quick summary: One arbitrator could effectively change the entire Wikipedia system, with high probability of success, by setting an example, creating an independent, efficient advisory structure, persisting through opposition. Below, I don't emphasize the outbound half of the advisory structure: such structures, properly designed, work bidirectionally, both inward and outward, and so the arbitrator could also marshal support as needed to assist in the implementation of the advice received.... the scale would not be limited.

Structurally, this wouldn't work unless the arbitrator were dedicated to representing true, fully-informed consensus (as well as helping form that consensus by facilitating structures that can efficiently negotiate it). As a personal power grab, it would be almost guaranteed to fail.

If we printed out the confusion on this point, it would weigh a ton. "Reliable source" is reliable for notability, not for scientific fact. No source is fully "reliable," in the ordinary sense, for fact without attribution, but it gets close enough under a number of conditions, such as a lack of controversy over a fact, or, with "scientific fact," acceptance of stated facts in articles under peer review, in mainstream journals. Most legitimate controversy over sources would be resolved with attribution; what happens with global warming, as with "other pseudoscience topics," -- hah! -- is that the spectre of "undue weight" is raised.

There is a classic solution to the problem, which is distinguishing between a topic as a scientific one, or as an historical or social one, and forking the articles into scientific and social/historical focus, each one citing the other in summary style. This solution only fails because there are factions of editors who clearly don't want a balanced presentation, they want to exclude opposing POV, and, on Wikipedia, that's what I called the Cab -- "cabal" without the need for conscious conspiracy.

As with all Wikipedia problems, the dispute persists because seeking consensus isn't actually policy, when push comes to shove. ArbComm gets easily stuck in finding guilt and blame, instead of acting to improve direct dispute resolution, and only sanctioning behavior in a preventative fashion, that is, only preventing disruption of dispute resolution process. I was pretty explicit about all this in my workshop proposals in the RfAr that banned me, but with the avalanche of disgusted comment from the Cab directly opposing, for example, the idea of seeking real consensus, it was buried. A competent ArbComm would use various organizational techniques to see through such a blizzard, but .... Wikipedia doesn't exactly select for competence. Several arbitrators are reasonably competent, but clearly not enough.

Still, my view is that one arbitrator, developing clear connections with the community as a whole, could shift the balance, possibly permanently. I don't see that any arbitrator understands the situation sufficiently to pull it off, unfortunately. If one of them, just one, wakes up and starts to seek the solution, instead of imagining that it's impossible, on the one hand, or that things aren't so bad, after all, on the other, the seeker would find what is sought. It's practically been poking them in the nose for some time.

By the way, the solution doesn't depend on me, personally. I can see it, but so could others; however, most who could see it don't happen to get sufficiently entangled with Wikipedia to care enough to propose it, that is the only reason I'm relatively unique. From history, it seems to take two active people to create major shifts in large organizations. It's counter-intuitive, we might, on the one hand, think that it takes many, and, indeed, in a way it does. But then many follow, crystallizing around the activity of the person who might be called the mirror. Behind movements such as abolition (of slavery) and suffrage (of women), there were theoreticians who formulated the movement's principles, but then popularizers who took the concepts and created actual organizations. The popularizers are often the ones who get the bulk of the credit, because the popularizers are highly visible, and, in a way, this is just, because the theoretician doesn't actually make it happen, or at least can't do it alone. Social change requires people connecting with each other, acting coherently, and, obviously, one person alone can't do that.

The shift will only occur when there is enough "saturation" of the necessary concepts, so that the structure will crystallize when a seed appears. As with any supersaturated solution, crystallization doesn't occur from a single molecule, it takes a seed crystal, which is a small number of molecules organized in the crystalline pattern, large enough that Brownian motion doesn't tear it up before it can align other molecules in its pattern.

