Jump to content

Task force/Community Health/Making Wikipedia a Happier Community

From Strategic Planning

This is largely a personal essay based on personal experience. It was written a week or so after the events narrated by way of illustration in the essay occurred, while they were still fresh in memory but some time had occurred for me to reflect on them."

I recently withdrew from en.Wikipedia in the wake of a physical speedy delete to a reasonable first stub draft of an article which occurred within 1 minute of the article having been created. The speedy was swiftly reverted but then drew a subsequent speedy request. This was politely but firmly rebutted. There then ensued a critical tag-fest of biblical proportions. With 9 years worth of editing WP and about 7 as an admin under my belt I was frankly appalled. If this was happening to me, how much more frustrating and hostile, to the point of abusiveness, this must appear to a new user creating one of his or her first articles. In brief, the article in question was on Authonomy, a web-site owned by HarperCollins, itself a subsidiary of NewsCorp, the largest (or second largest depending on how you do the numbers) media conglomerate on the planet. Hardly a candidate, I would contend, for a speedy.

I am unlikely to return. I have worked on the project since almost its inception, a tour of some of the contributions archived in NostalgiaWiki shows some of my edits going back into the spring of 2001. Things have now, in my opinion, reached a critical stage with deletionism and bad-faith drive-by tagging running rampant. In fact, in my pique, on exiting, I blanked my user page only to find that it too was speedily deleted within a couple of days per GSD7. This is unlikely to impress any editor who does similarly, and reconsiders, nor did it me, in fact it confirmed to me that once and for all the inmates had a cast iron stranglehold upon the running of the asylum. Not that I have any intention of returning, certainly until we move away from the current dirigiste regime of policy-driven content and back to one which is closer to the ideal of content-driven content. These two concepts, and a third, content-driven policy, are ideas which I will define and elaborate on since they seem fairly fundamental to the debate.

Policy-driven content

The term policy-driven content in the context of this essay has a very precise and specific meaning. It determines a process of content generation which is explicitly predicated by the definition of rules and policy. In its extremest sense, it represents a scenario in which no one other than those of a brutal and explicitly bureaucratic inclination would ever voluntarily work on a project since creativity, innovation, style, and all those other similarly fundamental characteristics of good writing would be (and are being) stifled. It further describes a condition to which Wikipedia, in my personal estimation, is very close to approximating. Moreover, as is evidenced by the decline in contributors, and the spillage of both new and old editors alike, it is one which is symptomatic of a hostile community in which reasonable and sensible middleground editors are more likely to be driven out than those at either end of the polarities of policy-abuse. Clearly having policy determining content is more consonant with paid employment than with something which a user will voluntarily and freely give of his or her time.

As an exemplification of the current state of affairs I will cite selectively and fairly extensively from a few of the supportive mails I have received from members of the community (and yes, permission was sought to quote and granted in all cases, and anonymity is guaranteed). If nothing else it illustrates the fact that I am not alone:

  • "[..] like you I think the whole thing now is more trouble than it's worth and you'll find plenty of better things to do with your spare time like I did. [..] You'd find it a lot quieter & generally more amenable in hell than in Wikipedia to be frank. [..]" (an ex-admin of about 4 -5 years standing)
  • "[..]The system actually punishes you for being reasonable in a dispute, and actually punishes you for compromising. You lose support, and you surrender your power. It rewards you for being consistently hard-headed and blindly loyal. You always have support and, worst case, you can stop anyone else from getting their way. It's gangland. There's no law, so the only protection is a gang. I left when I realized the gangs had taken over.[..]" (another former user/admin of long standing)

None of this happened in the old days. Frankly we were much too busy; there were articles which needed creating, servicing, beating into shape. This is not to say that disputes did not happen. They did. They were also frequently of a large order of magnitude. But by and large, with a few notable exceptions, and a few sadly missed departures, the real fallout was much less since the volunteer base was far more committed, and this again was by a large order of magnitude. There really wasn't the time for deletionism, and the likely probability is that the majority of users who currently ride shotgun over the carefully referenced, sterile, non-existent and blank pages which is their nihilistic vision of nirvana, would have been kicked for vandalism.

