Abstract Wikipedia talk:Notability

Suggesting even sharper criteria

While the current proposal is a bit sharper than the Wikidata notability criteria, I would like to propose to have it even sharper. I think that a Wikidata item should need at least one Wikipedia language sitelink to have Abstract notability. I think of two arguments for this. a) If no one bothered to write about it in any human language, is it really notable? b) Or perhaps they did, but it was deleted as not notable in all languages tried so far. Are we then opening some doors for people to still be able to "write" about very niche topics? ♥Ainali ^discussion_{contributions} 19:10, 18 May 2026 (UTC)

I'm inclined to agree. Writing an abstract article is much, much harder than writing a concrete article. If a topic truly is notable, the author can probably throw together an article in a Wikipedia of any of the languages they speak. If it does miss anything that is actually legitimate to include, well, we're not a bureaucracy, so we can just keep it. Perhaps a note that deletion should be avoided for articles that are obviously notable, but just don't have a concrete Wikipedia article. Feeglgeef (talk) 19:17, 18 May 2026 (UTC)

I don't know if this is a great standard, since it would mean that our notability standard would depend on two factors on which I would rather not rely: (a) the languages the article author knows, and (b) the notability standards of other projects. English Wikipedia has quite high notability standards, so I can understand why someone would be less willing to make an article there. And if they only know English, they wouldn't be able to make one anywhere else. JJP_Master (she/they) 19:44, 18 May 2026 (UTC)

This is true, we shouldn't discriminate against prospective contributors because they know fewer languages, nor should we allow people who know languages with small Wikipedias to write about whatever they want. Perhaps we could require that it be "possible to write a reasonable encyclopedic entry about the topic"? This, however, is incredibly vague, leaving a lot of wiggle room to sysops/RfD voters. Feeglgeef (talk) 19:54, 18 May 2026 (UTC)

It sounds like you are arguing for a right to be able to escape the notability criteria of Wikipedias by writing here. I'll flip that around to a question. If they only know English, and can't write about the topic on English Wikipedia, why should they be allowed to write about that specific topic here? ♥Ainali ^discussion_{contributions} 20:01, 18 May 2026 (UTC)

The question to ask is why shouldn't they. It's a wiki, so you're able to add content by default. Feeglgeef (talk) 20:03, 18 May 2026 (UTC)

Well, if they are allowed to write about anything technically possible, the same argument could go for what they write about it: "anything should be allowed, it's a wiki". I think that is a very weak argument in our movement, that's more of an argument for self-hosted wikis. ♥Ainali ^discussion_{contributions} 06:23, 19 May 2026 (UTC)

I don't mean people should be able to write whatever they want, I mean, philosophically, the burden of proof should be on those who want to restrict. Blog-like content takes up space and bandwith, detracting from our mission of allowing everyone to share in the sum of human knowledge. I mean content should be blacklisted, not whitelisted. Your argument about "if they only know English, and can't write about the topic on English Wikipedia, why should they be allowed to write about that specific topic here?" requires more explanation to show why there should be an active prohibition. Feeglgeef (talk) 15:17, 19 May 2026 (UTC)

I was mainly thinking that since this specific topic won't be imported to English Wikipedia, and that particular user only could verify the (effectively useless) output in English, it just seem like a waste of everyone's time and resources.

I'm not sure we need hard rules for checking the output, but I would really hope that when creating an article, there is a presumption that one should check that it looks okay in at least one language and not create them blindly and move on. ♥Ainali ^discussion_{contributions} 14:52, 20 May 2026 (UTC)

I think I agree with Ainali here. --Ameisenigel (talk) 18:40, 21 May 2026 (UTC)

I like the way the policy is right now. Abstract Wikipedia aims to be language-independent, so I don't think it should restrict itself to articles that have already been created in a specific language. Yes, it may abstract information present in existing articles some or even most of the time, but so long as "[the item] refers to an instance of a clearly identifiable conceptual or material entity that can be described using serious and publicly available references" I don't see why entirely new articles shouldn't be allowed. I think it's important to consider those without a Wikipedia in their language (albeit a minority) and more so those who come to Abstract Wikipedia to contribute information in a completely abstract way from the get-go. Some articles on the language Wikipedias are translated, while others are entirely new. Why should Abstract Wikipedia be any different? I'm not saying there should be articles for everything, but so long as a real article can be constructed, where's the harm? Ultimately it comes down to if the project should literally abstract and aggregate Wikipedia as we know it, or if it should become its own universal collaborative encyclopedia. Some helpful person (talk) 22:15, 22 May 2026 (UTC)

I think you are touching on an important aspect. And the question I wonder about is: Are articles created here primarily meant to be consumed in client Wikipedias, or do we want this project to be where readers go?. If the second, I agree that it doesn't make sense to be restricted by existing articles in other languages (although then I think we need to establish proper notability criteria like the Wikipedias rather than Wikidata). ♥Ainali ^discussion_{contributions} 09:35, 23 May 2026 (UTC)