And one of the concepts is the tremendous value of consensus. There are a million possible disputes on Wikipedia, and focusing on each one alone is necessary, but, obviously, this can't take place in large-scale discussion, even one dispute, discussed on a large scale, can heavily consume editorial resources, so attempts are made to create short-cuts; the classic solution is rules, so that disputes can be resolved by appeal to rules. It works when the rules can be made clear enough, and this is behind the faction that wants clear and strong rules, enforced neutrally. But it also needs that other magic ingredient: neutral judgment. Rules are useful, but there isn't any substitute for awakened consensus, in fact, rather disputes merely get shoved into arguments over the rules, what the rules are or are to be, and how to interpret them and apply them to specific situations.

Wikipedia was of interest to me because the basic policy of neutrality requires, if one follows it through to necessary conclusions, consensus. Thus Wikipedia, being in fundamental need of consensus, would seem an ideal place to test concepts for how to efficiently seek consensus on a large scale. I can say this much from what I did while active: the concepts work, to the extent that they are applied.

And organizational theory also predicts that opposition will arise, for any change to a system which distributes power equitably where it was previously inequitably distributed will be seen, correctly, as reducing the power of those who enjoyed the excess. In volunteer organizations, these are typically the most active members, and they fear, again with some justification, that distribution of power will result in loss of intelligence, as those less experienced than they exert power. This group therefore, classically resists increase in distributed, democratic control, and, because it has excess power by the terms of the problem, it will often succeed. Note that no evil, greedy motives are necessary for this highly conservative force to operate. It's normal. The Iron Law of Oligarchy isn't about greed and power hunger, per se, it is more functional than that. My conclusion is that oligarchy is necessary, but that the addition of another element, present in small organizations and largely missing in large ones, that will guide the oligarchy, is necessary. It will also restrain it, but my theory is that actual restraint is less necessary than we think. People will follow good advice if it comes from a trusted source.

So, to summarize in a nutshell how a single arbitrator could shift the entire situation: the arbitrator would set up independent structures for a single purpose: to advise the arbitrator. The personal power of the arbitrator remains, it is not turned over or surrendered to this advisory structure. The advisory structure distributes communication and analytical load over a large group, and is designed so that no individual is overburdened, yet every member experiences connection with the top, the arbitrator. If the structure is small, the connection can be immediate, but if it grows beyond a certain size, and to be fully functional on Wikipedia, while it could be highly useful even if small, to be maximally effecive, it would need to grow beyond the size where routine direct connection works, the connection might be, for most members of the advisory structure, indirect.

There was, until a few days ago, a Wikipedia article on an election method that could set up this kind of structure from the bottom, instead of from the top. It was at one time proposed as a method for electing a Wikipedia council, which it could do without fuss, easily, given a defined membership (and defined membership is already used for all Wikipedia elections). It was just deleted, as part of the activity of a block-evading sock, blatantly. The information came in about the sock puppetry, it even made it to AN/I, and it was, as such information often is, ignored. Wikipedia structure is highly unreliable, highly inefficient, it wastes vast labor resources.

The method is old, first proposed, AFAIK, by Lewis Carroll (Charles Dodgson) in about 1884. Asset voting, it was called when reinvented by Warren Smith in about 2002, as I recall. There was prior discussion on the Election Methods mailing list in the 1990s, where it was called Candidate Proxy. Very simple. Also very little written about it, the name Asset voting is a neologism, but because Carroll's work is covered in secondary source, the concept is probably notable enough for an article or for a mention in Single transferable vote, for Carroll invented it as a tweak on STV, to allow the enfranchisement of voters who didn't have enough knowledge to rank a series of candidates in order of preference, as STV requires for it to be a decent method, otherwise, even though STV, with full ranking, is a good proportional representational system, as long as enough seats are being elected per voting unit, it loses representation for ballots that don't have enough rankings on them, that name, say, an unpopular candidate. STV, then, only works well in party systems, for people generally have enough knowledge to rank parties. In Australia, in most areas, mandatory full ranking is the law; ballots that don't fully rank are discarded. This, then, leads to "donkey voting," where voters just rank in order on the ballot, after they vote for their favorite. Carroll's solution was both simpler and more complex. Allow the candidate in first place on the ballot to exercise the vote, later, as if it were the candidate's "property," if the ballot is "exhausted," i.e., the candidates on it have been eliminated as not getting enough votes to qualify for a seat. This would allow proportional representation without requiring parties at all, just vote for your favorite. And if parties aren't required, there goes the entire political structure.... no wonder it's never been tried!

