The loss of prolific contributors in Wikipedia | Hacker News Reader

The loss of prolific contributors in Wikipedia

semanticscholar.org

56 pointspolm2336 comments

Having contributed to the Danish Wikipedia, I was astounded by the arrogance and the accusations by the other contributors/mods. That was what made me not contribute anymore.

It is nothing short of astounding how small amounts of power corrupt otherwise intelligent people.

Sometimes I despair at the state of democratic politics, but looking at the edit wars of Wikipedia, it could have been worse. So much pettiness for nothing.

>So much pettiness for nothing.

well as the saying goes, 'the fights are so fierce because the stakes are so low', or slightly more technically 'In any dispute the intensity of feeling is inversely proportional to the value of the issues at stake'. Wikipedia is like a breeding ground for Girardian terror with people who tend to be very homogenous all competing for very similar things which often are only relevant because someone else wants to exercise control over them. Only place worse might be reddit moderation.

I would definitely take wikipedia politics over american politics. Wikipedia has its problems, but nobody there is violently storming the whitehouse.

As a long time, low volume contributor, I see 2 explanations for what happens:

1. There is not a lot to contribute as a lot of matters are already covered. This is not a bad thing at all.

2. Moderators do a very bad job. Last year I created an entire article about a popular vehicle, it took 6 months to be published and it was just about a page long with solid references. At one time it was rejected because it had "not enough external links", so I added half a dozen links to the dealers selling that vehicle, on top of the original manufacturer page. This discourages contributors and it is a serios problem.

There is plenty of room for more content but deletion focused contributors remove whole classes of articles and tend to turn off whole groups of contributors in the process. The biggest challenge for Wikipedia now is to find some way to tame the rise of deletion as a form of contribution.

Yep, had plenty of bespoke technical pages that explained some pretty involved network infrastructure from the 90s and various hardware families outside Cisco. The delete party would come in like locust and nuke all your work citing all forms of wikilaw. I just can't be bothered. Internet archive and EFF get my money each month. Not the wiki foundation.

Not to defend Wikipedia at all (and they certainly should not get your money), but, the solution that Wikipedia themselves would advocate is that you should publish those explanations on some site of your own, and Wikipedia could then cite it. That also means that the publications are under your control and yours alone and nobody can come in and delete them.

It's pretty great that Wikipedia is a centralized source of information, but I do sort of lament the decline of personal web sites on GeoCities or university web hosts or whatever.

> the solution that Wikipedia themselves would advocate is that you should publish those explanations on some site of your own, and Wikipedia could then cite it

The problem with "deletionism" is not lack of sources and citations. It's the fact some moderators don't want certain material there. Creating sources is not a guarantee you'll be able to add them back, quite the opposite.

In the past I've seen purges of all kinds of well-sourced material: law, electric engineering, literature, important CS/engineering figures. It's never because of lack of sources, it's always some subjective rule.

Actually, I've seen "the content is already available in another website, why do we need it here?" being be used as an argument against reinstating some very uncontroversial articles.

Do the sources even matter? I don't know how many articles I've read in the past that have had some very questionable "facts" so I check the source link and it's a dead link.

How does that work?

Great point.

In practice dead links should be and are mostly converted to web.archive.org links. It's not a problem for them to be dead, they just have to exist somewhere and the original source should be classified as reputable. It's hard.

My opinion is that it should be inclusionism and not deletionism. Include many thing as needing.

> Actually, I've seen "the content is already available in another website, why do we need it here?" being be used as an argument against reinstating some very uncontroversial articles.

Having it on Wikipedia is helpful. This way, you can find it in Wikipedia, can have a free license, mirrors (if there are any), can use MediaWiki API, and will have the wikitext you do not worry as much about the hostile JS/CSS and can also print out. And, you can also fix it if there is a mistake in the article, too.

They should also try to be neutral from corporate or other interests (although if there are multiple points of view about some subject, they can try to be included if you can avoid being biased, but in the total it should be neutral).

> My opinion is that it should be inclusionism and not deletionism. Include many thing as needing.

I agree with you, unfortunately we're losing the battle here.

There are already plenty of rules defining what fits or not in Wikipedia: not only it needs sources, but also notability and it can't be original research.

The issue is not that those rules aren't being followed: is that deletionist mods are deleting articles, metadata, lists, and everything else, even when those rules are followed. And they're also making it harder to create new articles.

The quality of a product is just as much about what it contains as what you leave out.

Deletionism vs inclusionism is a balance, and i don't know if wikipedia has the right balance, but i'm glad its not everything2. If "everything goes" content-wise, then you lose consistency and quality, which is a big part of what makes wikipedia great.