I would like the notability criteria still a bit sharper:

1. To be notable the Wikidata item should need at least one Wikipedia language sitelink to have Abstract notability (as proposed above).
2. If oldest Wikipedia link is newer than 30 days (or so) the Abstract Wikipedia article may be speedily deleted. This is to prevent promotional articles. 30 days is longer as usual. AW would be a perfect place to spread promotion in all languages. This does NOT imply that the article SHOULD be deleted. It is perfectly acceptable to add a range of articles on a Wikipedia, Wikidata and Abstract Wikipedia in one flow, if the community think that is OK.
3. All facts mentioned in the Abstract Wikipedia article should be present in the linked Wikipedia(s), Wikidata or other Wikimedia projects articles, or closely related articles. This is to prevent fan cruft, only to be maintained in AW.
4. The provided output should be checked by the author. It should be OK in at least one of the languages the article author knows. I am in doubt what to do when new additions will break the output in for example English, or any language the author knows. PS: Babel templates on the user page would preferably be mandatory.

These proposal is not because I am an Deletionist. In fact I consider myself an Inclusionist. I really think this is needed to prevent promotional articles, unverified facts, articles that do not produce a proper sentence etc. HenkvD (talk) 22:09, 26 May 2026 (UTC)

I have to respectfully disagree. I definitely share your concern about spam/promotional articles, but I think you're adding too much work for editors. This is a volunteer project, and the number of people willing to contribute is already low, plus this proposal would just outsource the problem to other wikis. I haven't been convinced we can't adequately prevent spam just like other wikis do, by identifying the spam, nominating it for deletion, and having a sysop delete it. I think this proposal is sort of like TSA in the US, it wastes tens of millennia of human time per year and still only has a 20% success rate. Feeglgeef (talk) 22:42, 26 May 2026 (UTC)

My concern is that one edit can have a large affect. Like one promotional article will be shown o on ALL Wikipedia's (showing AW content), Like adding a complex sentence that only works for English, but will break the whole paragraph in all other languages, A bot operator adding thousands of articles that are poorly formatted, etc. etc. But all is open for discussion. HenkvD (talk) 08:44, 27 May 2026 (UTC)

Do you anticipate humans mainly checking these criteria, or should a bot check on sitelinks for example? Arlo Barnes (talk) 22:17, 26 May 2026 (UTC)

My thought are that humans would check these. If bots can do it it would be very helpful of course. On the English WP any addition should be referenced, and the reference should contains the facts presented. This feels similar to me. HenkvD (talk) 22:34, 26 May 2026 (UTC)

I agree with Feeglgeef, this seems like something that would take a lot of manpower to enforce. Also, I think this reflects a "literally abstract Wikipedia" approach to the project. I'm curious if AW is solely meant to translate existing articles (the introduction suggests it's a replacement for machine translation) or if it should become its own language-independent Wikipedia (encouraging new information)? Some helpful person (talk) 01:51, 27 May 2026 (UTC)

I think both are true: it requires existing sources, and summaries/relevances of the sources, in order to start; but inevitably it will be independent simply due to being a separate wiki, even if it is to be regarded as being mainly 'for' imports to natural language wikis. Arlo Barnes (talk) 03:11, 27 May 2026 (UTC)

• Test case: One class of article I've been pondering is chemical compounds. I think they'd meet the current criteria, but not always the sharper criteria proposed above. A basic article for a molecule may be easier to write on AW than it is on any language, because many facts could be either drawn from Wikidata or templated as a calculation based on that data. This is the kind of article that even English Wiki might want to transclude. I'd be interested in hearing whether you think they are in or out. --99of9 (talk) 12:38, 27 May 2026 (UTC)

  Interesting point. I think this could even be avoided if we get to create "template" articles here, but to be honest, I haven't heard about those in such a long time I am not sure if they are still in the road map. (As I remember them, we just create one template to reuse for all chemical compounds, soccer players, human settlements, etc. so that we don't need to "write" them separately, and also can improve them all at once.) ♥Ainali ^discussion_{contributions} 14:04, 27 May 2026 (UTC)

  The only criteria that would be blocking here is to have at least one Wikipedia language sitelink. I see the chemical compounds to be an addition to the English Wikipedia I think it is very unlikely that English Wikipedia, or any large Wikipedia will consider to allow AW entries in their Wikipedia. And maybe a request for exceptions can be honored in case by case basis. In my opinion it is very useful to have this rule to prevent promotional articles. HenkvD (talk) 14:53, 27 May 2026 (UTC)

  I don't really see how natural language wikis could prevent copying and pasting from rendered output of an abstract article; as long as the attribution is intact, it would be analogous to copying from a public domain Britannica article. But probably most would turn off the option to import automatically. Arlo Barnes (talk) 15:06, 27 May 2026 (UTC)