But, of course, it's more complex as well, because then there is post-ballot process, where holders of votes negotiate putting them together to create seats. The original ballot can be secret ballot, but it's probably essential that the subsequent reassignment of votes to create quotas for seats be public. So "candidates" in an Asset Voting election actually become "electors," public voters, as with the U.S. Electoral College. And this leads to a whole world of transformative possibilities. It can shade into a quasi-direct democracy.... The U.S. Electoral College, a very advanced concept, was quietly dismantled, becoming nothing more, in actual practice, than a rubber-stamp, quite different from the original intention.

The rejected WP:PRX would have established a file structure for allowing open "votes" for "candidates" for a Wikipedia Assembly. It didn't mention the assembly, it was just about a proposed file format, allowing every editor to name a "proxy," but without any specified proxy powers, just to see what would happen.... The suggestion that this would be paradise for sock puppets was preposterous. The proposer, Absidy, named me as his proxy, and we were promptly checkusered. The last thing a puppet master wants to do is to call attention to the connection! The article on Delegable Proxy was deleted, of course, simultaneously with the attempt to MfD WP:PRX, even though it and its predecessor, Liquid Democracy, had persisted for years. Asset Voting finally followed, more than a year later. Delegable Proxy may be more notable, overall, but it's a field where little gets reduced to formal publication, election methods experts widely adopted the internet in the early 1990s.

Well, it's been tried within a party to select candidates in Sweden. There's a discussion of it in Proxy voting under the subheading of deligate voting where "liquid democracy" is classified as the operational system when you have an infinitely transferable fluid deligate-voting cascade (where proxies can transfer their votes to other proxies, who presumably know even better what the issues are, and what should be done about them).

These things are promising alternatives to fine-tune things so that something closer to direct democracy can take place, but still with many (most) voters having limited information. We presently transfer individual voter democratic power to their representatives in the government in just one or two steps in Republics, but obviously the more finely this is done in concordance with voter knowledge (the more the voter spends time finding about issues and sub-candidates, the later he surrenders his or her proxy), the better the information processing and result will be.

I think "liquid democracy" is a perennial proposal at WP. The worst problem with it, is the same problem with all "vote-like" things at WP, which is that there's no way to register and identify voters. So they are all doomed to be just shadows (and bad ones, at that) of any democratic system (even the worst ones) in the "real world."