Policy-driven content is in real terms, officially sanctioned vandalism on the grand scale. It offers valuable, sensible, middleground contributors little or nothing in the way of any sort of reward, tangible or intangible, whilst the deletionists and rabid inclusionists conduct their respective scorched earth policies of attrition on the once fertile soil of Wikipedia. Policy-driven content will inevitably result in a Wikipedia which is fundamentally useless. It will wither on the vine; if every assertion is necessarily sourced and attributable it is effectively nothing more than a cut-price aggregator with some spurious commentary.

Content-Driven policy

The term 'content-driven policy' also has precise meaning within the context of this essay. It is not the antithesis of policy, or indeed, of policy-driven content, it is something else. What it means is that the apportionment and application of policy to any article should be relevant and proportionate to the nature and maturity of the article. It is closer to policy driven content than content-driven content, since policy is still a primary driver in this scenario.

The corollaries of this approach are that speedy deletes on anything which looks remotely like it might turn into an article in the near future may not under any circumstances be speedy deleted. The onus of proof should be on the deleter and not on the article originator. This is fundamentally consonant with one of the central founding tenets of Wikipedia, that Wiki is not paper [1]. It was a sad day that this particularly important document was ever shipped off to meta and not made compulsory reading for every new user. The overall quality of Wikipedia would have been consequentially much richer. With all the good middleground (and some of the better extremists) being weeded out by the policy nuts, rabid deletionists and the inclusionist rabble, quality is the first to go by the board. As an example, the Authonomy article which was my moment of no going back, now has no tags plastered all over the page, less information and is less informative than at my last edit. It also has no more references. It is +/- the same article minus the offending tag spree. The article has not moved on substantially, Wikipedia is minus one deeply experienced admin/editor/user, and the only net winner is myself, with more time to spare on more amenable projects and fewer objectionable pedants to deal with. Had content-driven policy been the guiding light, the users involved would have had to have recourse to using their brains instead of their keyboards; in all probability there would have been no speedy because they would have to look and see wtf HarperCollins was all about, where this particular line of argument might ultimately lead (it's pretty significant imo from the point of view of how mainstream media gathers and disseminates information) and by leaving well alone would have allowed the article to develop, as in my view WP articles should, organically, naturally, with all bringing their knowledge and expertise to the party. I cannot emphasise enough how strongly I feel about this; it is at the very nub of what exactly is wrong with the current status quo, one in which policy defines content and not as should be the case the other way around.

Content-driven content

Given the fundamental hypothesis that Wikipedia is not paper [2], and that over-zealouness in inflicting policy appears to be a major problem, we need to do some drastic things to correct the marked imbalance between contributors and enforcers, and the negative impact that the latter invariably have upon the former, I am bringing a number of proposals to the table. Chief amongst these is a return to basics, which is to say that the emphasis should firmly be put upon content over policy. If the content is of sufficient quality, which, given another central wiki tenet, given enough eyeballs, all bugs are shallow, it inevitably will be, we can let policy take care of itself and move forward to a new golden age, one which was paramount in the early days of Wikipedia, one in which good content generates more good content.

How do we get there (again)?

Move away from evidence based corroboration and back to good faith writing

What Wikipedia is and always has been about is content. This should be the central focal point of all editors without exception.

Content, it should be noted, is only as good as the contributor can make it. A user should be able to and invariably can, justify an edit.

Finding and citing references is a notorious pain and has resulted in much editor dissatisfaction. We need to make citation seamless. At the moment creating a cite often takes longer than writing the sentence which it supports: this is the very opposite of content-driven content and more consonant with policy-driven content, or in fact an even more extreme trope towards which WP is inexorably headed without remedy, policy-driven policy. Simple, obvious statements of fact, which are unchallenged, should not need half an hour's worth of laborious evidence gathering and cross-referencing, yet time and again I see the citation needed tag splattered mindlessly across a page. If they are to be challenged a simple question to the user making the edit to corroborate and document the question on the article talk page should be more than sufficient, except in the few very rare extreme cases of considerable and controversial dispute, in which case all the citations in the world aren't going to change an antagonist's POV.