Part of the problem is that philosophically, Wikipedia wants to pretend that contributors are "thin" interfaces for pure knowledge. That there is a well-defined set of "reliable sources" and all contributors have to do is summarize and create hyperlinks to them.

Not surprising given the ~Objectivist philosophies of its creators.

What's the alternative? Acknowledge that there is no Truth, and sit down and cry? That may be technically more correct, but i fail to see how one can build an encyclopedia under that view.

An alternative is to admit that knowledge is social, and include a lot more data on who claims and interpretations are coming from. It should be a first class feature of such an encyclopedia that I can hover over any sentence/paragraph and immediately learn who is behind it, rather than herding cats with diffs and reference links.

"Wrong" interpretations are perfectly useful data as long as I know who is making it.

I don't think i have ever had a problem figuring out which citation corresponds to which sentence.

Unless you mean which user added it? I fail to see what a potentially ephemeral pseudonom really tells me. Its not like you need to present a drivers license to edit wikipedia. Anyone can be anybody.

But even ignoring that, i suppose i just don't agree. Seems like one big exercise in the genetic fallacy. I want people to show their work, and have their encyclopedic writing work be evaluated on its merits. I'm mistrustful of systems that put stock in people's reputations over what they actually do.

Yes, it requires contributors to have more skin in the game. Zero-cost contribution is the main reason why some topics on Wikipedia are outright disinformation "contributed" by special interest groups. And don't get me started on the "regulatory capture" of power users...

It was never and can never be sustainable to depend on a small number of super-contributors for a "crowd-sourced" knowledge store.

Maybe the Wikimedia Foundation should focus more on attracting and keeping a broader range and number of contributors instead of curating few it considered "good".

Why not? That's how the rest of the world works. Almost every human endeavour forms a small world graph.

How's that working out?

Decently well? Society has its problems, but its had some pretty cool accomplishments along the way.

More to the point, i think your argument would be more compelling if you had some examples of non-"small world graph" endeavours that worked out well. I'd love for some experiments, but i think we need some succesful experiments before saying that its a better model.

> Decently well?

There are millions of oppressed, desperately poor, and disadvantage people who would scoff at that assessment.

This study sounds like it's part of a strategy to establish corporate control of Wikipedia

I read the paper and I didn't come to that conclusion. Can you explain why you think that?

Here is their stated purpose:

The primary objective of this work has been to bring forth the issue of the growing depletion of editors, especially the experienced editors in Wikipedia.

One may be able to take their data and then determine if certain editors are near the quitting threshold. The data may also reveal operational and environmental conditions that could be changed to limit the loss of experienced editor.

From the abstract:

> a major concern for not only the future of this platform but also for several industry-scale information retrieval systems such as Siri, Alexa which depend on Wikipedia as knowledge store.

That doesn't imply that the paper is advocating for it, but given that Apple and Amazon now have built products on top Wikipedia, their bottom line depends on it. It's entirely reasonable to wonder if they would prefer to have more influence and control over it. Whether or not it was ever a good idea for Alexa and Siri to have a dependency on Wikipedia is a moot point. They do now, and it wouldn't surprise me to see them wanting to take an active part in keeping Wikipedia fresh.

Of course, because their revenue depends on it, they probably would want more control. In the same way the Amazon is working to exercise more control over the Rust language, Apple or Amazon could decide that taking over Wikipedia is the right move to protect profits.

> Of course, because their revenue depends on it, they probably would want more control.

I don't think this follows. They could have all the control they wanted if they set up their own product, but their revenue doesn't depend on control, their revenue depends on the product being good. It would take a lot of work to make your own Wikipedia-alike; it would take a lot of work to even start with the current Wikipedia (which they legally can, since its license permits commercial use) and keep it up to date.

> their revenue depends on the product being good

Yes, and what happens when the content of Wikipedia, through neglect, gets out of date or (negligently or maliciously) wrong? Imagine the following:

User: "Siri, who is the president of North Macedonia?"

Siri: "Macedonia is a geographic and administrative region of Greece, in the southern Balkans"

I don't know, but I wonder if the loss of these contributors is also why wikipedia became so biased.

The bigger Wikipedia gets, the less contribution is required, since the “static” content gets more and more covered.

that's been my impression as well. when I look at articles focused toward my field I struggle to figure out how I can contribute without needlessly adding complexity or excessive detail to--what amounts to--an encyclopedia entry. Wikipedia isn't done but a lot of the low hanging fruit has been plucked.

I think there is plenty room for more content, but a lot of it requires more than average level of expertise, or contributions from demographics that historically haven't contributed to Wikipedia.

Reply on news.ycombinator.com