Well, now that it's mentioned here, what is the half-life of that bit of text? Might survive, the Demoex material actually has reliable source. But that's not Asset Voting. Demoex used DP for a short time, abandoned it because they thought of it as a software thing. It's actually a communications networking concept, properly done. What they did find was that too many people chose one person, they thought that gave too much power to the one. They were also using direct democracy to "instruct" a representative on the city council how to vote. I had extensive correspondence with one of the principals, used to have a Wikipedia article, forget his name at the moment. Anyway, my suggestion was that the single person named by so many could easily suggest that some, maybe most, of those who have chosen him or her choose someone else instead, to provide some layering so the chosen advisor (that's what they called what I called the proxy, and it emphasizes the role, which functions both up and down). Further, if they moved away from the democratic control model and toward democratic advice, and stopped running a meat puppet for the council, which created some severe political problems, as might be anticipated! -- they might actually unify the town instead of dividing it. I haven't followed Demoex lately.... Yes, Milton has described liquid democracy or delegable proxy, used for voting. But voting is used in control models, and I propose DP for communication filtering. Very different requirements, the advisory model requires much less security.
I wouldn't even suggest delegable proxy for governmental usage. Asset Voting, though, could actually be done, and can be implemented to create a peer assembly with true proportional representation; what's ever more interesting is that the "electoral college" created -- the collection of all those who get votes in what may be a secret ballot election -- could continue to function. Because they are named voters, and it is known where their votes go, any elector could vote directly on any issue, and the "seat"'s vote would be devalued by the appropriate proportion. The problem of scale in democracy has to do with noise, not with voting, as such. We generally want voters to have participated in deliberation, which is why initiatives and referenda, with rare exceptions, are a Bad Idea, except perhaps as public ratifications of some more limited process, but what would happen with an Asset Assembly with direct voting allowed (by electors only!) is that most votes would be cast by those with seats, I'm sure, for efficiency reasons. But electors, some of them, might control enough votes to make it worthwhile paying more attention, and they can always communicate to the floor of the Assembly through their chosen seat -- or seats....

In the systems I've discussed, generally, there is a secret ballot election, and candidates must be declared or identified, any eligible individual may declare a candidacy. Any candidate who gets even a single vote, in systems where high security against vote coercion isn't critical, then becomes an elector. So if you do want to be able to vote on issues before the Assembly, you can. But you only get to participate in deliberation there, except of course for open hearings that an Assembly might hold, if it finds that useful, if you have received, directly or by reassignment, a quota of votes.

It is then possible to handle removals; if the electors can change their votes at any time (a bit like delegable proxy in that way). What I've suggested is latency. If you lose your quota, you can still particpate for a time, the only harm in having a few extra seats is a little extra discussion possibility. The Assembly can even decide to admit certain members based on a determination of usefulness. They don't get any votes other than what they might have obtained through the electoral/reassignment process, but they do get the right to participate in discussion. Basic democratic concept: any assembly can determine its own rules, and who gets to address the assembly is among the rules it can determine.
This is, I'm afraid, the first stupid thing Milton has said here. If an Assembly is advisory and all you get by being elected to the assembly is the ability to advise the community through a coherent structure, exact vote counts don't matter. Sock fraud could, at best, diddle with a very few seats, and revealing sock connections through voting is singularly stupid. In any case, elections are currently held for the WMF Board. Standard qualifications for voting are used. There are countless possibilities for how to qualify voters. I really don't think that the power of Asset Voting is recognized, it's not particularly sensitive to minor errors or manipulations, unlike plurality or even STV systems, which can be quite quirky.

It's far from clear how many Wikipedians would participate in an Asset election, which allows them to simply vote for the candidate they most trust, without worrying at all about the electability of that person, for their vote won't be wasted -- as long as the one receiving it remains active and doesn't toss it in the trash! TANSSTAFL, vote for someone not responsible and not willing to compromise to gain representation, yes, it might be wasted. But if you vote for yourself and become an elector, which means being willing to be public about whom you support, you are in charge of your own vote, you can delegate it to a seat, vote directly -- if you are willing to pay the price of paying attention! -- or, if you can gain enough support, participate directly in discussion. If you don't have a seat, you can still, of course, write to the one who represents you with suggestions. So you end up being filtered by someone you have chosen.

Suppose 3000 Wikipedians did vote. It's possible for the "vote" to be a standing process, so the numbers "voting" could accumulate. The Assembly could shift its own size according to what it finds most efficient and, note, any elector could vote on the changes. Vote, but not disrupt discussion by endless dilatory debate! If there were thirty seats, that would mean each seat would represent 100 votes. Okay, so Jimbo gets 500 votes. Jimbo then names five seats, or keeps one for himself, whatever Jimbo decides he wants to do....