Key Points:

  • Simplify citation.
  • Assume good faith edits.
  • If challenged the user must justify the edits or be prepared to take, or have them taken, down. Alternatively, if you're that concerned about corroboration, fix the citation yourself and do not put the onus on the editor.


We need to re-educate the serial deletionists and psycho-taggers. This in itself is not a particularly radical proposal (except maybe for the wannabe keyboard Judge Dredds out there, of whom there are not a few).

I would propose that any user (or indeed admin) may only tag or speedy or AFD up to the number of substantial edits that they themselves are responsible for, with obvious exceptions, caveats, and mitigation for occurrences where they are manifestly dealing with palpable vandalism; well I would except that this in itself is a form of policy creep. What is needed is an acceptance of this as a guiding principle, which is enforceable in extreme circumstances and where abusive and markedly antipathetic behaviour is perceptible to keep them honest. Obvious sanctions include removal of admin rights, removal of tagging rights (i.e. they are unable to commit an edit containing any tag if their account is flagged accordingly), or similar. This will considerably curtail the activities of those who are only working on Wikipedia to flex their e-penises by drive-by tagging or random speedying new articles. It might even make them spend some time doing something more constructive which is fixing the articles they so recklessly diss with the ease which they currently enjoy.

Key Points - Speedy/AfD

  • A new agreement, to be either explicitly or implicitly acknowledged by all users, that all users will endeavour to improve rather than destroy or mutilate any article by speedy/AfD or vandalise by tagging. First day speedy deletion should be a last resort for copyvios (and even here there are reasons why brains should be engaged before reaching for the delete key - see [3] where I was at one stage apparently in breach of my own copyright which I had implicitly relinquished in adding the contents of an article I had written previously to WP), and palpable vandalism - fortunately this occurred and was ironed out in the days when civility was a given and not a luxury afforded only to the members of the current ruling hegemony.
  • The rule should be to ask 3 questions before considering a speedy/Afd on a new article:
    • "Is this article going anywhere?"
    • "Is it a genuine attempt at an article rather than pure and simple self-promotion or vandalism?"
    • "Has it potential to be improved?"

If the answer to any of the foregoing is 'yes' then the article should be left at least 48 hours and a comment left on both the article talk page and the user page that a question mark about the validity of the article has arisen, and that an explanation is requested.

Key Points - tagging

  • The rule should be to ask 3 questions before considering a tag:
    • Can I fix it?
    • Might the article be improved organically, or with my assistance, without recourse to tagging?
    • Will my tag alienate the editor?

If the answer to any of the foregoing is 'yes' then the article should either be fixed accordingly, or a comment left on both the article talk page and the user page that a question mark about the validity of the article has arisen, and that an explanation is requested at least 48 hours prior to the affixation of a tag .


Wikipedia needs to be a more civilised place to work. It cannot afford the human wastage which it is extravagantly pissing down the drain because of the current obsession with policy-driven content. Once an editor is gone, they are almost always gone for ever. There is an almost endless list of talented people whom I have known who have been gradually driven out by their disillusionment with the project, among them my collaborators in forming an outline policy which would shape the early NPOV tenets of history policy, Dr Julia Hoffman Kemp, and Professor Michael Tinkler, whose wit and wisdom are sorely missed in that homogenous sea of largely turgid and uninformed prose which now graces some of the pages which their particular insights once adorned. I would correct some of the more egregious errors which have replaced much of my and their work but I would be up against it I fear, the policy-driven content lobby would have a fit at the excision of many of their fetish references (frequently POV, or just plain wrong) and replacement with a commonsense and pragmatic interpretation of events by people who have studied this stuff into the ground. Wikipedia, in its current form, has no place for people who know their stuff, who are interested or passionate about what they do. If WP is to survive it has to change, and this cannot be tinkering around the edges. It needs root and branch surgery. Every week that elapses, a handful more people with deep skills are lost and more potential newcomers are discouraged and abandon hope. This is not an infinite pool of resource and the pool from whence they came is dwindling to nothing as more civilised and hospitable communities emerge.


  • [4] Nicholson Baker in the New York review of Books on deletionism