(Standing process, though, might require all voting to be open. WP:PRX used an open file structure; it used a single central proxy analysis table that would collect all proxy assignment files from editors who register by adding a transclusion link to the table; however, that model is probably too vulnerable to disruption; I suggested a different model that allowed independent proxy tables to be constructed. Electors could, in fact, use delegable proxy to guide them in reassigning votes to create seats, thus reducing the decision complexity to the single choice that I predict is most effective: whom do you currently most trust?)

However, the path from here to there could well be advisory structures that would function like the bottom-up structures that would be built through delegable proxy or asset voting, but built top-down, by someone desiring to be advised. Starts very simple: arbitrator, say, starts personal advice mailing list. Want to advise the arb? Send an email to the arb asking to join the list. If the arb accepts, adds you to the list. Okay, list traffic grows, traffic starts to become a problem. Arb has already put some editors, perhaps like me , on moderation, and has permitted other members to approve posts; Arb then defines a carefully-chosen list as direct advisors, members of his or her advisory council, and each of these is encouraged to start their own advisory mailing list. New applicants for the Arb advisory council are asked to choose one member of the council, and if that member agrees, the applicant is assigned and becomes a member of that member's mailing list. Members of the Council pass on posts from their own lists which they consider ready for consideration and not redundant to what has already been said. Pass on too much, the Arb demotes you.... or moderates or arranges to moderate your submissions. Ideally, all this would be documented. You want to communicate with an arbitrator, sure, you can write directly. But as has been noticed, mail falls through the cracks, and if ArbComm were truly functioning as part of dispute resolution process, it would be much, much busier. As it is, taking on only a small portion of what goes on, it is overwhelmed with noise.

If lists are limited to 30 members (it could possibly be much more than that, with only 30 being reasonably active), and there are two layers, we are looking at 900 active participants, with each direct discussion that takes place happening in a relatively small group. The arb has fully authority at all times, over the direct council advisory structure, but if the arb overcontrols, the arb is the one who suffers, by damaging the advice. Each member of the council, with his or her own list, controls that list. Noise control. Each node in the network is independent. The lists can be public or private, that decision is up to the arbitrator or the individual list owner. ("Public" means readable by anyone, in this context. It's essential that posting rights be restricted, when the scale gets large.

An advised arb could ask for trusted editors to investigate directly, instead of using the passive American model where the judges depend on a contest between the parties. What I've noticed is that good arbs do, in fact, do their own investigation. How about having a veritable army of assistants, with the information filtered through a list structure? If other arbitrators imitated this, Wikipedia would start to have a committee that was far, far more intelligent and effective. Details would vary, and it starts very simply, with the concept of building up communication networks that are easy to use. Right now, every Arb has something like this list: their Talk page. But that model is easily disrupted, and blocked editors can't particpate. I'd say that any sensible arbitrator should be as thoroughly advised as possibly by banned editors.... it's important to understand the negative side!

Race and intelligence was just locked because of a dispute over the citation of a NYT article. That particular article has attracted a lot of editors that have been banned, one by Jimbo himself.

What Kato has written about the unreliability of science reporting in the media applies equally well here.

Only by sticking to peer-reviewed journals and the like can thinly disguised bits of race-hate be kept out of articles like this. Nevertheless people have used articles on books by Richard Lynn to post tables of highly dubious data on this matter. There was an ANI thread recently on the removal of all hint of controversy from the BLP of J. Philippe Rushton, one of the other dubious academic psychologists dabbling in the subject.

The article on 'Blonde' spouting some evolutionary psychology sexual selection stuff has only newspaper reports.

But it's all very well saying 'peer-reviewed journals are best' etc. (putting aside just for a minute my criticism of the ludicrous game of peer review 'trumps' in play on Wikipedia that happens when people are POV warring), however, most of these articles are not accessible to the public at large for various reasons. So, the public only gets the wikipedia editors' version of what the 'peer-review' journal article is saying, which is itself always at risk of being distorted and unreliable (and can often be shown to be exactly that), rather like a newspaper article, ironically!